IWmm III £ o* 1 1 Oo * 9 i -\ " V ^ ll % J " • * * A° iOo <0* \ ,0 o y r- < 'b. CL^ s A ° N '■ « % o «V &, N .1 <$* V \ 1 V # x v^ .<$<*. < u "V ./ practical Electric Railway Hand Book BY ALBERT B. HERRICK Consulting Electric Railway Engineer SECOND EDITION, REVISED AND CORRECTED NEW YORK ! McGRAW PUBLISHING COMPANY 1906 tr TF SWflMiHIH 1111 -ar M MHK i.t n ii. ■ LIBRARY of CONGRESS Two OotJtejs deceived FEB 23 J 906 J? Coayrlffht Entry CLASS CC XXc* No, t 3 1 >^Z COPY B. Copyright, 1901 by Street Kailway Publishing Company and 1906 by the McGraw Publishing Company New York ^ CONTENTS. PAGE Section I.— General Tables I to 27 :ction II. — Testing ...... 28 to 105 Section III. — The Track ...... 106 to 166 Section IV. — The Power Station . # . 167 to 299 CTioN V. — The Line . . . . . 300 to 362 Section VI. — The Car House . . . . 363 to 366 Iection VII. — The Repair Shop .... 367 to 371 action VIII. — The Equipment . . . . 372 to 441 :ction IX. — The Operation ..... 442 to 450 PREFACE TO THE SECOND EDITION, The reception which the first edition of the Electric Railway Hand Book received at the hands of its readers in the electric rail- way field was most gratifying and the author wishes to thank them for valuable suggestions and data which they have sent him. A considerable part of this matter has been incorporated in this edition. In the second edition a number of sections have been rewritten and expanded and new subjects have been introduced to accord with recent developments in the electric transportation industry. New methods of testing have also been described and data on new types of apparatus have been added. It has been the author's effort to develop this Hand Book along the lines originally proposed and to keep within the limits of what is accepted as conservative engineering. He has also restricted the use of formulae and mathematics as far as possible, so as to make the text useful to the greatest number of co-laborers in this field. ALBERT B. HERRICK. RlDGEWOOD, N. J., Jan. i, 1906. SECTION I.-GENERAL TABLES. TABLES OF WEIGHTS AND MEASURES. MEASURES OF LENGTH. 93 .2 on 2 9 fl M «8 B O o3 .2 03 u O 03 o3 .2 be 03 .2 03 93 1 Inch 1 12 36 198 7,920 63,360 .0833 1 3 660 5,280 15,840 3.2&08 .027- .333 1 5y 2 220 1,760 5,280 1.0936 .00505 .0606 .1818 1 40 3?0 960 .1988 095399 I Foot .001515 .000568 .025 1 8 24 .004971 .304b01 1 Yard .91440 1 Rod, Pole or Perch 1 F urloni? .003125 .125 1 3 .0006214 .001041 .0467 .333 1 5.0298 201.16 1 Statute Mile 1609.3 I League 1 French Meter 89.37 1 Mil equals one thousandth of an inch. 1 Palm equals three inches. 1 Hand equals four inches. 1 Span equals nine inches. 1 Military Pace equals 2^£ feet. 1 Fathom equals 6 feet. Geographical Mile fixed by U. S. at 6,080 feet, or 2,026 yards. 1 Degree of great circle of the earth, 69.77 statute miles. SURVEYOR'S MEASURE.— MEASURE OF SURFACE. Inch Link Chain (Gunther's) Furlong Mile Inches Links ma in a 1 .1262 7.92 1 792 100 7,920 1,000 63,360 8,000 Chains in a .001262 .01 1 10 80 Furlongs in a .0001262 .001 .10 1 8 Miles in a .000125 .0125 .125 1 ) 10 Square Chains equal 1 Acre. 1 Acre equals a square whose side is 208.71 feet long or 69.57 yards long. A strip 1 chain wide runs 8 acres per mile. / ELECTRIC RAILWAY HAND BOOK. SQUARE MEASURE— MEASURES OF SURFACE, lSq. Inch 1 Sq. Foot l£q.Yard 1 Sq. Rod, Pole or Perch IRood 1 Acre lSq. Mile Sq. Incnes Sq. Feet Sq. Yards Sq. Rods P. or P. Rods in a Acres in a in a 1 144 1,269 in a m a in a .006944 1 9 .000771 .11111 1 .0000255 .0036*3 .033059 .0000918 .000626 .0002066 89,204 1,568,160 6,272,640 227.25 10,890 43,560 27,878,400 30.25 1,210 4,840 3,097,600 1 40 160 102,400 .025 1 4 2,560 .00625 .25 1 640 Sq. Miles in a .0000097 .00039 .0015535 1 1 square inch equals 1.2732 circular inches. One circular inch is the area of a circle one inch in diameter having 0.7854 square inches. A circular mil is the area of a circle one thousandth of an inch in diameter. 1,000,000 circular mils equals one circular inch; 1 square inch equals 1,273,239 circular mils. ENGLISH. 10.764 square feet ) 1.196 square yards f 1 square yard 1 square foot .155 square inches 1 square inch .00155 square inches 1 square inch 1 are = 1 square decameter = 1076.41 square feet. 1 hectare = 100 ares = 107,641 square feet, 2.4711 acres. 1 square killometer = .386109 square miles, 247.11 acres. FRENCH. 1 square meter = 1 centiare, .836 square meter. .0929 square meter. 1 square centimeter. 6.452 square centimeters. 1 square milimeter. 645.2 square milimeters. SOLID OR CUBIC MEASURE. Measure of Volume. Cubic Inches in a Cubic Feet in a Cubic Yards in a CuMc Inch 1 1,728 49,656 .000578 1 57 .0000214 Cubic Foot .037037 Cubic Yard 1 1 Cord of Wood = a pile 4x4x8 feet = 128 cubic feet. 1 Perch of Masonry = 16^ xlj^xl foot = 24% cubic feet. 1 IT. S. standard bushel is a cylinder with a diameter of 18^£ inches and 8 inches deep, containing 2150.42 cubic inches, 1.2445 cubic feet. This is known as a struck bushel. A heaped bushel contains 1*4 struck bushels. The capacity of a cylinder in U. S. bushels = square of diameter in inches multiplied by height in inches and multiplied by .0003652. ELECTRIC RAILWAY HAND BOOK. UQUID MEASURE. 3 m PS 00 % 05 CD jo CO u u o3 m CD EH 30 03 CD o 1 a d -t-3 •a PQ u O & s lGill i 4 8 32 1,008 .25 1 2 8 252 336 504 672 1,008 .125 .50 1 4 126 168 252 336 504 .03125 .125 .25 31^ 42 63 84 126 .000992 .003968 .007936 .03174 1 1H 2 21 4 lPint .002976 .005952 .0238 .75 1 2 3 1 Quart .003968 .01587 .50 .6666 1 1H 2 .00297 .0119 .375 .50 .75 1 1H .00198 1 Gallon .007936 1 Barrel .25 1 Tierce .333 1 Hogshead .50 1 Puncheon .6666 1 Pipe or Butt 1 The U. S. gallon contains 231 cubic inches; 7.4805 gallons = one cubic foot. A cylinder 7 ins. in diameter and 6 ins. high contains one gallon or 230.9 cubic inches. The British Imperial gallon contains 277.274 cubic inches or 1.20032 U. S. gallons. The miners' inch varies in different parts of the country— from a delivery of 1.36 to 1.73 cubic feet per minute — due to the varying heads of water above the centre of the aperture. The most prevalent method is the flow of water through a slot 2 ins. high, and whatever length required cut in a plank IX ins. thick. The lower edge of the slot should be 2 incheg above the measuring box, and the plank extend 5 ins. high above the slot, making a 6 in. effective head. Each sq. inch of this slot deliv- ers one miners' inch, and equal to abont 1)4 cubic feet of water per minute. BOARD AND TIMBER MEASURE. In measuring boards and timbers they are estimated in equivalent lumber 1 in. thick. To compute the number of feet board measure in a board or stick, multiply its length in feet by its breadth in feet by its thickness in inches. To compute round timber when all its dimensions are given in feet, find the mean girth and diameter and multiply them together and divide this by four and multi- ply by the length of the timber which gives the result in cubic feet. On square timber, when all dimensions are given in inches, divide by 1728 to get cubic feet ; When two dimensions are given in inches, divide by 144 to get cubic feet; when one dimension is given in inches, divide by 12 to get cubic feet. ELECTRIC RAIL WA Y HAND BOOK. CONTENTS IN FEET OF JOISTS, SCANTLING AND TIMBER. LENGTH IN FEET. Size. 2x4 2x6 2x8 2x10 2x12 2x14 3x8 3x10 3x12 3x14 4x4 4x6 4x8 4x10 4x12 4x14 6x6 6x8 6x10 6x12 6x14 8x8 8x10 8x12 8x14 10x10 10x12 10x14 12x12 12x14 14x14 12 14 16 18 20 22 24 26 28 FEET, BOARD MEASURE. 8 9 11 12 13 15 16 17 19 12 14 16 18 20 22 24 26 28 16 19 21 24 27 29 32 35 37 20 23 27 30 33 37 40 43 47 24 28 82 36 40 44 48 52 56 28 £3 37 42 47 51 56 61 C5 24 28 32 36 40 44 48 52 56 30 35 40 45 50 55 CO e5 70 33 42 4S 54 60 66 72 78 84 42 49 56 C3 70 ■77 84 91 98 16 19 21 24 27 29 32 35 37 24 28 32 36 40 44 48 52 £6 £2 37 43 48 53 59 64 69 75 40 47 53 60 67 73 80 87 93 43 56 64 72 80 88 96 104 112 56 65 75 84 93 103 112 121 131 36 42 48 54 60 66 72 78 84 48 56 64 72 80 88 96 104 112 60 70 80 £0 100 110 120 130 140 72 84 96 108 1C0 1^2 144 156 1C8 84 98 112 126 140 154 168 182 196 64 75 85 96 107 117 128 139 149 80 93 107 120 133 117 160 173 187 96 112 128 144 100 176 192 208 224 112 131 149 108 187 205 224 243 261 100 117 133 150 167 183 200 217 233 120 140 100 iro 200 220 240 260 2F0 140 103 187 210 233 257 280 303 327 144 168 192 216 240 264 2P8 312 336 1GS 196 224 252 280 308 336 384 392 196 229 2G1 294 327 359 392 425 457 30 20 30 40 50 60 70 60 75 90 105 40 60 SO 100 1C0 140 90 120 150 180 210 160 200 240 280 250 300 350 360 420 490 MEASURES Or ' WEIGHTS. Grains in a Ounces in a Pounds in a Grammes in a Kilogrammes in a Grains 1 437.5 7000.00 15.432 15432.36 .00228 1 16 .03527 35.274 .000143 .0625 1 .00205 2.2204 .06479 28.349 453.59 1 1000 .000064 Ounces, adv , Pounds, adv Grammes .02835 .45359 .001 Kilogrammes 1 1 carat is 3.168 grains or .205 grammes. 1 stone is 14 lbs. 1 quintal is 100 lbs. 1 quarter is 28 lbs. 1 hundred weight is 112 lbs. There are twenty hundred-weight to one long ton or 2240 lbs.- Net or short ton is 2000 lbs. Metric ton is 2204.6 lbe. In shipping, 100 cubic feet is equivalent to one registered ton. V ELECTRIC RAILWAY HAND BOOK. MENSURATION. TABLE OF REGULAR POLYGONS. ■■ -■■< 3.1416 8.34i0 3.5466 3.7583 8.9761 2 2 2 ' 3 ELECTRIC RAILWAY HAND BOOK. AREAS AND CIRCUMFERENCES OF CHICLES.— Continued . Diameter, Circumference, Inches Feet. Area, Sq. inches Feet. Diameter. Inches. Feet. Inches. 5-16 3-8 7-16 1-2 7.2f f 4"3 7.4G128 7.657 3 7.85398 4.2000 4.4301 4.6664 4.9087 2 2 6 9-16 5-8 11-16 3-4 8.05033 8.24G68 8.44303 8.63938 5.1572 5.4119 5.6727 5.9396 2 2 9 13-16 7-8 15-16 8.83573 9.03208 9.22843 6.2126 6.4918 6.7771 2 mi 3. 1-16 1-8 3-16 1-4 9.42478 9.62113 9.M748 10.0138 10.2102 7.0686 7.3662 7.6699 7.9798 8.2958 •I 3 3 3 5-16 3-8 7-16 1-2 10.4065 10.0029 10.7992 10.9956 8.6179 8.9462 9.2806 9.6211 3 3 6 9-16 5-8 11-16 3-4 11.1919 11.3883 11.5846 11.7810 9.9678 10.321 10.680 10.045 3 3 9 13-16 7-8 15-16 11.9773 124757 12.3700 11.416 11.793 12.177 8 vu 4. 1-16 1-8 3-16 1-4 12.5664 12.7627 12.9591 13.1554 13.3518 12.566 12.962 13.364 13.772 14.186 4 4 4 m 3 3 5-16 3-8 7-16 1-2 13.5481 13.7445 13.9408 14.1372 14.607 15.033 15.466 15.904 4 4 6 9-16 : 5-8 | 11-16 3-4 14.3335 14.5299 14.7262 14.9225 16.349 16.800 17.257 17.721 4 4 7^ 9 - 13-16 "7-8 15-16 151189 15.3153 15.5116 18.190 18.665 19.147 4 i(H 5. .4-16 18 15.7080 15.9043 16.1007 19.635 20.129 20.629 5 5 M 10 ELECTRIC RAILWAY HAND BOOK. AREAS AND CIRCUMFERENCES OF CIRCLES.— Continued. Diameter, Inches. Circumference, Inches. Feet. Area, Sq. inches. Feet. Diameter. Feet. Inches. 3-16 1-4 16.2970 16.4934 21.135 21.648 5 3 5-16 3-8 7-16 1-2 16.6897 16.8861 17.0824 17.2788 22.166 22.691 23.221 23.758 5 5 4^ 6 9-16 5-8 11-16 3-4 17.4751 17.6715 17.8678 18.0642 24.301 24.850 25.406 25.967 5 5 7V4 9 13-16 7-8 15-16 18.2605 18.4569 18.6532 26.535 27.109 27.688 5 Wi 6. 1-8 1-4 ! 3-8 1-2 18.8496 19.2423 19.6350 20.0277 20.4204 28.274 29.465 30.680 31.919 33.183 6 6 6 6 6 I* * 5-8 3-4 7-8 | 20.8131 21.2058 21.5984 34.472 35.785 37.122 6 6 6 ft. 1-8 1-4 8-8 1-2 j 21.9919 22.3838 22.7765 23.1692 23.5619 38.485 39.871 41.282 42.718 44.179 7 7 7 7 7 3* 5-8 l 3-4 7-8 23.9546 24.3473 24.7400 45.664 47.173 48.707 7 7 7 10H \ 8. 1-8 1-4 3-8 1-2 25.1327 25.5254 25.9181 26.3108 26.7035 50.265 51.849 53.456 55.088 56.745 8 8 8 8 8 P F 5-8 34 7-8 | 27.0962 27.4889 27.8816 58.426 60.132 61.862 8 8 8 9. 1-8 1-4 3-8 1-2 28.2743 28.6670 29.0597 29.4524 29.8451 63.617 65.397 67.201 • 69.029 70.882 9 9 9 9 9 5-8 ft 30.2378 30.6305 31.0232 72.760 74.662 76.589 9 9 9 10^ II ELECTRIC RAILWAY HAND BOOK. II AREAS AND CIRCUMFERENCES OF CIRCLES.— Continued, Diameter, Circumference, Inches. Feet. Area, Sq. inches. Feet. Diameter. » Feet. Inches. ... 1-8 1-4 3-8 1-2 31.4159 • 31.8086 32.2013 32.5940 32.9867 78.540 80.516 82.516 84.541 86.590 10 10 1 10 10 10 3 9* 5-8 3-4 7-8 33.3794 33.7721 34.1648 88.664 90.763 92.886 • 10 10 10 11. 1-8 1-4 3-8 1-2 34.5575 34.9502 35.3429 35.7356 36.1283 95.033 97.205 99.402 101.62 103.87 11 11 11 11 11 5-8 M 3-4 7-8 36.5210 36.9137 37.3064 106.14 108-43 110.75 11 11 11 F 12. 1 37.6991 113.10 12 1<% EARTHS, ORES, STONES AND MISCEIXANEOUS. Weight, Material. lbs. per cubic ft. Asbestos, starry 192 Asphalte 150 Asphaltum 87 Belts, leather, per sq. ft., per ply thickness (13-16) Bitumen, red * 72 Bitumen, brown , 52 Borax 107 Brick, best pressed 150 Brick, common 112 Brick, fire 140-150 Brick, hard 125 Brick, soft, inferior 100 Brickwork, ordinary 112 Brickwork, pressed brick 140 Brickwork, coarse, inferior soft bricks 100 Carbon 219 Cement, hydraulic, ground loose, Rosendale 50-56 Cement, hydraulic, ground loose, Louisville 50 Cement, hydraulic, ground loose, Copley 54 Cement, hydraulic, ground loose, Portland 95-102 Chalk 95 Chalk 174 Clay 120 Clay, with gravel 155 Coal, anthracite, Pennsylvania 93 Coal, anthracite, broken to any size, loose 52-56 Coal, anthracite, broken, moderately shaken 56-60 Coal, anthracite, broken, 40-43 cu. ft. per ton. r 12 ELECTRIC RAILWAY HAND BOOK EARTH, ORES, STONES, ETC.— Continued. Weight Material. lbs. per cubic ft. Coal, bituminous 84 Coal, bituminous, broken to any size, loose 47-52 Coal, bituminous, broken, moderately shaken 51-56 Coal, bituminous, broken, 43-48 cu. ft. per ton. Coke, loose, of good coal f 3 Concrete, dry, 130-160, average , 150 Earth, common loam, dry, loose 72-80 Earth, common loam, cry, shaken 82-92 Earth, common loam, dry, moderately rammed 90-100 Earth, common loam, slightly moist, loose 70-76 Earth, common loam, quite moist, loose 66-68 Earth, common loam, quite moisl^ shaken 75-90 Earth, common loam, quite moist, moderately packed 90-100 Earth, mud dry, close 1 80-110 Earth, mud wet, fluid 104-120 Emery 250 Flint 162 Glass, window or flooring 157 Granite 160-180 Graphite 137 Gravel 109 Grindstone 134 Gutta-percha. 61.1 Ice at 32 degrees Fahrenheit 57.5 Leather 60. Leather belts, per sq. ft., per ply thickness (13 16) Lime, quick, ground loose or in small lumps 53. Lime, quick, ground well shaken 64. Lime, quick, ground thoroughly shaken 75. Lime, hydraulic . . . . i 171 Limestone 168 Magnesia carbonate 150 Magnesium 103 Marble, Dorset, Vermont 105 Marble, East Chester, New York 180 Marble, North Bay, Wisconsin 175 Marble, Italian, common 168 Marble, Mill Creek, Illinois, drab 172 Masonry, of granite or limestone, well dressed 165 Masonry, of sandstone, well drtssed 144 Masonry, of mortar rubble, well scabbled 154 Masonry, of mortar rubble, dry, well scabbled 138 Mica 171-193 Millstone 155 Mortar, hardened 67-118 Mud (see Earth). Paving stone 151 Pitch 72 Plaster of Paris 141 Plaster of Paris, ground loose 56 Plaster of Paris, ground well shaken 64 Plumbago 131 Porcelain 140 Quartz 1C5 Quartz, finely pulverized 90-112 ELECTRIC RAILWAY HAND BOOK. 13 EARTH, ORES, STONES, ETC.— Continued. Weight Material. lb*, pi r cubic ft. Rosin 69 Rotten stone 12 i Rubber 58 Salt, coarse 42-70 Salt, fine table 49 Sand, perfectly dried, loose, usually 90-1% Sand, naturally moist, loose, usually 80 i»0 Sand, perfectly wet ll.s-120 Sandstone, building 151 Sandstone, quarried and piled 86 Sewer pipe 141 Slate 168-181 Slate, purple 174 Snow, fresh fallen 5 12 Snow, compacted by rain 15-50 Sulphur 125 Tallow 58.6 Tar, coal 62 Terra-cotta, solid 122 Terra-cotra, hollow, \% in-*, thick, including air spaces 65-80 Terra-cotta, nollow, 12 x 18 ins. or larger on face 70 Trap rock 1 87 Trap rock, broken, in piles 107 Turf or peat, dry, unpressed ..: 20-30 Wax, bees „ 60 . 5 Lbs. per Tons per. Tensile Strength. sq. in. (2000 lbs.) eq. ft. Brick, 40 to 400.... 220. 15.8 Cement, hydraulic, Portland, pure, 7 days in water 300 21 . 6 Cement, 6 months old 450 32.4 Cement, 1 year old 550 39.6 Common hydraulic cements average 1-6 as much. The last.neat, adhere to brick and stone with from 15 to 50 lbs. when only 1 month old 82 2.3 At end of 1 year 3 times as much 96 6.9 Concrete 180 13 Glass, 2,500 to 9,000 5,750 414 Glue holds wood together with from 300 to 800 550 89 . 6 Granite 1/00 7 J Gutta-percha 3,500 252 Leather belts, 1,500 to 5000. Good ; 3,000 216 Marble, stron?, white, Italy 1 .034 74 .5 Marble, Champlain, variegated 1,666 120 Marble, Glenn's Falls, N. Y , blk., 750 to 1034 892 (4.3 Marble, Montgomery Co., Pa., gray 1,175 84.7 Ma ble, Montgomery Co., Pa., white 734 53 Marble, Lee, Massachusetts, white 875 63 Marble, Manchester, Vermont, 550 to 800 675 43. 6 Marble, Tennessee, variegated 1,034 74 . 5 Mortar, common, 6 months old, 10 to 20 15 1.08 Plaster of Paris, well set 70 6 14 ELECTRIC RAIL WA Y HAND BOOK. EARTH, ORES, STONES, ETC.— Continued. Lbs. per Tons per Tensile Strength. sq. in. (2000 lbs.) Material. sq. ft. Rope, Manilla, best 12,000 864 Rope, hemp, best 15,000 1080 Sandstone, Ohio 105 7.58 Sand-tone, Picton, N. S 434 31.2 Sandstone, Connecticut, red 590 42.5 Slate, Lehigh 2,475 178 Slate, Peach bottom, 3,025 to 4,600 3,812 275 Stone, Ransome's artificial 300 21.6 Compressive Strength. Lbs. per sq. in. Brick 550-4,100 Brickwork, ordinary, cracks with 280-420 Brickwork, good in cement 420-550 Brickwork, first rate in cement 700-970 Cement, 7 days in water, Portland, neat 1,050-2,100 Cement, 7 days in water, U. S. common, neat 210-420 Concrete. Portland, sand and gravel or broken stone. . . 165-260 Concrete, Portland, 6 months old 670-1 ,000 Concrete, Portland, 12 months old 1,000-1,670 Concrete, with common hydraulic cements about \ to | as much. Flagging, North River, N. Y 13,400 Glass, green crown and flint 18,000-32,000 Granite, U. S 13,000-28,000 Ice, pure, hard j. 290-900 Ice, inferior 220-820 Limestone, U. S 6,000-23,000 Marble, Lee, Massachusetts 23,000 Marble, Rutland, Vermont 10,700 Marble, Montgomery Co., Pa 10,000 Marble, Colton, California 17.800 Marble, Italian 12,100 Mortar, 1 of lime, 3 of sand, \i% months 118-135 Plaster of Paris, 1 day 550 Plaster of Paria, 4 months 1,980 Tons (2000 lbs.) per sq. ft. 40-300 20-30 30-40 50-70 75-150 15-30 12-18 48-72 74-120 960 1,300-2,300 940-2,000 21-64 16-59 430-1,660 1,660 770 720 1,280 870 8.8-9.7 40 142 Rubble masonry, good coursed is f % of that of the stone of which itis built. The strength of common rub- ble is not much greater than its^mortar. Sandstone, American 6,000-12,000 Sandstone, New York 10,000-42,000 Slate 5,500-11,000 Terra-cotta, solid , . . , '. . 5,200-7,000 430-860 720-3,000 400-bOO 375-500 ELECTRIC RAILWAY HAND BOOK'. *i 15 METALS. Material. Aluminum, bar ... Aluminum, ca*t .. Aluminum, rolled Antimony, cast . . . Bismuth, cast. Copper, bolts Copper, cast ! Copper, electrolytic. , I Copper, rolled plates, Gol^, cast, pure. Gold hammered. 7ron, cast Iron, malleable . Iron, structural. Iron, wrought . , Lead, cast . . Lead, pipe. . Lead, red . . . Lead, rolled. Mercury- Nickel Nickel, cast. Platinum^ hammered Platinum, roiled Silver, cast, pure , bilver, hammer d, Si eel, cast, from ., Steel, cast, to Steel, plate Steel, rails Steel, rivet .. Steel, shaft Steel, structural • , Tin ., Zinc, cast... Zinc, rolled, Weight, Cubic Inches. .0937 .0932 .0072 .0938 .351 .321 .314 .322 .318 .697 .704 .260 .278 .278 .411 .414 .324 .412 .491 .318 .299 .798 .379 .380 .284 .284 .284 .2^4 .284 .284 .284 .266 .248 .260 Lbs. per Cubic Foot. 162 161 1C8 162 607 555 542 556 550 1,204 1,217 450 450 480 480 711 716 560 712 849 542 517 1,271 1,379 654 657 490 490 490 490 490 490 490 459 429 449 Ultimate Strength, Lbs. per Sq. In. Tensile. 28,000 15 000 24,000 1,000 3,200 36,000 20,000 30,000 20,000 20,000 48,200 42,000 50,000 2,050 1,650 2,500 55,000 40,000 70,000 70.000 G0.000 70.000 54.000 85.000 65,000 4,600 8,°50 7,500 Comnreseive. 12.0.,0 100,000 117,0.0 100,000 100,000 50,000 7,350 105,000 250 000 120.000 100,000 15,500 1 16 ELECTRIC RAILWAY HAND BOOK. AIXOYS. Material. Weight, Cubic Inches. Lbs. per Cubic Foot. Ultimate Strength, Lbs. per Sq. In. Tensile. Compressive. Aluminum Bronze, 1^ per cent. Al .313 .261 .264 .297 .293 .293 .307 .316 .297 .314 .333 .333 .252 .252 541 451 456 514 506 506 530 546 514 543 576 576 436 436 25,000 100,000 Aluminum Bronze, 11 per cent. Al 130,000 Babbitt Metals Brass, sheet 31,000 18,000 18,000 36 000 36,000 23,500 23,500 22,000 74,000 81.700 92,200 Brass, cast, from 50,000 Brass, cast, to 160,000 Bronze, gun metal, from. . . Bronze, gun metal, to Bronze, ordinary, from. .... Bronze, ordinary, to Bronze, phosphor, from. ... Bronze, phosphor, to German Silver, from German Silver, to WIRES. Material. Aluminum, from Aluminum, to. . . . Bi-Metallic (Copper Steel) .. Brass, annealed Brass, hard Bronze, phosphor, annealed, Bronze, phosphor, hard Bronze, silicon, from Bronze, silicon, to Copper, soft, from.. Copper, soft, to ... . Copper, hard, from Copper, hard, to.. . . German Silver, from German Silver, to. . . . Goid, from Goid, to Iron, bright Iron, gal. line wire " B. B." . . . . Iron, gal. line wire "E. B. B.'\ Piano Wire, from Piano Wire, to Platinum, annealed Steel, bright Steel, gal. line wire Silver, annealed Pounds per Million cir. Mil. Ft. .919 .919 2.87 2.86 2.86 3.14 3.14 3.04 3.04 3.027 3.027 3.027 3.027 2.38 2.38 6.60 6.60 2.65 2.63 2.63 7.3 2.67 2.65 8.46 Cubic Foot. Tensile Strength. Lbs. per circ. mil. 167 167 526 524 524 576 576 558 558 555 555 655 555 436 1,210 1,210 486 482 482 490 486 634 .0236 .0511 .0511 .0386 .063 .0495 .118 .044 .118 .025 .030 .0354 o0534 .0642 .0725 .0195 .063 .046 .0416 .236 .267 .0416 .081 .0515 .0314 - ELECTRIC RAILWAY HAND BOOK. 17 WOOD. Common Name. Apple Ash, Amef. White, Bamboo Birch .. •Cedar, Amer. Cherry "Chestnut Cypress Elm W Hemlock Hickory, Amer. Iron Wood, Black Lignum Vitae, Amer. Mahogany Maple ;Maple, Bird's Eye. . . . Oak, Live ■Oak, White iOak, Red Weight per Cu. Ft. From iPine, White 1 Pine, Yellow, Northern Pine, Yellow, Southern, Long Leaf Spruce Tamarack Teak Walnut, Black ; White Wood . . Willow 45 37 19 35 31 26 34 30 22 43 40 35 35 60 43 45 22 30 41 25 41 31 '3l' To 49 52 25 46 47 45 41 41 44 26 59 83 66 49 78 54 47 34 39 51 31 61 37 Lbs. Mean. 47 45 •22 41 39 41 35 33 41 37 24 51 81 62 51 42 36 69 48 46 28 35 46 28 24 51 2G 34 Strength Lbs. per sq. in. Tensile. 12,700 11,000 6,000 10,000 7,000 8 700 4,000 4,000 6,700 5,800 7,300 7,300 5.300 6,700 6,700 6,700 6,700 6 700 10,000 12,600 6,700 10.000 5,300 8,700 4,400 5,300 4,000 3,600 4,000 4,500 3,500 3.500 5,300 6,700 6,000 5,300 o,000 4:700 4,700 3,B0O 5,700 3,000 8.000 5,300 3,000 i8 ELECTRIC RAIL WA Y HAND BOOK. LIQUIDS. Weight. Pounds per Material. Cubic Inch. Cubic Foot. U.S. Gallon. U.S. Bar- rel. (43.21 gal.) U.S. Hogs. head (63 gals.) Acid, muriatic Acid, nitric .0433 .0439 .0667 .0285 .0293 .0335 .0319 .0453 .0258 .0337 .0330 .0348 .0280 .0316 .0330 .0312 .0330 .0359 .0359 .0360 .0368 74.8 75.8 115.2 49.5 50.9 58.2 55.5 78.6 44.8 58.6 57.3 60.4 48.6 54:8 57.3 54.2 57.3 62.3 62.35 62.5 64 10.00 10.13 15.40 6.62 6.80 7.78 \42 >.51 5.99 7.87 7.66 8.07 6.50 7.33 7.66 7.25 7.66 8.33 8.335 8.35 8.56 432 438 665 286 294 336 321 454 259 338 331 349 281 317 331 313 331 360 360.2 361 370 630 638 Acid, sulphuric Alcohol, pure Alcohol 95g Alcohol, 50£ Ammonia, 27.9$ Carbon, disulphide.. Ether, sulphuric. . . . Oil, linseed 970 417 429 490 467 662 377 494 Oil, olive 483 Oil, palm 509 Oil, petroleum Oil, petroleum Oil, rape 409 462 483 Oil, turpentine Oil, whale 457 483 Tar 525 Water, standard . . Water, fresh Wa ler, sea, 525.1 > 526 539 ELECTRIC RAIL WA Y HAND BOOK. GASES. 19 PROPERTIES OF SATURATED STEAM. u O • *- '-' & K B O 3D a ft u -t-9 S3 U 0> P< B Total Heat above 32° F. II *J . — Relative Volume. Vol. of Water at 39° F. = 1. .2 £ § . O > cs as as 3 O 3d Is 2 "5 w . gg si 1— 1 29.74 29.67 29.56 .089 .122 .176 32. 40. 50. 0. 8. 18. 1091.7 1094.1 1097.2 1091.7 1086.1 1079.2 £08080. 154330. 107630. 3333.3 2472.2 1724.1 29.40 29.19 28.90 .254 .359 .502 60. 70. 80. 28.01 38.02 48.04 1100.2 1103.3 1106.3 1072.2 1065.3 1058.3 76370. 54660. 39690. 1223.4 875.61 635.80 28.51 28.00 27.88 .692 .943 1. 90. 100. 102.1 58.06 68.08 70.09 1109.4 1112.4 1113.1 1051.3 1044.4 1043.0 29290. 21830. 20623. 469.20 349.70 334.23 25.85 23.83 21.78 2. 3. 4. 126.3 141.6 153.1 94.44 109.9 121.4 1120.5 1125.1 1128.6 1026.0 10 0.3 1007.2 10730. 7325. 5588. 173.23 117.98 89.80 19.74 17.70 15.67 5. 6. 7. 162.3 170.1 176.9 130.7 138.6 145.4 1131.4 1133.8 1135.9 1000.7 995.2 990.5 4530. 3816. 3302. 72.50 61.10 53.00 13.63 11.60 9.56 8. 9. 10. 182.9 188.3 193.2 151.5 156.9 161.9 1137.7 1139.4 1140.9 986.2 982.4 979.0 2912. 2607. 2361. 46.60 41.82 37.80 7.52 5.49 3.45 1.41 11. 12. 13. 14. 197.8 202.0 205.9 209.6 166.5 170.7 174.7 178.4 1142.3 1143.5 1144.7 1145.9 975.8 972.8 970.0 967.4 2159. 1990. 1846. 1721. 34.61 31.90 29.58 27.59 Gauge Pressure lbs. 14.7 212. 180.9 1146.6 965.7 1646. 26.36 per sq. in. 0.304 1.3 2.3 15. 16. 17. 213.0 216.3 219.4 181.9 185.3 188.4 1146.9 1147.9 1148.9 965.0 962.7 960.5 1614. 1519. 1434. 25.87 24.33 22.98 3.3 4.3 5.3 18. 19. 20. 222.4 225.2 227.9 191.4 194.3 197.0 1149.8 1150.6 1151.5 958.3 956.3 954.4 1359. 1292. 1231. 21.78 20.70 19.72 6.3 7.3 8.3 21. 22. 23. 230.5 233.0 235.4 199.7 202.2 204.7 1152.2 1153.0 .7 952.6 950.8 949.1 1176. 1126. 1080. 18.84 18.03 17.30 9.3 10.3 11.3 24. 25. 26. 237.8 240.0 242.2 207.0 209.3 211.5 1154.5 1155.1 .8 947.4 945.8 94i.3 1038. 998.4 962.3 16.62 15.99 15.42 12.3 13.3 14.3 27. 28. 29. 244.3 246.3 248.3 213.7 215.7 217.8 1156.4 1157.1 .7 942.8 941.3 939.9 928.8 897.6 868.5 14.88 14.38 13.91 3 . si £ .00030 .00040 .00058 .00082 .00115 .00158 .00213 .00286 .00299 .00577 .00848 .01112 .01373 .01631 .01887 .02140 .02391 .02641 .02889 .03136 .03381 .03625 .03794 .04110 .04352 .04592 .04831 .05070 .05308 .05545 .05782 .06018 .06253 .06487 .06721 .06955 .07188 r V %o ELECTRIC RAILWA V HAND BOOK. PROPERTIES OF SATURATED STEAM.— Continued. id §1 o o M S.S CD ^ n. cq C^ O CQ ■a u -t-3 a o Eh Total heat above 32° F. ii i C3 &** IS- 1 ? O ii a ». r— I ^ . o -m V Si ES S \ \ X^ ta \ ^ v &e. .050 2500 ,025 625 ^x \ ^ 1 10 20 30 40 50 60 70 © 90 I 100 .s 110 2 o 120 130 140 150 160 170 180 190 11 16 21 31 36 41 .46 51 56 61 200; Twist DrilLSizes. Fig. 3a. ~ 24 ELECTRIC RAILWAY HAND BOOK. Curves showing the current carrying capacity of copper wire, both rubber covered aud weather proof , as allowed by the National Board of Underwriters, are I s<> shown. These curves are plotted with the numbered size of the wire as din tes, and the diameter in mills, the area in circular mils, and the current in nperes as abscissae. By means of the chart, the diameter, area in circular mils, and current carrying capacity of any given size of wire can be seen at a glance. The table on page 25, gives a number of electrical and mechanical units, and their conversion into terms of each other. Corresponding units, on opposite sides of the diagonal line are reciprocals of each other. THE RELATIVE RESISTANCE OF CONDUCTORS. Material. Resistance in Ohms of a wire 1 ft. long, 1 Mil in diam. Resistance of a Rod 1 sq. Mil in Section and 1 ft. long. Tempera- ture Coefficients v (1 x at) Authority for Temp. Coefficient. a Silver, annealed Copper, *• Silver, hard drawn Copper, " ..... Goid, annealed 9.65 10.3 10.48 10.5 13.28 13.52 18.73 36.0 56.69 63.21 66.29 74.33 84.57 126.0 126.1 147.6 205.4 230.2 258.7 419.0 577.6 643.6 845.2 6,734. 37,920. 7.58 8.09 8.23 8.25 10.43 10.62 14.71 28.27 44.52 49.65 52.06 58.38 66.42 98.96 99.04 115.93 161.32 180.80 203.18 329.08 453.65 505.48 663.82 52889 297.82 .00377 .00388 .00377 .00388 .00365 .00365 .00390 .00365 .00247 .00453 .0007 Riviere. Matthiessen. Riviere. Matthiessen. Gold, hard drawn Aluminum, annealed.. . Zinc, compressed Platinum annealed Iron ii Benoit. Matthiessen. Benoit. it Alloy 2, Gold 1, Silver. . Nickel Matthiessen. Tin .00365 .00044 .00387 .00031 .00021 .00389 very small .00122 .00887 very small .00354 .0009 .00052 Matthiessen. German Silver Ma8cart. Lead Matthiessen. Allov 1, Platinum 2. Silver ii Platinoid t« A ntimony, pressed Manganine A „.... Manganese, steel Mercury ti Helmholtz. Fleming. Mascart. Manganine B Helmholtz. Bismuth, pressed Graphite Matthiessen. Joubert. Arc Light Carbon ti Determining Resistance of Conductors. — Column 1 gives the mil foot constants for the conductors ordinarily used. To And the resistance of any con- ductor of circular section, square its diameter in thousandths of an inch, and divide this into the constant given in column 1, and it will give the resistance of a con- ductor one foot long. Example— Let aluminum wire 23 mils in diameter be given to find its resistance per ft : 23 squared gives 529 circular mils. The constant for an- 18 73 nealed aluminum is 18.73, therefore -r^g- = .0354 ohm per foot ELECTRIC RAILWAY HAND BOOK. 25 JF* gfe *■ _• O O ft CU^ \ 00 5 -a 5 00 3 gji §8. >ja «o 00 . O QQ 8 £ CO . •O CO p 3k§00 o . fc- . *- CO t- ~ OD E3 CO 3 .O 3 00 . £ O ^2 O — '.P — O CO .3 ac o' CO OQ S' 10 CO 02 ©t,a J> CO T-1,0 i» « CO *-" C- ** C§ « TH © Q CU •ft • a a s C £> 3^ p § § 00 u CO CU ft 10 •- c*g . © ft 8 CO ft ft "*'fl I 00*3 CO t- ft Si (1 * 1 p -* 3 co 2 1 *- u cu . CU ft p, 8 » CO H 00 JO TH 8 § o 1 55 £-* 85 ft 8 Eh P Ol O 10 O si 05 5 flj CO s i 0* t- t*,P CO t-. § 3 8 r-t O a> O CU 01 p, p. ft • IS . CO CP OQ g ^ CO !2 *> u, O O 1 ft cu u co 8 . cu n © a °Q ft . p cd o Tf-P *H CD ft cu ft St* . a ft <=> 2 ft 3 o o ft 55 55 55 o ^ CO 8 CO p OS o g-p 8 *« . cu ft tn 1 u \ m u £ t- P S3 I CO 3 O3 :,p ^.p g.p ^ 00 3 §0 h U u ® ^ 1 CO 22 8 fa 8 « P O rnP t4 co 3 t4 *-• 00003? r hou ft ft CU ft CU ft ^a 8& t4 P CM A +3 t* p n § -^ 0N -4-3 r2 P t-rl m 3 t- 1 . O a K "1 O H5 cu tt M H £ 1H •p+5 • d O 50 .O P cu 3— ■ X CU P«M 81 gfe tH©« o ft«_ £o£* 20 ELECTRIC RAIL WA V HAND BOOK. For conductors of rectangular section reduce the area to square mils, which unit will have less resistance than the circular mil ; by multiplying by the ratio of the area of the circle to its circumscribed square, .7854 will give square mil foot values, which are given in column 2. Example— Find the resistance of a bar of iron 49 64 1 x \y% inch, 1000 x 500 = 500,000 square mils and = .0009 ohm per foot. Engineers, as a rule, use 10 as a rough working constant for copper per mil foot, and thus introduce a factor to cover temperature variations, low conductivity and poor connections in the copper conductor. The weight of a copper wire 1,00 cm. section and 1,000 feet long weighs 3 lbs. Therefore the weight of any bare copper conductor per 1,000 feet is equal to the circular mils divided by 1,000 and multiplied by 3; or may be obtained by multiplying circular mils by .003. GALVANIZED IRON WIRE TABLE FOR GRADE B B. No. Washburn & Din DMn Cir. Mils. Iron Frame Ohms, per foot (legal). Feet Pounds Moen Gauge. Mils. Safe Capacity. per Ohm. per Foot. 3 No. 4 & 1 No. 6 434 188,356 200 .000488 2049 .500 2 No. 4 & 1 No. 5 & 1 No. 6 425 180,625 190 .000504 1984 .4792 4 No. 5 414 171,396 180 .000540 1852 .4540 3No. 5&lNo. 7 400 160,000 170 .000579 1727 .4236 3 No. 4 390 152,100 160 .000607 1647 .4026 4 No. 4 390 152,100 150 .000607 1647 .4026 3 No. 5 359 128,881 140 .000720 1389 .3405 2No. 6&IN0. 4 353 124,609 130 .000742 1348 .8288 2No. 6&IN0. 5 341 116,281 120 .000793 1261 .3081 2 No. 4 318 101,124 110 .000910 1098 .2684 1 No. 4 & 1 No. 5 306 93,636 100 .000988 1012 .2477 2 No. 5 293 85,849 90 .001080 925 .2270 1 No. 5 & 1 No. 6 262 79,524 80 .001157 864 .2108 2 No. 6 272 73,984 70 .001255 797 .1946 2 No. 7 250 62.500 60 .001480 675 .1662 4 225 50,625 55.6 .001820 549 .1342 5 207 42,849 47.5 .002160 463 .1135 6 192 36,864 34.8 .002510 398 .0973 7 177 31,329 30.1 .002960 337 .0831 8 162 26,244 26.6 .003530 283 .0695 9 148 21,904 23.2 .004230 236 .0580 10 135 18,225 19.7 .005080 196 .0483 11 120 14,400 16.2 .006430 155 .0383 12 105 11,025 13.9 .008400 119 .0292 13 92 8,464 11.6 .010940 91.4 .0224 14 80 6,400 9.28 .014470 69.1 .0169 15 72 5,184 6.96 .017860 56.0 .0138 16 63 3,969 5.8 .023380 42.8 .0110 17 54 2,916 4.29 .031760 31.4 .00868 k ELECTRIC RAILWAY HAND BOOK. 27 1 RESISTANCE OF INSULATORS AND THEIR SPECIFIC CAPACITY. Material. Benzine ♦ . < « • Insulation Resistance in Megohms per Cubic Centimetre. 14 x 10 6 Temperature. Degs. 46 Cent." 20 •• 20 " 24 " 20 Cent 30 Cent 20 Cent 20 " 46 " 28 Cent Specific Inductive Capacity. 2 20 Distilled Water 7 83 80 Ebonite ,, 28,000 x 10 6 2 56 Glass, Flint 20 000 x 10 6 6 72 Glass, Ordinary Gutta Percha 91xl0 6 450 x 10 6 5.83 4.20 Ice 2240 Mica Micanite Micanite Cloth 84 x 10 6 2490 x 10 6 310 xlO 6 5.00 Micanite Paper Oiled Asbestos 1240xl0 6 850 x 10 3 .... Olive Oil 1 x 10 6 4.78 Paper, Parchment 03xl0 6 Paper, Ordinary 0485 x 10 6 1.96 ! Sea Water 30 ohms. Shellac 9000 x 10 6 2.74 Walnut Wood 53 x 10 6 1670 xlO 6 « « . White Vulc. Fibre. . . . 14 x 10* . ! 4LA SECTION II —TESTING. ELECTRICAL UNITS. If the two terminals of a source of electrical energy, such as a battery, dyna- mo, etc., be joined by a copper wire or other conducting path a current of elec- tricity will flow through the completed circuit thus formed. The current manifests itself by causing neighboring compass needles to deflect from their natural posi- tion, by heating the wire, by the appearance of a spark if the wire is broken, by chemical action in an electrolytic cell placed in the circuit, etc. Fig. 4 shows a circuit containing a primary cell and an electrolytic cell. The Ampere.— The current flowing in the circuit may be determined by cutting the wire and connecting the severed ends to two silver plates immersed in cincv/T SILVER PLATES AA/O S0LVT/OM BATTERY iFio. 4. a nitrate of silver solution. It will then be found that the current in flowing through this solution carries with it silver from the positive to the negative plate, and if the battery gives a steady current the weight of silver carried over will be proportionate to the time that the current is passing through this solution. If for each second of time it is found that 0.001118 grammes of silver are carried over, then the flow of current will be one ampere; or the total grammes weight of silver divided by the seconds during which the current was flowing through the solution, divided by 0.001118 will give the total number of amperes flowing through the circuit during the test. This is the way in which the unit of current flow, the ampere, was given a definite value. There are many refinements necessary to carry out the above test in order to obtain reliable results. This method is the one used to determine the true value of the ampere, but it is not useful for practical work. ELECTRIC RAILWAY HAND BOOK. The Ohm. — If there is added to the above circuit, Fig. 5, a much longer wire of the same size, and the test repeated allowing the current to now through the sil- ver solution from one plate to the other for the same length of time, it will be found that the current has not carried as much silver across as in the first experi- ment, showing that the lengthening of the circuit has diminished the current flow. This was caused by the added conductor offering resistance to the current. This is a property of all electrical conductors and is measured by a unit called the ohm. If the circuit had been made of No. 30 wire, Brown & Sharpe gauge, and was 9 ft. 9 ins. long, then the copper circuit would have been nearly one ohm in resistance. The standard for the ohm is the resistance of a column of mercury 106.3 cen- timeters long (41.8503 ins.) of uniform cross-section, and weighing 14.4521 grammeB (.5098 ozs.) at the temperature of melting ice. This is known as the "Interna- Fig 5. tional Ohm" or " True Ohm." There are two other older standards, known as the British Associations and the Legal respectively, whose relation to the Inter- national ohm is shown in the following table. Legal = 1.0023 International. B. A. 1. = 1.0136 .9866 = 1. .9977 = 1.0112 1. The Volt.— Electrical pressure is required to force the current through the wire and the silver solution. Electrical pressure can be opposed by an equal elec- trical pressure, and there will then be no current flow in the circuit which contains the opposing electrical pressure, just as a water pressure can be acted against by an equal pressure of water, when no water will flow. Fig. 5 shows how these pressures may be equalized electrically. A standard battery of one volt is connected in series with a delicate current indicator, known as a galvanometer; the ends of this circuit are connected to the end of the 9% feet of copper wire, as shown in Fig. 5, so that the electrical pressure of the standard battery circuit opposses the fall of pressure in the main circuit. Then when the loss in volts, or electromotive force, in the main circuit is equal to one volt, which y 30 ELECTRIC RAILWAY HAND BOOK. 1.8 1.7 4 6 6 7 AMPERES Fig. 6. O 10 11 ELECTRIC RAIL WA V IIAXD BOOK. 3i 10 20 30 40 60 60 70 dO SO 100 110 1Z0 AMPEREd Fig. 7. 32 ELECTRIC RAILWAY HAND BOOK. is the pressure of the standard battery, no current will flow through the galvano* meter circuit. Since the circuit measures one ohm, there must have been one am- pere flowing through it to produce the loss of pressure of one volt. The unit of this pressure is known as the volt. The value of the volt is a little less than the e. m. f. of an ordinary gravity cell. It has been proved that if an e. m. f . of 1 volt act3 on a circuit of 1 ohm a current of 1 ampere will flow. The ampere being fixed as that current which will deposit 0.001118 grammes of silver per second the volt, therefore, depends upon the value of the ohm and we have International, B. A. and Legal volts which bear the same relation to each other as the corresponding ohms. The practical standard for e. m. f. is the Clark cell made according to speci- fications drawn up by the Electrical Congress of 1803. The cell consists essen- tially of pure zinc in zinc sulphate and pure mercury in contact with mercuroua gulphate the e. m. f . at 15° C being 1.442 International volts. OHM'S LAW. Ohm discovered that the current varied directly as the pressure and inversely as the resistance. If we measure these quantities in practical units, z. e, % in am- peres, volts and ohms, the relation given above that the action of 1 volt on 1 ohm produces 1 ampere gives us the law: Current = Electromotive Force Resistance This is known as Ohm's Law and is generally written C = =.. From this rela* R tiou, if we have any two of the quantities given, the third is readily found. This is accomplished graphically in Fig. 6. There volts are given on the vertical lines and amperes on the horizontal lines; the radial lines giving the ohms. In any circuit where the amperes and volts are known, if we trace these two values on their respective scales to their intersection, this intersection will occur at the radial line which is marked in ohms. In a circuit for exam- ple, with 8 volts and 2 amperes, we will find the intersection on the radial line marked 4 ohms, which is the answer. Supposing that we had a circuit of 2 ohms resistance and 6 volts potential, then follow the radial line down until it intersects the 6 volts horizontal line and also the vertical line for 3 amperes, which is the answer required. In the same way, when ohms and amperes are given in a circuit, the intersection of these values will fall on the volt line required. As in railway work 500 volts is the voltage commonly used, a scale, (see Fig. Ifa for 850 to 550 volts, and to 120 amperes is also given. METHODS OF CALIBRATION. Galvanometers.— The galvanometer is used in insulation and cable tests and in connection with the bridge method as a current indicator, also to make potential and current measurements. It is easily affected by external magnetism. It consists in general of a small permanent magnet suspended by a silk fibre, or mounted on a concave jewel having a needle point to support it. The suspen- sion should be such that only a very slight effort is required to turn the needle. This needle is free to rotate in a spool around which are wound many turns of fine wire. A pointer, usually made of aluminum, is attached to the needle to magnify the amount of deflection. The zero position of this pointer may be fixed by the earth's magnetism, but is often controlled by a local permanent magnet. ELECTRIC RAILWAY HAND BOOK, 33 The current flowing through the coils tends to cause a deflection, the magnetizing effect of these coils being at right angles to the suspended magnetic needle. The currents producing the deflections are related to each other as the tangent of the angles of deflection; if the needle is short and placed at the center of a circular coil the galvanometer is then called a tangent galvanometer. In the Thomson reflecting galvanometer the readings are taken by means of a beam of light reflected from a mirror on the back of which is secured the mag- netic system. When this beam of reflected light is read on a scale at right angles to the beam of light before reflection, the readings on the scale of the deflected beam are practically directly proportional to the currents deflecting the mirror. ELECTROMOTIVE FORCE. E. M . F. Direct Method.— To set up a galvanometer to read volts direct there is required a variable resistance-box, A % in series with a standard battery, i?, I 1 O I GAL VAH0ME7£X POT£ffT/AL $CAL£ Fig. 8. a proportional coil, C, a double-throw switch, Z>, and key, K. The connections are made as in Fig. 8. First such resistance is inserted by box, A , with the standard battery (if a Clark cell is used as that standard) that it will give a permanent deflection of 144.2 divisions. Then this setting up gives for each division T £„ of a volt; throwing the switch so as to connect in T ^ of the total voltage across the proportional coil, the galvanometer will read the main potential in 1 volt per division. A voltmeter to be standardized should be connected across the mains at X' V, The proportional coil is generally made of Xo. 32 resistance wire with a low tem- perature coefficient, preferably platinoid. With 50 ohms per volt to be measured a proportion of 1:100 is usually used for potentials up to 150 volts, 1:500 for rail- way work. For reading potentials lower than the standard cell the connections are changed as in Fig. 9. The standard battery is in this case connected across 34 ELECTRIC RAILWAY HAND BOOK. the proportional coils, and the resistance changed in series with the galvanometer until deflections (142.2) are again obtained; then each deflection is TIJ uua OI " a v0 ^ ^ a proportion of 1:100 is obtained from the proportional coil. The galvanometer can then be thrown over to the potential to be compared and read direct in njtjus °f a volt per scale division. Potentiometer Method. — Where a constant potential is to be maintained during a test, the potentiometer method is more convenient. This requires a stand- ard battery, a galvanometer, a variable resistence of a uniform wire of such a size that it will not be heated by the current passed through it; a portion of this wire is provided with a sliding contact over a scale which is divided into a thousand divisions for this length. This apparatus is connected up as shown in Fig. 10, the standard battery opposing the drop of potential along the potentiometer wire. The contact, JF, on this w 7 ire is slid along until a pcint is determined at which the galvanometer shows no deflection. Then the reading on the scale will be where the drop is 1.442 volts, or the e. m. f. pf the standard cell. Then the scale length I Q I GALWWOMETER WO i To C I u W 1 n FOTEHVAL TO 3EM£ASU*E0 SCALE Fig. 9. is to the total length of wire, as 1.442 is to X, or the terminal voltage to be meas- ured. Say the scale read 2 and the total length of wire was 100, then 2 : 100 : : 1.442 : X, or = 72.1 volts. The scale can be calibrated in volts direct if the same standard potential is used for all tests. As the Clark cell is easily injured by an excess of current it is important that a resistance of, say, 10,000 ohms be inserted in series with it. This will prevent its being short circuited through the low resistance of the potentiometer wire when the counteracting force of the e. m. f. under investigation has been removed, and the battery will not be short-circuited by the slide wire. This resistance will have practically no effect on the accuracy of the readings as there is no current flowing when balance is obtained. ELECTRIC RAILWAY HAND BOOK 35 Current.— To read amperes by galvanometer deflections the requirements are a shunt of known resistance, a standard battery, a double-throw switch and a variable resistance. The connections are made as shown in Fig. 11. The gal- vanometer is first brought back to zero by raising or lowering the current through shunt, C. If this shunt was .01 of an ohm and a standard Clark cell is used, the battery's potential being opposed to the drop of potential in the shunt, and if with the double-throw switch at A , the galvanometer shows no deflection, there are 144.2 amperes flowing through the shunt. If this current is held steady, the switch thrown to position, B, and sufficient resistance is added to give 144.2 deflections, then each division is equal to'l ampere passing through the shunt. By putting an ammeter in the circuit it can be calibrated throughout its scale. By shunts of higher or lower resistance any desired range can be secured. .1 ohm for 15 amperes, .01 ohm for 155 amperes, .001 ohm for 1,500 amperes and .0001 for 15,000 amperes give all the required ranges for checking up meters on a switch- board. The shunt can be arranged with terminals so as to plug into the switch GALYAN0MET£R W 3UDER CONTACT ^ t — r Fig. 10. jaws when the switch is open. The galvanometer can be located at any conven- ient part of the building, wires leading to it from the shunt at the switchboard. If the test drop wires are connected together on the gallery and form part of the cir- cuit between the shunt and galvanometer after the proper setting up has been obtained for the galvanometer to read amperes from that shunt the shunt can be removed to the gallery and the drop points connected to the ends of the pressure wire, and the pressure wire connected together where they took the pressure from the shunt in the testing room. In any shunt the connection for the pressure or galvanometer wire should be well within the contacts that carry the main current to the shunt, and should never be connected to the same contact, for then the contact resistance may be 36 ELECTRIC RAILWAY HAND BOOK. included in the shunt resistance, and accurate or constant results will be difficult to obtain. In calibrating ammeters in regular use it is best to find the average load read- ing, and have a single stroke bell on the gallery, the man in the testing room giving one stroke of the bell by a push button located at his hand when the current attains the agreed reading in amperes. The gallery attendant then will note the reading of the ammeter. By repeating this a few times and averaging, the error of the instrument at that point can be readily discovered. There may be a leakage which should be removed before calibration. This is detected by first connecting to the live jaw of the switch, and noting if there is any permanent deflection of the galvanometer. If there is, it may be due to leakage TO BANK SHUNT 'C SCALE <~> LAMP AMP.METER TO BE CALIBRATED yARIABLE RESISTANCE STANDARD .BATTERIES r J00O0 OHMS 6® A m 1, \\6AL VAMMETtift r& SWITCH Fig. 11. of test lines or instrument; these should be discovered and insulated before pro- ceeding with the check measurements. There is often considerable magnetic disturbance around a station by which the galvanometer will be influenced. This can be opposed by surrounding the galvanometer with two cylinders of % in. sheet iron separated in the middle so as not to interfere with the ray of light; two short sections of wrought iron pipe are still better. In a d'Arsonval galvanometer the magnet is stationary and the coil revolves. Thus the coil turns in a strong field and is not so much affected by changes in the magnetic condition of the surrounding space. ELECTRIC RAIL WA Y HAND ROORT. 37 PRACTICAL ELECTRICAL MEASUREMENTS. Resistance Measurements.— -Where it is required to compare the resistance of a conductor with a standard, such as field or armature windings, the simplest way is to connect the resistance to be adjusted in series with a standard resist- ance; then on passing the same current through both resistances they have the same value when the difference of potential is the same on both. This simple method is diagramatically shown in Fig. 12. Where the drop wires are lead out to a double-throw switch, so that the press- ure on both coils can be compared quickly by throwing the switch from one side INSULATION. TOLA, BANK stNSU (a 7 V Fig. 12. TEST f/ELO TERMINAL. Fig. 13. to the other, there should be sufficient current sent through the field coils to cause the voltmeter employed to read to nearly full scale in order to magnify any small differences in resistance that may exist between the standard and the coil tested. Care must also be taken that the pressure leads are distinct from the contacts through which the current is carried into the fields. A special connection is shown in Fig. 13 for this test, for fields, for clamping the ear projecting from the field coil, the two sides of the clamp being insulated; MILL I VOLTMETER TO GROUND FIELD TEST TO LAMP BANK Fig. 14. DIFFERENTIAL VOLT METER. Fio. 15. to one the current lead is attached, and to the other the lead from the voltmeter used in comparing. It can be readily seen that in Fig. 14 the testing current is measured ; then, with the drop in volts known, the volts across the coil divided by the current flow- ing through it will give the resistance directly in ohms. In railway work a constant source of potential is not usually convenient, and Vmder varying conditions of ypltage it is tedious to get reliable results, JTqx 3* ELECTRIC RAIL WA Y HAND BOOK. this case a method of finding the value of an unknown resistance in terms of the standard is shown in Fig. 15. Here the drop from both the standard resistance and the unknown resistance act against each other in their effort to turn a mag- netic system ; consequently the deflection of the magnetic system will only be the resultant of these two forces. If a differential voltmeter is used and connected as shown in Fig. 15, the deflection of the instrument will be due to the differences in potential drop in the two resistances compared; from these deflections the differ- r-nces that exist between the two resistances can be determined directly, the current variation due to changing line voltage averaging out. The Thompson method, the connections for which are shown in Fig. 16, is an improvement over the differential method, and is especially useful for the com- parison of low resistances. It requires a standard resistance A B, a galvanometer EF H, and four equal resistances or having the ratio — — . The function of these resist- ances, which need not be greater than 10 ohms , is to reduce the flow of current through the circuit leading to the galvanometer, so that all contact resistances can be neglected. In a conductivity bridge made for measuring copper only, A B can Fig. 16. be of copper wire, which can be made interchangeable, so that a wire of fixed length and standard gauge can be compared against a wire of similar dimensions whose conductivity is to be measured. The temperature of the standard and the wire under measurement must be the same. If the wire to be measured is strung by the side of the standard, it will shortly assume the same temperature; or a more expeditious way is to pass a current through both the standard and wire under measurement to heat them, the measurements being made while their tem- perature is falling. If the length of the standard be divided by the length of the wire under test when a balance is obtained on the bridge, the result will be. the conductivity in terms of the standard. If the standard is 100 units long, then the reading of the point of balance on the scale for the wire under test will be its con- ductivity in direct terms of its length. The resistance of a wire depends on its sectional area, and the average squares of the diameters should be taken in conductivity measurements. Weighing the standard length and dividing this weight by its length in feet and also by the weight of one mil foot of this conductor, will give the true average section, and this is the method most generally employed. There should be a very low resistance between the points of contact with the standard resistance and the rod under measurement. This is often not conven- ient to obtain. If the fall of potential along the standard be brought to one pair of terminals of a differential galvanometer or differential milli-voltmetcr and the potential lines from the resistance to be measured, this allows of considerable resistance in the circuit between the standard and the resistance to be measured. ELECTRIC RAILWAY HAND BOOK. 39 ^ For approximations of conductivity in commercial work, and where a gal- vanometer is not at hand, a considerable length of wire may be measured off; then, by increasing the current flow, the resistance can be measured by the volt- meter and ammeter method, as described in the field resistance tests. With in- sulated wire, it is best to submerge the wire in water in order to determine the temperature of the copper; the current should be applied for as short periods as is possible in order to obtain reliable readings, as otherwise the heating effect will reduce the conductivity of the wire. For temperature coefficients, and standard resistances see under "Line: Prop- erties of Conductors." Wheatstone Bridge.— The Wheatstone bridge primarily consists of a rheostat having two parallel circuits, one of which can be varied by cutting in X RESISTANCE 70 BE MEASURED. Fig. 17. tnown resistances, while in the other parallel circuit is placed the resistance to be measured. If the current flow is the same through both branches of these parallel circuits, the resistance in the rheostat is equal to the unknown resistance under measurement. When on the parallel circuits equal potential points are joined by means of a galvanometer (see Fig. 17), then, when the galvanometer shows no deflection, the sections of the two paths have the proportion, A : C::B: X. If X is the unknown resistance to be measured, C can be varied — if A and B are equal — until the galvanometer, //, reads zero ; then the current is equally divided between the two branches, and the resistance at Cis equal to that at X. A and B need not, however, be equal, but can be made of any known ratio; the same ratio will then exist between C and X when the bridge is balanced. If A is 10 and B 1000, then the resistance of C should be multiplied by their ratio, i. e. 100, in order to get the value of X. If Jf is large the value of A is made greater than B so that C may be able to balance X. The reliability of the bridge method is within the range of % OI> an onm ana 10,000 ohms. In the cases of low resistances the contact resistance where the resistance to be measured is connected to the bridge is also measured, causing a large element of error with low resistances, The current ii 40 ELECTRIC RAILWAY HAND BOO IT. bo small that a galvanometer of high sensibility has to be used in order to determ- ine these values with high resistances. Insulation Resistance.— For testing high resistances an arrangement sim- ilar to that shown in Fig. 18 is found to be very satisfactory. The connections are there made for testing the insulation resistance of a cable. A known length of the cable is placed in a tank of water, and the resistance measured between the conductor of the cable and the water. About 100 cells of battery are necessary, and the unknown resistance is compared with a standard megohm by means of a Thomson galvanometer. The deflection made by closing the key, A", when the double-throw switch is in the position shown by the solid lines, may be called A ; and B, that obtained with the switch thrown to the position of the dotted lines. Then 1__ : — — x, the insulation resistance sought. / MEG-OHM BO TOGHOIWD OR SHEATH OF TANK OR CABLE. Fig. 18. When the unknown resistance differs greatly from one megohm it will be found necessary to use a shunt with the galvanometer in order that the two read- ings, A and B, may be on a convenient part of the scale. The shunting ratio should be inserted in the above formulae. INSULATION TEST BY VOLTMETER METHOD. There are required a voltmeter, F, of known internal resistance, Fig. 19, and a source of constant potential such as M. Oue method is to first determine the voltmeter constant, which is obtained by multiplying the voltmeter resistance by the initial voltage used in measuring the resistance. If we open switch, A', throw- ing in series with the voltmeter, the unknown resistance, A 5 , the total electromo- tive force will be divided between the voltmeter and external resistance in \\, which is attached to this frame. This instrument can be calibrated so as to give a scale proportional to the differ- ent currents flowing through the coil. If a voltmeter, it is connected acroes the leads of which the potential is required ; and in series with the leads if an ammeter. (See Fig. 22 for the proper connections of voltmeter and ammeter.) A Weston portable voltmeter for 500 volts has approximately 55,000 ohms resistance, and requires about .008 of an ampere for full scale deflection or 110 ohms per volt. The construction of the Weston ammeter is nearly the same as that of the voltmeter except that there is a shunt in the main circuit, and the instrument takes the drop across the shunt. For small instruments this shunt is in the ammeter case, but in station types is separate. The terminals of the instrument are marked -f- and — , and the instrument will deflect over the scale when the positive terminal is connected to the positive side of the circuit. Care should be taken to see that the shunt leads and ammeter always bear the same shop number, for they are cali- brated together, and are not interchangeable. Nearly all station type ammeters have a constant resistance of .305 ohms; the current required to give full scale deflection averages .075 amperes. The resist- ance of the instrument and its leads being known, the length of the cable required as a shunt for the ammeter can be found in this way. The length of main or bus bar, Z, is equal to the product of the resistance of the meter, including leads, multiplied by the current, C, required in the instrument to give full scale deflection, divided by the resistance of a square inch of copper 1 foot long, R, This dividend is again multiplied by a dividend obtained by dividing the cross-section of the bus ELECTRIC RAIL WA Y HAND BOOK. 43 bar (or cable) in square inches, S, by the maximum current, C, to be measured on the bus bar (or the full range of the ammeter), or L = — ^ — X— • I f tne resist- K C ance, in case of a cable, is accurately known per foot, then the proper resistance, 7?, to be included between drop points can be found by multiplying the resistance of the instrument and leads by the current required by the instrument; dividing this by the maximum current to be read by the instrument gives the resistances required, or R VjlC , which resistance divided by the resistance per foot ef the cable to be used as the shunt, gives the length of cable required to give the correct drop for the meter to read amperes. This determination can be checked by a SHUNT TftOUEY AMPERE METER yoir ME7EK DYNAMO Fig. 22. reading on a meter temporarily in series with the feeder or bus bar on which the shunt has been adjusted. It is often required to know the current over a number of feeders from time to time without the expense of a separate ammeter on each feeder, especially so on ground return feeders. Permanent drop points can be adjusted on the cable at some convenient place where it enters the station, and another a point at such a distance as to give the correct drop ; then the ammeter with the leads with which these shunts were adjusted to read correctly, can be connected to the drop points, and the current read on any feeder desired. POWER MEASUREMENTS. In order to get the power delivered to any electrical device, the constant current now in amperes multiplied by the volts lost through the device will give the watts consumed. As 746 watts are equivalent to one horse-power, the product divided by 746 will give the horse-power absorbed. The continuous power taken can be de- termined by multiplying instantaneous readings of volts and amperes when both volts and amperes are steady, but this method does not give reliable results. Where these are varying, as in a railway load, a direct reading wattmeter should be used. Here the main current is carried through the instrument, and also the f 44 ELECTRIC RAILWAY HAND BOOK. potential across the terminals of the current under measurement; the combined efforts of these two currents are calibrated on a scale from which the instantaneous watts can be read directly. (See Fig. 23 for the connections to be made with a 70 TROLLEY WATT M£T£ft. Fig. 23. wattmeter.) But these readings must be multiplied by the length of time in min- utes in order to get the continuous record of output in watt minutes. For a continuous test of power consumption, such as in a dynamo or a street railway equipment, an integrating wattmeter giving a summation of all energy delivered, is used. This meter is practically a motor whose speed varies directly Fig. 24. as the energy passing through it; and the resultant revolutions of this motor are recorded on a summation dial which can be read directly in watt hours. In Fig. 24, A, A, are the field coils; B, armature coils; C, C, copper disc; A A retarding magnets; E, spindle; F, F % wires leading through armature coil. <_ ELECTRIC RAILWAY HAND BOOK 45 TESTS ON ELECTRIC RAILWAY SYSTEMS. TEST FOR RESISTANCE OF INDIVIDUAL, BONDS. The instruments required for this test are : one milli-voltmetcr with zero in the center, two resistances, one-half ohm each, a stand like Fig. 25, or a strai^it edge like Fig. 26. In testing for individual bonds with the stand shown in Tig. 85, wo fixed contacts bridge the rail-joint at a distance of about 12 ins. apart, and Fig. 25— stand for individual bond test. the variable contact is moved along the rail until a balance is obtained on the milli-voltmeter. The tcale of the stand will then read, when the keys C and D are both depressed, the resistance of the joint in terms of the rail length: that is, the length of the solid rail, which has the same resistance as the joint. To determine the current flow in the rail, carry the cord out until it registers 10 ft. on the scale, and p:ess down key D\ then the current in the rail in amperes p! as RAM. Fig. 26— straight edge tor individual bond test. will be the millivolts, multiplied by the weight of the rail in pounds per yard, divided by 8.7. This formula applies to steel rails not exceeding .49 of 1 per cent manganese. With the straight edge shown in Fig. 26 the voltmeter reading is first taken with D only depressed; then with both C and D depressed. The ratio in readings will then give the resistance of the bond as compared with that of straight rail. As usually constructed, the distance between the contact spanning the joint is 1 ft., and that between the contacts on the solid rail is 6 ft. This gives a ratio of 6 to 1 $M makes tho bar about *% ffc loag, < 46 ELECTRIC RAIL WA Y HAND BOOK. This test provides a more rapid way of determining defective rail^>ads than that given in test No. 25. The apparatus required is one special truck, made up of two pairs of old wheels and boxes, with one axle cut and insulated. The two axles should be insulated from each other by making the side framing of wood, and attached to this side frame should be four metallic track brushes each located tr/io SfEGVl/ITOR J\ AMP£fiE M£T£-* yoLTMers* yoi. r Atrrfc/r Fig. 27— -trailing truck for testing bond resistances. as shown in Fig.27. The other apparatus required is two voltmeters reading 3 volts full scale, one ammeter reading 200 amps., one motor dynamo 500 volts to 5 volts, and 200 amps., and a regulator to control the speed of the motor-dynamo. The connections are shown in Figs. 27 and 28. ?/?£ /Hers* ro row T/t4/*Sfd/rAf£/T h^j feg^^- ; yet T M£~T£/f TRACK BRUSHES Fig. 28— plan view of trailing truck showing connections. The current from the low-potential side of the motor-dynamo is taken to wheel A of the pair of wheels having an insulated axle. These wheels should be located furthest away from the tow-car. The metallic truck brushes are located as shown at D D D D % and should be as far apart as possible, but between the v ELECTRIC RAILWA Y HAND BOOK. 47 r heels of the truck. The path of the measuring current is from wheel A , through the rail to wheel E, then across the continuous axle through wheel F, then along the rail back to the other wheel, C, which is connected to the opposite brush of the ow-voltage dynamo. The voltmeters G and // measure the drop between the track crushes on each side of the truck. The testing truck is towed along by a car, in which are located the measuring instruments, the motor-dynamo and its regulator. An examination of Fig. 29 will show how the voltmeter readings, as taken in different positions of the trail car, will determine the condition of the individual bonds. Assume a uniform weight of rail and the joints staggered. Then in Position 1 (Fig. 29), the current passes through joint B and back through rail A. The difference in the readings between the two voltmeters on the A and B sides of the car will give the resistance of joint B, as compared with solid rail. In Position 2, both voltmeters should read alike, if there is no cross-bonding across the four tracks. In the case of cross-bonding the current would be shunted around through the rails on the other track, and all of it would not go directly back through the opposite rail of the first track. The current will be diverted through the cross bonding, and the voltmeter readings will be less than that ^1 Fig. 29— direction of current flow. required by the amperes flowing, but the ratio of the voltmeter deflection to cur- rent flowing will indicate the conductivity of the cross-bonding work, as compared ■with the cross-bonding made by the car and test truck. No car should follow the testing truck within 1000 ft. If the road is in operation and the rails are carrying current, the side of the track carrying the current from the motor-dynamo will be increased in voltage when the test current and working current flow in the same direction, and when the test current is flowing against the current in the rails it will be decreased in voltage. When the drop in the rail is zero the current flow- ing from the motor-dynamo will be equal to that flowing in the rail. By carefully watching the voltmeter as the car proceeds, joints can be meas- ured in the way described at the rate of about 4 miles an hour. As the bad joints are found they can be marked by injecting whitewash on the roadway, and can then be marked permanently for repair later. AUTOGRAPHIC METHOD OF TESTING BONDS. The author has invented a test car which utilizes the principle of the preceding method by substituting one of the car trucks for the trailing truck. The working speed has been increased to ten miles per hour by the use of certain recording devices described below : The testing apparatus, which is mounted on a table near one end of the car. consists of two recording voltmeters and the record chart, which is moved by being belted to the axle. The voltmeters are so arranged that the movement of their hands is recorded without in any way interfering with their sensibility. This is accomplished by means of a high-tension spark which passes from a plate to the moving hand of each instrument, from which it passes to a. semi-circular copper 48 ELECTRIC RAILWAY HAND BOOK. sectionalized scale under the. pointer of each instrument. This scale is electrically connected to a series of terminals placed at right angles to the direction of move- ment of and under the record sheet, and the spark is of such a character that it burns a hole in the paper of the recording sheet as the latter moves along. In this way the circular movement of the hands is rectified and all ordinates on the record are proportional to the voltage. In some cases an ink belt is run between the spark and the paper, and with certain aniline inks the spark will carry the ink on to the record sheet. The voltmeters are usually set for their full scale of 120 millivolts, but their shunts can be adjusted to correspond to the current flowing in the rails. For instance, the normal current in the local circuit in each rail— that is, from one wheel into the rail and back through the other wheel on the same side, and which is provided by the motor-dynamo on the test car is 200 amps. This current is, of course, increased by the return currents of all the other cars on the system, so that the voltmeter readings depend upon the total amount of current in the rail. For this reason the voltmeters are adjusted so that the full scale can be used for measur- ing the variations in the voltage around each joint. As a rule, each joint is measured on the record by the proportion which its resistance bears to 4 ft. of solid rail. The record sheet moves 1 in. while the car progresses 120 ft., giving the bond record the scale of 1 in. to 120 ft. of track. As the voltmeters might be injured by an excess of voltage caused by a defective bond, an automatic cut-out is inserted in the circuit of each so that it cuts out the instrument before it can swing to full scale. This automatic cut-out is also elec- trically connected to a pen which makes a continuous straight line on the record when the track is in good condition, but a side dash, when the car passes any joint that has over 150 millivolts drop, or is practically open. When this automatic cut- out opens it closes another circuit which operates the valve of an air pump, by which a jet of whitewash is squirted on to the roadbed adjacent to the defective joint. In this way an open joint can be located by the trackmen without reference to the autographic record, which is kept in the roadmaster's office. The autographic record gives more information than simply the true condition of each bond. In the first place, the direction and amount of current flowing in the rail can be determined at any instant. This is done by opening the local circuit in the car, which, as already stated, carries 200 amps. The reading of the volt- meter after opening this circuit bears the same ratio to the reading before opening the circuit that the current in the rail has to the original current flow. The inductance of the rail circuit can also be very clearly determined by taking 100 amps, from the trolly wire through a resistance on the car and noting the time required on the record chart for the current to rise to the normal. It is also easily possible to determine the total transmission losses. This is accomplished by the use of the 100-amp. circuit described in the previous test, and multiplying the difference of the reading on the voltmeter before the additional current is thrown on the line and after by 100, to give the ohms. That is to say, since the resistance in any circuit equals the volts divided by the ampers, each volt difference will correspond to 1-100 ohm when the amperes are 100. Of course, if other cars are in operation, this test should be repeated until a constant or average value appears for the reading, on account of the varying voltage. The advantage of making this test when the road is in use is that the resistance thus determined is always different and usually less than that obtained when the cars are not in operation, the reason being that poor bonding and earth leaks increase in resistance when there is no current flowing in the rails. The method described, however, gives the resistance under operating conditions, wWeb \% tkG proper Criterion of the losses ELECTRIC RAILWAY HAND BOOK. 49 E AGGREGATE BOND TEST FOR A SECTION OF TRACK. The instruments required for this test are one ammeter reading 200 amps., one ltmeter, 30 volts, one water-barrel rheostat, one snap switch for 150 amps., and yyO volts, and one long pole to reach the trolley wire. On a motor car place the water-barrel rheostat, in which have two iron plates about 14 ins. x 24 ins., sep- arated by slats. Connect as shown in Fig. 30. Use bicarbonate of soda in the water rheostat, so that with 500 volts about 140 amps, will pass. Connect the snap switch on the trolley side of the rheostat and the ammeter in series. Have the cir- cuit breaker opened in the station (by prearranged signals) on the feeder supplying the section of trolley over the track to be tested. With a No. 18 wire connect all four tracks w r ith the dead trolley. It is advisable to be sure first that the trolley Is dead by reversing the trolley-pole to the dead trolley; if the lamps do not light, the section is open. Then the determination of the track return resistance can be made by first reading the volts between this dead trolley and the track, as shown by the voltmeter, and by dividing the reading thus obtained by the amperes flow- ing in that rail. The relative resistance of each track, as compared with the total circuit, can be determined by the drops on a rail length of each track. The drop between each rail and the dead trolley can be taken. The drop between the different rails will give the cross-bonding conditions. Where there is a loop or there are inter- secting tracks which offer other paths for the return current than the one under test, the current flowing back over the section under test has to be measured by drop on rail lengths. The current in the two paths beyond and behind the test car is inversely proportional to the resistances of these two return circuits. This gives the individual rail return resistance, and the collective rail resistance over Live FEeoc* Q£AO *F££Q£A r*OLL£r tV/AA? SECT/O* /#SULA70* - V — Fig. 30— connections for aggregate bond test. the section under test, the cross-bonding conditions and the value of the tested track in its ratio with any other return circuit. If one rail is carrying less than the other, its bonding is poor, but effective cross-bonding with no current indi- cates one or more open rail joints between cross-bonds. # TEST FOR CURRENT FLOW IN WATER PIPES. The instruments required for this test are one voltmeter reading 5 volts, one ammeter reading 15 amps., 600 ft. of No. 6 B. & 8. cable, 600 ft. of No. 10 B. & S. cable, two plug clamps like shown at A, Fig. 32, and one portable reel, shown at B, provided with a commutator, as shown at C. The reel should have a shelf, to which the instruments and switch are secured. A $o ELECTRIC FAIL WAY HAND BOOK. If two adjacent water-plugs, which are on the same line of pipe, are connected together electrically, as in Fig. 32, through an ammeter and if a current is flowing through the pipe, a part of the current will be diverted through the external am- meter circuit A-B when switch E is closed. To determine the current flow in the water-pipe the following readings will have to be taken : volts with switch E open which can be called V x . Volts with switch E closed, which can be called F" 2 , also amperes flowing, A. If we call the normal current flow in the pipe X, then X: A : : V x : V x — F 2 . This is approximately correct. The results may be unreliable from the following causes: First, the two plugs may not be on the same water main, then the am- meter leads form a jumper between these two pipes, and there is a very slight change of voltage for considerable current flow, and apparently a very low resist- ance is shown. A number of adjacent plugs along a street should be measured in order to get the average current value. A bad pipe joint will show high voltage VOLT I* WATER PLU& §1 WATER PLUG 'HH Ull H INI II II 11 11 II II 11 l lll! ; ii 5 5 ii ii ii Sini ii ii ii ii ii En V UUUUUUUUU UUUUUU U" * Fig. 31— diagram of connections. en open switch and large current with small drop in voltage when switch E is closed. Again, there may be considerable resistance in the lateral pipe connect- ing the plug to the main. When this is the case, the closed circuit volts .will be low, no perceptible, or very little, current will flow, and adjacent pipe section readings will not approximate the values which they should show. Fig. 32— apparatus required in this test. ' ELECTRIC RAILWAY HAND BOOK. 51 TEST FOR CURRENT FLOW IN WATER PIPE. The specific resistance of cast iron varies considerably with different samples of pipe due tot! e variation of free carbon in it, and also due to the fact that the weight per foot of pipe changes with the shrinkage, size and displacement of the core around which the pipe is cast. For this reason, in assuming a given resistance for cast iron, satisfactory results are not always obtained. Tables of resistance of cast iron pipe are given herewith, and should be used only for rough appro sana- tions, as they vary from .00112 to .00163 ohms per pound foot. Fig. 31-a The table (see page 51) for wrought iron or mild steel pipe, which material is of a more uniform composition than cast iron, and can be relied upon. The so-called iron pipe which is most extensively used in underground piping plants is really a mild steel pipe, such as given in the above table. CALIBRATION OF PIPE FOR CURRENT FLOW. Corrosion of pipe, both outside and inside, reduces the weight and also affects its conductivity, so the following method is the only one in which any reliance can be placed, when the current which is flowing in the pipe is to be determined. The instruments required are a millivolt meter and an ampere meter, reading about 10 amps., a pair of drop leads, and a pair of heavy current leads with clamps haying amalgamated terminal*, which can be clamped to the pipe. The ampere meter leads should be long enough to include at least 10 ft. of pipe, and not smaller than No. 0B. & S. flexible. In this circuit with the ampere meter leads is also a single pole switch, as shown in Fig. 31a, the drop points to the millivolt meter r 52 ELECTRIC RAILWAY HAND BOOK. AVERAGE CURRENT WHICH WIIX GIVE ONES MIIXIVOLT DROP ACROSS STRAIGHT CAST IRON PIPE. Inside Diameter Weight per Foot of Distance between Drop Points. 4 Feet 8 Feet of Pipe. Straight Pipe. Current Flow in Amperes per Millivolt. 4 20 3.5 1.7 6 30 5.2 2.6 8 39 6.8 3.4 10 58 10.1 5 12 84 14.6 7.3 16 120 21 10 20 180 31 16 24 220 38 19 30 310 54 27 36 440 76 38 42 560 97 49 48 720 125 62 60 900 156 78 AVERAGE CURRENT WHICH WIIX GIVE ONE MIIXIVOLT DROP ACROSS STRAIGHT IRON WELDED PIPE. Nominal Inside Outside Weight of Pipe per Lineal Feet. Distance in Feet Between Diameter. Diameter. Drop Points. Inches. Inches. Pounds. 4 Feet. 8 Feet. M .54 .42 .18 .09 .84 .84 .36 .18 9£ 1.05 1.12 .52 .26 1 1.315 1.67 .72 .36 % 1.66 2.24 .95 .48 1.9 2.68 1.14 .57 2 2 375 3.61 1.54 .8 2^ 2.875 5.74 2.43 1.22 3 3.5 7.54 3.20 1.6 S]4 4 9 3.81 1.9 4 4.5 10.66 4.5 2.3 m 5 12.34 5.23 2.6 5 5.563 14.50 6.15 3.1 6 6.625 18.76 7.93 4 7 7.625 23 27 9.85 4.9 8 8.625 28.18 11.9 6 9 9.625 33 70 14.3 7 2 10 10.75 40.06 17.3 8.6 11 11.75 45.02 19 9.5 12 12.75 49 208 10.4 13 14 54 22 8 11.4 14 15 58 24 6 12.3 15 16 62 26.2 13.1 Beyond this size pipe is taken from the outside diameter. >. ELECTRIC RAILWA Y HAND BOOK. 53 being within the length of pipe bridged by the ampere meter leads. The drop on the pipe is first taken with the ampere meter switch open, and then with the switch closed, and at the same instant the switch is closed, read shunted amperes as shown by the ampere meter. If the shunt circuit is too high in resistance the flow will be too small to give reliable results. The pipe surface, where the contact plates make contact, should be thoroughly cleaned and amalgamated ; or a pipe joint can be bridged by the current leads only, to increase the current through the shunt circuit. The fall in millivolts divided by the current shunted is equal to a constant, which, multiplied by the total millivolts observed when the shunting switch is opened, gives the actual flow of current on the pipe under normal conditions. The effect of the shunt circuit can be neglected, as it is not appreciable in the total pipe circuit. Fig. 31-b Suppose there were 12 millivolts between drop points when switch was open, and 2 amps, diverted when switch was closed, but the drop fell to 11 millivolts, then 1 millivolt is equal to 2 amps, on total flow of 24 amps. With these same connec- tions another test can be applied which is necessary in order to trace currents in a ramified piping system. PROPORTIONAL DISTRIBUTION OF SYSTEMS. CURRENT ON PIPING Assuming the drop connections to be left as in the test for calibration of pipe for current flow, and a wire, about No. 6, brought from the rail of the street rail- way track. In this wire is inserted another ampere meter reading as high as 150 amps. When this wire is connected to the pipe as shown in Fig. 31b, a current will flow either from the rail to pipe, or from pipe to rail, depending upon whether the pipe is positive or negative to the rail. In either case the drop on the millivolt c4 ELECTRIC RAILWAY HAND BOOK. meter will be increased if the wire from the rails is connected to the positive end of the pipe, or the end of the pipe from which the current is flowing, as shown by the millivolt meter. If the increased deflection of the millivolt meter does not agree with the current shown flowing through the ammeter in the track lead, by applying the constant found in the test above mentioned, then the current has another path through the piping system. To find the relative conductivity of these two paths connect the wire from the rails to a point between the drop wires and in a line with them, and fled such a point where the millivolt meter reads zero; then the conduc- tivity of the two circuits formed by the piping system will be the inversely propor- tional to the ratio of the length from one drop lead to the track wire connection, as the length from the track wire connection is to the other drop lead, or A is to B, as the conductivity of path by A is to the conductivity of path by B. TO TEST FOR THE CURRENT DIVERTED TO AN UNDERLY- ING PIPING SYSTEM. There must be several conditions existing to cause this flow ; first, there has to be a drop on the rails back to the pow T er station ; second, the piping system must present a path of conductivity toward the power station; third, the earth resistance must be low to connect the rails and pipe so that current will be diverted to the piping system, and the rails must present a potential, in the location where the cur- rent is diverted, higher than the pipe relative to the power station, to cause a flow toward the pipe. In order to discover the aggregate flow of current on a piping system, the earth resistance where the current leaves the pipe must also be deter- mined, where the piping system is not drained by a ground return from the piping system to the negative bus in the pipe positive territory. Here we have an investigation of currents flowing in three directions, and the system must be completely tested, and data laid out, preferably by graphic methods, in order to discover the condition surrounding, and the conductivity of these aux- iliary return paths for the ground return current in railway systems. LAYING OUT CONTOUR MAPS. The potential contour map shows the fall of potential over the rails to the power station, and also the fall at the same point over the pipe line under test. The pressure wire generally used is the common ground of a telephone system which does not interfere with the service, only making the line noisy while the test is being made. Take a simple case, Fig. 31c r which shows a portion of a city with the lines drawn through points of equal potential, relative to the negative bus of the power station. The broken lines being the water and the full lines being drop on the rail. This gives the potential existing on each system in the same way, as the contour of the country would be laid out, using a potential instead of elevation as the altitudes. This gives a relative plan of the potential which tends to force cur- rent into the piping system. Another map can be made showing the difference of potential existing locally between the pipes and the rail by passing a line through all points in a system of the same difference of potential; also a contour map showing the resistance of the earth between the pipe and rail as determined by test. ELECTRIC RAIL WA V HAND BOOK. 55 Now if we have the resistance of area between pipe and rail, and know the difference of potential, the current diversion to the pipe will be that due to the drop on the track back to the power station, acting over the earth resistance between the pipe and rail. The sum of the pipe line resistance, and the resistance of the area where the current leaves the pipe adjacent to the power station, divided by the electromotive force caused by the mean total drop on the rail back of the power station, will give the current diverted to the piping system. To establish any auxiliary path to make these tests will disturb the normal relation existing between rail and pipe. The expansion of any voltage line and its relation to the point it cuts other diverting lines indicates the relative conductivity of these rails. This line may be distorted for two reasons, one the high current density on the rail, and the other, the character of the bonding of the rail. To determine the current flow J Ufr ^lUUU !_J UUMUMyUUUUl 3BBJ0g§BH"0BaropDIJ i czip ppaa qzi ego c^i Epqan uV Power Station. Street Car Lines. Fig. 31-c through a rail a device like that shown in Fig. 31a can be used with a millivolt meter. This spans when open, 4 ft. of rail, and the average drop on each rail of track at the point of test should be taken, and the approximate current for each rail deduced from the following table. The resistance of rails varies considerably, due to the percentage of carbon and manganese. The resistance of rails can be found under section on rails. TEST FOR DROP ON GROUND RETURN CIRCUITS. The apparatus required is the same as in test (Fig. 30). First open the circuit- breaker in the station and grourkl the feeder to be tested by connecting it to the negative bus by small fuse wire. The volts read at A , divided by the current at £, will give the ground return resistance, including all paths to station. If the return is metallic only, the current will follow Ohm's law; if the return is partly metallic and partly eaith the return resistance will fall with an increase in the 56 ELECTRIC RAILWAY HAND BOOK. measuring current. The most convenient way is to use the longest feeder for a pressure feeder, and employ a tapping clamp for the pressure wire, like that shown in Fig. 34. The relative values of ground returns can also be determined by employing the trolley current. This test requires a five-way shunt board, as shown at C, Fig. 35, and an ammeter, A, to read the main current, and capable of recording 200 amps., uy£ rr£0£/f Fig. 33— connections for testing the drop on ground return circuits. switch rheostat,one plug clamp, and four track clamps and leads and an ammeter, i?, to read off the divided circuit shunts. When the current from the trolley wire through the rheostat reaches the shunts, C, Z>, E, F, c7, it splits up in proportion to the resistance of these various circuits. The conductivity of each circuit can then be obtained by seeing the proportion of the current taking each path, as shown by the readings of A and B. This method, however, short-circuits the ground resistance between rail and -pipe and the apparent pipe conductivity is thus lower than the actual pipe return. Fig. 34— tapping clamp. \%\\%\\v\v\\\\\^\\m\\%v^\%\\^\\m\\%\\x\\^\\^\\m\\^\\m^ »y CK>0 *-ooo-cx>* \ME03TAt 4A 7$ WAfS Fig. 48. center of the scale, and a standard adjusted bridge reading to 20 ohms by T $ ohm divisions, and capable of carrying 4 amps, without any appreciable error. It is convenient to locate the bank on top of the bridge box. The principle on which this test is made is indicated in Fig. 53. Here the goes > Fig. 49. bank current splits through % ohm balancing arms and, when the current in both branches is equal, the milli voltmeter stands at zero. The variable resist- ance can be adjusted until this balance is obtained. Then the resistance in the rheostat is equal to the resistance in the car circuit. This method overcomes the ELECTRIC RAILWAY HAND BOOK. 67 variation of potential continually taking place when testing with trolley circuits; for the rise and fall of potential affects both branches equally, and the sensibility of the instrument and not the zero of the instrument is aCeeted. The measuring circuit to ground plate being always connected, there ia no danger of the operate! on the equipment receiving a shock. ► To find opens remove the ground connection from under the brake shoe and CONTACT HOOK for car test; Fig. 50. GROUND CONTACT £ETJV££/y BRM£< <$ W> >SWTCH VL Fig. 52. A car rheostat varies considerably in resistance especially when used on heavy grades. The G. E. rheostat as a rule falls in resistance with use, while the Westinghouse, Walker and Steel rise in resistance with use. Similar equipments may vary between each other 80# on the resistance steps. For testing the armature and field resistance when in the equipment, it is best to take two insulated volt test handles and stab the bars on the commutator, 68 ELECTRIC RAILWAY HAND BOOK. which project from under each brush, and read the volts; or a metal brush can be made having in it a pressure point lead like shown in Fig. 54, and substituted for the carbon brush and the resistance of armature thus obtained. For this test there should be a 15-volt scale on the 150-volt voltmeter connected across the test brushes, as the controller contacts and wiring give rise to too large a variable to include them in this resistance test. The fields can be tested by means of plugs 4 AMP lAAfP BAM. soo yoir CMCWX CALIBRATED RHEOSTAT 20 OHMS 8* '; OHM 2AMPEHES CAPAC/T? JT/TH .000/2 TEMP. COS* CAR WHEEL GHOOAfD TO TROLLEY MILL! VOLTMETER 0- JN CENTER TO GROUND PLATE TO TROLLEY WHEEL Fig. 53. shown in Fig. 84, clamped to the field wire leads with the connection shown in Fig. 44. Burnt-out fields are indicated by a lower resistance than the standard eet for each type of motor. When cold these fields may appear to be normal and they should be measured when hot if possible. In order to measure the temperature of a motor by field resistance, it should be borne in mind that the resistance of copper increases .21 of 1% for each degree rise in Fahr. A Westiughouse 12A field measures cold approximately .575 ohms at 60° Fahr. If the motor comes in hot and measures .620 ohms, the temperature ELECTRIC RAILWA Y HAND BOOK. 69 can be found by dividing the increase in resistance, .065, by the resistance of field at 60°, .575, which gives \\.Z% rise; dividing this again by .21 will give 53.8° Fahr. increase, or 113.8° Fahr. actual temperature. The armature of this motor measures about .303 ohms at 60; its temperature can be obtained in the same way by resistance measurements Curves 'aid out for temperature corresponding to hot resistances give the rise direct, and save com- putation each time. Again this is useful for locating poorly soldered armature leads, which will show high when hot. If the calculated temperature is higher than the temperature as shown by a thermometer placed on the body of the arma- ture and protected from external radiation by waste placed over the bulb, then the armature should be tested for faults. TEST FOR POWER CONSUMPTION IN STREET RAILWAY EQUIPMENTS. The power consumption of an electric street car passing over a given route at a specified schedule will vary when any one of a number of conditions are P/IE5SME LEAD t *%f/BEfi BUSHING METALIC BRUSH SEcrioiii F/BER BUSHWG Fig. 54. changed. Density of traffic (which will vary the number of stops), the location of stops, the loading of the car, condition of track and obstruction of headway, all introduce conditions in the operation of the car affecting the amount of power required to propel it over the route. In comparing the different types of cars among themselves the variable? are the motorman, trucks and length and weight of car bodies and method of controa. In comparing different sections of tracks the variables are grades, condition of road bed, potential of power delivery and track construction. Tests must be made to determine these variables in order that the result of the different tests on different roadways under the various condi- tions that arise in practice may be compared. Of course, a number of these variables need be determined only where definite values are to be fixed, but methods for making all the determinations will be given. The first variable usually determined is the power value of a start. This value varies with the grade, the time allowed for acceleration and the method of motor control. Where a mixed car equipment is used on the road one car is selected, which type represents the average conditions of all the types used. In order to make these determinations a portable integrating wattmeter should be connected so that all the current passes through the series winding, as shown in Fig. 55 and the armature connected between trolley and ground; also in series with the watt- meter is an ammeter, ^.through which all the current supplied to the motor passes and the line potential is read on the voltmeter, V. Switch 5" should also be connected so that the armature of the wattmeter can be disconnected from the line potential. In order to determine when the car has 7o ELECTRIC RAILWAY HAND BOOK. reached its maximum speed on level track, ammeter readings should be taken when the car is passing ever a level stretch of track with the controller on the last notch; the constant current taken under thesa conditions required by the equipment will be the maximum level run constant. This constant can be determined on different points of the controller where the car can be continu- ously operated. For the determination of the power value of a start from the state of rest to maximum speed, the car should be stopped at a marked position on the track where it will have at least 600 ft. clear headway and the reading of the integrating wattmeter taken. The pressure switch which connects the wattmeter armature across the line is now thrown in. The car is started up with regular time between controller points and the rise and fall of the current through the ammeter watched carefully. The instant the current has fallen to the maximum speed vaiue previously determined, open switch A , and at the same instant note the point that the car is passing. A number of starts will have to be made from the Fig. 55. 9ame point and the car brought to full speed within the same distance, until the total watts consumed, divided by the number of starts, will give the same con- stant. Maximum speed constants will have to be determined for the different grades on the road, for different loadings of the car and different points of the controller and curves plotted for these results. These determinations are used in connection with the following tests in order that the power consumption of the cars in practical operation at different times of traffic density may be compared. Different rates of acceleration can be determined in the game way, and the distance and energy required to get to maximum speed. From these tests, then, the best method of handling the controller can be developed, both for schedule required and economy in power consumption, by making test runs over track sections and varying the method of handling the controller, both in the series mul- tiple and loop positions, for different grades. To determine the kilowatt consumption per car mi le under the practical traffic conditions of a road and the power required by different kinds of car bodies, trucks and equipments, a section of tracks should be selected which represents' the average track condition of tho system, which should not be less than three HiWee lonff. A profile of this tection will help to Analyze the result^ but the foi* kx ELECTRIC RAILWAY HAND BOOK. 71 lowing precautions must be taken in getting data from these runs so that the different cars will be comparable as to power consumption. If the wattmeter is read only at fixed intervals of time or distance or when an equal number of watts have been metered, the values apply only for a level track. Where there are several grades in the test section, readings of the wattmeter should be taken at the moment the car reaches the top of the grade, going in both directions over the test track. This will obviate the error introduced by drifting, for an easy run- ning truck over a variably graded track may show a very high economy, with the car passing in one direction. This is due to the fact that the motors are working at full efficiency for a a short time while climbing a short sharp grade, and after reaching the top, the car may drift for a long distance; yet in the return trip the motors will not be worked economically while climbing along moderate grades to reach the top, and have only a short period of rest while drifting down the steep short grade. Also, if the car stops at the end of the test track at a level lower or higher than at the beginning, the complete run to the end and back is the only one that will give comparable values as to power consumption. These points must be borne in mind in arranging the test, so that seemingly contradictory results will not be obtained. The other data to be obtained on these runs are the times which elapse be- tween the start and finish of the test runs'. All stops should be noted, and all stops longer than thirty seconds should be timed. When the car is left running on a particular point of the controller, it should be noted. The maximum current rise and the running current values with the line voltage for both these values should be noted, where specific values are to be determined for a different car equipment; but these variables can be averaged out and a practical average car consumption value per car mile can be obtained if a large number of runs over the same test track for each equipment tested be averaged. A wattmeter in the station will show higher watt readings per car mile than those shown on a car test, as the readings will be increased by the line drop and ground return losses. These values will vary as the distance from the station increases and with the economy of the distribution system. Data used for figuring the power consumption of electric cars as usually given, are based on a false assumption, when these data are applied to the car mounting grades, as it is assumed that the car continues to mount the grade at the same speed as it travels on the level. The usual method is to take the cur- rent required for the level speed and add to it the energy in current necessary to raise the weight of the total equipment through the elevation attained by mount- ing the grade in one minute of time— mounted, as grades are in practice, with no resistance in series with the motors, in either a series or parallel combination. It is evident that the only way more energy can be supplied to the motors with a constant line voltage, is when they drop in speed and reduce their counter e. m. f., so that more current can flow. Consequently, the above assumption wilj lead to erroneous results, as the car will at all times adjust itself to such a speed as to obtain the maximum energy for mounting any grade when there is no external resistance in series with the motors. This makes the grade determina- tions of power consumption very important, as they cannot be figured with the degree of accuracy required for power consumption determinations, due to the large number of variables which affect the current flow through the motors ; but it is a wise plan to use the above approximate method in figuring the railway feeders as it introduces a factor of safety in railway feeder calculation which is Usually neglected. yniilo the wattmeter gives the operating economy of tb* equipment as regard! 72 ELECTRIC RAILWAY HAND BOOK. the demand on the power station, it does not indicate the best economy of operation with respect to the heating effect on the equipment ; and as the depreciation and repairs of motors are largely dependent on the temperatures at which they are operated, the integrating wattmeter is not a criterion of the best method under a varying potential delivery due to feeder and return drops. The reason of this is that the heating effect is a function of the square of the current, whereas the motor heating per car mile will increase a great deal more rapidly than the watts per car mile, when operated under potentials lower than those for which the equip- ment was designed. TEST FOR MOTORMAN'S CHARACTERISTICS. To obtain the motorman's characteristics for running his motors, the maximum ammeter readings have to be taken in connection with wattmeter readings. On starting it is also necessary to take the volts delivered to the equipment, for as the ft TH£KMQM£T£R PAP£# IRON SPOOL Fig. 56. volts fall, the efficiency of the equipment falls, and therefore a mile-run on a part of the system distant from the power station will not show the efficiency of a mile-run of the equipment under the same conditions near the power station. The average motorman's characteristics can be more clearly ascertained by the C 2 R effect of the current which results in heating the motors. They can be meas- ured by making a motor calorimeter (Fig. 5G) as follows: A wrought iron spool has a hole drilled at the top to receive a thermometer and is filled with mercury; around the spool is wound No. 6 B. & S. copper wire for double motor equipments, and No. 5 B. & S. copper wire for four motor equipments. The main current to the motor is carried through the calorimeter which is insulated to have practically the same rate of radiation as the motor; then the temperatuies obtained by the calorimeter correspond to the motorman's efficiency in handling the motori. ELECTRIC RAILWAY HAND BOOJC. 73 ACCELERATION AND BRAKING TEST. The same connections that are used in the above test are used in testing the efficiency of acceleration, with the addition of the apparatus shown in Fig. 57. This instrument, which is located in the car, consists of a long pendulum, A, arranged parallel to the rails, and a pencil carrier, E, which is moved over a sheet of paper, D, by means of a fine cord belt, F, passing over rollers, C,C % C,C. The paper may be fed along by hand, clockwork, or may be connected to the car axle by a belt, B. The zero line is made for the pencil carrier when the pendulum, A, is hanging plumb. On starting the car the pendulum bob will be deflected and as long as Fig. 57. the car is accelerating, will not return to the zero line; as soon, however, as it reaches the zero line again, the car has finished its acceleration. The wattmeter is first read before the car is started and again when the pencil carrier returns to zero. It is convenient to have a switch in the pressure line which can be opened the instant the acceleration has ceased, and the wattmeter is read. To find the distance required for acceleration, if the paper is fed from an axle which is not driven by a motor, its rate of feed can be calculated or calibrated very easily by a few test runs between known distances. In making a brake test the pencil travels in the direction opposite from zero. 74 ELECTRIC RAIL WA Y HAND BOOK. Here the time is noted as soon as the hand leaves zero nntil it returns again, and if the paper has progressed in a ratio to the movement of the car, this curve will give the relative braking effect of the shoes on the wheels. Where acceleration is to be determined on grades, the pendulum will draw curves on the paper proportional to the grade over which it passes; in this case, therefore, the zero is to be marked at the instant of applying the brake, and the braking effect refers to this zero and not to the level track zero. The same pre- cautions must be used in acceleration tests. When an ammeter is placed in series with both motors in an equipment (generally most conveniently done by taking out the fuse and substituting am- meter leads), the ammeter will show a large flow of current when the controller is put on the first point. This rush of current will not reach that which should be shown by dividing the line potential by the equipment resistance, due to the momentary inductance of the motors. As soon as the equipment moves, the cur- rent will be found to fall, due to the counter electromotive force of the motor armatures, that is, the armatures are revolving in their fields, and in them is induced a potential which in direction is against the potential of the current operating the motor; this produces a throttling action in effect like that of a resistance in the motor circuits. Due to this the resistance can be cut out of series with the motors as they rise in speed. The ideal acceleration is one in which a constant current flow would be main- tained through the equipment, and the resistance would be cut out as the accelera- tion increases the counter electromotive force of the motors (until the equipment has reached maximum speed). For methods of approximating these resistances for the different types of equipments and controllers, see under "Equipment Adjustments." There are three methods that can be used for acceleration tests. One, the stationary current values, the time and total distance; the second, the fixed time between controller points, reading the current and total distance; and the last, fixed distance, current and time variables. The test track can be marked with eleven numbered stakes 100 ft. apart for 1000 ft.. The track should be practically level. With high speed equipments a 1500 ft. stretch is necessary when the first method is used. The acceleration tests on an equipment with K 10 controller and two 12A West- inghouse 30 hp. motors give the following results of current on each step : Steps: 123456789 Equipment up to standard 68 35 26 23 20 68 71 63 52 " Wi ^w b ?hi y o 8 TSrnc d e and }-- 1 "» ^ 27 27 22 K0 70 65 60 " with baked fields in motor 120 97 85 84 82 130 100 95 90 Kunning alone with No. 1 motor 140 110 90 64 50 " No. 2 " 160 130 1C0 80 82 showing the location of the bad fields in No. 2 motor. With a stop watch bad fields can be located by first running between fixed points, from 1000 to 2000 ft. apart, with motor No. 1 alone and taking the time elapsed between passing both posts, then going over the route again with No. 2 motor. The time to cover the distance should be the same for both motors if they arc both good, or both burnt out (which is rarely the case without the fuses going so as to indicate trouble). This is also a good test to discover whether the fields are properly connected up. In K type controllers do not use the loop around fields as these loops vary in resistance enough to affect the running times and thus throw suspicion on the fields. ELECTRIC RAILWAY HAND BOOK. 75, * MOTOR TESTS FOR REPAIR SHOPS. The Prony brake method is sometimes used where the efficiency of the motor is to be determined. A double flanged pulley, as shown in Fig. 58, is belted to the axle shaft, and over it is fitted a short hard wood beam; a brake strap of rope or sheet iron lined with wooden shoes is arranged so that it can be tightened &D jj OAK BEAM "Z&^ G Fig. 58. or loosened by the nuts, D D, At a fixed distance from the center of the axis is a notch, F, in the lower side of the lever which rests on a knife edge, as at G; this is mounted on the platform of an ordinary weighing scales. This beam should be balanced so that the end of the beam will not rest with weight on the TAOLLEr SM/TCH 37AKT/A/G RES OR WATER RHEOSTAT AMPERE MET£f( VOLT, GROUND. Fig. 59. platform scales ; rubber washers, JF, are interposed to deaden the vibration from F to the scales. The motor is connected up as shown in diagram, Fig. 59, the voltmeter across the motor and the ammeter in series with it. In addition, a tachometer is required to read the speed of the motor. The water barrel rheostat is best where the voltage varies, so that the test can be made under identical conditions of line potential by varying the rheostat. T* ELECTRIC RAIL WA Y HAND BOOK. After switching on the current and starting the motor, any desired load can be obtained by tightening the nuts and drawing up the eye-bolts, thus increasing the friction between beam and pulley. If the horizontal distance from the center of the pulley to the bearing point of the beam on the scale platform were equal to the radius of the car wheel the pounds indicated upon the scale would be the pull at the periphery of the car wheel for the current passing; if this distance were equal to four times the radius of the car wheel, the scale reading should be TROLLEY Fig. 60. multiplied by four to obtain the pull at the car wheel. By placing a tachometer against the end of the axle, a speed reading, usually in revolutions per minute, at any load may also be obtained. To calculate the horse-power developed by the motor for any given amount of current passing, proceed as follows: Multiply by two the radial distance in feet and decimal parts from the center of the pulley to the center of the notch on the beam and this result by 3.141G, which gives the circumference of the sweep of the beam were it free to move; multiply th:s result by the revolutions per minute, as read from the tachometer, and the result is the ELECTRIC RAIL WA Y HAND BOOK. 11 >ed in feet and decimal parts per minute. Multiply this calculated speed by le pressure the beam exerts against the scale platform, when balanced, and [ivide the result by 33,000, which gives the horse-power exerted by the motor; and dividing the horse-power exerted by the horrc-powcr supplied (which is obtained by multiplying the amperes and volts together and dividing by 7iC) gives the efficiency of the motor. In this is included the friction of gearing and axle. TESTING MOTORS: MOTOR-DYNAMO METHOD. Here the Prony brake is replaced by another motor. Both motors should be coupled together by a chuck, which will slip over both pinions on the ends of the motor shafts with screws set down between the teeth on both pinions ; the TROLLEY SWITCH START//VG RES. OR WATER RHEOSTA 7 DY/VAMO FJELD. JUMJL AMPERE /HET£K VOLT METER COUPLED TOGETHER MECR/IM/CALLY GROUND MATER RHEOSTAT Fig. 61. motors will then rotate together. Fig. 60 shows a diagram of the connections required for testing two pairs of motors at once. With the switches closed, the Westinghouse 3 and the G. E. 1200 would be operating as motors and the G. E. £00 and the G. E. 1000 would be operating as generators through the w liter rheostat. This test is used more extensively to determine the insulation resistance of the armature windings and as a running test for armatures rather than to make efficiency measurements. The usual rule is to ran each machine ten minutes as a motor and ten minutes as a dynamo; if no excessive sparking or other faults arise the armature is put in jitock for use. The motor fields are best turned upside down, so that the brushes do not interfere with lifting the armature to be tested in and out of its bearings. The different ways of connecting up the motor are shown in Figs. 61 and 02. 7? ELECTRIC RAILWAY HAND BOOK. In Fig. 61 the field of the motor running as a dynamo is separately excited; the motor to excite itself as a dynamo has to be run in the opposite direction from that which it runs as a motor, or the field leads should be interchanged. Dividing the kilowatts output by the kilowatts input gives the total efficiency of the transforming system, including all losses in both motor or dynamo. The efficiency of the motor is greater as a rule when operating as a motor than when operating as a dynamo. To make detail tests on a motor, the dynamo should supply all the current required by the motor, and the power losses should be compensated for by a second motor, geared or belted to the motor generator so as to run it faster than TROLLEY SW/TCH STARTI/YG RES. OR WATER RHEOSTA 7 DYNAMO FIELD FfELD AMPERE METER VOLT METER COUPLED TOGETHER MECHANICALLY WATER MOSTAT GROUND Fig. 62. at its rated current and speed. A dynamo machine, when running at a given speed, will not produce an e. m. f . as great as that which it will require as a motor under identical conditions. Fig. 63 shows the diagram of connections employed. A (the dynamo) and B (the motor) have their armatures and fields in series and are connected so both revolve in the same direction. The necessary increase in speed of A can be approximately calculated if the current, C, and internal resistance of its armature and field, a, are known; then a C is the drop in the machine due to the resistance. Let V be the desired volts at the motor terminals; the counter e. m. f. will be equal to V — aC — E. Let S be the speed of the machine when running as a motor with V volts and current, C. Since as a motor it generates a counter e, m. f. of E Volts and as a generator it produces V volts, then K\V\ iSiS* •••••••■ • ELECTRIC RAILWA Y HAND BOOK 79 S' being the speed required by this motor-generator so that the motor can operate from the current supplied by the generator to which it is coupled. Mr. Parham gives the following method of carrying on this test: A and B have their fields and armatures in scries, as shown in the diagram, and included in the circuit are a switch, A", and a variable resistance, R, capable of carrying the machine's full current. R exceeds the critical resistance of the machine for the given speed, so that upon closing 7T the dynamo will not generate until part of R is cut out. Before starting a test it is well to determine the correct position of the rheostat handle for the dynamo to generate. On account of the ability of a series machine to pick up rapidly as soon as it begins to generate, it is well to provide belt guards to avoid the annoyance of losing the belt under sudden overloads. A further precaution is to insert a light fuse at the start, and then cut it out when the test is under way. If the motor shaft is arranged to be thrown in by a clutch the start is much smoother. In starting up, the machine is brought up to speed, K is closed and R is slowly worked F/eto IRUMWtfG MOT OR \r« °n^ Fig. 63. out, at the same time /.'a field is weakened by means of the shunt, r, shown in Fig. 63. As soon as the ammeter shows A to be generating, 7? must be very care- fully handled to avoid precipitating a heavy overload and throwing the belt. A will refuse to generate until a certain amount of R has been cut out, and will then pick up very rapidly. It is absolutely necessary that means be provided for weak- ening the motor field, otherwise since the same current must pass through both machines, and since they run at the same speed, the counter e. m. f. of B will be the same as the e. m. f. of A and a load cannot be worked on. The shunt affords the same regulation as obtains on a car, but has a different relation, ; n that on a car its value is constant and the speed is variable, while in this test the shunt is variable and the speed constant. ^r-Ts current passing through B runs it as a motor, and helps to turn the sys- tem, thus lessening the demand on the supplier, which then supplies only energy enough to cover the losses, which may amount to from 25 to 35 per cent of the motor's output. After running A for a stated time as a dynamo it is changed over aacl run as a motor, Tniu change is most rapidly effected by using a crossed 8o ELECTRIC RAIL WA Y HAND BOOK. belt to reverse the direction of rotation; it is then only necessary to move the shunt from Bio A. To separate the different motor losses both motors are run by the operating motor, first free with brushes out and no fields. The reading of an ammeter and voltmeter across the operating motor terminals will give the electrical losses in the operating motor, and the friction and air resistance losses of all the machines. Running the operating motor free with belt off will give some of the losses which should be subtracted from the total input of the operating motor. Fig. 64. To determine more nearly trie true friction losses of the motors (the belt Jobs is still included in the motor friction') care should be taken that the two motors under test are properly aligned with no undue friction from coupling the shafts. Put in the brushes on both motors and read power input, and the differ- ence is the brush friction, which should be divided by two to find the brush friction on one motor. If the field is excited with different currents independently, and trie power required for each degree of excitation is plotted between power taken and excita- tion current, the core losses at different outputs will be obtained. If a voltmeter is placed across the brushes and the voltage is read for each change in excitation ELECTRIC RAILWAY HAND BOOK. 81 a characteristic curve can be plotted between excitation and open circuit voltage developed at the armature; the set can then be run as a dynamo-motor com- bination up to overload, the auxiliary motor supplying losses, with a constant current flowing through the system. Another important point is to have the auxiliary motor calibrated so that from the current input the actual power output will be known, as the efficiency and losses in the auxiliary motor are not proportional to the current changes. This motor should be calibrated with the Prony brake as shown in "Motor Tests," so that a curve of actual output in horse-power will be known from the input in watts. ARMATURE TESTING. The armature may have the following faults, which can be located by testing: An open lead from the commutator to armature coil ; an open armature coil ; bars on commutator short-circuited; coils on armature short-circuited; ground on commutator ; grounds in armature. rv-/ -5 ^ , DEVELOPED J ARMATURE YOLT METER GROUND Fig. 65. Fig. 64 show a contact bracket rig which slips over the armature shaft and can be revolved so that the different bars can be tested; besides this rig the test requires a low reading voltmeter, reading about 1.2 volts full scale, and two amperes from a test bank, which is all the current necessary. If the brushes on the test rig are so located that four commutator bars are between these brushes, the best average conditions for testing the different railway armatures in use are secured. The current is carried through two brushes and the other two take the drop to the voltmeter. In case of open lead at A (Fig. 65) the bank will go out and the current brush will flash when passing to that commutator bar, but it will also be seen that the circuit is completed to ground through the low reading voltmeter so that armatures tested for broken leads to commutator should be tested around the commutator by the lamp bank only first. Then the pressure brushes are lowered and the armature is again slowly revolved and the deflection on the voltmeter watched. If there is a break in the armature winding, as at A^ the drop on the volt- meter will increase greatly beyond the normal drop for four bars, for the reason that all the test current has to follow from one brush to the other through the windings of the armature external to the brushes. 82 ELECTRIC RAIL WA Y HAND BOOK. The voltmeter will fall again to normal readings when the bar connected to the broken coil passes beyond the test brushes. The bar can be located in this way and marked. A short circuit between two commutator bars or windings will show a lower reading than normal when the commutator bars to which the windings are con- nected are between the test brushes, and can be located when passing from under 2-2 Fig. 66. the test brushes by the voltmeter reading jumping back to normal again. To determine whether the short is between the bars or the windings, the adjacent bars should be short-circuited by a copper bridging piece, when the defective part of the armature is between the brushes. If this changes considerably the reading on the meter, the short between bars is in the armature. If no change is made the short is in the commutator. For thc3c two teste ««^ »*»t* test, which is the most expeditious and searching for shorts in annatuio*. 'ELECTRIC RAILWA Y HAiVD BOOK. 83 For the location of grounds the connections, as shown in Fig. 65, are changed by carrying the current through the armature and then to the ground with a 500-volt voltmeter in series. A green armature should not measureless than 75,000 ohms (See Insulation Test by Voltmeter Method), and a baked arma- ture should be over V/% million ohms, or \\^ megohms. If the insulation rises on the application of the testing current it indicates the presence of moisture, while if it falls, a leak through a charred insulation is probably present. If the voltmeter shows nearly normal volts then the ground can be located in the fol- lowing way : Pass the bank current through the commutator to the shaft and also connect across these points a voltmeter reading the drop between the commutator and shaft; carry the contact and voltmeter contacts slowly around the commutator and watch the voltmeter drop, when this has reached its lowest point, and rises in volts again then the commutator bar connected to the grounded coil is located. Several grounds may exist, and can be picked out in this way by following up each one separately. INDUCTION METHOD OF TESTING SHORTS ON THE ARMATURE. For this apparatus working drawings are given in Fig. 66, which shows a framework made of angle iron mounted on three rollers. The testing trans- former is made of laminated iron about J% in. thick and of the shape and size shown in Fig. 66; this is adapted to test G. E. 800-1000-1200 and No. 57, and Westinghouse No, 3 and No. 12. The magnetizing coil is made as shown in Fig. 66, and wound with 1210 turns of No. 13 B. & S., D. C. C. wire. The curved face of this transformer is adjusted by a hand wheel and screw so that it can be shoved up against the armature before the latter is removed from the winding bench. The body of the armature completes the magnetic circuit of the transformer; the armature is then rotated by hand in this field. If any two windings are short-circuited and are w r ithin the influence of this varying magnetism, a local induction circuit is created which causes a vibrating magnetic flux in the teeth of the armature included in the short circuit; this is discovered by passing a thin strip of iron around the armature, which when over a short-circuited coil will vibrate in unison with the alternating current supplied the transformer. There will be two such points in the armature at a quarter from each other when two adjacent armature coils are short-circuited, but at four points when the short is between the commutator bars in a four pole cross-connected armature. The current for this testing device may be obtained from a railway motor (an old style one will do), changed over for this purpose. Two slip rings should be secured to and insulated from the shaft, and connected to the windings of the armature. The connections in a two-pole motor should be located one- quarter of the circumference of the armature from each other and in a four- pole motor, one-eighth of the circumference apart. The motor should then be wound with a shunt field of fine wire. With a two-pole motor a speed of 1400 r. p. m. gives a good frequency to detect these crosses. The armature room should be wired for this current over the winding benches, and flexible cords with attaching sockets located at points convenient for connecting to the testing device. While on the winding bench an armature can be tested by this method in less than one minute. 84 ELECTRIC RAILWA Y HAND BOOK. FIELD TESTS. In fields the fault most generally looked for is a short circuit between the layers. This may be caused by charred insulation from overheating, or by a breakdown in the insulation between the windings. A field cOil should never measure more than 5 per cent under its standard resistance; annealing will account for this difference in some copper wire. A field may be low for the reason that turns are shy or have been cut out in repairs, both of which are bad practices, and should not be used where the best service from the armature is desired. The most treacherous defect in the whole equipment is a baked field. These will test O. K. when cold, but when hot will show the defect; this is due to the expansion by heat; the convolutions are brought into more intimate contact and turns shorted out. For testing a cold equipment the connections are shown on page 31. While the current passes through the field, if mechanical pressure can be brought to bear on the cover or spool face so as to bring the wire in the field windings into close contact, and the reading on the voltmeter changes, this will indicate at once that the field is baked. When testing separate field spools that have been used, always stand on them, when, if they are baked, they will show change in resistance due to shorts set up in the coil due to pressure. The method of testing for field grounds in the equipment, is already given In Equipment Test pages 50 and 51. CONTROLLER TEST. The test on the controller consists of locating grounds and shorts. A ground can be located after first disconnecting the controller from the equipment and then testing for grounds the same as in the case of the equipment ground test. When the controller is disconnected from the equipment each clip has to be tested separately for grounds and different portions of the controller cylinder to see that the contact rings are not connected with each other. The test can be carried out by a magneto or a series of five lamps or a Weston volt meter in series between the line and the clips, having the other side of the line connected to the controller back. In some types of controllers the ground is permanently connected to this back frame of iron, which connection should be removed before attempting to locate other grounds. RHEOSTAT TESTS. Testing the rheostat for resistance is done in exactly the same way as that shown for fields. The resistance is measured on each step by connecting the voltmeter leads to the terminals of each step when a known current is flowing through the whole rheostat. For locating an open, connect the rheostat between the lamp bank and ground. Then take a piece of wire and bridge or jump around the open, reducing the span of the bridge until the break is located between the bridging wire. Grounds on the resistance can be discovered in the same way as given for controller test. TEST ON RAILS. According to D. K. Clark the usual tests for steel tramway rails are : — Breaking stress (tensile), 3? to 43 tons per square inch. ELECTRIC EAILWA Y HAND BOOK. 85 Elongation in length of 8 inches, at least 15 per cent. Contraction of sectional area, at least 30 per cent. A piece of rail 5 feet long, on supports 3 feet apart, to resist a blow from weight of one ton falling on center from given height without causing more than 1 inch deflection; a second blow from another given height without exhibit- ing sign of fracture. The height of drop of first and second blow are determined by the following table. Weight of Rail per yard. Height of First Fall in feet. Height of Second Fall in feet. 60 to 70 70 " 80 80 " 90 90 " 100 6 7 8 9 15 20 22^ POLE TESTING. The method usually given to test a pole is to set it the proper depth in the I ground, apply a tackle to the top and draw it up with a given tension, noting the deflection. The tension is usually applied at an angle from the ground, 1 and is borne partly by the pole longitudinally and partly by the spring of the Fig. 68. pole. The ratio of the two depends on the angle of application. Any yielding in foundation is liable to be charged to elasticity. If the pole is planted in cement, several days should be allowed the latter to set before the test, which makes a rather lengthy affair. I have found that the following will give all the practical results necessary and can be readily constructed and calibrated on the spot. See Fig. 68. For this testing rig two of the largest and most sym- metrical poles are selected; they are then laid by the side of each other and separated far enough apart to allow any of the poles to be tested to lie between them. They are braced together at two points, the distance «, from the bottom of the foot brace to the top of the head brace, to be the depth of setting to be employed for the pole. Sufficient area to these braces should be given at // so that the strain applied will not crush the fiber of the pole under test. A 12-in. turnbuckle, and a 1:20 steelyard, with a 100-lb. weight, completes the outfit. 86 ELECTRIC RAILWA Y HAND BOOK. One arrangement can be made as above. The weight on the steelyard is fixed, and, after the bridle is put over the pole, the turnbuckle is tightened until the steelyard is balanced. "When the specified tension is applied, the deflection of the pole is measured by a mark on a board opposite the pole under test. This deflection is composed of the yield of the braces and the flexibility of the crib, but the pole on returning will show the permanent deflection which will be the difference before and after strain. Iron poles can be tested by the same method by substituting an iron pole for the wood, and the steelyard should be provided with a set of weights, so as to reach a 2,000-lb. strain, where extra heavy poles are to be tested. CEMENT TESTING. A rough way to test the quality of cement is to take two batches of about a handful each, mixed with as little water as possible, and make them into cakes. Put one of these cakes in water and the other in air. If the cake in the air dries with a light color without any particularly well defined cracks, and the water cake sets with a darker color and without cracks the cement is probably good. If the cement cracks badly in setting, or if it becomes contorted (sometimes called blowing), it is poor and should be rejected. Kidder gives another simple test for the soundness of cement. This is to place some cement mortar in a glass tube (a swelled lamp chimney is excellent for this purpose) and pour water on the top. If the tube breaks the cement is un- fit for use in damp places. Any natural cements that give satisfactory results with these simple tests will answer for making mortar for any ordinary building construction. A good cement will not expand, contract, check or crack when setting. Where great strength is required in the mortar it is better to use Portland cement, but if for any reason Portland cement cannot be obtained or its price prohibits its use, the strength of the natural cement should be carefully tested in the manner described. Clear Rosendale cement one week old in water should have a tensile strength per square inch of at least 60 lbs., and the best brands should average 100 lbs. Measuring Fineness: "The degree of fineness of a cement is determined by measuring the per cent which will not pass through sieves of a certain num- ber of meshes per square inch. 1 ' A cement that will pass through a sieve of 2500 meshes (No. 35 wire gage) with only 5 to 10 per cent residue is sufficiently fine for any building construction. TEST FOR OVERHEAD LINE INSULATION. Testing for overhead line insulation is best done by clamping to the trolley pole, near the harp, two blocks of fiber to which are attached two strips of phosphor bronze. The latter should project beyond the trolley wheel, far enough to come in contact with the span wire, and as the car passes along these flexible strips should make contact with the span wire. Then connect a 500-volt voltmeter between a lead from the strips and the ground so that only such potential as can leak through the trolley insulator can deflect the voltmeter. If the voltmeter is calibrated in insulation resistance in a way similar to that described in the section on testing equipment, the resistance can be read in megohms as the car passes slowly along, striking each span wire. - ELECTRIC RAIL WA Y HAND BOOK, 87 HIGH TENSION INSULATOR TESTING. As the potential on a line increases, the tendency to loss by leakage increases in a ratio varying nearly as the square of the potential. This is the case with bare conductors, but when insulated conductors are used losses by leakage increase less rapidly, on account of the combined effect of the two insulations. The in- sulation afforded by an insulator varies with the moisture in the air, dust, tem- perature, and other climatic conditions. If the material forming the body of the insulator is a good insulator, the loss on each insulator is a matter of surface leakage. In the design of insulators and the securing of the conductors to the insulator, the subject of areas of leakage is not usually given proper attention. The external surface of an insu- lator once determined, the value of this surface as an insulator can be computed approximately by ascertaining what is the cross-section and length of a film of moisture which could be deposited thereon. The linear distance from the point of connection with the conductor to the point of contact with support- ing pin multiplied by the mean circumference of this path, will give a compara- tive value for insulators of the same material which will vary as their insulating qualities. The insulating values of these leakage surfaces vary with the exposure to the weather. The external surface of a bell has in damp weather no appreciable insulating value, but the petticoats provided underneath the bell maintain the insulation. It is hardly possible to pierce insulators made of glass by increasing the potential, as they will withstand a potential which will flash over the external surfaces and arc between the pin and conductor before the glass is actually pierced. The glass surface on being exposed to rain is serrated and grooved, due to the solvent action of rain water on silica, and these roughened surfaces allow the lodgment of dust and soot, which forms a partial conducting medium, and reduces greatly the effective insulation. A test made on forty glass insu- lators, representing a mile of line, after being dipped in water once and a little dirt removed, gave a resistance of 23,223 ohms per mile. After being dipped four times this resistance increased to 56,400 ohms. With new insulators and pins 66,600 ohms per mile was found. With high tension currents, particles of moisture are repelled from the conductor electrostatically, and foggy weather, for this reason, does not bring down the insulation of the line as much as on telegraph and low potential lines. When a leak over the surface of an insulator is established, the current flow- ing over this leak raises the temperature of the insulator and dries up the con- ductor moisture. Power transmission lines require the stringing of wires of considerable weight per foot, and the fragility of glass has made it an uncertain mechan- ical support for these conductors when under tension. It is very important that the conductor should never touch the cross-arm or pin, as the leak will probably burn up the pole if of wood, and if of iron, it will cripple the conductor system. Porcelain, when used for the body of the insulator, possesses more mechanical strength and the surface of porcelain when good, weathers exposure without deterioration. In the clay from which these insulators are made, a large proportion of American kaolin should be avoided where the insulators are to be used for high tension work, as these clays are too refractory to completely coalesce or vitrify when fired, and consequently they lack homogeneity, an essential quality in a high potential insulator. Semi-porous insulators can be easily detected by applying aniline ink to a 88 ELECTRIC RATLWA Y HAND BOOK fractured surface. If porous, this ink will be absorbed into the porcelain, but if thoroughly vitrified, the ink can be washed off without leaving any stain. This test should be tried on the thicker portions of the body of the insulator, as they may not have been vitrified, while the thinner portions reach sufficiently high temperatures to be vitrified. The specific insulation of porcelain is less than that of glass, and being opaque, it affords a good harbor for the nests of insects inside the insulator; but on account of its superior mechanical properties, it is used on nearly all the transmission lines in America employing hi~h potential. Insulators for high tension lines should be tested individually at at least four times the electrical pressure under which they are to be used. This will probably give a factor of safety of 16. The method of testing generally adopted is to insert the insulator head down in a shallow pan, into which is poured sufficient solution of bicarbonate. of soda and water to reach above the groove for the tie wire. The same solution is poured in the hole for the pin, and in this is inserted a metallic wire which forms one terminal of the high potential circuit, the other being connected to the pan on which the insulator rests. The testing potential applied should be of the same character as that to be used on the transmission lines. A metallic con- ductor connected on the outside of the insulator, in the same manner as used in the transmission line, and the insulator screwed down with a metallic pin, reduces the area of contact, so that the insulator will stand a much higher poten- tial than in the test given above. When in actual practice this conductor is sup- ported in the rain on this insulator, the area of contacts will be much greater than in the dry test. The insulator should be submitted to the testing potential for some time, as for a few moments it may stand a much higher potential; but it is under a stress which reduces its insulating values, and if applied long enough may break down the insulator. Ten minutes should certainly determine whether this fatigue would reach the rupturing point. It is hardly advisable to place any insulators on a high potential line without first testing them individually, for the reason that in the formation of the insulator in the mould, the clay has to be moulded under a uniform pressure, If this pressure is not distributed through- out the clay while in the mould, it will cause the clay to be of unequal density throughout the body of the insulator, which will result in unequal shrink- ages while firing. These differences will cause small fissures through the body of the insulator, and this inequality will lead to a breaking down under the potential test. This condition is not evident from external inspection. The final glaze on an insulator should be entirely burnt into the porcelain itself. If too much glaze is put on it is worse than none at all, as this glaze has all the characteristics of a glass surface. When an insulator breaks down under a high potential, it is actually punctured by the current, which usually pierces between the pin and the external surface of the insulator. Under 60,000 volts, a poor insulator will explode with some vio- lence. The open double petticoat insulator was found to dry more rapidly than the closed single petticoat insulator, but during actual rainfall the insulation lost >^y the double petticoat form is greater and more rapid than that of the single form. In order to break the conducting film of moisture on the surface of an insulator, several methods are used; one is to have an internal groove in which oil is poured. This interrupts the continuity of the moisture film and improves the insulation. Lining the top of the petticoat with paraffin has also been tried with partial success. ELECTRIC RAIL WA Y HAND BOOK. 89 Another feature of design of these insulators for higa tension work is to make them helmet shape, the rim of the helmet being over tho cross-arm, so that water dropping from this insulator will not fall directly on the cross-arm, and spattei moisture underneath the insulator. The greatest leak on transmissior lines occurs during fogs, and the greater the change in temperature, under foggy conditions, the greater the leak. The insulation of a line rises as soon as rain begins to fall. The method of securing high tension lines to their insulators is to provide a groove in the top of the insulator, in which the conductor rests. The conductoi is held in place by a tie-wire, the object of this wire being to hold the conduc- tor \-\ the groove, and yet allow of contraction and expansion of the conductor, without bringing additional strains on the line. Tins. — Wooden pins are as a rule preferred for supporting these insulators. They should be made from split locust and their values as an insulator are in created by being boiled in paraffin. £c::;e tests on the leakage of cross-arms, made in New York City, are given, but in llie test the length and dimension and method of test of the cross-arm are not ttut.d; consequently, the results are only comparable among themselves. Ohms. All four surfaces wet with sponge 3,120 Soaked one day, left to dry one day, and then wet 2,680 Painted three years before test 6,150 Same washed 9,166 Very dry 77 11,000 to 330,000 Newly painted 7,214 Unpainted for many years 4,300 Same after being well washed 13,653 Same after being well dried 80,000 Arms and pins together (wet) 3,686 TESTING DYNAMOS. After the erection of a dynamo and before it is put into service, the following teste should be made in order to locate a misconnection, an improper field spool or defective insulation. Without this precaution of testing the dynamo, faults may develop which will seriously injure the machine when put into service. Taking a multipolar direct-current generator for example, proceed as follows: Pass a known current through the field coils and measure the potential drop across each coil. These drops, in a properly constructed machine, are equal. The current in adjacent magnet coils should circulate through the windings in opposite direc- tions, and, if facing the south pole of the magnet, the current will pass around the convolutions of the field winding in a clockwise direction. In order that these con- ditions may obtain, the inside layers of two adjacent coils are connected together, then the outsides of these same coils are connected to the outside layer of the next adjacent coils, and so on around. If the field windings are connected in multiple, qo ELECTRIC RAIL WA V HAND BOOK, the inside end of every alternate magnet coil should he connected to one of the main field terminals, and the intermediate field magnets should have their outside ends connected to the same field terminal. If these connections are made, current from some foreign source can be passed through the magnet coils and their polarity determined with a compass. Care should be taken not to bring the compass too near the the field because the magnetism in the compass needle may be reversed. In a properly connected multipolar machine with a compass taken around the out- side of the frame, the needle should make a complete revolution when passed over the ends of any three consecutive magnets. If there is any question about the inside and outside ends before connecting the field magnet coils together, it can be determined by the same means, i. e., sending a current through each magnet coil independently and using the compass. The in- sulation resistance of the field magnet winding should be determined by disconnecting one end of the exciting circuit and connecting a voltmeter of known resistance between this open end and the frame. With a known initial potential, a known voltmeter resistance and a given deflection use the formula on page 41 to determine the actual insulation resistance. In generators of 500 volts and under, two megohms is a fair value for insulation resistance, but in a good field winding it should attain values as high as five or eight megohms. If it is lower than two megohms, which is possible in newly erected machines, due to the moisture which the wrapping and insulation absorbs; the damp spools can be placed over a boiler, if kept dry, and their temperature not be raised over 120 degs. Fahr. ; or a current can be passed through them for a time sufficiently long to drive off the moisture. After one or two hours of this treatment, their insulation to base should be tested again. If this insulation resistance has fallen it may be due either to inherently poor insulation or to the moisture not being completely driven out. If the insulation resistance has risen, the coils should be allowed to cool, and the resistance again measured ; if it is greater than two meg- ohms, it should be allowed to pass, but if still low, the field coils should be discon- nected and each one tested for a ground, and if the poor insulation is located in any one field coil, it should be baked. If the insulation in all the field coils is low, all should be baked again for another hour, and then another test be made. This bak- ing process will not be of much service unless the generator is in a dry and protected place. Where there is escaping steam or in a new building, it is hardly possible to raise the insulation until the building and dynamo have thoroughly dried out. Field windings can be tested for insulation by a break down test as follows : With the normal current flowing through the field winding, connect one leg of the circuit to the frame, and break the circuit connection with the field magnet. This test should be performed upon each leg of the circuit. A field discharge produced from severing the field circuit will tend to puncture the insulation on the field magnet spools, and it is a condition which may arise at any time when the field circuit is broken on a machine. In a multipolar machine the armature should be symmetrically placed with regard to the bore of the fields, that is, there should be an equal clearance between the field magnets and the armature body. The brushes of the same polarity should not at first be connected together, but all of them should be let down on the com- mutator with a weak field on the machine. Potential measurements should now be made across each pair of adjacent brushes, and these potentials should be equal for each set of brushes. If they are not equal there is an unsymmetrical distribution of magnetism, and if this amounts to more than one per cent, from the normal voltage of the machine, the armature should be moved towards these pole faces be- tween which the low potentials were found, until the difference is equalized, pro- - ELECTRIC RAIL WA Y HAND BOOK. 91 ^ vided that the previous measurement of the drop on all fields showed them alike, and that the construction of the machine allows of this being done. If a current is passed through the windings of a symmetrically connected arma- ture, the difference of potential between an equal number of consecutive commuta- tor bars should be the same. If the brushes are now set on the commutator, and current passed from one terminal of the dynamo to the other, it is evident that if the sections of the armature, which are included between the brushes, are equal in resistance, the distribution of current from one brush to the other, either in a multi- polar or bi-polar machine, will be equal ; and the drop between an equal number of segments, between any brushes, should be equal if the brushes are symmetrically set. A milli-voltmeter will show whether the setting is one bar out, if for genera- tors up to 800 k. w. a testing current of five amperes is used. The insulation test of the armature can be made in the same way as that described for the field test, i. e., by inserting a voltmeter in one lead of the dynamo and con- necting the other lead to the frame of the dynamo and applying and e. m. f. across the terminals. The brushes should be down, and field connections and all leads that connect with circuits external to the dynamo, should be disconnected, so that only the insulation of the armature, brush rigging and connection board are under test. In first starting a new machine, it is advisable to run it for several hours at full speed in order to get the bearings in shape and also to dry out the armature. During this run the bearings should be watched carefully, because poor oil, grit in the pillow block, poor alignment, sprung shaft, all tend to make the bearing heat from undue friction. To charge the field magnet, where there are sources of potential external to the generator, the current can be passed through the field magnet windings which should be disconnected, and independent of the armature circuit. If this dynamo is to be connected in multiple with others, one leg of the dynamo switch should be bridged with a piece of fuse wire, and the other side should be bridged with a lamp or voltmeter. If, when the potential rises on the machine, the lamp fades out or the voltmeter point falls to zero, the potential of the machine has the proper polarity and value for throwing in with other machines. If the lamp increases in brilliancy or the voltmeter reads beyond 'the initial voltage, either the field terminals have to be reversed (providing the machine is separately excited) or the dynamo leads will have to be reversed, in order to obtain the proper polarity. Where there is no current available that can be used to excite or charge the field magnet the following methods can be used, if the machine does not excite when the fields are connected directly across the brushes and the armature is up to full spee I In placing the field windings directly across the brushes the best way is to cut out the rheostat by looping the wires to the rheostat and the lead resistances. In case the machine does not excite recourse may be had to the following: first move the 1 brushes away from the neutral point, then slowly move them back again : strike the field with a hammer ; press the brushes down hard, especially if of carbon; sand paper the commutator; and if none of these methods succeed, flashing will have to be resorted to. In order to perform this operation hold one end of a piece of wire on one brush, and strike the brush of opposite polarity with the other end of the wire. This will in some cases start the machine. Holding the brushes on the loop and suddenly withdrawing the connecting wire will sometimes jump the current through i the field circuit. Sometimes short circuiting one coil of a bipolar field magnet, or several field coils in a multipolar machine, will reduce the resistance of the field Circuit, so that the residual magnetism will be sufficient to start the building up ot the flux. Large machines will very often build up slowly, taking ten or fifteen 9 2 ELECTRIC RAIL WA V HAND BOOK. minutes before a noticable rise in potential has been created. After the machine has come up to voltage, bring the brushes to the point of least sparking, and let it run in this position for several minutes, even if the commutator sparks badly, which it will probably do. The loop between the leads to the rheostat can now be cut and the rheostat resistance inserted and the machine voltage controlled. If none of the above methods are successful the indications are that the e. m. f. generated across the brushes tends to demagnetize the field and that the fields are not properly connected with reference- to the brusnes. Reverse the field connections and repeat the above methods of inducing excitation. If the field test, as described above, has not been previously made, there may be a poor field connection which can be eliminated by looping out the different fields. If the generator has been assembled so that the magnetic circuit contains imperfect magnetic joints difficulty will be added to the starting of the machine. All joints in the frame should be clean and bolted up tightly. Curves obtained by plotting the relation between the current and the voltage generated by the machine are called dynamo characteristics. These characteristics give an indication of the performance of the dynamo, in the same way as steam indicator cards indicate the performance of steam in the cylimder of the engine. An analysis of these characteristics gives some of the electrical constants of the machine. In a shunt machine two characteristics are usually taken, namely: the internal and the external characteristics. The former is a curve plotted between the current which flows through the field magnet coils and the resultant potential gene- rated across the brushes of the machine when running at a constant normal speed with no load. A number of readings of the voltages and the resultant e. m. f. are taken while gradually increasing the current: if the voltages are laid out along the vertical lines on cross section paper, in height proportioned to their values, and the currents laid out along the horizontal lines in the same manner, for any set of simultaneous readings, there will be located on the cross section paper a point at the intersection of these two values from the horizontal and vertical scale. A num- ber of these points plotted out with the field current gradually increasing to the maximum and then gradually reducing to zero, will form a curve which is known as the internal characteristic of the dynamo. The external characteristic is a curve plotted from values obtained when the machine is worked on external variable resistance. These values are the current delivered by the armature and its e. m. f. This characteristic has no great.practical bearing, as a dynamo is usually required to deliver a variable current at a constant potential, or a constant current at a variable potential. There are two characters of tests carried out on dynamos: one to determine whether the dynamo has been properly designed, and the other to determine whether a properly designed dynamo has been built to meet the specifications under which it is sold. A dynamo is generally delivered on the testing floor completely assembled. The voltage, current output, efficiency and heating limits are known. Assume first that the simplest form of test is to be made: a generator driven by a motor, to which it is belted or directly coupled. In a complete test the first in order is the insulation test of field and armature windings, according to the methods given on page 90; next is to determine the field winding resistance, which is accom- plished by passing a known current through the field coils and taking the drop of potential across the terminals. Current in the field coils causes the poles to become magnetized, and should the normal field current be suddenly broken, the high self- induction of the coils would cause a strong current to flow for an instant in the op- posite direction and might result in a complete demagnetization of the poles or ELECTRIC RAILWAY HAND BOOK, 93 perhaps a reversal of the magnetism. In order to prevent this the rheostat or other regulating resistance is gradually cut in and the current reduced to about one-fourth its normal value before opening the field switch. Where it is desired to have any particular brush positive, the terminal from the field winding which will connect to that brush should be conneted to the positive terminal of the test circuit, which can readily be determined by applying the end to a Weston direct-current voltmeter, the right hand terminal always being positive when the needle swings across the scale. The next question is what current will the field windings carry ? The current is turned on, and a wattmeter, where possi- ble is connected in as shown in Fig. 23. The temperature is taken with four ther., mometers: one for the air near the generator; one on the covering of the field coil ; one on the metallic surface of the pole piece; and one on the back of the frame. The thermometers are best secured by binding tape to the different portions of the machine placing a small bit of waste between the thermometer bulb and binding tape. The glass bulb of the thermometer should be in direct contact with the surface of which the temperature is to be taken. The exposed surfaces of the fields should be measured and calculated. This static field test may be tabulated in the following form : -Temperature. - Time. Volts. Amperes. Watts. Air. Field Surface. Pole Face. Yoke. If the temperature is also to be measured by the increase in resistance of the m field windings, then the voltmeter and ammeter can be substituted for the watt- meter. A rheostat should be provided so as to keep the watts lost in the field circuit constant, because this loss will gradually decrease as the temperature rises. It takes about three hours for machines under 10 k. w. capacity to rise to their maxi- mum temperature; four hours for machines up to 250 k. w., and five hours for 1000 k.w. and upwards. In a dynamo with ventilated type of armature, the tempera- ture of the field structure and armature does not rise to its maximum value until ten I to twenty minutes after the machine has been shut down ; accordingly for the purpose J of determining the proper amount of power that can be dissipated in a field circuit and the temperature kept within stated limits, the static test is the most reliable. i The readings should be taken at frequent intervals during the test, and can be : plotted in the form of four curves between watts and temperature. The area in i square inches of the field spools being known, the watts delivered to the field circuit 1 can be divided by this area and the watts per square inch of spool surface obtained, which is the basis upon which field spools are generally designed. It is also advis- j ble to plot the curve while the test is in progress, for as soon as it becomes a .straight line the temperature of the spool has reached the point at which it can 1 dissipate the energy as fast as received, or the maximum temperature has been attained. This condition is also indicated by not requiring any change of the rheo- ! stat to keep the watts constant. To calculate the temperature by the resistance increase of the field circuit due to the temperature co-efficient of copper (see table), and, note between the tempera- ture cold and the maximum temperature what per cent, increase the copper would have; then the resistance cold to the resistance hot. To take temperature alone by Sthe resistance increase for copper, the rise in resistance is 0.22% for each degree rise Fahr. and 0.4% for each degree Cent. It will be found that as a rule resistance I tests will show a slightly higher temperature because the inner convolutions dissi- pate the heat more slowly than the external convolutions. The field spools should always be tested with the wrapping and finish that they receive when sent from the thop, as they materially retard radiation of heat generated in the field winding. 94 ELECTRIC RAILWAY HAND BOOK. If it is required to find the number of turns in a field winding, the wire being drawn to gauge and the resistance determined as above, the number of feet of wire on the coil can be found, and if the inside and outside diameters of the winding on the spool are known, the diameter of the mean turn is the sum of the outside and inside diameter divided by 2, and this diameter multiplied by 3.1416 will give the length of the mean turn, which length divided into the total length of wire will give the number of turns of wire on the coil. For the inspection of field coils for symmetry see page 89. The neit step is to operate the dynamo. Its perfect mechanical operation being assured, the field circuit can then be connected across the brushes or brush busses. If it is a multipolar machine, inserting in series a field switch and a field rheostat of full current carrying capacity, a resistance equal to one-half the resistance of the field magnets is generally ample. When the brushes are adjusted the field switch can be closed, and if the commutator sparks at the brushes the brush can be shifted to the point of least sparking. After the generator has been excited, the next question is the method of obtaining a load, or a method of absorbing the energy developed by the generator. This may be taken up by a wire rheostat, preferably made of galvanized iron wire, (for capacity see p. 25) where the output of the generator is only several kilowatts. This rheostat can be further increased in capacity by mounting on a wooden form and immersing in running water. A lamp bank can be used where a number of lamp sockets are arranged in convenient multiple arc circuits, terminating in plug switches, so that they can be cut in or out to adjust the load. Lamps give the most steady load attainable. When the test is to be carried out where the general methods of taking up a load are not convenient, the water barrel rheostat, when heated by the passage of current, becomes very steady and the current can be varied in several ways. The trouble with the fumes is greatiy reduced by putting in common salt or bicarbonate of soda. If two plates of ^-in. sheet iron about 10-in. x 24-in are used as electrodes they can both be mounted together on a wood frame that will go into an oil barrel, or one sheet can be secured to the inside of the barrel and the other sheet be pro- vided with a handle so it can be moved up and down for regulation. If supplied with enough water to keep it at boiling point one barrel will have a capacity of 120 amperes at 500 volts. A number of these barrels can be used in multiple to take up larger loads. Where a permanent rheostat is required for factory testing the form shown in Fig. 69a is that most largely used. A long length of %-in. iron pipe with water flowing through it makes an excellent method of dissipating energy for large railway machines, where the current goes up into the thousands of amperes. The above are methods where the current output is absorbed. Several methods will be explained later where the generator output is used to help drive the generator under test. Continuing the Test on the Generator. — After it has been excited and the voltmeter shows that it responds to the changes of the field rheostat, a load can be gradually worked up on the machine until full load is reached and the brushes adjusted to the least sparking point; or if the design is poor, to the point where the spark gives the least wear on the commutator. Under potential measurements the method of finding this point with the voltmeter is given, as is also the method of determining the distribution of potential around the commutator. Another test should be made on the armature to find its resistance from commutator bar under one brush to the commutator bar under the next brush of opposite polarity. In a bi-polar machine the current can be passed from one brush ELECTRIC RAILWA Y HAND BOOK, 95 .y^Iron Rollers -s. Movable Iron Plate .X"Thick Plank 23* TUick of II J{ Iron h Plates \ H To Dynamo ■ I Stationary Iron Plate Iron Band2x>6 \ Around Sides And Bottom , „ B — 3 10 Fig. 69- a 96 ELECTRIC RAILWAY HAND BOOK. to the other and drop wires held on the bars through which the current is intro- duced into the armature, and the drop leads on being connected to a low reading voltmeter will give the volts lost in the resistance of the armature, and this, divided by the current flow, which will be shown by an ammeter in the test current circuit, will give the effective resistance of the armature. To ascertain the temperature of a hot armature the rise can be estimated by taking the resistance cold and then hot and computing in the same way as that given for field coil tests, page 93. In taking data for the saturation curve connect in the field circuit in series with a rheostat a field switch and an ammeter. Across the field terminals connect a voltmeter, also connect a voltmeter across the brushes of the machine. With no load on the machine it is brought to full speed with all the resistance of the regu- lator in circuit. This resistance must, as a rule, be higher than the working rheostat supplied with the generator in order to reduce the field and get the low points on the curve. The log for the test is as follows : Speed. Volts (field). Amperes (field). Volts (armature). The speed is kept constant and the first reading taken with the field circuit open. If there is a permanent field the e. m. f . across the brushes is plotted on the volt scale above the origin. The field circuit is now closed with all resistance in series and all instruments read, and then the resistance is reduced slightly by manipulating the rheostat and readings taken again and noted. This process is continued by suitable steps so as to get enough points to form a satisfactory curve up to the full e. m. f . for which the machine is designed ; and sometimes the test is carried on until the field rheostat is completely cut out. The form of the curve gives the magnetic permeability of the field frame and armature and is important in determining the regulation and leakage on the machine. POWER STATION TESTING. It is important to determine the economy under which a station operates Tinder the various loads, management of boilers and engines, and the loading of different units. Such data are essential in order to determine how to best operate the plant for maximum efficiency. Coal.— The weight of coal that is burnt under the boilers can be readily de- termined, and when only a temporary test is to be made, a platform scale large enough to hold a wheelbarrow, can be used, Fig. G9, The scale is generally set so that an even number of pounds is weighed each time — either adding or taking off coal, until the scale balances. The coal handler should make a record each time he weighs, and if the coal is to be used moistened, it should be weighed before wetting. For continuous records of station operation, a number of methods are used. Fig. 70 shows one method of supporting a hopper where the coal is stored above the boilers in bins. The hopper is filled by opening the chute, and when nearly balanced, the coal can be throttled until a perfect balance is obtained. The bottom part of the weighing hopper can then be opened, and the coal delivered on the boiler room floor, convenient to the boilers. It is also suggested to have an electrical contact on the top of the arm of the weighing device, so that each weighing can be recorded on a dial magnetically. ELECTRIC RAIL WA Y HAND BOOK. 97 Firing. — Before starting a boiler test, all coal should be cleaned up from the floor around the boiler, bo that only the weighed coal will be fired. In making comparative boiler tests, the coal for each boiler should be kept separate. The proper method of firing a boiler depends upon the coal, the furnace, the grate and the draught. An expert will change his methods to suit different steam de- mands on the boiler. There are three distinct methods in hand firing: (1) Spreading, which is the common method, where the coal is scattered evenly over the whole surface of the grate, commencing at the bridge and spreading toward the door. (2) Alter- nate firing, in which the charge of coal is laid along one -half of the grate at a time, from the bridge to the door, each side alternately; with a double door furnace, this is usually the method used. (3) Coke firing, which is more specially applicable to bituminous coal, here the charge of coal is first thrown on the dead plate or front part of the grate, where the volatile matter is burned out and the coke coal gradually pushed back to the bridge, where it is completely Fig. 69.— hand coal weighing. burned. The steaming advantages of the different methods of firing can only be determined by their application to individual cases. In regard to the wetting of the coal before firing, this has advantages in iome few cases of slow burning furnaces. The action of the excessive water in the coal is to decompose into hydrogen and oxygen in the intense heat of the combustion of the coal, which gases combine again to form water in the cooler parts of the furnace; in combining they raise the temperature of the gas pass- ing through the furnace. The effect is to transfer the active heating of the gases from the furnace fire to other portions of the furnace whose normal tem- perature is lower. The energy required to raise this additional water to the temperature of the gas leaving the boiler will be lost, and through this range the capacity of water for heat is great. In the case of wet coal, the temperature of the gases issuing from the boiler maybe reduced over the dry coal, but the actual number of thermal units escaping up the chimney may be increased. Water used under the grate to wet the ashes is evaporated by their heat, and the heat radiating downward through the grate bars; this steam passes through the grate up with the draft and re- duces the intensity of the heat of the glow fire, and most of the energy used to raise the water to steam when used this way would otherwise be wasted. Ashes should not be wet if they are to be weighed. 9 8 ELECTRIC RAILWAY HAND BOOK. Combustion.— Coal in burning combines with the oxygen of the draft, giving up its carbon; first, to form carbonic oxide, CO, and then further com- bining with oxygen to form C0 2 , or carbonic acid, the presence of which indi- cates complete combustion. Insufficient air supply or incomplete combustion of the coal will change the ratio of carbonic oxide to carbonic acid in the gas issuing from the boiler. The carbonic ozide in uniting with oxygen will give up one-third more energy than if passed out as carbonic oxide. The condition of combustion is indicated by the percentage of carbonic oxide that exists in the gas leaving the furnace. There is an instrument made called a composimeter, Fig. 70.— automatic coal weighing. which indicates and records continuously the percentage of carbonic oxide, or CO, in the chimney gas, and indicates the condition of combustion. This is connected directly to the uptake of chimney, and the indicator can be located at any convenient place for the firemen's inspection. The above is only strictly true for anthracite coal; bituminous coal increases in smoke as the draught is increased, or the temperature of fire falls Each pound of coal requires 21.3 lbs. of air for complete combustion, or at 60 degs. Fahr., 280 cubic feet of air. In coal the carbon, hydrogen and oxygen are the heating elements, and water, nitrogen and ash the waste. For the an- alysis of some of the American anthracite coals, sec under Fuels. Where only an approximate determination of the heating capacity of coal is required, it can be figured from its analysis where the percentage of free carbon in the coal is knowno ELECTRIC RAILWA Y HAXD BOOK. 99 Exampte : Take Lehigh anthracite coal which contains 3.7 p»-r cent moisture. 6.3 per cent ash, 84.6 per cent carbon, and 5.4 per cent vola f ile matter. re- ducting moisture and ash, which make a total of 10 per cent, from the 100 p-r cent, gives 90 per cent; fixed carbon is 84. G per cent, which, gives the fixed car- 84 6 bon ratio of the coal — — - which equals 94 per cent. 90 ' if The table below gives for this ratio 15,120 B. T. U. Ten per cent of this is ash APPROXIMATE HEATING VALUE OF COAES. d^ » o ~ o d O £> . 5* "£— >>d E§3 « *3 oa £ *i &'Z w Id «s x 7 ^, p" r* * x ^-« ^3 53 +* c *- «- 9 fc£ . 2 c* 3 <- """"a *-■ — _r 1?1 > d 2 t- d C- ® c a) .2^ ^ ^ e o S Pi Kp; u 100 14.500 15.00 68 15,480 16.03 97 14,760 15.28 63 15,120 15.65 94 15,120 15.65 60 14.580 15.09 90 15,480 16.03 57 14,040 14 53 87 15,660 16.21 54 13,320 13.79 80 15.840 16.^0 51 12.600 13.04 72 15,660 16.21 50 12,240 12.67 and moisture, having no heating value, and consequently the coal would only have 90 per cent of this value, which would be 13,C08; as it takes 9C6 B. T. U. to evap- orate one pound of water from and at 212 degs. Fahr., at the pressure of the air, the evaporative efficiency of this coal r if used with perfect combustion and a 13 608 perfect boiler, would be ■ ' , which would be 14.08 pounds of steam at 2112 degs. 966 Fahr. Ashes. — There is always considerable difference between the weighed ashes and the ash found by analysis of coal, caused by unconsumed carbon being car- ried away with the ash and clinker, and the ash will absorb considerable moi c ture on being exposed to the air. In analysis, care is taken to prevent any absorption of the moisture by the ash. In temporary tests ashes can be weighed in the same way as provided for coal. Where continuous records are kept the ashes are gen- erally weighed as they are hauled away for disposal. The weight of coal sup plied to the grate in a given time, divided into the weight of ash taken from under the grate, will be the commercial percentage of ash, which will vary with different coals, and will be affected by the skill in handling the fire. The impor- tance in knowing the percentage of ash in the different coala used has a bearing on its steaming values, as the wasted ash costs as much as the consumed carbon, "it is not always true that the coal that gives the least ash has the highest evap- orative efficiency, as bituminous coals are very low in ash, yet may waste their carbon in smoke. "Water.— The amount of water entering the boiler is a third quantity which has to be known in the boiler room, in order to determine the efficiency of steam 1 LOFC. IOO ELECTRIC RAILWAY HAND BOOK. production. Each boiler should be provided with a water meter attached to the feed pipe near its entrance to the boiler, and it should be so connected to the piping system with flange couplings, valves and a by-pass that it can be readily removed f or recalibration. The temperature of the feed water should be known, and this is readily determined by means of a feed water thermometer, (Fig 71). The form made for this purpose can be screwed to a Y connection in contact with the water in its passage to the boiler. These thermometers are graduated from 60 degs. to 2G0 degs., where feed water heater is used, and from 100 degs. to 400 degs. where an economizer is used. The heat units that are added to the feed water before its introduction to the boiler above the normal temperature of the water should be deducted from the total units required to evaporate the water into steam at the pressure used. For Fig. 71. —feed wateb thermometer. this allowance see Table of Properties of Saturated Steam, giving the thermal units in a pound of water at different temperatures, and the thermal units given up by coal combustion will be the difference between those in the feed water and those of the steam issuing from the boiler. The temperature of the draught of the up-takc in the chimney, which can be measured by a draught thermometer reading up to 700 Fahr., is useful information in order to determine the management of the dampers and drafts, especially where the forced fire is used for any period of the station load, and also indicates which method of firing gives the best results. Losses.— Having arranged the above apparatus, the losses which will occur in this utilization of coal in the form of steam are as follows: First.— Heating draught air to temperature of up-take. As it takes 21.8 pounds dry air at 60 degs. Fahr. to burn one pound of coal, and, assuming the ELECTRIC RAILWAY HAND BOOK. 101 temperature of up-take in the chimney as 560 degs., and each pound of air requires .238 B. T. U. to raise it one degree; then as the air is raised 500 degs. Fahr. the heat units lost per pound of coal are: 21.3 X .238 X 500 = 2,534. Second— If the relative humidity of the air is GO per cent., then there will be .007 lb. of moisture in each lb. of air, which is delivered to the chimney up-take, at an elevation of temperature of 500 degs. Fahr. As it takes .48 B. T. U., per pound of moisture for each degree, then the heat required for the moisture in 21.3 lbs. of dry air will be : 21.3 X .007 X .48 X 500 = 36. Third— The weight of the moisture in the pound of coal is taken at .029 lb., and is first heated from 60 degs. Fahr. to 212 dc^s. Fahr., = 4.4. As it takes 966 B. T. U. per pound of water to change from water at a temperature of 212 degs. Fahr., to steam at the same temperature, the .029 lb. of water will require .029 X 966 = 28. Fourth— .029 lb. of steam heated from 212 degs. Fahr. to 560 degs. Fahi., will be .029 X 348 X .48 = 4.8. For properties of saturated steam see pages 19 and 20. There is in the ash .02 lb. combustible carbon wasted, which has a value of 14,544 B. T. U. per lb., and which will give in wasted energy, 290.9 B. T. U. In the draught .0237 lb. C burned to CO, which by incomplete combustion will be .0237 x (14,544—4,451) = 290 B. T. U. The rest will be radiation and unaccount- able difference, and the total losses assembled in this way, with coal, having 14,- 245 B. T. U., are:— Total per B. T. U. cent of B. T. U. Heating, draught and moisture in draught to temperature of up-take 560 degs 2,570.0 18.04 Heating water in coal 37.2 .26 Heating water formed by combination of hydrogen in coal 97.2 .68 Loss by incomplete combustion 239.2 1.68 Combustible loss in ashes 290.9 2.04 Hadiation and other losses by difference 712.0 5.00 Total 8,946.5 27.70 Beat units utilized in making steam, equivalent evapora- tion 10.66 lbs. from and at 212 degs. F., per lb. coal . ... . 10.298.8 72.30 14,245.3 100.00 In connection with the above equivalent, it is important to determine, first, whether the steam is saturated or contains the quantity of heat due to the pressure; second, whether the quantity of heat is deficient so that the steam is wet; and third, whether the heat is in excess of the pressure or the steam superheated. The quality of steam given off by the boiler bears directly on the work being per- formed by the boiler, and its efficiency, and if it is not taken into consideration, the evaporation performance of 'the boiler can be made to show any efficiency desired. A simple method of testing steam for its condition is to use a barrel calorime- ter which will give fairly accurate results within 2 per cent of the true quantity, when carefully operated. The steam to be tested should be taken from the steam pipe near the boiler by means of a perforated J^-in. pipe inserted into the pipe leading from the boiler, so that no condensed steam can enter the test piptj 102: ELECTRIC RAILWAY HAND BOOK, and provided with, a valve. The steam is carried through a hose which is well wrapped in felt to prevent condensation, and led to a barrel set on platform scales holding about 400 lbs. of water, and provided with stirring vanes, so that the water can be kept in rapid circulation. After carefully weighing the barrel and the water, steam is turned on through the hose and allowed to blow on* until the pipe is thoroughly warmed. The hose is then inserted in the barrel so that all steam is condensed, and the water is kept in rapid circulation. The steam pressure on the boiler tested should be noted and kept uniform. In order to determine the temperatures, a thermometer is inserted in 'the water and watched until the temperature arises to about 110 degs. F. The hose is then quickly withdrawn and exact temperature noted, and the barrel carefully re- weighed. An error of ^ of a pound in weighing the water or % a degree in tem- perature will cause an error of over 1 per cent, in the calculation of moisture in the steam. The original weight of water is, say, 404 lbs c , and its temperature 34 degs. F.; the final weight of water is 435 lbs., and the temperature 106 degs F.; the boiler pressure is 60 lbs. To find the percentage of moisture, proceed as follows:— H = total heat of 1 lb. of steam at 60 lbs. pressure. . „ 1,175.6 T = total heat of 1 lb. of water at temperature of steam at observed pres- sure '. ,.... 307.10 N = temperature of condensing water in barrel, original 34 degs. F. N x • = temperature of condensing water in barrel, final 106 degs. F. W = weight of condensing water, corrected for water equivalent to ap- paratus p 410 lb. IT = weight of steam condensed 435—410 lbs. 25 lb. 1 r-W -l Percentage of moisture— 100= =— , |--(N 1 — N)— (T— N x ) | Substituting values: Percentage of moisture-100= . . * QA ^ . f^(i06— 34)— (307.1-106)1=122.7 1,1/0.0 — Q\)i.l l -/£Q «•* -This shows that in the case assumed above, there was 122.7—100 == 22.7 per cent moisture in the steam, indicating heavy priming in the boiler. The appearance of a steam jet will indicate roughly with a little experience the quality of steam; if a jet flows into the air a change of 1 per cent saturated steam is easily discernable. If the jet i3 transparent, close to the orifice or even a light grayish color, it may be assumed to be nearly dry, and the ordinary methods will not determine the water in the steam, but if the jet be strongly white, with experi- ence the amount of water may be judged up to about 2 per cent ; beyond this, a calorimeter only can determine the exact amount. Ordinarily a boiler should not give more than 2 per cent moisture unless foaming or priming; the water level should not be carried too high, or the boiler forced beyond its capacity, which generally increases the percentage of moisture. If a "boiler givca normally more than 1% Per cent moisture, there is something wrong in its construction or connection, and it is very uneconomical to use wet steam in the engine cylinder, as it increases greatly the losses due to cylinder condensation and the danger of entrained water in the cylinder. These tests have so far referred to continuous boiler tests, but actual condi- tions that arise in station practice change greatly the demands on the boiler dur- ing deferent periods of the dav, and there are hold-over losses on boilers Kot ELECTRIC KAIL WA Y HAND BOOK. 103 delivering any steam whose values, for economical management of the boilers, should be known. Coal burnt to keep a boiler in steaming condition is lost, and it has been proposed to U3e a cheaper grade of coal for the purpose, which will keep a uniform low fire and give to the boiler those heat units which are lost in conduction and radiation. The report of the committee on data of the National Electric Light Associ- ation, gives the following figures for hold-over losses. In this case the boiler is shut oil from the main steam supply. No water is added and *he coal is simply to supply the constant losses. A boiler runs 16 hours a day at an average rate oi 12 lbs. of coal per square foot of grate per hour, and stands over the othei eight hours with a consumption of % lb* of coal per square foot of grate surface; while idle it will consume 2.04 per cent of the whole. If it runs twelve hours and stands twelve hours, the coal cost idle will be 4 per cent of the total expense. The data given for different boilers is as follows;— A Philadelphia station requires 1 200 lbs. of coal to keep up a pressure oi 125 ibs. on two water tube boilers, each having 59 sq. ft. of grate surface; that is, .424 lb. per sq. ft. per hour. A five days' te3t on a horizontal tubular boiler showed a consumption of .35 lb. per sq. ft. of grate; another water tube boilei showed .5 lb. of coal per sq. ft. of grate per hour. Waste in the form of leakage, whether from wet steam or actual escape, has reached in three stations 3.500 lbs., 2,000 lbs., and 500 lbs. Auxiliary uses of steam, such as heating and feeding water to boilers, are drags on the boiler, but have to be considered in the total boiler room efficiency, and bring down the total plant efficiency. Boiler room records are usually kept on printed forms by the foremen of the different shifts of firemen, and the form adopted by the different stations will depend on the data obtained. A form largely used is given below: Bate 9 . Weather Foreman.., Time op coming 0N....0 Boiler Pressure, Maximum, 161: Min., 142. 1. Boiler number. 1. 2. 4. 5. 6. 2. Condition Banked 110,684 1,306 5.10 p.m 11.06 122.384 Fired 117,680 7,141 Fired 111,466 6,400 Sh'td'wn 111,480 Banked 18,705 1,685 6 P.M. 123,240 Sh't d*wi 8. Wa'ter meter, lbs. 4. Coal fired to each boiler 18,971 5. Started up at 6. Shut down at 7. Water meter, end of run 12 M.N. 163,760 178,600 'llY,485* * "is.Vtr 8. Temperat. of feed water, 21s *'., 1U6° a\ 9. Weight of ash Remarks. 10. Natural draught, inches 11. Forced draught, inches 12. When started up 5 13. When shut down 9 W—tyo .45 P.M. .30 P.M ..Pump ....618° .15£— 6* 14. How fed 15. Temperature of uptake 16. Per cent of C0 2 Item 9 is usually taken at stated intervals when ash is hauled away and not for each boiler, but is entered on this form. Items 10, 11, 12 and 13 (where forced draught is used), are important to know when forced draught is started up to <&* 104 ELECTRIC RAILWAY HAND BOOK. termine if it was not put in operation too soon for the power demand on the sta- tion; also to know the natural draught conditions due to the weather. Item 14, the method of feeding water, whether injector or pumps were used, affects the economy of operation of the boiler plant. Item 15, temperature of up-take, gives a value depending upon firing and arrangement of dampers. Item 16 gives from the composimeter the condition of draught for the proper consumption of the coal burned. A recording pressure gage is the only means for showing whether the proper maintenance of boiler pressure has been uniform throughout the run, but from the above figures all the commercial efficiencies and losses in the boiler room can be computed. LUBRICATION. Instead of a separate oil cup for each bearing which requires individual atten- tion, systems of oiling are now generally installed, which consist of a piping system leading from a fountain head or reservoir and through a sight feed with a controlling valve to the bearings. The oil is taken away from the bearings by an oil drainage system, all pipes leading to a sump. The advantages of an oil circulat- ing system are that it allows the bearing to be abundantly supplied with oil, thus reducing the maximum rise in temperature and the co-efficient of friction of the rubbing surfaces and in case a bearing heats up it can be flooded, the oil acting as a cooling medium. There are several systems of circulation in use— gravity is always used. That is, the fountain head is above the level of the highest bearing, but this pressure may be increased by compressed air, either by having the reservoir larger than the maximum volume of oil and allowing the oil pump to pump the oil from sump against this confined air, or, compressed air can be piped directly to the top of the reservoir. In any case the pressure due to gravity should be sufficient to cause the oil to flow even if the compressed air line fails. The. oil after entering the sump is passed through a filter and then to the cooling tank from which the oil pump forces it to the oil reservoir above. If oil is kept cool its wearing qualities are increased. The reserve oil in the reservoir is generally 10 times one hour's feed supply. In large plants, duplicate systems of oil supply are used and in any plant the reserve supply should not be less than one half operating supply. Oil piping requires great care in installation, brass and welded wrought iron are both used but the thread should be cut so as to make up tighter than is the case in ordinary piping. A plumbago paint gives better results than white lead for treating the joints before screwing together. The system is generally laid out in a loop and fed on opposite sides, from the loop the supply pipes are taken to the different bearings throughout the station and valves are so placed that one section can be cut out without interfering with the rest of the system. The oil pipe is often fitted on top of the oil cup, and there is screwed into this cover a plug which can be removed and the bearing oiled by hand when the oil circulating system fails. The size of pipe used on the pressure side of the system ranges from % in. to 1% in. and the area of the pipes on the drainage side of the system should be at least twice that of those on the feeding side. The oiling of crosshead guides both in horizontal and vertical engines causes an oil spray. Some engines produce considerable spraying. The oil mist can be readily detected by hanging a sheet of paper near the engine and out of the way of any direct oil throw. Jn direct connected units which have the ventilated type of armature this spray is sucked into the armature and some of it is deposited on the interior, one effect of this is to form a good binder for the carbon dust from the brushe* and the other is that the oil coat forms an oxydizing agent which reduce* ELECTRIC RAILWAY HAND BOOK. 105 the carbonizing point of the armature insulation and leads to their rapid deteriora- tion and final short circuiting. The practice of covering the cross head guides and crank with riveted galvanized iron housings provided with convenient slide doors so that the moving parts can be inspected, has materially reduced the rate of deterioration of armature insulation. In some plants air is blown through the housing and conveyed out of doors through a pipe. Operating engineers object to the unsightliness of such arrangements with pipes leading to the atmosphere, but nevertheless the apparatus is efficient. If oil has access to rubber covered cables it soon rots the insulation and renders it useless. Circulating oiling systems when properly installed are oil and labor savers. LUBRICATING OIL.S. There is nothing more extravagant in an engine room than poor oil. The use of lubricating oils that are liable to oxidize from heat or use and become gummy due to chemical changes is poor economy. The gummy property of an oil is a relative quantity. It can be tested against two oils. Linseed oil which has great drying and gummy qualities and sperm oil which does not become viscous under normal con- ditions. Cut three V grooves in a plate of cast iron parallel to one side and to each other and about a foot long and incline the plate with a slant of one foot in six, now with a medicine dropper, drop six drops of linseed oil in the first groove, six drops of the oil under test in the second and the same quantity of sperm oil in the last. The oils will gradually flow down the groove and the oil under test will stop flowing somewhere between the linseed and sperm; the nearer it stops to the sperm the less liable it is to gum; oil that does not flow more than half way down between the linseed and sperm should be regarded with suspicion and if it does not flow half way down it should be rejected as its gumminess would seriously interfere with its lubricating qualities. Acidity is another undesirable quality in lubricating oil. This property is tested by pouring a little oil in a test tube and adding a few scales of copper dioxide Cu 2 and shaking; then heating gently, if at the end of an hour there is fatty acid present the oil will burn with a greenish color, if a vegatable acid is present it will burn with a blueish tinge, in either case the oil will react on the bearings and piping and increase the coefficient of friction. Oil should also be tested for the flash point, because a hot bearing with oil having a low flash point will start a fire that may lead to serious trouble. The test for the flash point will reveal the presence of lighter oils used to increase the viscousity or which were not removed in the refining process. In testing for the flash point, heat the oil gradually in an open vessel in which a thermometer reading to 600° Fahr. is immersed, and after the oil has reached 120° apply a flame to the oil at every 10° rise in temperature. The oil should be heated at a rate not greater than 15° per minute and for safe lubricating purposes it should not flash under 300° Fahr. Providing that an oil, in addition to safely undergoing these tests, presents a uniformly colored body containing no suspended matter and does not cloud after standing in the sunlight for several days, it may be accepted as suitable for engine lubrication. SECTION HI —THE TRACK. LOCATION. The location of street railway tracks in streets and highways is largely con trolled by ordinances, which specify the distance from the rail to the curb 01 from the center of the track to the curb, the distances between track centers of a double track road, the type of rail head, its location regarding the street level, etc. In country and interurban roads, however, a greater latitude as regards track construction is usually allowed the engineer. The question of crossings, trees to be removed, roadway obstructions, character of roadbed, and cubic yards of embankment necessary to cut or fill have all to be considered with regard to each section of the road, to ascertain which location had best be adopted; and where the right of way has to be secured through private property its cost is another factor. LEVELS. As the possible speed and energy required to make a trip between two term! nar points are largely determined by the grades over which the road must be built, the profiling and determining of grades and levels is one of the necessary duties of a railway engineer, particularly in interurban railway construction. The details of the methods employed in laying and profiling can be found in any text book on surveying, but the simpler methods are given here to assist the street-railway engineer in plotting grades. The elevation of any part of a road is always given as higher than some Jevel surface of known or assumed elevation, and in order that the elevations may all be positive, this surface should be selected below any elevation to be measured. This surface is called "The Datum." The elevation of the datum is always zero and the elevation of any point is its vertical height above the datum. The point selected for a datum should be permanent and is called a bench mark; in a long route several bench marks are taken for convenience, but all the elevations are figured from the datum, and the elevations of the other bench marks are determined in reference to the one first selected. The instrument gen- erally used for levels is a spirit level, mounted on, and parallel to, a telescope, the field of which is provided with cross hairs, so that the line of sight through the intersection of the cross hairs is horizontal when the bubble stands in the middle of its tube. Then any point in the line of the horizontal cross-hairs through the telescope is on the same level as the cross-hairs= To ascertain levels, the instrument should be set up and levelled at a point higher than and in full view of the first bench mark, on which a surveyor's rod should be held vertically. When the line of sight is turned on the rod the point of the rod covered by the horizontal cross-hair is then on a level with this cross- hair, and the cross-hair is higher than the bench mark by the distance on the rod from the bench mark to the point where the horizontal cross-hair crosses the rod. Add this distance to the height of the bench mark, and we obtain the height of the instrument, technically known as the •' Height of Instrument." find sometimes designated by H. I. ELECTRIC RAILWA Y HAND BOOK. 107 Having obtained the height of the instrument above the datum line, any point lower than the cross-hairs can be easily found by taking the reading of the rod upon it; the reading subtracted from the height of the instrument gives the elevation above the datum of the point on which the rod is set. Is o elevation of a point can be obtained if the rod, when placed on that point, is not in the line of eight of the instrument; in this case the instrument must be moved to a new position, either higher or lower than before as the case may require. Before the instrument is removed a temporary bench, called a *• turning point" and designated by T. P. or Peg, must be established, and its elevation determined; as for any other point the reading of elevation on the rod being taken eti the turning point, the instrument can be removed to another position. After it is properly levelled up, the new height of the instrument is obtained by a new reading on the same turning point; since the cross-hairs are higher than the point, this reading, added to the elevation of the point, gives the height of the instrument. Records are kept of the levels of the different stations as they rise above the datum or zero level selected, and the distances between stations. These stations can be 400 feet from the instrument and accurate work done. These distances and elevations are then laid out on profile paper to any desired scale. For con- venience, in drawing the profile, the entire length is considered as straight. The limiting grade, up which it is safe to carry an electrical equipment, except under special conditions, is about 15 per cent. Where traffic is heavy, it is questionable whether considerable investment could not often be profitably made to obviate a heavy grade on account of the slow time made, together with the heating and consequent depreciation of the equipment. The cost of operating over grades is governed by local conditions so that it is not amenable to any general treatment. The elements that enter in this calcu- lation, however, can be indicated by an assumed case: Suppose that a grade affects such a reduction in the speed that an extra equipment has to be operated all the time to maintain the proper schedule between cars, then the cost per year to operate this equipment would be the interest on the amount of money which would be profitable to expend to avoid or reduce the grade, the other considera- tion being the greater depreciation of all equipments ascending the grade and the additional hazard on descending the grade. Cuts and Fills. — To estimate cubic contents of excavations or fillings as- suming parallel end faces, parallel top and bottom surfaces, and uniform section e / = length. dj> — perpendicular height or depth. ivd = width at bottom of cut or top of fill. Dimensions in linear feet, results in cubic yards. Slope 1 to 1 : Cu. yds. = .03704/ X dJ>(ivd-\- df). Slope 1*4 to 1 : Cu. yds. = .03704/ X dJ>{ivd-\- V&df). Slope 2 to 1 : Cu. yds. = .03704/ X dX^d^-2d/>). Slope 2}4 to 1 : Cu. yds. = .03704/ X dj>(wd+VL%d/>). To estimate cubic contents of wed^e-shaped end of cut, assuming horizontal base and uniform decrease in depth from maximum to zero. io8 ELECTRIC RAILWAY HAND BOOK. Slope 1 to 1: Cu. yds. = .006172/ X dj>{%wd4-2dj>). Slope 1% to 1 : Cu. yds. = .006172/ X dj>(%wd '-f-3#> = .018516/ X dj>(wd -f dp). S.ope 2 to 1 : Cu. yds. = .006172/ X dp{Zwd -f 4dj>). Slope 2]4 to 1 : Cu. yds. = .006172/ X dp(Zwd-\-*>dj>). The above formulas are true for fill having horizontal top surface, and uni form decrease in depth from maximum to zero. To estimate cubic contents of wedge-shaped end of fill, assuming horizontal base and uniform decrease in depth from maximum to zero. Slope 1 to 1 : Cu. yds. = .006172/ X dXSwd-\-4dp). Slope 1\4 to 1 : Cu. yds. = .006172/ X dp$wd+§dj>). = .018516/ X dj>(wd + 2dj>), Slope 2 to 1 : Cu. yds. = .006172/ X dj(Zwd+%djp). Slope 2% to 1 : Cu. yds. = .006172/ X dj>$wd+\0dp). Shrinkage. — In estimating the relative amounts of excavation and em- bankment required, allowance must be made for difference in the spaces occu- pied by the material before excavation and after it is settled in embankment. The various earths will be more compact in embankment, rock less so. The difference in volume is called shrinkage in the one case, and increase in the other. SHRINKAGE IN 1000 CU. YDS. Material. Of Excavation. Of settled embankment. Sand and gravel 80 cu. yds. 87 cu. yds. Clay 100 " 111 Loam 120 " 136 " Weteoil 150 " 200 " INCREASE IN 1000 CU. YDS. Rock, large fragments 600 cu. yds. 375 cu. yds. Rock, medium fragments 700 " 413 " Rock, small fragments 800 " 444 " Thus an excavation of sand and gravel measuring 1000 cu. yds., will form only about 920 cu. yds. of embankment; or an embankment of 1000 cu. yds. wil. 1 require 1087 cu. yds. of sand or gravel measured in excavation to fill it ; but will require only 587 cu. yds. of rock excavation, the rock being broken into medium- sized fragments; while 1000 cu. yds. of the later, measured in excavation, will form 1700 cu. j r ds. of embankment. Tor Weights of Earths and Stones, see pages 11 to 13. TRACK LOCATION. Track is made up of straight and curved track; the straight track is called ;c tangent " and the curved may be a simple curve, that is, a circle struck from the center so as to be tangent to both tracks which it connects. The length of this radius can be found by erecting at the adjacent ends of the tangent track vertical lines, which will intersect at the center of the curve to be struck. To ELECTRIC RAIL WA V HAND BOOK. 109 correctly join these two tangents in street railway work the curved position of the track is defined by the radius of this curvature. Steam roads usually adopt a different nomenclature, which is the number of degrees of curvature, included in an arc 100 ft. long. With the short curves used in street railways it is readily seen that this designation would not be suitable The survey of a line of track is always made from the middle of the track by setting a row of stakes where the center line of the track passes, and the rail is gaged both ways from this center. Before these center points are located by the engineer in city streets, all knowledge possible should be obtained regarding existing subterranean struct- ures, which may be beneath the surface or on the street surface, in order that they may not be disturbed nor their usefulness interfered with. All city maps of the city, gas, water, subway systems, sewage, etc., become useful if reliable. Usually all obstructions that are in direct line of the rail can be relocated, but in the case of water and gas mains, any street surface structures leading to, or mctallicly connected with, this piping system should be removed at least two feet from the rails, and if possible cement should be interposed between these two structures. Otherwise current may be deflected into these piping systems from the metallic connections made in this way and damage from electrolysis may result. In city work where curves occur, it is desirable, before commencing work, first to lay out each curve on a fairly large scale, say 1 in. = 5 ft., drawing in the tangent tracks which meet at this corner, also any obstruction in the street such as sewer covers, man holes, gas or water gate boxes, also the curb line and any obstruction on the corner if it should offer any possibility of hazard to a passenger standing on the running board of an open car in passing around the curve. After all these details have been drawn in the map the curves can be thrown in, as ex- plained later. In laying road on paved streets with traffic on them, it is best to set stakes at an offset from the line of track out of line of the traffic, about 50 ft. apart on straight track, and close enough together on curves to have at least two points opposite each rail. In special work a point should be set opposite the heel and toe of each switch. TYPES OF ROADBED. Specifications for Street Railway Track in Cities.— Pratt & Alden give the following specifications for railway track, located in city streets. Construction.— Nine inch girder rail on wooden ties, broken stone ballast and granite block pavement. 1. Work to he Done. — The work to be done consists of the construction of a simrle track on , , and Streets, between Street and Street, in the City of , State of 2. Tools and Labor. — The contractor is to furnish all necessary tools appara- tus, and other means of construction, and do all the work required for the above construction. 3. Material. — The company will furnish and deliver to the contractor, at its yard located on Street, all material required for the above construc- tion except such as are not to be part of the finished construction, which will he furnished by the contractor. HO ELECTRIC RAILWAY HAND BOOK. 4. Interference with Traffic— -The street must not be torn up for a greater distance than 500 ft. in advance of the finished paving. The contractor must so arrange his work and deliver the material upon the street as to obstruct public travel as little as possible, and a roadway must be kept free on at least one side of the street for public travel. The contractor shall use all necessary precautions to prevent accidents by maintaining suitable barriers and by keeping lights burning at night. 5. Grading and Excavation. — The roadbed is to be excavated to sub-grade, which will be 24 ins. below the finished grade of the street. This excavation is to extend.... ft. each side of the center line of track. If any further width of excavation is required, it will be directed by the engineer in writing, and paid for under clause . (e) Price per running foot for laying tile-drains and connecting to sewers or drains (/) Price per ton per 1000 ft. for hauling material from the excavation a greater distance than ft. from the excavation 18. Estimates.— It shall be understood and agreed by the parties hereto that due measurements shall be taken during the progress of the work, and the estimate of the engineer shall be final and conclusive evidence of the amount of work performed by the contractor under and by virtue of this agreement, and shall be taken as the full measure of compensation to be received by the con- tractor. The aforesaid estimates shall be based upon the contract prices for the performance of all the work mentioned in these specifications and agreement, and when there may be any ambiguity therein, the engineer's instructions shall be considered explanatory, and 6hall be of binding force. SPECIFICATIONS FOR EXPOSED TRACK. On interurban lines, when the track is exposed, the steam railroad practice can be followed very closely. The Pennsylvania Railroad Company has de- veloped the most complete set of specifications, which indicate that company's method in grading, ballasting and draining. Below is given an abstract from the Pennsylvania Railroad's general specifications, covering such structural features as would be well to follow in cross-country electric track construction. Figs. 72 and 73 show the cross section of single and double track as designated by the specifications. P. R. R. Specifications for Laying Roadbed.— Roadbed.— The surface of the roadbed should be graded to a regular and uniform sub-grade, sloping gradually from the center towards the ditches. Ballast.— There shall be a uniform depth of 6 ins. to 12 ins. of well broken stone or gravel, cleaned from dust by passing over a screen of J4-in. mesh, spread over the roadbed and surfaced to a true grade, upon which the tics are to be laid. After the tics and rails have been properly laid and surfaced, the ballast must be filled up as shown on standard plan ; and also between the main tracks and sid- ings where stone ballast is used. All stone ballast is to be of uniform size and the stone used must be of an approved quality, broken uniformly, not larger than a cube that will pass through a 2>£ in. ring. On embankments that are not well settled the surface of the roadbed shall be brought up with cinder, gravel or some other suitable material ELECTRIC RAILWAY HAND BOOK ii3 H4 ELECTRIC RAILWAY HAND BOOK. Cross-Ties. — The ties are to be regularly placed upon the ballast. They must be properly and evenly placed, with 10 ins. between the edges of bearing surface at joints, with intermediate ties evenly spaced: and the ends on the outside on double track, and on the right-hand side going north or west on single track, lined up parallel with the rails. The ties must not be notched under any circum- stances; but, should they be twisted, they must be made true with the adze, that the rails may have an even bearing over the whole breadth of the tie. Line and Surface.— The track shall be laid in true line and surface; the rails are to be laid and spiked after the ties have been bedded in the ballast; and on curves, the proper elevation must be given to the outer rail and carried uni- formly around the curve. This elevation should be commenced from 50 ft. to 300 ft. back of the point of curvature, depending on the degree of the curve and speed of trains, and increased uniformly to the latter point, where the full eleva- tion is attained. The same method should be adopted in leaving the curve. Joints. — The joints of the rails shall be exactly midway between the joint ties, and the joint on one line of rail must be opposite the center of the rail on the other line of the same track. A Fahrenheit thermometer should be used when laying rails, and care taken to arrange the openings between rails in direct proportion to the following temperatures and distances; at a temperature of Odcg., a distance of T 6 5 in.; at 50 degs., 5 6 2 in.; and in extreme summer heat, of, say 100 degs. and over, T * s in. must be left between the ends of the rails to allow for expansion, The splices must be properly put on with the full number of bolts, nuts and nut-locks, and the nuts placed on inside of rails, except on rails of 60 lbs. per yard and under, where they shall be placed on the outside and screwed up tight. The rails must be spiked both on the inside and outside at each tie, on straight lines as well as cm curves, and the spikes driven in such a position as to keep the ties at right angles to the rails. Switches. — The switches and frogs should be kept well lined up and in good surface. Switch signals must be kept bright and in good order, and the distant signal and facing-point lock used for all switches where trains run against the points, except on single track branch roads. Ditches. — The cross-section of ditches at the highest point must be the width and depth as shown on the standard drawing, and graded parallel with the track, so as to pass water freely during heavy rains and thoroughly drain the ballast and roadbed. The line of the bottom of the ditch must be made parallel with the rails, and well and neatly defined, at the standard distance from the outside rail. All necessary cross-drains must be put in at proper intervals. Earth taken from ditches or elsewhere must not be left at or near the ends of the tics, thrown up on the slopes of cuts, nor on the ballast, but must be deposited over the sides of embankments. Berm ditches shall be provided to protect the slopes of cuts, where necessary. The channels of streams for a considerable distance above the road should be examined, and brush drift, and other obstructions removed. Ditches, culverts, and box drains should be cleared of all obstructions and the outlets and inlets of the same kept open to allow a free flow of water at all times. Road Crossings. — The road-crossing planks shall be securely spiked; the planking on inside of rails should be % in., and on outside of rails it should be % in., below the top of rail, and 2% ins. from the gage line. The ends and inside edges of planks should be beveled oil as shown on standard plan. ELECTRIC RAILWAY HAND BOOI? 115 COST OF IN1ERURBAN ROAD CONSTRUCTION. The cost of an interurban railroad varies within wide limits, depending upon the part of the country it is located in and also upon the type of road required. The best type of road, which is built to steam standards, on private right of way, with rock ballast, 70 to 80 lb. steel rails, costs from $12,000 to $18,000 per mile for single track, not including any electrical equipment except rail bonds. The former figure applies to a section built on comparatively level ground, the latter where there are numerous cuts and fills. In detail, this cost is as follows: Per Mile. Grading. $2,700.00 Ballast • 2,000.00 Ties 1,300.00 Rails 3,500.00 Fencing 500.00 Joints and bonds 700.00 Spikes 300.00 Labor 1,000.00 Total $12,000.00 The cost of overhead construction also varies widely depending upon the num- ber of motor cars, and their weight and speed. The cost of heavy construction such as would be appropriate for the track outlined above would be about $4,354.00 per mile. In detail this costs as follows : Per Mile. Poles $150.00 Cross arms, pine braces, etc 25.00 Brackets 208.00 Strain wire 10.00 Hangers and ears . .- , 28.00 Insulators, etc ,i 15.00 Trolley wire No. 0000 » ..... 643.00 Feeder wire 2-500,000 CM ,. 2,800.00 Labor 475.00 Total $4,354.00 Copper wire at 18 cents per pound. A three-phase high tension feeder line, consisting of 3 No. 5 B. & S. G. copper wires including poles, cross arms, insulators, labor, etc., costs about $600.00 per mile. The cost of light interurban railways such as are laid on the highways will aver- age $8,500.00 per mile, exclusive of overhead work. The cost of overhead construc- tion for above using No. 00 trolley and No. 0000 feeder is a bout $2000.00 per mile. 1 he cost of power stations of course depends upon the rated output. In general they can be built for $90.00 to $100.00 pei kilowatt exclusi/e of the cost of the land. Rotary transformer sub-stations cost aboi t $40.00 per kilowatt, depending on the cost of the building. w 116 ELECTRIC RAIL WA Y HAND BOOK. EXAMPLES OF TRACK CONSTRUCTION. Track Construction on Concrete Girders This form of track con- struction does away with the frequently spaced tie and substitutes a lateral bearing for the rail. The rails are tied together at intervals, but depend upon the concrete foundation for their support. This construction is especially useful BASE OF CONCRE tKCL TIE 63 GIRDER RAIL VCACEO.IOVCENTERS ■~i T- r I | BASE OF CONCRETE Fig. 74.— scranton, pa., track construction. where paving such as concrete, brick or asphalt already exists on the street to be tracked, as trenches only have to be cut in the paving for the rails and their foundations with occasional cross cuts for the tie rods, thus reducing the cost for renaving. ^ ELECTRIC RAILWA Y HAND BOOK. 117 Scranton, Pa., Construction.— The rail is a 5-in. T, weight 57 lbs., in 60-Jt. lengths and with a 6-bolt joint. Underneath each joint is an inverted section of same rail, 4 ft. long, extending 2 ft. each side of the joint and riveted to tho \z\\ by eighteen %-in. rivets ; four of these rivets are copper for bonding. The concrete VOIDS UNCER • RAIL HEADS J FILLED WITH CEMENT MORTAR » j/r» K >» &•;- ••.•,•■•£ ■ .: CONCRETE/,. Fig. 75. — Detroit track construction. Is laid 6 ins. below rail except for 2 ft. 6 ins. each side of the joint where it is 12 ins. below the rail. The ties which are old 52-lb. girder rails, are spaced 10 ft. apart and the rails are bolted to them. There is also a tie rod % ins. x \% ins, between each tie. For details of construction see Fig. 74. .- A n8 ELECTRIC RAIL WA Y HAND BOOK. ELECTRIC RAIL WA Y HAND BOOK. 119 Detroit Track Construction.— The rail is a9-in., 100-lb. girder with steel ties 5-ft. centers. The ties are channels 7 ins. wide, 7 ft. long and % in„ thick, with flange 1% in, deep. Concrete was laid 6 ins. below base of rail and tumped nnder tie, also laid on top of ties and carried up to within \y z in. of the base of the paving. Where concrete stringers were used, the trench was 15 ins. deep and 1 ft. wide. A layer of concrete composed of 1 part Portland cement, 4 parts Louisville cement, 8 parts sand and 16 parts broken stone was then laid in bottom of the trench to depth of 6 ins. The rail used in this construction was 7 ins. high; the space between the base of rail and base of concrete was grouted with 1 part Portland cement, 1 part sand and 3 parts clean fine gravel. Tie rods were used 10 ft. apart. For details of construction see Fig. 75. Kansas City Construction.— The foundation trench is 20 ins. wide on top, 16 ins. on botton and 15 ins. in depth, so there will be 6 ins. below the rail when it is on grade. At 10-f t. intervals are placed wooden blocks 8 ins. x 10 ins. x 16 ins. to which the rails are spiked. After gaging and aligning the track, the trenches are filled with concrete made of 2 parts sand, % P art Portland cement, y 2 part domestic cement and 5 parts crushed stone small enough to pass through % in. ring, ail by measure. Temporary splice bars are bolted on at rail joints which eventually are cast welded; the metal for the cast weld being composed of % pig iron and % scrap iron. See Fig. 76 for this construction. L,os Angeles Track Construction.— The rail is a 6-in., 60-lb., 60-ft. T. The managers of the Los Angeles Railway Company were the pioneers in the Fig. 77.— los angeles track construction. bringing of theee long rails across the continent, and report finding no difficulty or additional expense in transportation. In a recent shipment of 500 tons only three rails had to be straightened. For details of construction see Fig. 77. Milwaukee Track Construction.— The rail is a 6-in., 72-lb. Shanghai section in 60-ft. lengths and is laid on cedar ties 6 ins. x 8 ins. «: 7 ft., 2-ft. centers. Under pavements which have concrete foundations the ties are laid on a 6-in. bed of cement, and in other streets broken stone ballast is used. The Falk cast-weld joint is used. In exposed track which is cast welded slip joints are provided every 500 ft. The contraction and expansion in exposed track has been found to amount to about 1J4 in. per 100 ft. of track, so at the slip joints the rails are sometimes 6 ins. apart. The standard suburban roadbed of this company con- sists of a 56-lb. T, 434 in « 60-ft. rail laid on broken stone or gravel ballast, with Weber joints. See Fig. 78 for these constructions. J I 1 20 ELECTRIC KAIL WA Y HAND BOOK. U U LHJ u HI L *•«*- CtHSttr /T£»r*W~ PlO. 78.— MILWAUKEE TUACK construction. ELECTRIC RAILWAY HAND BOOK. 121 RAILS. Sections.— Bails have been rolled in nearly every conceivable shape, which would serve the purpose of a rail. The forms illustrated are those most used in modern street railway practice. Fig. 79 gives some general sections, the box girder being" nearly obsolete; the T and the girder are the prevailing rail- 7. FAIL CENTER BEARING ML Fig. 79.— general rail forms. forms now Jn use. Fig. 80 gives the nomenclature for the different parts designated. Taking up the grooved type of girder rail, Fig. 81 shows the Crimmins or original rail adopted by the Metropolitan Street Railway Company of New York City. The peculiarity of this rail consists in its long lip, extending beyond the guard. This allows of the pavement being laid adjacent to the rail, and carries the street traffic which tracks on the rail, thus preventing to some extent the wearing of grooves along the pavement adjacent to the rail. Fig. 82 shows the later section of the girder rail adopted by the Metropolitan Street Railway. The head tVAPO GAUGE LW£ HEAD tram or tread 'track bolt spucesarbolt 'fillet "jO/NTPLATE SPLICE BAR CHANNEL PLATE PISH PLATS 1H(rE r AH jin. to 1 in. To allow play to the wheels and for the difference of J4 in., which is the undergage usually allowed for setting wheels, the groove should be at least 134 ins - wide at top. It will be noticed that the groove is usu- ally sloping with a large guaid angle so that the wheel flange will throw out dirt that accumulates in the groove. In a straight sided groove the dirt is packed by the rolling flange, and the power required is increased. A grooved rail which has a groove not wider than 1J4 ins. will not form a tramway for vehicles, and they will follow more the street to the side of the track and reduce the wear on the pavement adjacent to and between the rails. Pavement. — In pavement with asphalt laid against the rail, the full-grooved rail is largely used, as any other attracts traffic and concentrates vehicular traffic on or parallel to the rails, which soon wears the asphalt in deep grooves and breaks up the surface of the pavement. Also in cold climates the iron rail against the asphalt makes it brittle and depreciate rapidly. In order to reduce this wear at these points, a granite toothing stone is first laid against the rail, consisting of alternato headers and stretchers, and the asphalt pavement is brought to this surface. There have been made bricks which overlap the tram of the rail, forming a grooved rail and reducing the breadth of the tram. "Where a T-ra:l is laid in paving, the paving is laid directly against the rail and a car truck with larger flanged wheels than ordinarily used and heavily weighted is drawn over the road to form the groove in the paving stone. The depth of the rail is regulated somewhat by the character of pavement used. Brick or asphalt require a sub-base and should not rest on the ties, as this will soon give the pavement an unequal setting. Where this pavement is used, therefore, a 7-in. girder, at least, is required. With a 6-in. Belgian block there is required 1 in. cf sand for a bed and a deep rail must also be used. In macadam pavement the T section forms the best rail, as the grooved rail would be continually filed with dirt, and the flat tram would attract traffic which would put excessive wear on the pavement adjacent to the track. KAILS: SPECIFICATIONS, COMPOSITION AND TESTS. Illinois Steel Company Standard Specifications for Steel T-Rails.— 1. The section of the rail throughout its entire length shall conform to the American Society of Civil Engineer's Standard lbs. per yard. The fit of the fishing or male templet shall be perfectly maintained. When the rolls are new, the section of the rail may bo 6 T 4 in. low. As the rolling pro- ELECTRIC RAILWAY HAND BOOK. 125 ceeds, a variation not exceeding 5 x a in. in excess of height over templet may be permitted in a delivery of 10,000 tons of rails, after which the rolls must be reduced to standard height of such sections. The standard of measure to be Brown and Sharpens United States standard steel vernier caliper rule. 2. The weight of the rail shall be kept as near to lbs. per yard as is practical after complying with Section 1. The rails shall be accepted and settled for according to actual weights. 3. The standard length of rail shall be 30 ft., at a temperature of 70 deg., Fahr. Shorter rails, having length of 22 ft. to 29 ft. inclusive, shall be accepted to the extent of 10 per cent of the entire order. A variation in length of % in. over or under the specified length will be allowed. 4. Care to be taken in cambering the rails so as to reduce the amount of work in the straightening press to a minimum. The rails must be kept straight in all directions as to both surface and line, without twists or kinks. 5. The rails must be smooth on the head and base, and free from all mechani- cal defects and Haws, and mr.st be sawn square at the ends; the burrs made by the saws must be carefully chipped and filed off, particularly under the head and on the top of the flange, to ensure proper fit of the angle-bars. 6. The drilling for the bolts to be in strict conformity with the blue print at- tached, or the dimensions given. Holes imperfectly drilled to be filed to proper dimensions. All holes must be accurate in every respect. 7. The section number, name of maker, year and month, to be rolled on the side of the web. The number of the heat to be stamped in the side of the web. 8. The chemical composition of standard rails under 70 lbs. per yard to be as follows : Carbon 37 to .45 of 1 per cent Sulphur not to exceed 05 " " Phosphorus not to exceed 10 " '* Silicon 07to.l5 " " Manganese 70 to 1.10 '■ •' The chemical composition of standaid rails 70 lbs. and over per yard to be as follows: Carbon 45 to .55 of 1 per cent Sulphur not to exceed 05 " " Phosphorus not to exceed 10 " ** Silicon 10to.20 " " Manganese .80 to 1.00 • " " 9. From each heat one test ingot shall be cast 2J4 in. x 234 in « s: 6 in. long. This to be drawn down at one heat by hammering to a test piece % ins. square by 18 ins. to 20 ins. long. The same when cold to be required to bend to aright angle without breaking. This bar must be bent by blows from a hammer. 10. After cutting off or allowing for the sand on the top end of the ingot, at least 12 ins. more of seemingly solid steel shall be cut off that end of the bloom. If after cutting such length the steel does not look solid, the cutting shall be contin- ued until it does. 11. The inspector representing the purchaser shall have free entry to the works of the manufacturer at all times while his contract is being filled, and shall have all reasonable facilities afforded to satisfy him that the rails are being made in accordance with these specifications. The manufacturer shall furnish daily the carbon determinations of each heat, and a complete chemical analysis of at least one heat of each day and night turn in which each element is to be de- termined. 126 ELECTRIC RAILWAY HAND BOOK. 12. The requirements for No. 2 rails shall be the same as for No. 1 except that they will be accepted with a flaw in the head not exceeding % in.» and a flaw In the flange not exceeding y% in. in depth. No. z rails to the extent of 5 per cent of the entire order will be received. Composition. — The composition of steel rails is a much mooted question for the reason that the different processes used in the manufacture of the ore cause the rail to vary in its chemical composition, but do not necessarily change markedly the physical properties of the metal. Mr. E. W. Richards suggests the following composition after considering the matter from both the point of view of the manufacturer and the user: Minimum. Maximum. Carbon 35 .5 of 1 per cent Silicon 05 .1 Sulphur 04 .08 " " Phosphorus .08 " " Manganese 75 1.00 •■ " The American Committee of the International Association for testing materials for rail from 50 lbs. to 75 lbs. per yard gives the following composition : Carbon 35 to .50 of 1 per cent Silicon Notover.20 " " Sulphur Phosphorus Not over .10 ** " Manganese 70 to 1.05 «« »• The strength of the rail is affected by the temperature at which it passes through the rolls. It is proposed to introduce into rail specifications that the shrinkage after they leave the finishing roll till they attain normal temperature shall be expressly stated in per cent. Regarding the wearing qualities of a rail A. J". Moxham states that the life of a rail is now determined by the life of the joint. The question is not how much the rail has worn, but how much has the hammering at the joint destroyed its usefulness. For heavy traffic he states that what is wanted is a hard ductile rail which can only be produced by low phosphorus and high manganese. Anything RESULTS OF THREE STEAKS' TESTS ON RAILS. (a. j. MOXHAM.) a d o d O ® OQ £ o >-> e3 "-• a .2 to o . fe d 1 1*'" < ar in ins. per Millio zsliq passing oVer rails. run d ars pa d o .a t-> a O d o o 02 U O P< o d .d a a 03 d' o M > c3 M o o jj ■H o CD P, C/2 a d &o 3 d o o «d m o~ LENGTHS OF RAILS. Gross Tons per Mile. Feet of Track per Ton of Kails. ted c ^ PL, g Gross Tons per Mile. Feet of Track per Ton of Kails. O U ^ ft Gross Tons per Mile. Feet of Track per Ton of Kails. 12 13 14 18.86 20.43 22.00 280.0 258.46 240.00 48 49 50 75.43 77.00 78.57 70.00 68.57 67.20 84 85 86 132.00 133.57 135.14 40.00 39.53 39.07 15 16 17 23.57 25.14 26.71 224.00 210.0 197.65 51 52 53 80.14 81.71 83.29 65.88 64.62 63.40 87 88 89 136.71 138.29 139.86 38.62 38.18 37.75 18 19 20 28.29 29.86 31.43 186.67 176.84 168.0 54 55 56 84.86 86.43 88.00 62.22 6109 60.00 90 91 92 141.43 143.00 144.57 87.33 36.92 36.52 21 22 23 83.00 34.57 36.14 160.00 152.72 146.09 57 58 59 89.57 91.14 92.71 58.95 57.93 56.95 93 94 95 146.14 147.71 149.29 36.13 35.75 85.37 24 25 26 37.71 39.29 40.86 140.00 134.4 129.23 60 61 62 94.29 95.86 97.43 56.00 55.08 54.19 96 97 98 150.86 152.43 154.00 35.00 34.64 34.29 27 28 29 42.43 41.00 45.57 124.44 120.00 115.86 63 64 65 99.00 100.57 102.14 53.33 52.50 51.69 99 100 101 155.57 157.14 158.71 83.94 33.60 33.27 30 31 32 47.14 48.71 50.29 112.0 108.39 105.00 66 67 68 103.71 105.29 106.86 50.91 50.15 49.41 1*2 103 104 160.29 161.86 163.43 32.94 32.62 32.31 33 34 35 51.86 53.43 55.00 101.82 98.82 96.0 69 70 71 108.43 110.00 111.57 48.70 48.00 47.32 105 -106 107 165.00 166.57 168.14 32.00 31.70 31.40 86 37 88 56.57 58.lt 59.71 93.33 90.81 88.42 72 73 74 113.14 114.71 116.29 46.67 46.03 45.41 108 109 110 169.71 171.29 172.86 31.11 30.83 30.54 89 40 41 61.29 63.86 64.43 86.15 84.0 81.95 75 76 77 117.86 119.43 121.00 44.80 44.21 43.64 111 112 113 174.43 176.00 177.57 30.27 80.00 29.73 42 43 44 66.00 67.57 69.14 80.00 78.14 76.36 78 79 80 122.57 124.14 125.71 43.08 42.53 42.00 lit 115 116 179.14 180.71 182.29 29.47 29.22 28.97 45 46 47 70.71 72.28 73.86 74.67 73.04 •51.49 81 8-2 83 127.29 128.86 130.43 41.48 40.98 40.48 117 118 119 120 183.86 185.43 187.00 188.57 28.72 28.47 28.24 28.00 ELECTRIC RAILWAY HAND BOOK. 129 in. Taking a track laid at a low temperature of 40 degs., and subject to a maxi- mum of 120 degs. or a variation of 80 degs. the stress is equal to less than 12, COO lbs. per sq. in., much less than the elastic limit. It would therefore appear th-t the effect on the steel would be harmless. It is well known to track men that the heavy rails do not show as much e:: pansion and contraction by heat as the lighter sections do. A report was razZj to the Road Master's Association in 1899 by Mr. V. T. Douglass of the Chora- peake & Ohio R. R., on exposed track construction. He gives the following coefficients for different weights of rails. Contraction from -f 5 degrees to —20 degrees F. Rail. Coefficient. 56-lb 00208 75-lb 00139 85-lb... 00101 Expansion from -f 5 degrees to 4-70 degrees F. Rail. Coefficient. 56-lb 00107 75-lb 000C9 85-lb 00065 A66-lb. rail, if supported by the proper number of cross ties, will answer every engineering demand, even of high speed electric cars. Anything over this goes to the debt of bad joints. Rail in use by different companies varied between 2 ins. and 2]4 ins. head A large number of roads used % in. to 1 in. wheel flange in width at tread of wheel and % to % ins. deep. Useful Formulae For Rails.— The number of tons of rail for one mile single track equals approximately 1} of the weight of rail in pounds per yard; the sectional area of rail in square inches equals approximately X V of the weight of rail in pounds per yard ; the maximum safe weight for rails properly supported on ties is one ton for each 10 lbs. weight of rail per yard. TIES. The life of ties is largely affected by the earth in which they are buried, and raising on ballast and drainage increases their life. The life of ties as given by Prof. Roth is as follows: Black locust, cypress, red cedar 10 years White oak, chestnut oak, chestnut 8 ■• Tamarack 7 to 8 " Cherry, black walnut, locust , 7 ■■ Elm 6 to 7 " Long leaf pine f5 " Red and black oaks 4 to 5 4 4 Hemlock 4 1 ) 6 " Spruce 5 " Ash, beach, maple 4 *• Mr. Hough gives the following table : Oak 7.4 years White oak 7.3 " Post '* 7.0 " Burr " 7.4 " Rock " 7.0 " Red * l 5.0 " Chestnut oak 7.1 " Black oak 4 5 '* Southern pine 6.5 " White «* 6.5 * 4 Cedar, red 11 .8 " Cedar, white 7.5 ** Cypress f , 8.7 M - ~o ELECTRIC RAIL WA Y HAND BOOK. Ash, black 3.8 years. Ash, white 4.3 " Cherry 6 to 10 •• The Railroad Gazette^ Dec. 26, 1884, gives the following percentage of the various woods used upon 90,900 miles out of 121,592 miles in operation of steam track: White oak 58.2 per cent. Cedar 10.4 " " Yellow pine 8.7 " " Northern pine 6.9 " " Hemlock 5.9 " " Chestnut 4 4 " " Fir 1.7 " " Spruce 1.6 " " Cypress 1.0 " " Miscellaneous soft woods 0.6 " " Miscellaneous hard woods 0.6 '.' " Total 100.0 (< (« Climatic conditions play a large part in the depreciation of ties. In low moist country cypress ties last fully as well as cedar; in a dry climate their life is reduced to seven years. Where a tie is covered with earth in its entirety it will decay much more rapidly than where it is only partially covered. It has been noticed that a large percentage of lime in soils will produce premature decay. Yellow pine tics have been found to be preserved by salt used in thawing the snow at guard rails and frogs, while they were badly rotted on adjacent portions of track not salted. When ties have to be laid on ground, the action of which on the different woods is not known, an examination of fence posts along the route, noting the kind of wood, and obtaining the length of time planted, will suggest the best kinds of woods to use for ties, There is no economy in putting down cheap ties. The cost of labor alone in ten years will be more than double that of the most durable tie that can be se- cured. The essential feature of any railway is the permanency of its rail sub- structure and without good ties this cannot be obtained. The treatment of tics primarily consists of heating the tie to evaporate the sap out of the cells, and afterwards filling or lining these cells with some com- pound or chemical which will preserve the mechanical characteristics of the tie, and hermetically seal the cellulose of the tie to protect it from attacks of fungi, or dry or wet rot. The three principal methods used are Burnettizing, Creosoting and Kyanizing. Burncttizing consists of partially impregnating the wood with zinc chloride. The preparation being soluble loses its value when exposed to rain or water. Oak, pine and fir cannot be thoroughly treated as the preparation only reaches Y/± in. in hard woods, and in soft wood penetrates the sap wood but not the heart of the wood at all. The Barschall treatment is the Hasselmann method largely used in Germany. Here the cellulose of the wood is chemically acted on during the treating process and forms a direct chemical combination with the woody fiber cellular tissue and cell contents. The treating liquid consists of a combination of sulphates of iron, copper and alumina and kainit (which is a natural salt of sulphate of potash and magnesia and the chloride of magnesia.) This treat- ment chemically impregnates the whole macs when the timber is boiled in it at a temperature of from 100° cent, to 1 40° tent., and under a pressure of 15 to 45 lbs. ELECTRIC RAIL WA Y HAND BOOK. 131 per square inch and this treatment is said to prevent decay and rot effected on exposure and does not change the physical characteristics of the w.ood except to reduce its inflammability. The treatment by creosote or dead oil of tar is largely used, and reports of tests show that the life of ties can be greatly prolonged by such treatment. The cost of treating ties should show a profit over labor of renewals and cost of new ties during the life of the treated tie. Ties per 1000 ft. and per Mile, SPACING 10 ties to 30 ft. rail 11 " it it t> 12 " II 44 u 13 " *t 44 it 14 " 44 (I it 15 " 44 V. 44 16 " 44 44 44 CENTER TO CENTER 18 ins. 21 " 24 " 27 " 30 " PER 1000 PT, 3331^ 400 433^ 466% 600 533^ PER MILE 1,760 1,936 2,112 2,288 2,464 2,640 2,816 TIES PER MILE 8.520 8,017 2,640 2,348 2,113 1,760 Board Feet, Cubic Feet, and Square Feet of Bearing Surface per Tie. SIZE 5 ins. x 5 ins. x 7 ft. 6 7 8 6 pv t 8 9 10 8 9 xlO x7 x7 x7 x7 x7 x7 x7 x7 x8 x8 x8 BOARD FEET 14 56 17.5 20.41 23.33 21. 24.5 28. 31.5 35. 32. 36. 40. CUBIC FEET 1.213 1.458 1.7 1.P44 1.75 2.041 2.333 2.625 2.916 2.666 3.00 3.333 BEARING SURFACE 2.91 3.5 4.08 4.66 3.5 4.08 4.66 5.25 5.83 5.33 6.00 6.66 The inspection of ties is largely a matter of experience and judgment. The hewn tie should have flat surfaces. There should be no bark or knot holes or in- dications of rot. Ties are graded and placed according to traffic, or the rails which they support; the largest and best proportioned ties are used for the joint on main line traffic, the second selection, for general main line work. It is nearly impossible to draw a specification for a tie, as the wood adopted, the location and local timber possibilities to produce good ties and the price advisable to pay, fix the character of tie which it is possible to obtain. The selection of the ties should be in the hands of a competent and skilled inspector. Steel Ties. — Steel ties have come into use, especially in hot countries where wood is attacked by insects. The steel tie or its equivalent, as shown in the Scranton construction, is becoming more extensively used, as concrete is now largely employed for foundation under roadbeds. Fig. 92 gives one form of steel tie, which consists of an inverted channel iron 7 ins. wide, 1% ins. web and T 5 5 ins. thick. The rail is secured ^0 the tie by means of an angle bar and bolts. The total weight of tie is about 55 ibs. The spacing 132 ELECTRIC RAIL WA V HAND BOOK. of these ties should be arranged from 5 ft. to 11 ft. according to the weight of the rail and the character of sub-construction in concrete work. Steel ties which rest on ballast are usually provided with concave under- Fig. 92.— steel tie. surfaces to prevent ballast from working from under the tie. This character of tie is largely used for steam railroads in southern countries. Spikes.— The size of the standard spikes for rails from 35 to 40 lbs. is 5 ins. x^ in.; from 40 to 52 lb. rails, 5 ins. x ? s ins. ; from 45 to 85 lb. rails, 5]^ ins. x ft ins. Spikes Required per 1000 ft. and per Mile Single Track, with Four Spikes per Tie. SPACING OP ties PER 1000 FT. PER MILE. 10 ties to 30 ft. rail 133^ 7,040 11 " ' t it tt 1466^ 7,744 12 " c 4 tl tt 1600 8,448 13 " 4 t tl tt 1733^ 9,152 12 « * I it tt 1866% 9,856 J5 " ' t It tt 2000 10,560 16 " 4 t tt tt 21333^ 11,264 SPIKE TABLE. Size of Spike. 3^2 in. x ft in. 4 in. x ft in, 4% in. x /gin. in. x^oin. Ui ... « in. 4KJH.X}, 6 in. x 5 in. x ft in. b\& in. x ft in. a •rH QJ U M o * "2 & M Pi %A u %-. © "" •*-> 83 Q •a > < 900 0.2222 780 0.2564 675 0.2963 600 0.33*3 530 0.3773 500 0.4000 890 0.5128 350 0.5714 KEGS PER MILE OF TRACK. Tib Spacing 4 Spikes per Tie. 2 ft. 6 in. 9.39 10.83 12.52 14.08 15.04 16.89 21.66 24.14 2 ft. 3 in. 10.43 12.04 13.91 15.65 17.71 18.78 24.07 26.82 2 ft, in. 11.73 13.54 15.64 17.60 19.92 21.12 27.08 30.17 6 Spike&per Tie. 2 ft. 6 in. 14 08 16.24 18.78 21.12 23.91 25.33 82.50 86.21 2 ft. Sin. 15.65 18.06 20.86 23.47 25.56 28.17 86.10 40.23 2 ft. Oin. 17.60 20.31 23.46 26.40 29. S8 81.68 40.61 45.25 ELECTRIC RAILWA Y HAND BOOK 133 It has been found that it takes 4281 lbs. to draw a ^-in. spike driven 4*4 ins. into a seasoned oak tie; the same spike in unseasoned oak took 6523 lbs. On seasoning the wood the spike loses in holding power. Experiments on J^-in. spikes driven 4% ins. into yellow pine, showed 3000 lbs. and for oak, CXO lbs. The force is considerably more in hard wood to pull the spike out ; in softer woods the force is about £ less to pull the spike than that required to drive it. Tie Rods.— These take all lengths and sizes depending upon the service and gage of road. The form generally adopted is shown in Fig. 93, % ins. x \% ins. being Fig. 93.— tie rod. a section of iron commonly used. The thread should be cut far enough back so the tie rod can be inserted after the rails are in place, and the hole in the rail should be large enough so as not to mar the thread in passing the tie rod in. The flat section requires very little space between the blocks in brick or granite pave- ment, but the tie rods shonld be so spaced that they can accomodate between them a convenient number of paving blocks or bricks without loss of time in cut- ting the pavement to fit. Round rods are used in macadam construction. Tie Plates.— The tie plate is more generally used on elevated than on surface roads and is interposed between the base of the rail and tie so as to present a larger surface to the tie than the base of the rail, Fig. 94; it is usually secured independ- ently to the tie so the rail movement will not chafe and wear the tie at the point of bearing w-ith the rail. On curves tie plates have an additional advantage of distributing the canting effect and lateral strain on the rail over a large area, and, in addition, where the spike passes through the tie plate the efficiency of the spike is increased, preventing the movement of rail away from gage line. Fig. I. — TIE PLATE. Fig. 95.— old horse- car rail on stringer. Chairs.— In horse railways, the flat rails emp/oyed were. mounted on wooden stringers. This was necessary to raise the rail above the tie and thus provide room for paving and sufficient filling under the pavement to prevent un- equal settling due to the pavement bearing directly on the tie. This construction is shown in Fig. 95. With the low rails first used in electric railway work chairs were used to supplant this wooden stringer, as it gave trouble from rapid decay, especially under the joints. They were originally of cast iron, but on account,o_f the variation in castings their fitting to the rail section was not satisfactory, and the fragile character of the chair led to the introduction of chairs made of drop 134 ELECTRIC RAILWAY HAND BOOK. forgings of iron and steel in various forms, Figs. 96, 97, 98 and 99, taking the form of the box girder rail to which is fastened the base of the rail by bolts or clips as shown. To overcome the canting e£ort of the rail on the pacing of a load, chairs combined with braces were used, see Figs. 100, 101 and 102. These also Figs. 96, 97, 98 and 99.— rail chairs. transfer directly to the tie the side thrust caused by the car wheel flanges bearing asrainst the side of the rail. RAIL JOINT FASTENERS. In no part of the track has more thought or ingenuity been spent than on the proper mechanical joining of the rail lengths together. The ideal joint is one which is as strong and substantial as the rail itself. If an opening is left between the ends of the rails, say of % in., to accomodate the changes in the length' of the < ELECTRIC RAIL WA Y HAND BOOK. 135 rail due to differences in temperature, an opportunity is afforded to start a pound when the wheels pass over it; each wheel in passing contributing its quota toward the destruction of the joint. Bolted joints take the form of a plate bearing against the side of the rail and bridging the joint. The common form consists of an arched plate having a top and bottom bearing rolled to fit the rail, and secured in position by bolts passing through the joint plate and rail. Figs. 81 to 91 show sections giving some dilTer- ent forms of joint plates. For rails 6 ins. and over in height the bolts can be drawn so as to buckle the joint plate thereby destroying its bearing contact with the rail. Figs. 82, 83. 84 and 85 show an intermediate rib rolled in the joint plate which is normally out of contact with the web of the rail, but is brought to bear on the web before the bolts are tightened sufficiently to buckle the plate. Pratt &Alden give the following recommendations regarding joint plates: M For 6-in. rails they should not be less than ft ins. thick at the center; for 7-in. Figs. 100, 101 and 102.— rail chairs and braces. rails % ins., and for 9-in. rail not less than % ins., to prevent buckling under the bolt pressure.'* They advise a double row of bolts located as near the bearing sur- face as possible for the reason that the channel or joint plate as shown depends entirely on the compression given them against the rail. Several railroads have hot riveted the plate to the rails, instead of using track bolts. To strengthen this weak part of the track numerous track joints have been devised to afford a bearing to the rails independently of the track bolt tension. The lengthened chair at the joint was the first attempt in this direction for im- proving the joint. The "Continuous " rail joint shown in Fig. 103 is an extension of the joint plate, which includes the bearing of the base of the rail on the plate; here ten- sion and compression are set up within the joint plate, and do not act directly against the bolt heads. If the fit was perfect around the lower flange of the rail it could transfer the strains across the joint without movement of the rail head. Fig. 104 shows the " Churchill " rail joint, which provides a bearing for the- rail on a plate secured between the projecting sides of the joint plate,th«- lower bolt 13^ ELECTRIC RAILWAY HAND BOOK. acting as a locking device for this bearing plate. The Atlas joint shown in Fig. 105 embraces the joint with two pieces which are clamped together by bolts, and ex- tra metal is used where the ioint is suspended. Fig. 106 shows the Weber rail joint which consists of an L-iron, on which rests the rail base, and is secured to the rail joint by lengthening the track bolts to pass through the angle, as well as through a wood filter interposed between the regular joint plate and the angle. Fig. 103.— continuous rail joint. Fig. 104.— chuechill rail joint. There are a number of track joints of the bolted and keyed type, some of which should give good service. But the track joint question is one that every railway man has to study for himself in order that the conditions of his own special problem of tracks can be fully considered. The spacing of the bolts and the length of the joint nlate are matters on which there is a diversity of opinion, n r Fig. 105.— atlas rail joint. Fig. 106.— weber rail joint. and are governed largely by local conditions. Track joints vary in length from 20 ins. to 38 ins., and are fitted with from four to twelve bolts Electrically Welded Joints.— Tn the original method of electrically welding joints the adjacent rail ends were abutted and a current of about 20,000 amps, was passed through the joint. This heating effect brought the contact surfaces up to welding temperature and while at this temperature the rail ends were forced together and welded. The heating of the rails evidently reduced the Carbon in the steel for the rail was softened and when the contraction took place ELECTRIC RAILWAY HAND BOOK. 137 due to air temperature changes, the rail fractured at the points of welding in a number of cases, and the rail being softer at these portions low spots alao developed. The latest method is to electrically weld on each side of the joint bars with bosses which confine the heat to small areas. The results are said to be very satisfactory and a large amount of track welded in this way is in use in Buffalo. Cast Welded Joints.— In this case the joint is surrounded by a matrix to hold molten iron in such form that when the mass is cooled the additional strength afforded by the metal surrounding the joint compensates for the loss in strength of the rail due to its rise in temperature. These joints are poured weighing from 120 lbs. to 250 lbs. each and have given good results in service where the rail has been of such a compo ition as not to have its hardness changed by the heat from the molten iron. In some rails the property of self hardening is not as marked as in others. A high manganese rail gives the best results, showing less failures of welded joints. In some cases care has not been used in raising the ends enough, so that the head of the rail can be ground to a surface in alignment with the surface of the rail. If there is a low spot at the joint, the action of the car wheels tends to aggravate the trouble. The thermit joint is a welded joint in which the heat for welding is generated by an ex-othermic reaction. Thermit is a composition of iron oxide and finely divided aluminum, and when p'aced in a special crucible and ignited by a special powder, a powerful reaction lakes place, accompanied by a large evolution of heat which raises the iron far above its melting point. The aluminum takes up the oxygen and floats as clay, while the pure iron at a temperature of about 5500° F, collects at the bottom. The crucibleis then tapped and the molten iron introduced into the bottom of a mold around the joint. The Nicolls joint is a combination of the integral and the fish-plate joint. The fish-plates are so placed that a T 3 g in. space is left between them and the rail, then this space is filled with molton zinc. The zinc by virtue of its mallea- bility is well able to resist shocks and retain the electric continuity of the joint. (For electrical conductivity of rail joints, see Return Circuit). Bolts. — Bolts for plain channel joint plates should be % ins. in diameter with ribbed plates. Bolts may be 1 in. in diameter. The nut may be either square or hexagon, the square nut giving the largest surface against the joint plate, but in many cases the hexagon has to be used in order to obtain clearance. The portion of the bolt adjacent to the head for a length equals the thickness of the joint plate. It is oval in form and fits into an oval hole punched in the joint plate. SPECIAL, WORK. This term is used to cover all portions of track requiring any special design- ing, such as curves which cannot be sprung into place by the track foreman, crossings, turnouts, switches, etc. A plain curve, Fig. 109, connects two straight sections of track at an angle to each other, both of the straight sections being tangent to the connecting curve. Fig. 110 shows the reverse curve which connects two tangent sections of track parallel to each other. Fig. Ill shows the right hand crossover, Fig. 112, the left hand crossover. Figs. 113 and 114 give3the right and left hand branch-offs respectively. Fig. 115 shows the crossing and Fig. 116, the connecting curve and crossing. Fig. 117 shows the plain Y; Fig. 118, the three- part Y and Fig. 119, the three-part through Y. r38 ELECTRIC RAILWAY HAND BOOK. Turnouts, Switches, Etc.— Turnouts are illustrated in Figs. 120tol22. The diamond turnout, where the main track is central to both turnout tracks is given in Fig. 120. Fig. 121 shows the side turnout where the turnout is thrown over to the side of the main line, and Fig. 122 shows the turnout where the center of the main line track is displaced by the distance between the centers of the turnout tracks. In Fig. 123 tne names of the different parts of special work as they are gener- ally known r.re given, although their nomenclature varies in detail in different parts of the country. The point of crossing of two rails is commonly known as a frog. The initials of the diUcrent parts are generally used: e. g., L. H. T. S. for left-hand tongue switch, etc. \ Where special work is ordered the parts that are shown in Fig. 123 together are generally made in one piece. The "hand " is always determined by the side to which the curve turns off as it appears to a person facing the point of the curve. The method of construction of special work varies with its uses. Bolted work is largely used in exposed surface track, but in city streets a more perma- nent structure has to be used in order that the life will be longer, on account of the large expense of repaving and renewal. Here the rail parts are cast together after being fitted. WEIGHTS OF STANDARD TRACK BOLTS. Bolts with Square Nuts. Pounds per Thousand, .2 o p. a cm .2* s CM .5 CM 288 .2* 5 CM 302 .2* CO 316 .2 5 CO 330 .2 5 co .2* CO .2 .2 .2* .2* .2* is ^ 200 274 344 358 372 386 400 414 428 ^ 112 iV 352 370 388 406 424 442 460 478 496 514 532 550 568 i 9 s 146 i 454 476 498 520 542 564 586 608 630 652 674 696 718 54 218 626 658 600 722 754 786 818 850 882 914 946 978 1,010 Ps 245 858 901 944 987 1,030 1,073 1,116 1,159 1,202 1,245 1,288 1,331 1,374 374 i 1,155 1,210 1,265 1,320 1,375 1,430 1,485 1,540 1,595 1,650 1,705 1,760 1,815 1 525 M. 1,595 1,666 1,737 1,808 1,879 1,950 2,021 j2,092|2,163|2,234 2,305 2,376 2,447 lfc 747 Bolts with Hexagon Nuts. Pounds » per Thousand. a d d d d d ' d d d d & ® » c3 £1 p . "™ i ■ •■-( ."- 1 £ • r-l • >-c ."-' d • r-c •p-i d OS rd -?^ ■5| CM cm cu CM CO CO CO C \ CO ^ 5! 379 ^ 7 393 407 tf5 421 5 « p. 2 H 253 267 281 295 309 323 337 351 365 9S ft 327 345 363 381 399 417 435 453 471 489 507 525 543 1% 122 % 436 458 480 502 524 546 568 590 612 634 656 678 700 1 182 i 597 629 6G1 693 725 7'-7 789 821 853 885 917 949 981 216 822 865 908 951 994 1,037 1,080 1,123 1,166 1,209 1,252 1,295 1,338 316 1 1,087 1,132 1,187 1,242 1,297 1,252 1.407 1.462 1,517 1,572 1,627 1,682 1.737 l 46? a* 1,513 1,584 ,1,655 1,726 1,797 1,868 1,939 2,010:2,081 2,152 2,223 2,294 2,365 m GH6 ELECTRIC RAILWAY HAND BOOK 139 fLAIN CURVE. Fie. 109. /f£V£RS£ CUM£ Fig. 110. / / jhghthand cross-over* Fig. 111. U FT HAM CROSSOVER, F». 112. RIGHTNAND BRANCH -OF6! Fig. 113. J.EFT HAND BRANCAS Fie, 114. crossing: Fig. 115. CONNECTING CURV$ and crossing. Fig. 116. PLAINS Fig. 117. THREE PART % Fig. 118. THREE-PART TRROU&H K Fig. 119. DIAMOND TURNOUT. Fig. 120. tf/0f TURNOUT. Fig. 121. THROWN OVER TURNOUT. Fig. 122. 140 ELECTRIC RAILWAY HAND BOOK Fie. 123.— nomenclature of special work, ELECTRIC RAILWAY HAND BOOK. 141 Fig. 124 shows one of the methods used for built-up frog work, and the nomenclature of the different parts. The fitting of pieces of rail together is done by using a templet to obtain the proper track angle between the rails and making a pattern which fits between the heel and flangeway. These patterns are then cast in iron and bolted and riveted in position. THEORETICAL POINT- ^ACTUAL POINT. MTRACK .'THROAT jJflA^ALQ^ SPREAD\ A0RTi0lNl_aAik, '"TURNOUT AINPTRAIL* ■**■ - - HEEL BLOCK OR RAISING 9LX Fig. 134.— built-up frog work and nomenclature or different parts. The wear on special work occurs at the switch and mate points and frog cross- ings. In order to maintain these points and have the flange of the wheel ride so that the tread of the wheel will not bruise the switch or frog points, they are made of harder metal than the main part of the special track. This result is generally obtained by an inset of manganese, or nickel, steel, formed to fit as in Fig. 125, or by some special hardening process. Fig. 125. — special work with hardened points. The spring frog, Fig. 126, is used on the main line where it is normally con- tinuous; at track crossings angle plates are used, as shown in Fig. 127, in built- up frogs, which are bolted or riveted to the intersecting rails. In addition a sole plate is usually bolted underneath the track crossings to maintain the alignment of rails. . Guard Rails.— At curves there is required a guard for the outside of the flange of the outside w r hcel on the curve for curves of radius of S00 ft. and under. It is the practice for curves under 70 ft. radius to have guards on both raili for both wheels. The distance to space a guard depends upon the depth and thickness of the car wheel flange, and whether the curve is laid wider than gage or natural gage. In some track construction the curve is laid to a slightly nar- rower gage than the main track gage for the reason that the gage line is a radius of the curve while the wheel axles are at an angle to this radius, making their gage line across rails shorter. 142 ELECTRIC RAILWAY HAND BOOK In relation to curved rails the clearance between guard and rail can be de- termined by making a section of the wheel flange tangent to the wheel tread in celluloid and passing this around the gage line of the curved rail; allowing the Fig. 126.— spring frog (main line). guard rail to clear this at least one-quarter of an inch will give the proper spacing between the guard and the rail. Fig. 128 shows one method to be used on long radius curves where a cast iron spacing piece and strap iron guard is bolted to the rail. The spacing pieces are from 14 ins. to 24 ins. on centers depending upon the curvature of rails. Figs. 129, 130 and 131 show other approved forms. Fig. 127.— built-up frog. Where tracks cross trestle work and at dangerous crossings it is compulsory in some states (and also an advisable construction) to place two continuous guard rails between the rails, bent to nearly reach each other ten feet before approach- ing the hazardous crossing, in order to throw a derailed car toward the track. ; ELECTRIC RAILWAY HAND BOOK. 143 Curves. — Where the track changes its direction the introduction of a curve is necessitated. The center line of this curve may be struck joining the tangents, the radius being determined by the local conditions. In single track, in streets, the center from which the curve is struck may be in the curb. Sufficient clear- ance must be allowed for the obstructions at street corners so as not to endanger alighting passengers, or, in the case of open cars, passengers standing on the Figs. 128 and 129.— guard bails. running board. An accident arising from such a condition is evidence in itself of improper construction, and renders the railway company liable for injuries sustained. The surface of a street presents obstructions in the way of manholes, and subsurface structures belonging to other companies. If none of these exists the diagram given on page 123 of curves for 90 degs. when main line tangents are at right angles, can be used in the following way: Required to find from the tangents the largest radius possible for the track; suppose that the road passes from one intersecting street 50 ft. wide to another 30 ft. wide at right angles to each other, this will bring the center line of track 25 ft. from one curb and 15 ft. from the other curb. From one scale terminating at A on the diagram page 123 follow the line frcm 25 until it meets the line from the other scale terminating at A from 15; these will be found to intersect near the curve 60 which gives the radius of the largest curve that can be used under these conditions. Plotting these re- sults on section paper shows, 71/, Fig. 132, that the dotted center line just strikes Figs. 130 and 131.— guard rails. the curb line. The radius of the proper curve- must be reduced by such an amount that the center of the track will be so far away from the corner of the curb that the widest car will pass with sufficient clearance. The car body forms a portion of a moving polygon, the side of which is permanently fixed through the centers of the car axles in a single truck, and through the truck pivots in a double truck car. All car body movements due to play in trucks tend to throw the car body away from its shortest curve of motion. The usual way that this proper curve is located graphically is to cut out of translucent celluloid the horizontal projection of the outline of the car body on the 144 ELECTRIC RAIL WA Y HAND BOOK. ELECTRIC RAILWAY HAND BOOK 145 same scale as is used for the special work layout, and through the fixed center of axles or trucks as the case may be drill a small hole. "With this templet the position of nearest approach can be found, and it will also locate the points through which the truck centers must pass for the largest available curves. Thus, with fixed tangent positions, the two points which represent the truck centers will locate the simple curve to be used for each case. 12 3 4 5 6 7 Fig. 132.— graphic method of laying out curves. If there are any obstructions, such as poles, lamp posts, or water plugs along the curb, allowance should be made for clearance when passengers are standing on the running board or step of car. For economicnl reasons the tendency is to increase the length and width of the car body. It is therefore important that the best possible compromise should be made between the longest curve radius that can be used and the maximum clearance, many roads to day being compelled to change curves in order to operate larger cars. In the consideration of curves on double track, each track can be treated in- dependently, but as in this case the car on the other track is the obstruction to be cleared, the car fender has to be considered as part of the car, as well as the i4t> ELECTRIC RAILWAY HAND BOOK. movement of tiie carbody en its bearings due to centrifugal force which will dis- place the inside car toward the outside one in passing. Two templets should be used and in no positr n of either should there be less than the allowed safe clear- ance, which varies with the speed and radius of curves. The curves in Fig. 132 have been worked out for 28 ft., single over all, car bodies having trucks 7 ft. between wheel centers, to show the method of applica- tion. In many cases this curve has again to be shifted to avoid obstructions G 1 2 FlG. 133.—METHOD OP laying out easement curve. which cannot be moved, such as gate boxes, man hole covers, etc., all of which are difficulties which confront the railway engineer at many curve locations. The simple curve only has so far been considered, but the ones really laid down in modern railway practice are what are variously known as spiral, transi- tion and easement curves. These are compound curves which change their direction near the tangent less rapidly than the simple circle. They are com- posed of a number of curves with varying radii, the longer radii being at the switch point and gradually reducing in length until the central portion of the curve becomes a plain circle. The effect of a car passing around one of these ELECTRIC RAILWAY HAND BOOK. T47 1 compounded curves is to gradually increase the angular motion of the car around the curve so as to make the change in direction less perceptible. An easement curve of the form of a parabola can be laid out on the ground in the following manner: Continue the track tangents to their point of intersection, and at the point, which will allow ot the clearances in the center of the curve, drive a stake in a symmetrical curve, at the point marked /in Fig. 133, which will / 148 ELECTRIC RAIL WA V HAND BOOK. be at equal distances from the point where the track center leaves the tangent. Having located these points, marked A and A by driving stakes, stretch a string (3^ in. cotton cord will do) between C and B and C and A which will form the track tangents ; then between these two lines stretch a cord and carry it toward A and B from C nntil it touches the stake /. This cord should be secured to stakes Fernd D both located the same distance from C, when the cord is touching stake A To locate the other points in the curve, divide distances C-F, C-D into an equal number of parts and also F-A and D-B, numbering the divisions from A to Cin order and from Cto B in the same order. Stakes should be driven at each of the points and cords should be drawn for stakes of like number on the two tangents, when the points on these curves will be at the center of any cord be- tween two adjacent intersecting cords: e. g. /is located midway between 10 and 12, M is located midway between 9 and 10, etc. When work is ordered for curves of this character, the points A, B, C, D, F and /should be plotted, as these points will determine the form of the curve from which the rail manufacturers can bend the rail. The dotted line shows the departure of this curve from the plain circle connecting the two tangents. This curve presents difficulties in double tracks, but curves derived from a succession of decreasing arc lengths from the point of tangency of the track until a simple curve can be struck, and again becoming a symmetrical spiral until the other tangent track is reached, give better center clearances. Fig. 134 gives the method of laying out; the center line of track tangents are laid out, and from them are computed the displacement of the track from this line for the different radii of curves forming this spiral. Supposing that spiial No. 1 was required to fit between two track tangents at right angles to each other, the first curve would have a radius of 210 ft., and would include an arc of 42 minutes, as shown in the table under the heading "angle.'* The center line would depart .015 ft. from the center line, column "x." and this point would be 2.5G5 ft. from the point of starting. These points can be measured off for each point of departure along X and Fas shown in Fig. 134. The column headed " S° " gives the total angular deflection at each point of the spiral. SPIRAL, NO. 1. Ead. 210 105 70 52^ 42 35 Angle. 0°42 / 1°24' 2° 6' 2° 48' 3° 30' 4° 12' x. 0.015 0.078 0.219 0.469 0.8G0 1.420 y> 2.565 5.130 7.692 10.245 12.780 15.283 S°. 0^42' 2° 06' 4° 12' 7° 0' 10° 30' 14° 42' Ver. Sine. .00007 .00067 .00269 .00746 .01675 .03273 Sine. .01222 .03664 .07324 .12187 .18224 .25376 SPIRAL NO. 2. Ead. Angle. X. 0.011 y- S°. Ver. Sine. Sine. 1 300 0°30' 2.618 0°30' .00004 .00o73 2 150 POO' 0.057 5.235 1°30' .00034 .02618 3 100 1°30' 0.160 7.851 3° 0' .00137 .05234 4 75 2° 00' 0.342 10.463 5° 0' .00381 .08716 5 60 2° 30' 0.627 13.065 7° 30' .00856 .13053 6 50 3° 00' 1.036 15.651 10° 30' .01675 .18224 7 42^ 3° 30' 1.587 18.187 14° 0' .02970 .2M92 8 37^ 4°00 / 2.309 20.703 18° 0' .04894 .30902 ELECTRIC KAIL WA Y HAND BOOK 149 SPIRAL, NO. 3. Had. Angle. ^r. y- s°. Yer. Sine. Sine. 1 300 1°0' 0.046 5.236 1° 0' .00015 .01745 2 150 2°0' 0.229 10.468 3°0' .00137 .05234 3 100 3^0' 0.G39 15.688 6 3 0' .00548 .10453 4 75 40 / 1.368 20.871 10° 0' .01519 .17365 5 60 5°0' 2.501 25.982 15° 0' .03407 .25882 6 50 B ^ 4.118 30.959 21° 0' .06642 .35837 7 40 7 o / 6.143 35.403 28° 0' .11705 .46947 SPIRAL, NO. 4. Rad. Angle. X. y- S°. Yer. Sine. Sine. 1 420 3 42' 0.031 5.131 0° 42' .00007 .01222 2 210 1°24' 0.157 10.261 2° 06' .00067 .03664 3 140 2° 6' 0.439 15.384 4° 12' .00269 .07324 4 105 2° 48' 0.939 20.490 7° 0' .00745 .12187 5 84 3° 30' 1.720 25.561 10° 30' .01675 .18224 6 70 4°12 / 2.839 30.567 14° 42' .03273 .25376 7 60 4054/ 4.352 35.469 19° 36' .05794 .33545 SPIRAL NO. 5. Ead. Angle. X. 0.023 y* S°. Yer. Sine. Sine. 1 600 0°30' 5.236 0° 30' .00004 .00873 2 300 1° 0' 0.114 10.471 I°30 / .00034 .02618 3 200 1°30' 0.320 15.703 3° 0' .00137 .05234 4 150 2° 0' 0.685 -20.926 5° 0' .00381 .08716 5 120 2° 30' 1.255 26.130 7° 30' .00856 .13053 6 100 3° 0' 2.073 31 .302 10° 30' .01675 .18224 7 85 3° 30' 3.175 36.374 14° 0' .02970 .24192 SPIRAL, NO. 6. Ead. 900 450 300 225 180 150 128 Angle. 0° 20' 0°40' 1° 0' 1°20' 1°40' 2° 0' 2° 20' X. 0.015 0.076 0.213 0.457 0.837 1.385 2.125 5.236 10.472 15.706 20.936 26.158 31.365 36.524 S°. 0°20' 1° 0' 2° 0' 3° 20' 5° 0' 7° 0' 9° 20* Yer. Sine. .00002 .00015 .00061 .00169 .00381 .00745 .01324 Sine. .00582 .01745 .03490 .05814 .08716 .12187 .16218 Where a switch is to be located at the beginning of a curve the radius of the switch tongue limits the first radius to be used, and the easement cannot be as great as in plain track, 100 ft. being a common radius for switch points. The Union Traction Co., Philadelphia, has developed for its own work spirals ': for 90 deg. curves which fit its track gage, 5 ft. 2*4 ins. The company always take ; these measurements from the gage line of the inner rail, the first easement radius being greater than the above tables, and the center radius less, which gives greater clearance between cars at center of the curve. Fig. 135 and table on page 129 give data of the standard plain curve. Fig. 136 and the table give < curves with 100 ft. radius switch. Fig. 137 gives the combination of these two I curves in the standard branch-off: curves. It is not within the scope of a handbook to go into the details of the treatment A of complicated cases of spirals and curves. The matter can be found treated ISO ELECTRIC RAIL WA Y HAND BOOK. fully in Pratt & Alden's " Street Railway Roadbed," Tratman's "Railway Track and Track "Work," Searle's " Field Engineering," and many articles on special work in the Street Railway Journal. Fig. 135. — standard plan curve. Fig. 136. — easement tor branoh-ofp curve with switch. ELECTRIC RAILWAY //AMD BOOR-. ki 0> t s O c OS o PS *V 03900 <| -0^> [( Z V + X T?) 8A— ( C « + Z t? + l V) BAJ C H + [ T T? BA — ( z b + x u) ba] z 5i -f- t b sa l £ }- = Z UBJ J •v boo 8 u«x— " 3i P [( z ^ + l v) urpr— c B-f z «-f- x «)iits] e }£ 4-[ x buib — ( z « + 1 b) his] Z H i* x « uis *H = i ubj j ' * V nmoo Z .C — v oosoo , JC + z x : 8 n«j, * v uu;oo x a — v jebOO z JL l x= t wgj j O qj w>^0co05^00i-<0i«— t> 1— ■ OS t> IO T* 1 OiOi «T-ll-HHrl r-.V .osfc^cocoioww P^ x/i 00 i> ©i ^ CO Oi Oi"* 1-H 00 iC 00 O O* C5 0* Tf< -r-t 00 o CO GOCiOCOCOi-iCOCDaOCO ©COOOiCr-tCOCDCCCO g^CC00C0O50Ci> •—•cOOiNOiOiOiOi CO N05lOiH«0 .COOrjirjiOiOM © I I I I I . I \ OHHrlNWMN!C O H»OOONMHOiO ca O OS O NWhO aaaooooi- T-iTTgo55i>Tti^tiiOCi^r' OOOrtrHCJCO^ON 00 0i05£~C0O' tj< CO © Tf CO iO OS 00050iOJ0505Xl-N 00 IO Oi 00 CO b- © CO W CO t1< -J"*iflWWn(i-(Oi-iTr i>t>«OlO>Tf«OCOOii-iOC lOiOT-iCClOrHCOlOT-lOi I I I I I I I I I I w 00 hhOWOWiOOC: O occiOi£>coio^J , TrcO' CO TH 1—t •Kind AOCOCSCSOiOiOiOicO JONiHrt«N««iO Oi ONNONMthOO MXOWhOOOO W*«»OOOOi>. 10NrtT-((M«N«0 O-^OOWtNiOO OH^r(^T)00 }>-i— (OJrfCOCOiOOSiO COOt^OCOOCOCOOO i-ico^rcooscoooTrTft Triocscocooicooico CD £- CO CO J> C5 00 O CO CO CO Oi t-< r-i S3: CO COOO?-<*t-iOtHt}4 I I I I I I I I "<*l0.i>0S0iO0C'rH ? « O Oi CO CO Oi t-t CO COOiCOT-iCOiCT-iOi I I I I I I I I 8 o o oo oi io o co co ©XCOWa-TjicOcO X c3 Cnci :r£ K3' Ssj " r- f 92 W x o ,; C c3 O S3 M € ^1 coo wo?© CO "* "^"lO CO ■r-ilDiOOCO CO lO CO 00*0 Ci iO i> »o •^ "T CO £-00 2 c o o o. o J>OS CO CDC CS iO "^t Tf i> lOTT COOii-' OOiONO CO "Ct"tf 1C CO i-i o o* t- o oo^rooco £> rt< rf CO 0* i-H CQ^OOSOO nnrl -H-hH- Mill 3 Oi -r-i O* CO rr o? co Oi co WOiiNO T-i CO r-l CO Oi COOiOlT-lT-l coco c-00 tHCO + i-i CO CO IO coo Oi- o co»o fc: ^ J w Oi OSCO TJH00O t-OOCi lONn CO i-l CO t-00 00 o cOOtOOiO CO -rji Tf O CO S o COCOOlOO? COCO-^^iO •HOXT^S J « | 'KlVId + co P3 Ph T I 00 O CO 00 O CO a s O Ci iO Oi IO CO 3* *-i + 4- i ti •Mf? 152 ELECTRIC RAILWA Y HAND BOOK. MIDDLE ORDINATES ON TEN TOOT CHORDS. M. O. Radius. M. O. Radius. M. 0. 341 3| Radius. M. O. Radius. \ ft. 800 266 240 in. 8A A 2ft m 2| ft. 64 64 63 in. lift 1ft 3ft ft. 42 41 41 in. 3ft ion 6| 411 4 2 s l 41 ft. 31 31 30 in. 4f ? 11 218 200 184 m 2ft 71 2|f 2ft 21a 62 61 60 5§f m 10ft m m m 41 40 40 21 10 51 iff 41f 4§1 30 30 30 * 171 160 150 1 21 m 2ft 60 59 58 11 4| 7| 3! 3§f 40 39 39 If 6ft 5 8 30 30 29 1 10ft 1ft 1ft 141 133 126 4| 2|f M 2§1 57 57 56 HI 3 7 9* 31 311 39 38 38 2ft loft 6| 6A 51 5ft 29 29 29 51S 311 1ft 120 114 109 41 It 2H 2| 55 55 54 HI 8ft m 3?§ 811 4 38 37 37 31 5ft 5ft 51 29 28 28 t if 1ft 104 100 96 416 3® 2|f 2*1 2§5 54 53 12 a 5ft 101 4ft 4ft 4ft 37 37 36 41 Si 5ft 5ft 511 28 28 28 7i 51 31 If 11! 92 90 88 41 7ft 2| 2P 2il 52 51 51 m 21 41 4 4 36 36 35 6ft 3ft HI* 5f 5 3 § 5ft 28 27 27 f 1? 1H 87 85 84 4| 9r 7 s 3ft 3ft 50 50 49 *A 4ft 35 35 35 8f 5J1 2ft 541 51 5*1 27 27 27 4ft m ift 82 81 80 10 A 3ft 3ft 31 49 48 48 n 7| ift 4ft 411 4| 34 34 34 HA 8t 9 5 51 5ft 541 5| 27 27 26 2| ft 1011 ill 78 77 76 91 6 31 33 s * 3 A 3ft 47 47 46 71 2 5 5 8ii 4H 4ft 4H 34 33 33 211 Hi 9ft 511 511 511 26 26 26 9ft 7ft 5ft 2 2^ «A 75 73 72 I lift 9^ 31 3ft 3ft 46 45 45 31 10ft 5ft 4| 441 4ft 33 33 33 61 31 11 5f 5§| 5ii 26 26 26 K »A 3ft 71 70 69 8| 81 311 3f 34§ 45 44 44 21 44! 4| 411 32 32 32 101 71 4| 5| 5|| 25 25 25 mi 9ft 7U 2$ 2ft 68 67 65 65 7M 10ft 1 oik 3f 811 43 43 43 42 91 ft. 411 411 32 31 31 31 2f 1HI 9A 61 i 25 25 61 4ft- In order to determine the radius of any simple "curve of track, a straight edge 10 ft. long is laid against the rail on the inside of the curve or gage line, and the distance between the middle of the straight edge and curve on gage line, perpendicular to the straight edge is measured. This will give what is called the 44 middle ordinate." From this length the radius of the curve can be determined ELECTRIC KAIL WA Y HAXD BOOK. 153 by reference to the table of middle ordinates on page 131. For example, if the distance was 2„ 9 Z ins., then the radius of the curve is 65 ft. 10^ ins. Rail bending to any radius is determined in the same manner. Fig 1C7. — easement — standard branch-off switch at both ends of curve. Super Elevation of Rails. — "Where cars run rapidly around curves it is the practice where possible to elevate the outer rail in order to reduce the pres- SUPER ELEVATION OF OUTER RAIL. Radius 3 Ft. Gage. 3 Ft. 6 In. Gage. 4 Ft. 8^ In. Gage. of Curve in Speed of Train in Speed of Train in Speed of Train in Feet. Miles per Hour. Miles per Hour. Miles per Hour. 10. 20. 30. 40. 50. i 10. 20. 30. 40. 50. 10. 20. 30. 40. 50. 40 4 1 ci 50 3fV 3! 5 60 m SS 4 3 - 90 if 2r l « w 120 1t 3 « 1 p 150 1& m 200 *§ ?T 3 B il s 3 f u 5 300 9. 2* s I * i s 83 400 I If 3i 9 6 H H* 2| 5| 600 1A- 23 V n 2 T § 7. IS ii 1 31 61* 800 3. 16 A Hi 3i 3 6 y 2* | 3f I 5 . 1 m 5 154 ELECTRIC RAILWA V HAND BOOK. sure against rail and wheel flanges due to the centrifugal force exerted by the moving car. The table on preceding page gives the elevation allowed for (.liferent gages and speeds. Sometimes only one-half the elevation is given to outer rail and the the inner rail is depressed by the same amount. For Bonding and Rail Connections see "Line Work, 11 Section 5. TRESTLE WORK. Where trestles show less cost than filling, especially over marshy or soft ground, and over railway tracks, they may be built either of wood or iron. The combination of wooden approach trestle and iron lattice or girder spans over SIDE ELEVATION SHOWING POLE SUPPOrt ^6" — ^S-O^H^-O Fig. 138. tracks where locomotives pass underneath is often used on account of danger from fire from sparks. Fig. 138 gives an excellent form of trestle for marshes. (Designed by Ford, Bacon & Davis.) The following is the general data: The bents shown in Fig. 138 are 15 ft. apart for cars not weighing, loaded, over 23,000 lbs. at a speed not exceeding 16 miles per hour. There are six piles in each bent located as shown, in this case from 30 ft. to 45 ft. long, 10 ins. to 12 ins. top. The framing is long ELECTRIC RAILWA Y HAND BOOK. 155 leaf yellow pine. The caps are 12 ins. x 12 ins. There are four girders under- neath each track, 8 ins. xl2ins.; the cross and longitudinal braces are 3 ins. x 10 ins.; the braces are tlioroughly bolted, and the caps are secured by a 1 in. dowcll pin from 18 ins. to 24 ins. in length driven in a 2£-in. hole, the dowell being covered with white lead before driving. The tics are 6 ins. xSins. x 9 ft. yellow pine, spaced 18 ins. on centers; the guard stringers, one on each side cf rail, are also yellow pine; the track centers are 11 ft. V/ % ins. apart, allowing two 9 ft. 6 in. car bodies to pacs. The mean given for the ultimate resistance to compression for white oak used as a post is 3470 lbs. per sq. in., and 4544 lbs. per sq. in. for yellow pine. 600 lbs. per sq. in. is the figure generally used, giving the proper factor of safety, for yellow pine. The length of the post which will not yield, as found from tests made at the Watertown Arsenal is given in the table below. These tests were made upon rectangular yellow pine posts with flat ends having a length of from 5 ft. to 28 it., and ranging in sectional area from 27 to 140 square inches. The results may be generalized as follows, calling — the ratio of length of post to least side of cross-section, and f the ultimate resistance to compression, in pounds per square inch: I f RATIO OF DECREASE, : S i5 4000 1.00 15 : 30 3500 0.88 30 : 40 8000 0.75 40 : 45 2500 0.63 45 : 50 2000 0.50 50 : 60 1500 0.38 WOODEN BEAMS. The following is a general summary of the results obtained by Prof. Lanza from numerous experiments upon wooden beams. They were of an average section of about 12 ins. x 4 ins. and were tested for mean span lengths of about 18 ft. _. , mAM . •»,■,, - M __ (Moment of forces causing rupture.) Kind of timber. Modulus of rupture =— = -£- — — — £ — *— - — '- /c (Moment of resistance of cross section.) MINIMUM. MEAN. 2995 4884 8438 4808 4984 6075 5092 7292 The above statement of the maximum and minimum values does not con- eider the results obtained in a few isolated cases for which the conditions were radically different than for the others. It was found that the beams frequently gave way through longitudinal shearing near the neutral axis, though this was not as common a source of failure as breaking across the grain. For spruce the mean intensity of the shearing strains, for beams that failed in this manner, was 191 lbs. and for yellow pine 248 lbs. For beams that failed otherwise, the mean intensity of shearing strains at the moment of rupture was very nearly the same. & The conclusion appears, therefore, to be warranted that for soft timber there is an almost equal tendency for beams to fail by shearing longitudinally at the neutral axis, as by the tearing of the outside fibers. MAXIMUM Spruce White Pine Oak Yellow Pine 5878 6415 7659 11360 156 ELECTRIC RAILWAY HAND BOOK. Owing to the wide range of the results obtained and the generally erratic behavior of timber subjected to strains, Prof. Lanza recommends the following values for moduli of rupture to be adopted in practice. Spruce and White Pine 3000 lbs. Oak 4000 " Yellow Pine 5000 " These values are lower than heretofore in use, and a safety factor of 4, on the basis of these values may be assumed as ample for all cases. The following table has been calculated for extreme fiber strains of 750 lbs. per square inch: SAFE [LOADS, UNIFORMITY DISTRIBUTED, FOR RECTANGU- LAR SPRUCE OR WHITE PINE BEAMS. One Inch Thick, (For oak, increase values in table by one-third). (For yellow pine, increase values in table by two-thirds). Span in Feet. Depth of Beam. 6" 7" 8" 9" 10" 11" 12" 13" 14" 15" 16" 5 6 7 600 500 430 820 680 580 1070 890 760 1350 1120 960 1670 1390 1190 2020 1680 1440 2400 2000 1710 2820 2350 2010 3270 2730 2330 8750 3120 2080 4270 3560 3050 8 9 10 • 380 330 800 510 460 410 670 590 530 840 750 670 1040 930 830 1260 1120 1010 1500 1330 1200 1760 1560 1410 2040 1810 1630 2340 2080 1880 2670 2370 2130 11 12 13 270 250 230 370 340 310 490 440 410 610 560 520 760 690 640 920 840 780 1090 1000 930 1280 1180 1080 1490 1360 1260 1710 1560 1440 1940 1780 1640 14 15 16 210 200 190 290 270 260 380 360 330 480 450 420 590 560 520 720 670 630 860 800 750 1010 940 880 1170 1090 1020 1340 1250 1180 1530 1420 1330 17 18 19 180 170 160 240 230 210 310 290 280 400 370 360 490 460 440 590 560 530 710 670 630 830 780 740 960 910 860 1100 1040 990 1260 1190 1130 20 21 12 150 140 140 200 190 190 270 260 240 340 320 310 420 390 380 510 480 460 600 570 540 710 670 640 820 780 740 940 890 850 1070 1020 970 23 24 25 130 130 120 180 170 160 230 220 210 290 280 270 360 350 330 440 415 410 520 500 480 610 590 560 710 680 660 810 780 750 920 899 860 26 27 28 29 110 110 110 110 160 150 140 140 210 200 190 180 260 250 240 230 320 810 300 290 390 370 300 350 460 440 430 410 540 520 500 490 630 610 580 560 720 690 670 640 820 790 760 740 To obtain the safe load for any thickness: Multiply values for one inch by the thickness of beam. To obtain the required thickness for any load: Divide by safe load for 1 in. ELECTRIC RAILWA Y HAND BOOK. 157 For use in trestle work the load of car is treated as a live load, the bearing centers of load being the distance between wheel centers; using beams under this condition the beam will only take one-half on high speed roads, and two- thirds on moderate speed roads, of the loads given. CAR HOUSE TRACK. This is built generally on a slight grade toward the main track in order to facilitate the movement of the cars in case of fire. Several methods are used to i 1 ■ 1 ■ 1 j | : 1 1 ! : : : ; : . ! Fig. 140. 158 ELECTRIC RAILWAY HAND BOOK. avoid breaking the main line rail when it is used for regular traffic especially where the switches have to face the direction of traffic. One is, where space is available, to run a track parallel to the main line track and to have all the car house tracks' switches on this auxiliary track. Another method, a compromise from the parallel track, is to run a gauntlet track, Fig. 139, 6 ins. or 8 ins. from the main line track, and have the crossings all jump over frogs so that the main line track is unbroken to traffic. This re- quires two switches. What is known as the ladder method (which is used extensively where the car house sets back from the track) shown in Fig. 140, is to run a spur from the main line track at an angle to the car house, and from this spur take the Fig. 141. entrance tracks to the car house. Fig. 141 gives a compromise on the laddej method, and requires only half the switches on the main track over the direct curves, but does not give so much room for the cars when in front of the car house. COST OF TRACK AND PAVEMENT. The following estimates were made by John A. Beeler of the Denver City Tramway Company in 1893. Section A. (Fig. 142)— This shows a 70-lb. T-rail (Shanghai) doing away with chairs, having a tie rod every four feet, which would make a very durable and serviceable track construction. This road is ordinary stone block pavement with one inch sand cushion and six inches of concrete for a base, as per city specifications, with a gravel foundation for track. ELECTRIC RAIL WA Y HAND BOOK 159 Fig. 142.— section a. Block Stone Paving on Concrete and Gravel. Ties 21" C. to C. Cost per Mile Single Track. 110 tons rails (including freight, inspection and hauling) at $37.50 per ton $ 4,125.00 18.000 lbs. aiigle bars (360 per 50 lbs. each) at $2.off>er 100 lbs 361.80 1,700 lbs. track bolts (% x3% ins.) at $3.01 per 100 lbs ttLi7 6,050 lbs. railroad spikes (5x^ ins.) at $2,46 per 100 lbs 148.83 1*4 M nut locks at $6.50 per M 8.12 3,017 hewn red spruce ties at 55 cts. each. . 1,659.35 360 bonds (placed complete) at 25 cts. each. 90.00 1,320 tie rods at 20 cts. each ." ; 264.00 2,347 cu. yds. excavation (trench 8 ft. wide 18 ins, deep, all hauled off) at 30 cts. percu. yd 704.10 Track laying, including blocking, etc 1,000.00 $ 8,412.37 Quantities per Lineal ft. 46.667 lbs. 3.409 " .322 " 1.146 Cost per Lineal ft. $0.7813 0.0686 0.0997 0.0282 0.0015 0.3143 0.0171 0.0498 0.1334 0.1893 $1.5932 Stone Paving 7.5 Feet Wide (Including 1 in. Sand Under Blocks.) Cost per Mile Single Track. 4,400 sq. yds. (stone $1.50, laying 75 cts., sand, tar, etc. 50 cts.) at $2.75 per yd.. .$12,100.00 14,085 cu. ft. concrete (10 per cent cement 6 ins. deep between ties) at 15 cts. per cu.ft 2,112.75 800 cu. yds. gravel under ties at 50 cts. per cu. yd 400.00 22,000 ft. B. M. lumber (2 ins. x 14 ins. pine, retaining concrete, etc.) at $14. per M. . ft 308.00 Carpenter work, nails, hauling, etc 60.00 Total cost per mile of paving $14,980.75 Cost of paving per sq. yd. $3.40. Total cost per mile single track $23,393.12 Quantities per Lineal ft. 0.833 yd. Cost per Lineal ft. $2.2917 2.669 C. ft. 0.4001 0.151 C. yd. 0.0758 4.167 ft. 0.0583 0.0113 $2.8372 $4.4300 i6o ELECTRIC RAIL WA Y HAND BOOK. Section B. (Fig. 143)— Same track construction as Section A, and same pav- ing, "but a good foundation for track is provided by a continuous bed of concrete six inches deep under ties. This would make the most serviceable and durable construction for streets where stone blocks are to be used and, he thinks, would give best satisfaction. The additional cost of a foundation is very little when compared to the total cost. Block Stone Paving on Concrete Foundation. Cost per Mile Single Track. 110 tons rails (including freight, inspection and hauling) at $37.50 per ton $ 4,125.00 18,000 lbs. ansrle bars (360 per 50 lbs. each) at $2.01 per 100 lbs 361.80 1,700 lbs. track bolts (%ins. x3^ins.) in- cluding freight and hauling at $3.01 per 100 lbs 51.17 6,050 lbs. railroad spikes (5 in*.x T ^ ins.) including freight and hauling at $2.46 per 100 lbs 148.83 1*4 M nut locks at $6.50 per M. . . 8.12 8,017 hewn ties (6 ins. x8ins. x 7 ft.) red spruce, including hauling and inspec- tion, «t 55 cts. each 1,659.35 360 bonds (placed complete) at 25 cts. each 90.00 1,320 tie rods at 20 cts. each 264.00 2,357 cu. yds. excavation (trench 8 ft. xl8 ins. deep all hauled awajO at 30 cts per cu. yd 704.10 Track laying including blocking, etc 1,000.00 $8,412.37 Ties 21 ins. Quantities per Lineal ft. 46.667 lbs. 3.409 .322 1.146 C to C. Cost per Lineal ft. $0.7813 0.0686 0.0097 0.0282 0.0015 0.3143 0.0171 0.0498 0.1334 0.1893 $1.5932 Stone Taxing 7.5 ft. Wide (Including 1 in. Sand Under Stone Blocks. Cost Cost per Quantities Mile per Single track. Lineal ft. 4,400 sq. yds. (stone $1.50, laying 75 cts., sand, tar, etc. 50 cts.) at $2.75 per yd. . .$12,100.00 0.883 11,734 cu. ft. concrete (10 per cent cement 5 ins. deep between ties) at 15 cts. per cu. ft 1,760,10 21,120 cu. ft. concrete (10 per cent cement 8 ft. wide 6 ins deep under ties) at 15 cts. per cu.ft 3,168.00 22,000 ft. B. M. lumber (2 ins. x 12 ins. pine for retaining concrete and pavement, at$14 perM 308.00 Carpenter work, nails, etc 60.00 Cost of paving per sq. yd. at $3.95 $17,396.10 Total cost per mile single track $25,808.47 Section C. (Fig. 144)— This track construction is good, heavy, 60-lb. steel with joint boxes. This rail is especially adopted for the Blake asphalt. In this section, track rests on a concrete foundation, with concrete to the top of the ties, 2.222 4.000 4.16^ per Lineal ft. $2.2917 0.3334 0.6000 0.0583 0.0113 $3.2947 $4.8879 ELECTRIC RAILWA Y HAND BOOK. 161 Fig. 143.— section b. Fig. 144.— section o. Fig. 145.— section d. Fig. 146.— section g. 162 ELECTRIC RAILWAY HAND BOOK. with the Blake Asphalt paving. This section is not perfect, however. He states that the earth and dust are pounded in the crevice between the rail and asphalt by the wheel flanges, and works its way between the asphalt and concrete at the line at the top of the tics, bulging the paving, letting in the moisture and eventually destroying the asphalt. Blake Asphalt Pavement. Ties 21 Ins. C to C. Cost per. Mile Single Track. 94^ tons steel rails (including freight, in- spection and hauling) at $37.50 per ton.. $ 3,536.25 10,800 lbs. angle bars (360 per 30 lbs. each, including hauling, etc.) at $2.01 per 100 lbs 217.08 1,150 lbs. track bolts {% ins. x3^ ins., in- cluding freight and hauling) at $3.01 per 100 lbs 34.62 6,050 R. R. spikes (5 ins. x T %ius. including freight and hauling) at ^2.46 per 100 lbs. 148.83 1^ M nut locks at $6.50 per M 8.12 3,017 hewn red spruce ties (including haul- ing and inspection) at 55 cts. each 1,659.35 360 bonds (placed complete) at 25 cts. each. 9Q.00 360 cast iron joint boxes at 50 cts. each. . . . 180.00 2080 cu. yds. excavation (trench 8 ft. wide 16 ins. deep, all hauled away) at 30 cts. each 624.00 Track laying, including blocking, etc 1,000.00 $7,498.25 Quantities per Lineal ft. 40.000 lbs. 2.070 .218 1.146 Cost per Lineal ft. $0.6697 0.0411 0.0066 0.0282 0.0015 0.3143 0.0171 0.0341 0.1181 0.1893 $1.4200 Asphalt Paving 7.5 ins. Wide, 4 ins. Thick. Cost per Quantities Cost per per per Single Track. Lineal ft. Lineal ft 4400 sq. yds. at $2.60 $11,440.00 0.833 14,085 cu. ft. concrete (10 per cent cement, 6 ins. deep between ties; at 15 cts 2,112.75 2.669 21,120 cu. ft. concrete (10 per cent cement, 6 ins. below ties 8 ft. wide) at 15 cts 3,168.00 4.000 25,700 ft. B.M. lumber (2 ins. x 14 ins. pine) retaining asphalt, etc., at $14 359.80 4.867 Carpenter work, nails, hauling, etc 75.00 Cost of paving per sq. yd. 83.90 $17,155.55 Total cost track laying and paving. . . .$24,653.80 $2.17 0.40 0.60 0.07 0.01 $3.25 $4.G7 Section D. (Fig. 145)— This section shows the same track construction with foundation, etc., but the asphalt is 5 ins. thick (1 in. deeper than in Section C). By this means the asphalt is bedded all around the rail, completely encasing it; hence the dirt cannot work in and deposit between the asphalt a::d concrete. This construction, however, is expensive, as too much asphalt is used, ELECTRIC KAILWA Y HAXD BOOK. 163 Blake Asphalt Pavement, 1 in. Below Top of Tie. Ties 21 ins. C to C. Cost per Mile track. Single Track Same as Section C $ 7,498.25 PAVINO. (7.5 ft. wide, 5 ins. thick.) 4400 sq. yds. Blake asphalt at $2.90 12,760,00 11,734 cu. ft. concrete (10 per cent cement, 5 ins. deep between ties) at 15 cts 1,760.10 21,120 en. ft. concrete (10 per cent, cement, 6 ins. deep below ties) at 15 cts 3,168.00 26,700 ft. B.M. lumber (2 ins. x 14 ins. pine) retaining asphalt and concrete at $14. . 359.80 Carpenter work, hauling, etc 75.00 Cost of paving per sq. yd., $4.12 $18,122.90 Quantities Cost per Lineal ft. per Lineal ft. $1,420 0.833 sq. yd. 2.417 0.333 6.222 cu. ft. 0.600 4.867 ft. 0.070 0.010 $3,430 $4,850 Total cost of track and paving $25,621.15 Section G. (Fig. 146) — We will take this section up next, as it bears upon the two immediately above. Track construction same as above, but economizes upon the asphalt. Here we have a concrete foundation, 6 ins. deep, under ties, and carry up the concrete above the ties, except for a space averaging 10 ins. wide directly under the rails, thus cementing the whole structure together, and giving a weariug surface of asphalt paving 3 ins. deep. (Barber asphalt i3 only 2% ins.) The rails and tics where exposed, should be coated with tar or liquid asphalt just previous to laying the pavement, thus making it air and water tight. This would be the ideal construction, and its cost will certainly be in its favor from the start. Blake Asphalt Pavement. Cement Concrete Foundation. Ties 21 ins. C to C. Cost per Quantities Cost Mile per per track. Single Track. Lineal ft. Lineal ft. Same as Section C $7,498.25 $1,420 PAVING. 4400 sq. vds. Blake asphalt (7.5 ft. wide 3 ins. thick) at $2.25 9,900.00 1.875 - 36,178 cu. ft. cement concrete (see below) at 15 cts 5,426.70 1.027 25,700 ft. B.M. lumber (2 ins. x 14 ins.pine) at $14 359.S0 0.0C8 Carpenter work, nails, hauling, etc 75.00 0.010 Cost per sq. yd., $3.58 $15,761.50 $2,980 Total cost track and paving $23,259.75 $4,400 concrete. Cu. Ft. Cu. Ft. Cu. Ft. Below ties (52S0 ft. x 7 ft. x 0.5 ft.) 18,480 Between ties and 6 ins. from ends (5. 280 ft. x 8 ft. x 0.5 ft.) 21,120 Less cu. ft. in ties (3017 ft. x 2.33% ft.) 7,040 Less cu. ft. in space below rail 817 7,857 13,263 Above ties (5280 ft. x 6 ft. x 0.14 ft.) ...... 4,435 36,178 164 ELECTRIC RAIL WA Y HAND BOOK. Section E.— This is a 60-lb. steel rail on chair construction with the necessary tie rods. Chairs are a little heavier than used formerly and if there is any error in these figures the principal one would be the price of the chairs, which would be very apt to cost more. A 6-in. foundation of concrete below ties is calculated, and the space between the ties filled in with concrete; above this the Barber asphalt and stone toothing. With this foundation this is the same practical construction as used on portions of Stout and Arapahoe Streets, increased to a 60 lb. rail construction to compare with other proposed sections. This is very expensive, and the chairs are an unmitigated nuisance. They should be avoided hereafter. Barber Asphalt Paving. Stone Toothing, T Kail on Chairs. Hewn Ties 21 ins. C to C. Cost per Cost Mile per TRACK. Quantities Price Single Track Lineal ft. 60-lb. T rail (including freight in- spection and hauling 110 tons $37.50 $3,536.25 Angle bars (360 per 50 1 bs.) includ ing freight.in^pection and hauling. 13, 000 lbs. 2.01 217.08 Track bolts (% in. x 3% ins.) includ- ing freight, inspection and hauling 1,700 " 3.01 34.62 R. R. spikes (4 ins. x A ins.) 750 to the keg, 200 1 bs. each, 37^4 kegs 6,450 " 2.56 165.12 Nntlocks.. 1J4M 6.50 8.12 Hewn ties (6 ins. x 8 ins. x 7 ft.) in- cluding inspection and hauling. 3,017 .55 1,659.35 Bondsinplace 300 .25 90.00 Tie rods 1,320 .20 264.00 Wrought iron chairs (4 ins. high).. 6.034 .60 3,620.40 Excavations (5280 ft. x 8 ft. wide x 20^ins.deep 2,672cu.yds. .30 801.60 Track laying (including blocking) 1,250.00 $11,646.54 TAV1NG. Concrete (7 ft. wide, 6 ins. deep) under ties, 10 per cent cement.. 18,480 cu. ft. .15 2,772.00 Concrete (8 ft. wide 6 ins. deep) be- tween ties, 10 per cent cement. 14,085 " .15 2,112.75 Barber asphalt paving, stone tooth- ing (8^ ins. deep) 4,400 sq. yds. 3.15 13,860.00 Cost of paving per sq. yd., $4.26 $18,744.75 I II Total cost of track and paving $30,391.29 $5.75 Section F.—Thls section has the same style of chair and track construc- tion as in Section E with a stone block pavement instead of Barber asphalt; and includes the foundation under tics. The first figures are an estimate of cost based on figures for concrete and paving, and the second are based on the price paid for the paving on Wazee Street by the Board of Works. ' ELECTRIC RAIL WA Y HAND BOOK. 165 Block Stone Paving, Concrete Foundation. T Rail on Chairs. Quantities. TRACK. Same as Sec. E, less excavation Excavation (trench 5,280 ft. x 8 ft; wide x 19 ins. deep) 2,477 cu. yds. PAVING. Stone block paving, 7.5 ft. wide, in- cluding 1 in. sand under blocks.. 4,400 sq. yds. Concrete 8 ft. wide, 6 ins. deep be- tween ties, 10 per cent cement. . . 14,085 cu. yds. Concrete 7 ft. wide, 6 ins. deep under ties, 10 per cent cement 18,480 cu. ft. Lumber for retaining concrete and 22,000 ft. B. M. paving Carpenter work, nails, etc 60.00 5 rice. Cost per Mile Single Track. $10,844.94 Cost Lin. ft. .30 743.10 $11,588.04 2.75 12,100.00 .15 2,112.75 .15 14.00 2,772.00 308.00 Costpersq. yd $3.94 $17,352.75 Total cost per mile, track and paving $28,940.79 $5.48 Section Fa.— Block Stone Paving, Concrete Foundation. Quantities. Price. Cost per Cost Mile Lin. ft. track. Single Track. Same as Section F _._ $11,588.04 PAVING. Stone block paving 7.5 ft. wide (in- cluding: 1 in. sand and 6 ins. con- crete under blocks 4,400 sq. yds. $3.50 15,400.00 Concrete under ties (7 ft. wide, 6 ins. deep) 10 per cent cement 18,480 cu. ft. .15 2,772.00 Cost per sq. yard $4.13 $18,172.00 Total cost per mile track, and paving $29,760.04 $5.63 Summary. Fonnda- Sec. Cost per Cost Cost tion. Mile Lin. ft. per Single Track. Sq. yd. Block stone paving, 70-lb. Shanghai Concrete rail &Gravel A $23,393.12 $4.43 $3.40 Block store paving, 70-lb. Shanghai rail Concrete B 25,808.47 4.89 3.95 Blake asphalt paving, 60-lb. T rail.. " C 24,653.80 4.67 3.90 Blake asphalt paving, 60-lb. Trail.. " D 25.621.15 4.85 4.12 Blake asphalt paving, 60-lb. T rail.. " G 23,259.75 4.40 3.58 Barber asphalt paving, 60-lb. T rail on chairs ' " E 30,391.29 5.75 4.26 Block stone pavinsi, 60-lb. T rail on chairs " F 28,940.79 5.48 3.95 Block stone paving, 60-lb. T. rail on chairs (contract price) * Fa 29,760.04 5.63 4.13 For cross country roads the pavement should be left out of the estimate. The price of material will depend upon freightage and local costs. If the tics are. ^ 1 66 ELECTRIC KAIL WA Y HAND BOOK. spaced at a greater distance than 21 ins. between centers, the number of ties per mile for the different spacing is given in Table, Page 110. The cost of laying a road parallel to country roads varies from 35 cents to 45 cents per running foot. Excavation costs from 23 cents to 45 cents per cubic yard depending upon local conditions. The price of bonds varies from 35 cents SECTION OF THIRD AVE. CABLE T/fACK Fig. 147. SECTIOff OF BROAD W CABLE T/fACH. Fig. 148. to 65 cents per joint depending upon the current density in the rail. Bonds can be inserted and applied at an additional price of 14 cents to 30 cents per joint. The effect of laying asphalt paving against the rail is shown in Fig. 147. Fig. 148 shows the methods used in New York in connection with cable tracks. m¥ JM A9ETHOD OF* « LAr/NG- HAILS JH_ASPHALK Fig. 149/ METHOD OFLAW& BAILS IN ASPHALT. Fig. 150. Fig. 149 shows method of paving with asphalt against the rails. Fig. 150 shows method of paving with asphalt with granite toothing block against the rails. Paving Cost.— -The following figures are from Washington, 1892: Trinidad asphalt with 7-in. concrete base, $2.25 per sq. yd. ; Trinidad asphalt with 4 ins. concrete base, $2.00 per sq. yd.; Asphalt block, $2.00; asphalt surface, $17.00 cubic yard Bituminous base, $3.00 p er cu ^ V( ^ j n place. Hydraulic cement and concrete, $5.00 per cu. yard in place. Asphalt surface, $1.02 per sq.yd. Annual average for repairs 3 cents per sq. yd. ; resurfacing, $1.50 per sq. yd. SECTION IV —POWER STATION. I Power Station Location.— The factors affecting the proper location for the power station are the cost of land, the cost of copper for distribution, the cost of coal and the value of condensing water. The price that can be given for a piece of property in a central location is determined by the way in which this location affects the other investments and station economics. Take, for example, a 15 mile stretch of road with cars uniformly placed requiring 20 amps, per mile average and 40 amps, per mile maximum, and assume 20 per cent drop in voltage on 7% miles of road. With the station centrally located the copper will cost about $20,000; if the station is 2 miles from the center of distribution the installation cost for copper will be increased $6700, Property in the central location would, therefore, be worth this much more to the railway company as better distribution could be obtained from a station on that site. Where the coal can be delivered directly from the cars to the coal bins of the station the cost for handling is the lowest. Where there is any rehandling, the price depends upon the distance traversed. To load and move 1 ton 1 mile or less costs about 25 cents per ton; 1% miles, 30 cents; 2 miles, 32 cents. These figures are taken from average prices paid for hauling over a variety of roads. A station with the capacity mentioned above would require, on an average, about 11 tons of coal per day ; if hauled 1 mile this would cost per year with shrinkage in coal weight due to moving, about $1000, or 6 per cent on an investment of $16,666. The value cf condensing in a street railway plant of the size cited above can be roughly estimated at 18 per cent saving in coal. At $2.80 per ton this would be $2023 per year or 6 per cent on an investment of $33,700. This station would take about 700,000 cu. ft. of water per annum for boiler, use. If the water had to be bought, at say, $1 per 1000 cu. ft., a site would be worth $11,600 more where free water could be obtained. Too often stations have been located on property owned by the lailway which it would have been a great deal more economical to have given away, and located the station with reference to the least operating cost. The saving thus effected would, in many cases, pay interest on the investment on both properties. The location of the power station near the car house reduces in many ways the labor item, and the insurance hazard will not be increased if care be taken in the design. Another point to be considered is the liability for damages due to smoke nuisance, pollution of streams and external fire hazards. FOUNDATIONS. In locating the power station building the character of the soil and subsoil and its effect upon the c6st of proper foundations for building and machinery should also be carefully determined. Every endeavor should be made to dis- cover the character of soil on which the station foundations will rest. Where there are adjacent buildings these can be inspected and data as to the character, 168 ELECTRIC RAILWAY HAND BOOK. depth and weight per square foot of surface, obtained from the builders. Four borings should be made on the different sides of the site by using a post auger. If four borings show at the same depth the same character of soil, it can be as- sumed that there is no great dip to the strata. For those foundations which carry large weights, such as those under the chimney, the ground should be bored to a depth of 20 to 25 ft. This can be done by two men using a lever with a C-in. or 8-in. auger. In soft soils a pipe must first be driven; a 4-in. pipe is a convenient size and a smaller auger can be used to bore the core out of the pipe. EilTcrcnt soils have greatly different bearing power and their safe loads are changed when the soils are wet. The table following gives the bearing power of soils as given by Ira O. Baker. Bearing Power of Soils in Tons per Square Foot. MINIMUM. MAXIMUM. Rock, hard 25 30 Rock, soft 5 10 Clay on thick beds always dry 4 6 Clay on thick beds moderately dry 2 4 Clay, soft 1 2 Gravel and coarse sand well cemented 8 10 Sand compact and well cemented 4 6 Sand, clean and dry 2 4 Quick sand, alluvial soils, etc 0.5 1 There are many peculiarities of soils which should be thoroughly understood before applying the above table, as there is no more important point of the station building than its foundations. The following remarks of a general character will serve as a guide to the street railway engineer in preliminary estimates. Rock, when extending entirely under the building site, makes the best foundation bed. The softer rocks will sustain more weight than the walls resting on them can safely carry. Water is usually met with in substrata rock founda- tions, due to the seeping over them of the surface drainage; outside drains should be thrown around such a site. If the ledge of rock slopes to one side, a tile or stone drain may be built from the lowest to the highest point of the founda- tion footings and the water drained from the rock in this way. When nearly level, a hole should be blasted at the lowest point of the foundations and meas- ures taken to dispose of the drainage water. The surface of the rock on which the foundations rest should be prepared, and all loose and decayed portions of rock in line of the foundation footings should be cut and dressed to a surface. If the rock surface is uneven, the surface should be cut in steps or plane sur- faces: in no case should a wall rest on a sloping surface. All fissures or depres- sions should be carefully cleaned out to the hard rock surface and filled with cement to the level of the adjacent step. If deep cavities or fissures appear too large to fill, they may be bridged by arches of brick, stone or cement. In rock it is important that the different footings around the foundation of the buildings be as nearly on a level as possible. Where the building is to rest partly on rock and partly on soil the footings on the soil should be made very wide so the pressure per square foot will not be enough to cause an unequal settling of the foundations. These conditions of unequal sustaining power of the foundation bed should be avoided if possible as it is risky at best. Clay. This designation covers soil conditions varying from slate and shale to soft, damp material, which will squeeze out in every direction when pressure It brought to bear upon it. Clay soils which can be kept dry and compact carry ELECTRIC RAIL WA Y HAND BOOK. ^ the usual loads without trouble, but clay as a rule gives more trouble than sanely gravel or stone. The top footing in this case must be carried below the frost line, which "varies from 6 ft. in'Xorthein States to 2 ft. below the Kason & Pixon line. Freezing ailects clay more than other soils, so that the drainage of foundations on this character of soil must be taken care of, and when on a slope the founda- tions are considered hazardous. If the clay contains coarse gravel or stone its retaining power is greatly increased. Gravel gives less trouble than any other material for foundation bed if the foundations are properly proportioned, and it is not affected by water provided the gravel and sand cannot wash away from under the foundation footings. Sand, if confined from lateral movement, makes an excellent foundation. If no water can move it, it is practically incompressible. Dry, clean river sand sc confined has been known to carry 100 tons to the square foot. All foundation footings in sand should be carried to the same level, and when the engine and boiler foundations are separate from the building structure, the building foot- ings should be carried below the bed of any internal foundation, thus making the building wall the retaining wall for the internal foundation stress on the sand. Soils Containing- Vegetable Matter. No foundation should be laid on soils containing vegetable matter or land that has been filled. The original virgin bed of the soil should be reached unless the filling is made of clean beach sand, that has been made compact by drenching with water as it was filled; and this should be treated and retained as required for foundations on sand. Mud or Silt can only be used by extending the foundation area on the sur- face by spreading the footings on wooden or steel beams, by sinking beams or pillars until hard soil is reached, or by driving piles distributed over the founda- tion bed so as to take the weight of the structure uniformly. Some Data on Soils.— The Capitol at Albany rests on blue clay containing 60 per cent to 90 per cent alumina, the remainder being fine sand containing 40 per cent water. The safe load was taken at 2 tons per square foot; a load of 5.9 tons per square foot produced an upheaval of surrounding earth. The Congressional Library at Washington, D. C, rests on yellow clay mixed with sand; 13}^ tons were required to produce settlement and the footing was proportioned for 2^ tons per square foot. Hard indurated clay under the piers of the bridge across the Ohio River at Point Pleasant, West Virginia, carries 2% tons per square foot. The Cincinnati Bridge foundation bed is of coarse gravel 12 ft. below water and carries 4 tons per square foot. The Brooklyn Bridge foundations are 44 ft. below bed of river and rest on bed rock and a laj'er of sand, 2 ft. thick. This material resists a maximum pres- sure of 5^ tons per square foot. Methods of Testing Foundations.— One method suggested is to con- struct a platform about 4 ft. square with 4 legs each 8 ins. square, the platform being set on the bottom of the foundation trench and carefully leveled. A level should be set up so that the levels on each leg of the bench are taken, a level is also taken to a bench mark on a stake or some fixed point. A uniform load is then gradually placed over the platform until a settling is noticed. From one- fifth to one-half of the load that produced the settling of the platform can be allowed on the foundations, the proper factor of safety being governed by cir- cumstance. When the footings for foundations are soft or treacherous, piles are largely 170 ELECTRIC RAILWAY HAND BOOK. used in station construction. The following data is taken from Kidder's "Build- ing Construction and Superintendence. 1 ' Piles should always be driven with small ends down, with all bark and branches trimmed close. The pile driver should strike the pile squarely on i is head. The usual weight of hammers are from 1200 to 1500 lbs., the hammer fall- ing from 5 to 20 ft., the last blows being given with quick strokes in succession and not over 5 ft. fall. Do not continue to drive piles when they sink only 1% ins. under five blows of a 1200 lb. hammer falling 15 ft. Bearing Value of Piles. Character of Soil. Pile Length Average Penetra- Load in in Feet. Diameter. tion. Tons. Silt 40 10 6 234 Mud 30 8 2 6 Soft earth with boulder or logs... 30 8 1% 7 Moderately firm earth or clay with boulder or logs 30 8 1 9 Soft earth or clay 30 10 1 9 Quicksand 30 8 H 12 Firm earth 30 8 y& 12 Firm earth into sand or gravel .... 20 8 34 14 Firm earth to rock 20 8 20 Sand 20 8 20 Gravel.., 15 8 20 "When the penetration is less than that given above for soft soils the safe load may be increased. Foundation Courses.— The foundation may be either stone, brick or concrete as local prices and character of foundation dictates. A lower course of Fig. 151.— concrete footing. :,.• / CONOfeTE; £e^#3 I-BEAM #--' Fig. 152.— concrete foundation. concrete having a spreading base, like Fig. 151 gives a good bearing for any superstructure, as the concrete, when well tamped, conforms to the contour of the earth, and as large stone should be used for the footing courses, concrete makes the cheaper foundation. Concrete foundations as a whole are largely used in power station construc- tion, and where the supporting power of the soil varies, I-beams or rails are imbedded in the concrete to bridge the inequalities of bearing surface, Fig. 152. The piles should in no case be driven closer than 2 ft. on centers. The usual spacing is 2 ft. 6 ins. across the foundation trench, and 3 ft. along the line of wall. ELECTRIC RAILWAY HAND BOOK. 171 1 The capping of piles may be of cement, see Fig. 153, well tamped with concrete. The earth should be excavated 1 ft. below the top of piles and 1 ft. outside of them, the space around and between them being filled with a rich Portland cement deposited in layers. Piles should always be surfaced below the water mark, where water stands on the foundation, to prevent decay, and in no case should piles be used in dry soil. Materials for Foundations. — The materials used in foundation construc- tion are lime, cement, sand, broken stone, brick and building stones; these mater- ials are specified as to quality and the proportions which will be used in making mortar, cements and concretes. Lime. — Common lime, sometimes called quick lime or caustic lime, is pro- duced from limestones by heating to redness or calcination. These vary in compo- sition in different parts of the country. Good lime should show the following char- acteristics: Entire freedom from cinders and clinkers, and the other impurities should not exceed 10 per cent; it should be in hard lumps with little dust; it STONE CAPPING ^___ sl s j^Sr^r 1 jf^%-s!- _s->'i\ y PILES CONCRETE Fig. 153.— pile foundation. should slack freely in water forming a fairly smooth paste with very little or no residue ; it should dissolve in rain water. Hydraulic lime should harden under water after it has been made in a cake and has commenced to stifi!cn in air. Natural Cements.— Xatural cements arc made from natural rock composed of carbonate of lime, carb'onate of magnesia and clay. Care is required in selecting the stone, calcining to the proper degree and inspecting after calcination; it is then finely ground between mill stones. The natural cements are very quick in setting, but have less ultimate strength than the artificial, or Portland; they attain their full strength sooner, and are sufficiently strong for ordinary build- ing operations; they cost less than Portland and are used almost exclusively for cement mortar. They weigh less than Portland, being about two-thirds as heavy. The locations where natural cements are made on a large scale are Roscndale, N. Y., Louisville, Ky., Utica, N. Y., La Salle, 111., Milwaukee, Wis., Maukato, Minn., Cement, Ga. and Fort Scott, Kan. In Eosendale cement a light color usually indicates an inferior, underburnt rock. Eosendale varies in weight from 49 to 5G lbs. per cubic foot or CO to 70 lbs. per bushel. Artificial Cements. — The artificial cements are usually known as Portland, and require a homogenous mixture in the proper proportions of carbonate of lime, alumina, silica and iron; this mixture is subjected to heat sufficient to pro- duce a vitrified, dense and hard clinker and is afterward ground to powder. The American Portland cements have been used in the largest engineering works in this country. Good Portland cement is slow setting in comparison with the 172 ELECTRIC RAILWA Y HAND BOOK. natural cements, and in setting forms a crystalline structure similar to the natural zeolites. Portland costs approximately three times as much as Rosendale, but its strength makes its use in stone or brick foundation footings necessary, as it carries loads of over 1^ tons per foot, and should be preferred for these uses over any of the natural cements. Mortars.— Mortars are made of lime slacked in a water-tight box; water is then added. Different limes take different volumes of water. The water is rapidly absorbed, and with a rise in temperature the ultimate volume of slacked lime is about S}4 times the original lime; sand is then mixed with this, the pro- portions being about one part of lime to five of sand. Rich mortar contains a larger proportion of lime than above. Mortar of good quality^ si ides readily from the trowel; if it sticks there is too much sand in its composition. The nsual practice is to mix the sand with the lime as soon as it is slacked, and let it stand until ready for use. Better results are obtained if the sand is not mixed with the slacked lime until the mortar is needed. Sand.— Sharp sand should show angular formation of granules of various sizes. If there is any doubt as to the cleanness, the sand can be tested by putting some in a tumbler of water. If there are any impurities present they will rise to the top. On squeezing moist sand in the hand the sand should fall loosely down; if it retains the impression of the hand it should be rejected, as it contains loam which greatly weakens the binding power of the mortar. Cement Mortar.— This should be used for all work which is exposed to damp- ness or to the weather. The sand and cement are thoronghly mixed dry, and water added and mixed until the proper consistency is reached. This mortar works better when stiff. For natural cements it should not exceed 3 parts of sand to 1 of cement; for structures bearing heavy weights 2 to 1 should be used; Portland cement can be used in the proportion of 4 or 3 to 1, for first class mortars. For foundations under water a greater ratio than 2 to 1 should not be used. "When a cheaper cement will answer, slacked lime is added instead of more sand. Kidder gives the following estimates: Lime Mortar: 1 barrel of lime weighs 270 lbs., a bushel of lime weighs 75 lbs., 1 barrel of lime equals 3 bushels; 1 cu. yd. of sand will make i yd. of 1 to 3 lime mortar and will lay about 80 cu. ft. of rough brick work or common rubble. Cement Mortar: 1.8 barrels, 540 lbs. of natural cement, and .94 cu. yds. of sand will make 1 cu. yd. of 1 to 3 mortar; 2 lbs. of Portland cement and .94 cu. yds. will make 1 cu. yd. of 1 to 3 mortar; 1 cu. yd of mortar will lay from 67 to 80 c.u. ft. of brick work or rough rubble, and from 90 to 108 cu. ft. of brickwork with >g-in. to 34 in. joints. A cubic foot of brick work contains about 18 bricks. The following safe crushing strength of mortars per sq. ft. is usually divided by 8 for safe loads. Portland cement mortar, 1 to 8 3 months, 40 tons 1 year, 65 tons Rosendale »• •• " " " " 13 •« " " 26 " Lime mortar " •• " " 8.6 " " " 15 '• Lime mortar should not be used under piers that are to receive their full load within six months. Grout is a very thin mixture of cement mortar used to fill interstices in stone work, and usually poured on the courses of masonry, or run between stones to be bonded. Concrete. — This consists of cement mortar to which is added crushed stone. Granite and other hard stones make the best aggregates. It is essential that the ELECTRIC RAILWAY HAND BOOK. 173 crushed stone should be free from dirt. The sizes vary from those that will pass through a 1-in. ring up to the size of a hens egg. Clean gravel is also used largely in some sections. The usual way that the proportioning of parts is ac- complished is by the wheelbarrow load where the mixing is done by hand, one barrel of cement being taken in two barrow loads. The materials are dumped on a water-tight platform; the sand is first spread and then the cement is laid over it; these two materials are thoroughly mixed and on this mixture is dumped the broken stone, which is mixed in dry; then water is added, still continuing the mixing, until all portions are thoroughly coated. Many machines have been devised for mixing cement to save labor. .The Pittsburgh power station was built largely of concrete using an automatic mixer. The proportion of the parts vary with the size of the broken stone used, and the crushing strain on the concrete structure. There should always be enough cement to fill all voids in the stone. Concrete for foundations, bearing only a moderate weight, can be made of 1 part natural cement, 2 parts sand, and 4 parts gravel or broken stone. Portland concrete to take heavy weights should have 2 parts cement, 5 parts sand and 9 parts broken stone. Where a larger proportion of stone is used, the cement should be carefully tested as the building progresses, and close inspection is necessary to see that the proportions are carefully maintained. The concrete is delivered to the foundation trenches, which may be cut out in clay, and the concrete rammed into position; when sand or yielding earth com- pose the trenches, wooden cribs have to be constructed against which the con- crete is rammed to give the size and shape of the foundation required. The layers should not be more than 6 ins. thick, and the concrete should not be dropped from a greater height than 4 ft; each layer is to be rammed with a wooden, 20-lb. rammer until the top surface shows a flush of water and all inter- stices are completely filled. General Remarks. — The strains that Portland cement can take are from 1 to 5 tons per square foot; natural cement concrete, 1 to 6 tons per square foot. "Where the proportions are 1 part cement, 3 parts sand, 5 broken stone, size not exceeding 2 X 1]4 X 3 ins., one barrel of cement will make 22 to 24 cu. ft. of concrete. Concrete : 1 part of cement, 2]4 of sand, 3 of gravel and 5 of broken fetone, yields 1.18 yds. of concrete per barrel of cement. With labor at $2.00 per day, mixing and depositing should not exceed $1.00 per cubic yard. The cost of Port- land concrete will vary from $6.00 to $8.00 per cubic yard. STATION WALLS. These may be of brick, stone or concrete, iron and brick, terra cotta with iron beams, iron beams with concrete walls, wooden post and siding of novelty or shingle. The choice of material depends upon the cost and character of structure required. The walls of a station over the engine room and boiler room do not have to carry more than the roof weights unless the crane, steam piping or office floors are above. The prevalent form of brick station structure is to build brick piers or buttresses, the distances between centers being the same as between the centers of the roof trusses. Between the piers are thin curtain walls of brick, largely taken up by windows. These piers are bonded at the top by girders, The thickness of continuous brick Walls are often fixed by ordinance; nearly all building regulations requiring approximately the following thickness: lor buildings carrying heavy floor weights, two stories, brick 16 ins., stone 20 174 ELECTRIC RAILWAY HAND BOOK. ins.; three stories, brick 20 ins., stone 24 ins.; four stories, brick 24 ins., stone 28 ins. Station walls of cut stone may be constructed either with a plain face or but- tresses, the space between buttresses as a rule being arched for window open- ings, and the surface stone work and facings being left to the taste of the architect. Brick, either red or terra cotta or glazed, or concrete can form the external walls. The structure is sometimes supported by the buttresses of masonry, and in some forms of* construction contains within the brick walls, columns or pillars, which carry the roof load. This latter construction is used where space has to be economized, and where the roof weights to be carried would require larger foundation areas than could be well distributed by a masonry buttress. The building may be an iron or steel skeleton with thin double walls of brick, terra cotta or even concrete. In some cases a single brick wall faced on the inside with the numerous forms of compositions for ceiling and interior work is used. For temporary work corrugated galvanized iron or tin, or galvanized iron stamped with brick tiling, having an inside sheathing of wood or asbestos mill- board has been found satisfactory. For further protection the space between the inner and outer walls, when wooden, can be filled with dry clean cinders or mineral wool. In wooden structures the walls can be of novelty siding or dipped shingles laid over 1-in. spruce boarding. The inside wall can be of adamantine plastering laid on metallic or wooden lathes, the spaces between being filled with mineral wool or cinders to make a slow burning structure. Another form is to have all timber dressed; on the outside are nailed 2 in. hemlock planks dressed on the inside; over this the building paper and novelty siding or shingles are nailed. Eight feet can be carried between the posts. This character of construction reduces the insurance rates on wooden buildings, and makes what is known as the " slow burning " construction. The weights that the buttresses, pillars or struts have to bear in a power station engine room are usually the roof weights, and the moving crane and load. The distance between the spans varies from 6 ft. to 20 ft. depending upon the character of the roof trussing employed. In the smaller stations it is much more economical to use horses or cranes tracked on the floor than to strengthen the roof truss for rigging machinery. For strengths of building materials see tables on pages 11 to 15. ROOFS. Weights and strains thrown on the roof are due to wind pressure and snow. The wind pressure allowable depends upon exposure of building; 32 lbs. per square foot should not exceed the ultimate strength of the structure. Pressure of "Winds on Roof. (Unwin). a = Angle of surface of roof with direction of wind. F — Force of wind in pounds per square foot. A — Pressure normal to surface of roof. B — Pressure perpendicular to direction of wind. C = Pressure parallel to direction of wind. Angle of roof = a 5° 10° 20° 30° 40° 50° 60° 70° 80° 90° A=FX .125 .24 .45 .66 .83 .95 1.00 1.02 1.01 100 £ = FX .122 .24 .42 .57 .64 .61 .50 .85 .17 .00 C =FX .01 .04 .15 .33 .53 .73 .85 .9(5 .99 1.00 ELECTRIC RAILWAY HAND BOOK. Angles of Roof s as Commonly Usedo Proportion Angle Length of of Rise to Rafter Span Beg. Min. to Rise. % 45 .... 1.4142 H__ 33 41 m 1.8028 2/3 30 .... 2.0000 M 26 34 2.2361 1-5 21 48 2.6926 1-6 18 26 3.16.3 Velocity and Pressure of Winds. (Hurst.) Velocity in Miles Pressure in Lbs. Designation. per Hour. per Sq. Ft. V. P. Scarcely perceptible 1 .005 Perceptible 2 . c 020 Slight breeze. 4 .080 Moderate breeze 8 .320 Fresh breeze 15 1.125 Brisk wind 25 8.125 Strong wind 80 4.50 High wind 40 8.00 Storm 50 12.50 Violent storm 60 18.00 Hurricane 80 32.00 Violent hurricane 100 50.00 Gust observed at Liverpool Observatory iu 1868 , . . 126 80.00 The weights of the different kinds of roofing are as follows: Lbs. per Sq. Ft. Cast iron plates 15 Copper 8 to 1.25 Felt and asphalt 1 Felt and gravel 8 to 10 Iron, corrugated 1 to 3.75 Corrugated sheets, unboarded , 8 Iron galvanized, flat 1 to 3.50 Sheathing, pine 1 in. thick, yellow 3 to 4 Shingles on lathes 10 Spruce, 1 in . thick 2 Spruce, if plastered below rafters 12 Sheathing, 1 in. chestnut or maple 4 Slate on lathes 13 Slate on boards 1% in. thick 16 Sheet iron ^ in. thick 3 Sheet iron and lathes - 5 Skvlights, glass ^ in. to ^ in 2.50 to •}' Sheet lead 5 to 8 Tin 7 to 1.25 Tiles, flat 15 to 20 Tiles, grooved and fillets 7 to 10 Tiles, pan 10 Zinc 1 to 2 For spans over 75 ft. add 4 lbs. per square foot to the above loads. Snow weighs 5 lbs. to 12 lbs. per cubic foot depending upon the humidity of the atmosphere; 1 cu. ft. of snow compacted by rain weighs 15 lbs. to 50 lbs. It is customary to add 30 lbs. per square foot to the above for snow and wind when separate calculations are not made. The weight of any load upon a roof is taken as unvrormly distributed over the surface of the roof. The total weight on each pair of rafters, couple or truss, is equal to the sum of the weights of the truss itself, and as much of the roof as is carried between two trusses. i 176 ELECTRIC RAILWA V HAND BOOK. Safe Loads, in Tons of 3,000 L.bs. for Hollow Cylindrical Cast Iron Columns. i OCT* LENGTH OF COLUMNS IN FEET DO H .2-2 >s. of sper ngth ■d t « ft M PS « F IM 61) o 4> ft Ph QQ Tjt CO •O OS W N ri H t-4 rl6 M 35 THCO tH 0* JO O *>T* ©* o o> • . . .... ^ ooo TT CO 0* a 1-1 T-l T-t y-i © . .... thco C* jo iq os-** th oo ^d^cqt-j co oi c* oi th T-5i-4T-ir-i COJO CO i> OS ffl 005 CO 00 id U^co co oi JO C<> OS i> jo n< co ■ OJoirHT-i THi-ir-i OSt* 00 tH jo oh »o c* os COO* TH T-t J> CO JO »q OS xj5 CO TjJOt-TlJ TH©St>0 COCOOioi ©J r-I tH -r-J £3 JO OS tH b- » d o w C* (M TH TH CO iOi-H tH ©* © o6i> «o id «oqw« tf ^ co co OS CD CO T-t cioioioi •;99j m e^ioddiis U99AV -:jaq 90irB r si(i JOCO N 00 OS © H« CO "rf JO«O*>00 OJOi-i(M T-( t-(H r-1 i-t rlrtrlrt tH frl Oi Gi CO OS 00 OS OJ o> i-^ «d jo id coT-ji>cq © Ti* ^ CO CO CO b-jqcoTH © oqi>«qjq ©i©*oi©i ©* t-5t-5t-Jth K. THCOOSt- 00 ojoodi> so O"<^oor)j os «qcqooq eq tjjo?thos ©* os jo co^o oo OStHCOI* ©I NWON TjJ iHOSl>jq i>i>cdid id n^^co co cooioioi o? « COrO cq«qjqb» co TMT-jcq«q OOJOCOtH o OSCOi>0 o iqocqw os ©mh© «d ididTj5Tj5 oo cococooi T* t-ooo o CO OS t>Tf CO Rrlnrt rl jO CO CO CO thOOSOO jo oscqoocq os «qcoost-. fc-i (DSOiQta Tl* "^TfCOCO th 5 OSOOJOOS t* osnj--o th cq«qOTjj wwh'6 d odt>i>?d OS JOt-iGOiO jo id id ^ ^ osi>^co «o riHjqOjq tij Tijjq^TH i>id"Tji'oiTH oosoood ^ osttjo «q i> «d«dcDo o m tjjo?o?o? os o? din i Tf CO CO I thOSOSCO 00 «O«O«O00 O Tj.T-« •}90J Ul ejioddns uaoAV -pq 80UB T 9I(I JO O JO GO CO b.oqjqt>. co cqjqosjq co ththcojo jOTHCOidco r^osNO jo "^cooi"-* COCO^VC^OJ WnnH rl T-l— (r-.^-. O? CO Tf JO CO N00OO H 0?CO'^0«o!lO£ gives the boiler horse-power. The common heat unit is the British Thermal Unit, known as B. T. U., and is that quantity of heat which 13 required to raise the temperature of one pound of cold water one degree Fahrenheit. The boiler horse-power is the evaporation of 34}| ins.-% in. 80.72 f. c, 11.67 a.; pea, % in. -% in. 79.05 f. c, 14.66 a.; buckwheat, y 2 in.-*4 in. 7G.02 f. c, 16.62 a. The semi-anthracite and semi-bituminous coals contain 12^ per cent to 25 percent volatile matter in the combustible, usually run low in moisture, ash and sulphur, and have high heating value per pound of combustible. These form the best steaming coals in the United States, and can be burned at a higher rate on a grate surface without clinker. Bituminous coals contain 25 per cent to 50 per cent volatile matter, and vary considerably with the coal-bearing areas in the United States, west of the Alle- ghany Mts. In a general way the volatile matter increases as they go westward and northward of the Alleghany Mts. The percentage of moisture also increases as they go westward ranging from 2 per cent near Pittsburg to 14 per cent in some of the Illinois coals. From a chemical analysis of coal its heating value may be calculated within a limit of error of 2 per cent by the application of DeLong's Formula which follows. Here C stands for carbon, H for hydrogen, O for oxygen, and S for sulphur in the coal. The heat units per pound = .01 T 14,600 C -f 62,000 f H — — \ -f 4000 S 1 The proximate analysis of coal is also largely used, giving the volatile matter, fixed carbon and ash in the coal. The probable heating value can be figured within an error of 3 per cent by the use of the following table: APPROXIMATE HEATING VALUE OF THE COMBUSTIBLE PORTION OF COAI*. Composition. Heating Value per lb. Equivalent Water Evaporated from and Fixed Carbon. Volatile Matter. Heat Units. at 212 degs. per lb. of Combustible. 97 94 90 3 6 10 14,940 15,210 15,480 15.47 15.76 16.03 87 80 72 • 13 20 28 15,660 15.840 15,660 16.21 16.40 16.21 68 63 60 32 37 40 15,480 15,120 14,760 16.03 15.65 15.28 57 55 53 51 43 45 47 49 14,220 13,860 13,320 12,420 14.73 14.35 13.79 12.86 ELECTRIC KAILWA Y HAND BOOK. 191 The practical way that operators of central stations can exactly determine this matter for themselves, under their own conditions, is to secure sufficient coal for several days' run and burn this coal under average practical conditions, thus finding the output in watts for the weight of coal or the cost of coal per kw output. In changing from one coal to another, especially in the case of the different sizes of pea coal or from hard to soft coal, the grate or furnace may not be prop- erly constructed to utilize to the best advantage the heat units in the new coal. The unconsumed carbon in the ash, the temperature of the uptake, the smoke Issuing from the chimney and the draft should all be taken into consideration when testing a new coal in order that the test be carried on under conditions giving most accurate results. The fireman is often puzzled at first to get the best results from a change in the grade of coal. Unless these points are carefully watched in a coal test*of this kind, a coal which has capabilities of producing greater output for the same cost may not have a fair trial; and this is especially so in changing from anthracite to bituminous coal, as the furnace for one is unsuitable for the other, these two coals requiring considerable difference in fur- nace construction. Below is a table giving heating values of some well-known coals in heat units per pound of combustible and the equivalent, evaporation from and at 212 degs. Fahr. HEATING VALUES AND EQUIVALENT EVAPORATION OF VARIOUS KINDS OF COAL. Anthracite, Pa Semi-anthracite, Loyalsock and Bernice, Pa.. Semi-bituminous, Broad Top. Clearfield, Cam- bria, and Somerset, Pa.; .Cumberland, Md., New River, W, Va., and Pocahontas, Va. Close average Bituminous, Connellsville, Pa Youghiogheny, Pa Jefferson, Pa Pittsburg, Pa Brier Hill, Ohio Upper Freeport Seam, Pa. & O Vanderpool, Ky Middle Kittanning Seam, Pa Thacker, W. Va Jackson Co., O Hocking Valley, O Big Muddv, 111 Streator, 111 Mt. Ollive, 111 Lignites, la., Wyoming, Utah, Oregon >- Heat Units per pound evaporation from and at combustible. 212°. 14,900 15.42 15,500 16.05 15,700 min. 15,800 max. 15,750 aver. 16.30 15,300 15.84 15,000 15.53 15,200 14,800 15.32 14 300 14,800 15.32 14,400 14,500 15.01 15,200 15.74 14,600 15.11 14,200 14.70 14,700 15,22 14,300 14.80 13,800 ' 14.29 11,000 11.39 to to 12,900 13.35 With a boiler properly designed in all its proportions it is found that the maximum economy is obtained when the boiler is driven at an equivalent rate of evaporation of three pounds of water from and at 212 degs. per hour per square foot of heating surface. When the evaporation falls below this rate, the 192 ELECTRIC RAILWA Y HAND ROOK. radiation and other boiler losses rapidly grow in the percentage of the output as they are in a large degree constants. A3 a rule, the economy falls with a rapid rate of driving the boiler, yet among the di£erent boilers, changes in the economy for the different rates of evapora- tion even with the same kind of coal, will not follow the same curve between the water evaporated per square foot of heating surface and the pounds of water evaporated per pound of coal. It may, however, be expected that the evaporation per pound of coal will be approximately IS per cent less when the evaporation is forced up to 6 lbs. of water per square foot of heating surface than when it is at the average rate of maximum economy of 3 lbs. per square foot of heating surf ace. But the above general rule must be used with considerable allowance. A Babcock & Wilcox boiler in the power station of the Hestonvillc, Mantua & Fairmount Park R. H.> Philadelphia, Pa., shows an evaporation of 6.34 lbs. of water per square foot of heating surface, and the water evaporated per pound of combusti- ble was 10.97; on the Staten Island Electric E. It. Co.. New Brighton, Staten Island, the same type of boiler had an evaporation of 2. 06 lbs. of water per square foot of heating surface with 11.78 lbs. of water per pound of combustible. The ratio of heating surf ace to grate surface in the first case was 46.87, and in the Second, 60.28. The grate surface of a boiler can be roughly assumed to be ^ sq. ft. per hp- hour. The table given herewith of several types of boilers shows the grate sur- face used by different makers per hp-hour; it will be noticed that the larger the boiler unit, the smaller the surface required to produce a hp-houn The heating surface figure Is given as 11.5 sq. ft. per hp-hour. GRATE SURFACE PER HORSE-POWER HOUR. Type op Boiler. Hobsb-power op Boiler. HEINE. BABCOCK & WILCOX. CLONBROCK. Grate Surface per hp-hour. Grate Surface per hp-hour. Grate Surface per hp-hour. 150 200 250 800 860 400 600 1000 .1S6 .185 .176 .182 .164 .244 .222 .201 .32 .25 .24 .213 .1925 .183 .171 The ratio ol the heating surface to the grate surface varies largely in boilers of different types. Reports of tests on some thirty Babcock & Wilcox hollers show a variation in ratio of from 62 to 37: 1 of the heating surface to the grate surface. The Stirling boiler shows a variation of 76:1 on a 600 hp. boiler, 44.7: 1 on a 250 hp. boiler, and 37: 1 on a 125 hp. boiler; the Climax varies from 61 to 33:1 ratio. For further data and methods of conducting boiler tests see pages 77 to 88. ELECTRIC RAILWA Y HAND BOOK. I93 TYPES OF BOtLEKS. The Fire-Tube Boiler. — This boiler is manufactured largely throughout the country by boiler manufacturers, the fchell ranging from 73 ins. to 51 ins. in diameter. The table of dimensions below gives the practical, standard pre portions for cne of these boilers. Following is a report of a test of horizontal return tubular boiler for New Bedford & Fair Haven Traction Co., New Bedford Mass. Test made by H. L. Butler, M. E., of Wm. Sellers & Co., Philadelphia. Dimensions of Boilers on which test was made* Diameter of Shell 72 Ins. Length of Shell and Tubes 17 ft. 6 ins. Number of Tubes 132-3 ins. diam. Area of Healing Surface 1877.68 sq. ft. »' 4k Grate Surface 36 " 4 * 44 through Tubes 5.55 " " Ratio of Heating Surface to Grate Suiface 52 to 1 • 4 " Grate Surface to Tube Area 6.4 to 1 Stack 48 ins. Hia. 90 ft. high above Grates. Boilers set in Pairs. One Boiler tested, the other banked. Report of Test. Manner of Start and Stop Running Kind of Run Continuous Duration 11 Hours Quality of Coal used, "Pocahontas," bituminous, run of mine. Coal Consumed (11 hours) 4556 l^s. 44 44 (per hour) .T. 414.18 lbs. 44 per square foot of Grate per hour 11.5 lbs. ( Ashes not considered, non-combustible reduced to j Percentage of Ash ■< powder, weighing not more than 3 lbs. for whole run, > .00 ( Everything practically consumed. ) Percentage of Moisture in Coal 6.66 per cent Combustible used in (11 hrs) 4252.57 lis. (perhour) 386.59 44 Water Evaporated (11 horns) 52,236.5 " 44 44 (perhour) 4748.77 " 44 ** per square foot of Heating Surface 2.52 44 44 44 per pound Commercial Coal 11.46 " 44 4t 44 44 Combustible 12 28 " 44 44 4t 44 4i from and at 212 degs 13.23 4W Boiler Pressure (averase gage) 130 " Temperature of Water before enterin s Heater 3* dc " 4 * Feed Water before entering Boiler 180.93 l ■ 44 4t Escaping Gases, Maximum 402 u 44 44 4 ' 4 * Minimum 320 4t Percentage of Moisture in Steam 1.3 por cent. The heating surface in different types of this boiler, varies somewhat with the setting; but the test by n. L. Butler on this type of boiler, manufactured by the Harrisburg Foundry a c3 5 •-H .£ o3 ft ft -*j •«-» u B t— t O H < 0) 48 12 56 9 23 42 1°2 24 24 48 9 4 48 48 50 14 . r 3 9 23 b-b- » 3 « M g H O m « 5 ^ o o 88 88 8 CM CM (MM Stl OO OO OOO 00 OO OOO 5DN NCO OOO CM CM cmgm coooco g 88 88 88 .a § be 3 d ats ass •^ (OO O r-« OO 3^ 00 00 000 4j WJO CO '"i* ^ CM "*f CO COCO CO CO ^" c 00 00 00 00 OO OOO £ cm cm ©1 ex NCO CSO y-> CM Tt« T-t CM CM CM CM ■Saoi: •uit?t(I a C l>i> l>t* N^f SiSi CO TO CMCM N^< — « CM ^N i>b-t> CM CM CM bt 7^-h* OOO "^00 BD CO O O CM CO CO CCCO'f HH i-irt i-HCM t-iCM CMCM CMCMCM •Siioi 00CO COCO GOO COCO OTO OCOOCO •qSiH Od CO CR o> CS N ca 00 OO XCCi '3PL& •jnoH J^d hot* •ttiodsAg ranraitrjj\[ ON i>t>- CO "t Ci CM S* CM coo COS0 00 OS CM ON O O O QD CM lO 00 00 OOO no o o . o 00 rf N T-t N CO CCO-* ■^lO OO ONN •I«u;ov 'd *H S3 §5o1 O CM OCM CMCM ON OStH ?OCO «CMOl 9 J 'jRS ^S S8 sa M fe«* t3 a 00 101a O"* N^ TfT>« OOO & 5— ON NN COO COO 1-t 00 rHrlH J 198 ELECTRIC RAILWA Y HAND BOOK. The boiler is suspended at front and rear from wrought iron supporting frames, entirely independent of the setting, to allow for contraction and expan- sion without straining either the boiler or the brickwork; and to allow of repairs to, or renewal of, the latter without disturbing the boiler or its connections. Steam is taken from a dry pipe perforated on top side and connected into a flanged steam opening on top of drum. When two or more steam drums are used on one boiler, the drums are connected by a balance pipe upon which the safety valves are mounted, and a cross pipe for main steam valve. iTlG, 169— BABCOCK & WILCOX BOILER. The hand holes are closed on the outside by a cap which Is held in position by a forced steel clamp closing the hand hole opening from the inside and secured by a bolt. The mud drum is 12 ins. in diameter and of proper length to be connected with all of the sections in the boiler by means of nipples expanded into scats. It i 3 tapped for blow-oil connections on its rear side, and furnished with hand-holes for cleaning. The space occupied by a boiler of 250 hp. is shown in Fig. 169, and the sizes of a boiler of the W. I. F. typo from 1C3 hp. to 213 hp. are given in the. table on page 176. These dimensions should only bo used for approximately locating these boilers; a space should be left of at least 19 ft. in front of the boiler so the tubes can bo withdrawn in case of repairs. The table on the following page gives a test on two Babcock & Wilcox boiler8 taken from their catalogue on railway plants. ELECTRIC RAILWAY HAND BOOK 199 OWNER OF PLANT WHERE TEST WAS MADE. Engineer conducting test Kind of coal Bate puration, hours Coal burned per sq. ft. of grate p-. r hour Water evaporated persq. It. of heating surface per hour from and at 212 degs Water evaporated from and at 212 degs. per pound of coal Water evaporated per pound combustible from and at 212 degs Per cent of refuse iu coal , Draft Temperature of flue gases Per cent of moisture in steam Ratio of heating s urf ace to grate Hestonville, Mantua & Fair- mount Pk.R.R. Phila., Pa. J. J. BeKinder Henrietta, Pa. Mar. 29, 1895 9.50 6.34 10.28 10.97 6.32 .87 750° .12 46.87 Staten Island Elec. R. R. Co. New Brighton, Staten Island. W. N. Sheaff Victor, Pa. May, 23, 1897 10 15.17 2.66 10.70 11.78 9.13 .48 485° "66.'28 # J7-9 iJLL / / ' / .. //////// /////i ^*~<7/A/ // //////// ..__ /S - 3 »| FlS. J70— STIRLING WATER TUBS BOHJSB. 100 ELECTRIC RAILWAY HAND BOOK. Stirling "Water-Tube Safety Boiler,— The boiler consists, briefly, of three nppcr or steam drums, and one lower or mud drum, all connected together by means of tubes, which are bent slightly, so as to allow them to enter the drums normal to their surf aces. See Fig. 170. All the upper or steam drums are connected by steam circulating tubes, but the front and middle drums only are connected by water circulating tubes. The tubes used are 3J4 i ns « in diameter, and are made of lap-welded, mild steel. The circulation of the hot water is between the two forward drums and the wer mud drum, the feed water being introduced into the third drum. See Fig. Blow off Fig, 171— circulation in Stirling boiler. 171 tor path of circulating water. The equalizer pipes connecting the drums introduce steam into the middle drum, which is used as a heater for the numerous pipO sections. In order to produce dry steam the end of the steam pipe terminates in a/ T inside the middle drum to which are secured pipes, near the top of the drum and paralleling it, in which are cut slots in order to throttle the steam, so TABLE OF STIRLING BOILERS. Type. Height. Depth. "Width for Each HP. Class A 18' - 9" 16' - 0" .056' B 12' - 0" 14' - 0" .087' E 15' - 2" 16' - 8" .005' F 20 - 7" 10' - 9" .043' G 20' - 5^" 19' - 7" .056' H 18' - 3" 37' - 6" .051' I 21' - 3" 19' - 6" .056' K 21' -10" 17' - 7" .04 ' L 22' - 4" 18' - 3" .036' ELECTRIC KAILWA Y HAND EOOAT. 201 that dry steam will be produced for sudden steam demands; this is especially used for railway work where the steam demands will be frequently of this nature. In setting this boiler, it is sustained independently of the brickwork surrounding it. Th2 three upper steam drums are supported by wrought iron beams resting on wrought iron columns, while the mud drum is suspended and left free to allow for contraction and expansion. The dimensions of the Stirling boiler are given in the table on page 179. The Stirling boiler is built to meet the varying requirements of height, width and depth and the different types lettered as in the table* Some tests made on these boilers are given below. TESTS ON STIRUNG BOJXERS. Location. Engineer , v Eating . Duration test ; Temperature of feed Evaporation per lb. coal from and at 212 degs Evaporation per lb. combustible from and at 212 degrees Coal consumed per square foot grate per hour Draft Moisture in steam per cent Temp, of esca ping gases Per cent developed above rating. Below rating. . . ,.\ :..... Kind of fuel Lehigh Valley Trac. Co., Alleutowu, Pa. P. Hansen 600 10 191.8 8.47 __ 11.27 9.91 .3 1.5 Buckwheat Union Ry. Co., Provi- dence, R. I. Thos. Evans 250 9.7 66. 10.57 11.631 14.-8 '.75 1.75 Cumberland .Lindell Ry. Co., St. Louis, Mo. W. H. Bryan soo 9 47 7.82 8,92 448 15 .77 .14 Illinois Lump The Morrin Climax "Water-Tube Boiler is manufactured by the Clon» brock Steam Boiler Company. Fig. 172 shows the general construction. It consists of a vertical cylinder and loop-like tubes, which form the principal heating surface, extending the entire length of the boiler. The main cylinder shell, A y is constructed similarly to any cylindrical boiler shell, and is made perfectly steam tight as well as strong enough to resist internal pressure. It is provided on Cop with a manhole plate. The extremities of the loop-like tubes extend and are expanded into the cylinder, A, forming a steam tight con* nection; these tubes re-enter the cylinder about 18 ins. above their initial entraace. As the boiler increases in size, the water and steam spaces increase in the same ratio. In some boilers the feed water has been introduced through a spiral pipe located above the loop tubes, and used as a water heater for the boiler. The fire box is annular in form, and enclosed in a casing of iron, which is bolted together in segmental sections and lined with fire brick. Three or more fire doors are provided for the boiler depending upon its size. According to the data given by the Clonbrock Steam Boiler Company, the heating surface of its boilers evaporates t> lbs. of water per square foct of heating surface per fcour with -a rate of combustion oi 12^ lbs. of coal per square foot of grate per hour. The table herewith ^ives the sizes of these boilers and the spaces they occupy. 202 ELECTRIC RAILWAY HAND BOOK. Fig. 172— climax water tube boiler. ELECTRIC RAILWA Y HAND BOOK. 203 DIMENSIONS OF CLIMAX BOILERS. HP. Diameter. ' Height. Grate Surface. Heating Surface. 100 8' 14' 2" 34 sq. ft. 1025 sq. ft. 125 9' 17' 6" 42 M 1280 " 150 10' 18' 48 " 1675 " 200 10' 3" 19' 10" 50 •* 2000 " 250 10' 11" 21' 6" 60 " 2500 " 300 11' 2" 23' 1" 64 " 3000 " 350 11' 6" 27' 2" 69 " 3475 " 400 12' 27' 6" 77 " 3650 •« 500 12' 8" 31' A%" 89 " 4550 •• 600 14' 33' 6" 110 " 5600 " 800 15' 36' 10" 128 " 7000 " 1000 17' 42' 8' 171 " 9200 M Abendroth & Hoot Boilers. — The construction of this boiler provides for the circulation of the heated gases, not only among the tubes, but around the overhead steam and water drums; the heated gases also circulate around the, cross steam drum, maintaining the temperature of the steam. The general con. struction of this boiler is shown in Fig. 173. The setting is peculiar to this boiler. The entire weight on the front ends of the tubes and drums rests on a swung beam ; the weight is supported from under- neath and the beam is supported by rods attached to the top of the front setting of the boiler. The tubes forming the heating surface are connected together at the ends by a IT-pipe, which joins the ends of two adjacent tubes. Each vertical set of tubes starts from a header, and each pair of vertical tubes enters a separate steam header on top and these various steam headers enter one equal- izing header, as shown in Fig. 173. General Information.— Boiler setting plans, specially designed for the boiler to be installed, are usually furnished by the boiler manufacturer; and these should be adhered to closely, as any deviation from this design may seriously interfere with the guaranteed efficiency. When the boiler has been completely set, and the lime mortar and fire-clay luting has hardened so that a knife blade will not penetrate it more than ^ in., and not sooner than 12 hours after the masons have finished, the boiler should be slowly filled with cold water up to the high water gage ; then a slow fire of shav- ings and wood may be built on the grate surface, covering not more than one- quarter of the grate surface. This should be kept going until the boiler and setting are warmed up; the safety valve cr header, should be wide open during this firing. When steam issues from the boiler, the openings may be closed and the steam pressure gradually raised. A great many boiler settings have been ruined by heating them up to quickly; 48 hours should be required for a 200-hp. water-tube boiler to reach a temperature cf 212 degs. Fahr. Pyrometers for measuring flue temperatures also aid in the proper handling of the dampers and draft, but thi3 temperature of uptake cannot be taken as a criterion of firing economy; for with a slow fire, the flue temperature and effi- ciency may both fall together. Where steam blowers are used, these may reduce the temperature of the uptake, but, due to the additional heat required to raise the temperature of the draft moisture, the temperature may be lower in the up- take, while the heat units lost passing up the chimney may be increased. 204 ELECTRIC RAILWAY HAND BO OUT. Grates are made stationary, consisting usually of V-shaped cast-iron bars; the air space between the bars is generally about 25 per cent of the grate area, but this is less with screenings and coal smaller than pea. The bars are usually placed parallel to the furnace sides so the slicing bar and poker can readily clean between the bars. These grate bars are supported on trancvere beams, which are secured to the sides of the furnace. There are a nuirbsT 02 forms of shaking and dumping grates to reduce the time required to c/.s "mi ine fires, and assist in disengaging the ash from the burning fire bed. Their coi> struction embodies a pivoted section of a grate bar, movable by a handle pir« r'tect ■> Fig. 173— abendroth & root boiler. Ing from the boiler; in some forms they are arranged so that by further pulling out this handle, the grates are tipped sufficiently to dump the fire. The best height for grates is 30 ins. above the floor level. Where hot air is used for draft or in down draft furnaces, these grate bars are hollow and water kept circulating through them to keep them from yielding or melting under the increased temperature. MECHANICAL STOKERS AND TRAVELING GRATES. Here the crate bars are not made continuous and solid, but are cametwci- ward by proper mechanism away from the fire door, where coal is 1 Ctu02nfc»icti&f fed with a variable depth on the traveling grate from a hopper, /.'jta tfpeed .? driving this grate is under control; the coal should be completely ju*&e<. >ciCi" ELECTRIC RAILWA Y HAND BOOK, 205 it reaches the end of its journey through the furnace. Such a grate is nractically self cleaning with a coal that docs not slag badly. Mechanical stokers do not handle well either a very hard variety of anthracite nor a bituminous which calzes and melts badly; but with the exception of these two classes of coal, the mechanical stoker has proven a coal and labor saving device in a number c railway plants. In some of the types of mechanical stokers, the rate of feed-' the coal is automatically controlled by the steam pressure of the boiler. Record of Six Tests to Determine the Comparative Economy of Itoney Mechanical Stoker and Hand Firing on Hartford Retri Tubular Boilers, 60 ins, x 20 ft. Burning Cumberland Coal wit Natural Draft. Bating: of Boiler at 12.5 Square Feet. 105 HP. Note.— The same man fired on all six tests. First three tests, hand fired ; last three tests, stoker. HAND FIRED. Temp, of Feed Water Degs. Fahr. Total Total Evapora Evapora- HP. de- Duration Steam Coal plus Water tiou per tion per veloped of Pressure. Wood at Evapor- l\>. dry lb. dry above Test. Lbs. 40%. ated. Coal, Coal from rating of Hours. Lbs. Lbs. actual. & at 212°. Boiler. Lbs. Lbs. Per Cent 123.5 145.7 107.5 134,459 1,256,240 9.34 10.36 5.84 132.0 143.2 104.6 135,33S 1,270,753 9.39 10.44 13.52 64.25 152.2 66.1 31,214 310,960 10.02 n.oo 68.00 MECHANICAL STOKER FIRED. 65.5 145.4 63.1 28,121 288,781 10.81 11.89 54.65 64.5 146.0 68.0 29,794 303,887 11.06 12.25 66.68 65.5 145.2 65.2 29,000 320,034 11.35 12.54 84.26 . The American Stoker Company's mechanical stoker burns effectively botn bituminous and anthracite coal. The mechanism is simple and easily operated and under the combustion principles used in its construction, it attains a more economical use of coal than by hand firing. The gases leaving the chimney are totally consumed, leaving no free carbon. . In tests made in Cleveland, Ohio, with mechanical furnaces, the poorest of the cheap coals consumed was 4.93 lbs. per hp-hour. The economy of the cheap coai mechanically fired over that for the high priced coal, hand fired, shows an earning power of capital investment equal to 30 per cent. Data of tests on slack coal hand fired, and like coal mechanically fired show the following results : the economy favored mechanical firing 20 per cent plus the factors of lessened labor, cost in fire room and smokelessness. MANAGEMENT OF BOILERS. . Firing:.— In firing hard coal the grate bars should be such as to allow the least possible amountpf unburnt carbon or coals to drop through before they. 206 ELECTRIC KAIL WA Y HAND BOOK. are fully consumed, and yet sufficient draft area to consume the coal. The most economical firing, where buckwheat, pea or rice coal is handled and where the coal is not very high in ash, is to fire uniformly with a bed of coals not deeper than 3 ins. The skill of the fireman is shown by his ability to maintain the whole fire surface at a uniform color or temperature; dark spots or thin fire in spots indi- cate that he has not the control of the shovel necessary to distribute the coal exactly where required for uniform combustion. A poor fireman will show bad corners where ash is allowed to accumulate, in this way reducing the available combustion surface of the furnace. Dark spots indicate poor combustion or too thick a fire, and open spots will decrease the draft possible to maintain. It requires less skill to fire a heavy fire ranging from 5 ins. to 6 ins. thick, but the draft will be throttled and not sufficient air can pass through the fire bed to obtain total combustion, and the gases will pass off as CO instead of C0 2J not com- bining with the last molecule of oxygen which increases the temperature of combustion considerably. It requires less labor and attention to maintain thick fires than thin ones, but the coal cost is larger for the same amount of water evaporated under the same boiler. It also should be borne in mind that 'a high draft gage does not indicate the best firing conditions, for the more the fire on the furnace throttles the draft, the higher it is possible for the draft gage to show, but the volume of air passing may not be sufficient for complete combustion. A thin fire will re- quire less slicing than a thick one, and there will be found less unconsumed carbon in the ash as a rule. Cleaning the fires ought not to be done oftener than necessary to maintain the best firing conditions, and the fire should be jockied into condition to meet increasing loads on the station, and generally cleaned after heavy loads. From experience in testing and inspection of power stations it is found that there is no point iu the station plant where more money can be saved than in the proper burning of coal. A poor fireman can waste many times his salary per year by lack of intelligent attention to handling the fire, Merit systems have been established in some plants, which rebate to the fireman a percentage of his savings in coal costs. Soft coal firing is generally done by first firing near the door on the dead- plate and allowing the heat of the fire to gradually coke the coal, expelling from it the volatile part of the combustible. This is coked here and gradually shoved back to make room for new coal, and when it reaches the end of the grate it is almost entirely consumed. Lumps larger than 4 ins. in diameter should be broken up in order that this coking process may be properly carried on. Soft coal fires reach their maximum temperature at a more distant point in the draft than hard coal fires. It is necessary for the surfaces, over which the gases pass to be of higher temperature, in order that the unconsumed carbon may be raised to sufficient temperature to combine with the oxygen and not pass out in the form of smoke. The furnace arrangements should be such that the high temperature gases Mall not impinge against portions of the boiler unsuited to withstand them. The smoke passing out of the chimney from a soft coal fire is a criterion of the effi- ciency of its combustion. There are a number of arrangements for the fireman, by which he can determine the completeness of combustion. One is the Gas Composimeter, which indicates the percentage of CO carried off with the gases in the flue, and registers automat- - ELECTRIC RAIL WA Y HAND BOOK. 207 ically the amount of this gas present. The draft and dampers can be regulated to keep the carbon dioxide as high as possible, and aids in the correct firing of the boiler. Care of Boilers. — The following rules are compiled from those issued by various Boiler Insurance Companies in this country and Europe. 1. Safety Valves. Great care should be exercised to see that these valves are ample in size and in working order. Overloading or neglect frequently leads to the most disastrous results. Safety valves should be tried at least once every day to see that they will act freely. 2. Pressure Gage. The steam gage should stand at zero when the pressure is off, and it should show the same pressure as the safety valve when that is blowing off. If not, then one is wrong and the gage should be tested by one known to be correct. 3. Water Level. The first duty of an engineer before starting, or at the beginning of his watch, is to see that the water is at the proper height. Do not rely on glass gages, floats .or water alarms, but try the gage cocks. If they do not agree with water gage, learn the cause and correct it. 4. Gage Cocks and Water Gages must be kept clean. Water gages should be blown out frequently, and the glasses and passages to gage kept clean. The Manchester (Eng.) Boiler Association attributes more accidents to inattention to water gages, than to all other causes put together. 5. Feed Pump or Injector. These should be kept in perfect order, and be of ample size, No mu.ke cf pump can be expected to be continually reliable without regular and careful attention. It is always safe to have two means of feeding a boiler. Check valves and self-acting feed valves should be frequently examined and cleaned. Satisfy yourself frequently that the valve is acting when the feed pump is at work. 6. Low Water. In case of low water, immediately cover the fire with ashes (wet if possible) cr any earth that may be at hand. If nothing else is handy use fresh coal. Draw fire as soon as it can be done without increasing the heat. Neither turn on the feed, start or stop engine, nor lift safety valves until fires are out, and the boiler cooled down. 7. Blisters and Cracks. These are liable to occur in the best plate iron. When the first indication appears there must be no delay in having it carefully exam- ined and properly cared for. 8. Fusible Plugs, when used, must be examined when the boiler is cleaned and careruily scraped clean on both the water and fire sides, or they are liable not to act. 9. Firing. Fire evenly and regularly, a little at a time. Thin firing must be used where the draft is poor. Take care to keep grates evenly covered, and allow no air-holes in the fire. Do not clean fires oftener than necessary. With bituminous coal, a " coking fire," i. e., firing in front and shoving back when coked, gives best results, if properly managed. 10. Cleaning. All heating surfaces must be kept clean outside and in, or there will be a serious waste of fuel. The frequency of cleaning will depend on tne nature of fuel and water. As a rule, never allow over ^-in. scale or soot to collect on surfaces between cleanings. Iland-holes should be frequently re- moved and surfaces examined, particularly in the case of a new boiler, until S08 ELECTRIC RAILWA Y HAND BOOK. proper intervals have been established by experience. In water tube boilers, for inspection remove the hand-holes at both ends of the tubes, and by holding a lamp at one end and looking in at the other, the condition of the surface can be fully seen. Push the scraper through the tube to remove sediment, or if the scale is hard, use the clipping scraper made for that purpose. Water through a hose will facilitate the operation. In replacing hand-hole caps, clean the surfaces without scratching or bruising, smear with oil and screw up tight. Examine mud-drum and remove the sediment. The exterior of tubes can be kept clean by the use of blowing pipe and hose through openings provided for that purpose. In using smoky fuel, it is best to occasionally brush the surfaces when steam is off. 11. Hot Feed- Water, Cold water should never be fed into any boiler when it can be avoided, but when necessary it should be caused to mix with the heated water before coming in contact with any portion of the boiler. 12. Foaming-. When foaming occurs in a boiler, checking the outflow of steam will usually stop it. If caused by dirty water, blowing down and pump- ing up will generally cure it. In cases of violent foaming check the draft and fires. Water tube boilers should never foam with good water, unless the water is carried too high. If found to prime, lower the water line. 13. Air Leaks. Be sure that all openings for admission of air to boiler or flues, except through the fire, are carefully stopped. This is frequently an unsuspected cause of serious waste. 14. Blowing- Ojf. If feed water is muddy, or salt, blow off a portion fre- quently, according to the condition of the water. Empty the boiler every week ! or two and fill up afresh. When surface blow-cocks are used, they should be ■ often opened for a few minutes at a time. Make sure no water is escaping from the blow-off cock when it is supposed to be closed. Blow-off cocks and check valves should be examined every time the boiler i3 cleaned. 15. Leaks. When leaks are discovered they should be repaired as soon as possible. 16. Emptying. Never empty the boiler while the brickwork is hot. 17. Filling Up. Never pump cold water into a hot boiler. Many times leaks and, in shell boilers, serious weakness, and sometimes explosions are the result of such an action. 18. Dampness. Take care that no water comes in contact with the exterior of the boiler from any cause, as it tends to corrode and weaken the boiler, beware of all dampness in seatings or coverings. 19. Galvanic Action. Examine frequently parts in contact with copper or brass, where water is present, for signs of corrosion. If water is salt or acid, some metallic zinc placed in the boiler will usually prevent corrosion, but it will need attention and renewal from time to time. 20. Rapid Firing. In boilers with thick plates or seams exposed to the fire, steam should be raised slowly, and rapid or intense firing avoided. With thin water tubes, however, and adequate water circulation, no damage can come from that cause. 21. Standing Unused. If a boiler is not required for some time, empty and dry it thoroughly. If this is impracticable, fill it quite full of water, and put in ELECTRIC RAIL IV A Y HAND BOOK. 209 a quantity of common washing soda. External parts exposed to dampness should receive a coating of linseed oil. 22. General Cleanliness. All things about the boiler room should be kept clean and in good order. Negligence tends to waste and decay. BOLLER WATER. The character of water obtainable for steam making is too important a matter ,n station location and operation to be overlooked, both in regard to the cor- rosion of the boiler and to the scale forming a barrier between the heating surface and the water, reducing the efficiency of the boiler. The following waters are available for boiler uses : Rain water collected in the open country is usually nearly pure, but in the city it is objectionable because containing many impurities. Surface water is usually well adapted for boiler use except that it is usually turbid. This turbidity can be removed by settling tanks or filters. It contains a small amount of dissolved solids and is low in carbonic acid. Subsoil water obtained from springs and wells is clear, usually low In mechanically suspended matter but high in solids in solution. In periods of drought the soluble matter in subsoil water increases rapidly and water that will not give a troublesome scale ordinarily will cause trouble during a drought. Artesian well or deep water varies greatly in its character, even in a given locality. It is apt to be rich in dissolved solids. Iron compounds and sodium chloride are often present in such considerable quantities as to make the water unsuitable for boiler use. Waters of very high purity are liable to corrode the boiler badly, pure water having a corrosive action on the iron, either due to the carbonic acid or oxygen. Waters taken from marshes and where brought in contact with masses of organic matter are liable to contain acids, which when introduced into the boiler corrode it, due to the presence of this organic matter in the water. This water may be neutral to iron at normal temperatures, but on raising the temperature of the water they may become active agents. Water of this character can be tested by heating with iron filings or clean suspended iron plates, and the precipitate can be noticed; but it is to be remembered that a boiler that has scale on it already does not present an active iron surface to this organic matter. Corrosion in the boiler due to free acids may be overcome by neutralizing acids with an alkali such as caustic soda or soda ash. Corrosion due to the de- composition of magnesium salt can be benefited by almost any of the methods adaptable to prevent scale. Corrosion due to dissolved oxygen can be materially reduced by heating the feed-water before introducing into the boiler. The mag nesium and calcium carbonates can be held in solution by carbonic acid precipi- tated when that acid is removed or neutralized. This may be accomplished by heating the water or exposing it in thin layers to the action of the air; or the neutralizing of carbonic acid may be brought about by the addition of slack-lime or calcium hydroxide, which converts the carbonic acid into an insoluble car- bonate precipitating both itself and carbonates in solution. The precipitation of calcium sulphate is more difficult. This is soluble in water to the extent of 100 grains to a gallon, and is much less soluble at boiling point. This precipitation can be accomplished Dy heat alone, but it must be under decided pressure. Calcium sulphate is the most objectionable ingredient, as it forms a hard scale. One rem a dy is to use organic matters in the boiler which act by interfering with the crystallization of the sulphate, and thus j k. 210 ELECTRIC RAIL WA Y HAND BOOK. render the deposit in the boiler more easy to be removed. But the modern tend- ency is to use a direct precipitating agent for calcium sulphate, such as trisodium phosphate and sodium fluoride ; these substances convert calcium and magnesium compounds into insoluble, flocculent precipitates, yielding also highly soluble and non-corrosive sodium salts. Crude oil, kerosene, soda and tannic acid (many of the boiler compounds contain in gicater or less quantities these chemicals) are used for removing scale from boilers or making a precipitate of a muddy character, so that it can be blown off by the regular cleaning of the boiler. The engineer may make the following tests in order to determine roughly the character of water he has to deal with : Take a large, tall, clear glass vessel filled with the water to be tested; add to it, while stirring, a few drops of ammonia until the water is distinctly alkaline— this can be tested by litmus paper— next add a little phosphate of soda, the action of which is to change lime or magnesia into a phosphate which forms a deposit at the bottom of the glass. The water can then be filtered through a paper filter, leaving the precipitate in the filter, which can be weighed. This gives a relative idea of the quantity of sediment and scale- making material in the water. Water, which will turn litmus paper red before boiling, contains acid; and, if the blue color can be restored by heating, the water contains carbonic acid. If water has a foul odor and gives a black precipitate with acetate of lead it contains sulphur in various combinations. The hardness of water can be determined in the following way ; Dissolve castile soap in a glass of water and then stir into the water to be tested a few teaspoonfuls of this solution; the matter deposited will show the comparative amount of scale-making material contained in the feed water. The following chemical tests will indicate the character of the impurities in the water, by causing a precipitate: Carbonic acid is indicated by byrata water. Sulphates are indicated by chloride of barium. Chlorides are indicated by nitrate of silver. Lime salts are indicated by oxalate of ammonia. Organic matter is indicated by chloride of mercury. Heaters and Purifiers.— As many of the impurities are removed by heat- ing water, water heaters act as purifiers. Magnesium and carbonates are thrown down out of the heated water in the form of scale on the heater surface. In the open heater the water is spread out in open pans and the exhaust steam comes in direct contact with the agitated water; the shell as well as the pans is gener- ally made of cast iron, as this is not as liable to corrosion as steel and rolled iron. There are a number of forms of this type of heater, involving the spraying of the water through which the exhaust or live steam is driven, and the water is allowed to settle in pans or troughs, where the precipitate of the impurities which will not stay in solution at these temperatures is thrown down. The heated water is held in a reservoir until taken from it by a pump into the boiler. In the closed type of heaters, the cold water is conveyed through a nest of pipes, around the outside of which the exhaust steam circulates. In some cases the steam passes through the nest of pipes surrounded by the feed water to be heated. The dimensions per horse-power for feed water heaters were rated as follows at a meeting of the feed water heater manufacturers : It was decided that a heater should be rated in horse-power for each % Bc i' 't. ELECTRIC KAIL WA Y HAND BOOK. 211 of tube heating surface in the heater. The horse-power of the boilers at normal load was taken as the size of the heater, but heaters larger than required give additional capacity to the boilers to stand sudden overloads. One of the principal points of construction is, that sufficient flexibility be given the tubes so that leaks will not start, due to their expansion and contraction with changes in temperature. They should not be contracted in the steam areas so as to create a back pressure on the engine; this can be ascertained by taking cards on the engine, exhausiing through the heater and then exhausted directly to air. The temperature to which a heater raises the feed water is with exhaust steam less than 212 degs., and the temperature should be taken while the heater is delivering its full supply of water to the boilers. PERCENTAGE OF SAVING IN FUEL BY HEATING FEED- WATER, STEAM AT 70 1LBS. GAGE PRESSURE, 13 f-4 O TEMPERATURE TO WHICH FEED IS HEATED. 35° 40° 45° 50° 55° 60° 65° 70° 75° 80° 85° 90° 95° 100° I 100°,110 c 5.53 5.12 4.71 120 I I I ' 130° 140°;i50 c 6.38 7.24 5.97 6.84 5.57 6.44 4.30 5.16 6.03 3.89 4.75 5.63 3.47 4.34 5.21 3.05 2.62 2.19 1.76 1.30 0.89 3.92J4.80 3.50 4.38 3.07,3.96 2. 65 '3. 54 2.22;3.11 1.7812.68 0.45 1.34 2.25 0.00 0.90 1.81 160< 8.09 8.95 9.89 10. 6C 7.69 8.56:9.42 7.30 8.16J9.03 6.89 6.49 6.08 7.76'8.64 7.3718.24 6.96,7.84 5.67|6.56 5.2616.15 4.845.73 7.44 7.03 6.62 4.42J5.32 6.21 4.00 4.90|5.80 3.58 4.48j5,38 3.15 4.05[4.96 2.713.62 4.53 10.28 9.90 9.51 9.11 8.72 8.32 7.92 7.51 7.11 6.70 6.28 5.86 5.44 170° 180° 11.52,i2.38 ll.14jl2.00 10.76 11.62 190° 200 c 10.38 9.99 9.60 9.20 8.80 8.40 8.00 7.59 7.18 6.77 6.35 11.24 10.85 10.47 10.08 13.24,14.09 12.87 13.73 12.49 13.36 12.11 11.73 11.34 10.96 9.68 10.5? 9.28 10.17 8.88 8.48 8.07 7.66 7.25 9.78 9.38 8.98 8.57 8.16 12.98 12.60 12.22 11.84 11.45 11.06 10.67 10.28 9.88 9.47 9.07 210< 14.95 14.59 14.22 13.85 13.48 13.10 12.72 12.34 12.95 11.57 11.18 10.78 10.38 9.98 220< 250 c 15.81 19.40 15.45 15.09 18.89 18.37 14.72 17.87 14.35 18.38 13.9810.86 13.6016.35 13.2215.84 12.84 15.33 12.46 12.07 11.68 11.29 10.88 14.82 14.32 13.81 300° 29.34 28.78 28 23 27.67 27.12 ^6.66 26.02 25.47 24.92 24.37 23.82 23.27 13.31 22.73' 12.80 22.18 The heater should be placed between the pump and the boiler, so cold water can be handled by the pump. Hot water gives trouble in a number of places by eating and wearing the working parts of the pumps, so that they leak and the heater in this case has to stand full boiler pressure. Economizers —Here the flue gases pass around cast-iron pipes, containing the feed water, and the temperature of the water can be brought up to boiler water temperatuie. This form of heater removes the scale-making solvents from the water more effectively than heaters deriving their heat from steam. The construction usually employed consists of a battery of vertical cast-iron pipes connected with headers, which are large enough for both water circulation, and containing deposit from the water. The economizer is built into the flue, which is enlarged to accommodate it; a by-pass is also provided so the gases can be passed directly to the chimney, so that cleaning and repairing can be made without shutting down. The use of economizers increases the actual steaming capacity of the boiler and tends to hold the steam pressures constant. With varying steam demands 212 ELECTRIC RAIL WA Y HAND BOOK. they can improve the efficiency of the boiler plant from 10 per cent to 18 per cent depending npon local conditions. Where placed in a plant already installed, they reduce the effectiveness of the chimney draft, and may from this cause decrease the available heating value of the coal burnt; but with artificial draft they can be operated with undoubted economy. The soot and ashes should be cleaned by automatic scrapers from the tubes about one hour in twenty -four, depending upon the character of smoke passing through the economizer. ■ The table herewith gives the results of tests on nine plants using mechanical drafts and economizers. TESTS OF ECONOMIZER AND MECHANICAL DRAFT PLANTS, SHOWING INITIAL AND FINAL TEMPERATURES OF FLUE GASES AND FEED WATER IN DEGS. FAHR. ._ u *! 2 m CD N •r ^ 03 M H h B 3 ft 1 610 340 110 287 177 16.7 2 505 212 84 276 192 17.1 3 550 205 185 305 120 11.7 4 522 3:20 155 300 145 13.8 5 505 320 190 300 110 10.7 6 465 250 180 295 115 11.2 7 490 290 165 2S0 115 11.0 8 495 190 155 320 165 15.5 9 595 299 130 311 181 16.8 Boiler Feeding Methods.— There are several methods employed for feed- ing boilers, one by directly forcing water into the boiler by city water pressure, where the pressure is high enough to overcome the boiler pressure ; others by injectors or high-pressure pumps, either steam driven, belted or electric driven. The relative economy of the different methods of feeding boilers is given in the following table. This relation does not show the true operating economy in pumps as usually employed in street railway work ; for here the pumps, where steam driven, are run at a slow speed much under their maximum capacity to make up for steam consumption in the boiler, and they take as high as 160 lbs. of steam p^r hp-hour, under this method of feeding; and makes the showing of belt and elec- tric driven pumps 20 per cent to 36 per cent more efficient than steam driven pumps. With triple cylinder pumps, provided with by-passes so one or two cylinders can be thrown out of service to vary the water supply to the boilers, 33 per cent to 45 per cent greater efficiency is secured over ihe steam driven pump, doing the same duty. The amount of watrr required by a battery of boilers is usually taken as 3.6 gals, of water per hp hour, or nearly \ j cu. ft. of water pjr hp-hour. Two inde- pendent methods are required to feed the boiler, the general arrangement is to use the injector and steam driven pumps. For boiler plants of over COO-hp capacity two pumps are generally installed, each capable of taking care of the whole bat- tery. The pumps should be arranged, if possible, to take water from two sourceg ELECTRIC RAILWA Y HAND BOOK. 213 )f supply; it is advisable to arrange a storage capacity to carry the boiler for renty-f our hours in case of breakdown, where the city water system is de- >ended on. RELATIVE EFFICIENCY OF VARIOUS METHODS OF SUPPLY- ING FEED WATER TO BOILERS. Temp, of feed water as delivered to the pump or to the injector, 60° F. Rate of evapo- ration of boiler, 10 lbs. of water per lb. of coal from and at 212 degs. F. Relative amount of coal required per unit of time, the amount for a direct acting pump, feeding water at 60°, without a heater being taken as unity. Saving of i nel over the amount required when the boiler is fed by a direct acting pump without heater. Direct acting pump feeding water at 60 degs. without a heater 1.000 .985 .938 .879 .868 .838 .819 .0 Injector feeding water at 150 degs., without a heater. Injector feeding through a neater in which the water is heated from 150 to 200 degs 1.5 per cent 6.2 «« •• Direct acting pump feeding water through a heater m which it is heated from 60 to 200 degs 12.1 " •• Geared pump run from the engine, feeding water through a heater, in which it is heated from 60 to 200 degs 13.2 " " Geared pump run from motor, rheostatic control. . . . Geared pump, 3 cylinders with by-passes for regula- tion 16.2 " *' 18.3 '* " FEED-WATER PUMPS AND INJECTORS. The location of the pump where there is a natural head of water is one of convenience and shortest length of supply and delivery pipes, but where the pump has to take water from a well the suction pipe should be as short as possible. In horizontal runs of pipe the pipe should dip toward the supply end at least ^ in. in a foot to prevent an air trap, which will affect the proper working of the pump, and the foot valve and strainer on the end of the suction pipe beneath the water should be so large that the accumulation of trash w r ill not throttle the suction of the pump. When the pump is required to handle hot water, the water must be delivered to the pump by gravity. Hot water pumps give more trouble than those for cold water, and their depreciation is as much as 40 per cent greater in handling ordi- nary boiler waters, w r hile with waters containing sulphur, they are a continual source of annoyance. Cold water should be forced into the heating apparatus where possible. Injectors.— The injector is capable of feeding water into a boiler, if the water is under 100 degs. in temperature. The injector consists of a tubular brass casting having three openings: the first one, A, in Fig. 174, is for the delivery of dry steam from the boiler; the second opening, B, is the inlet for the water to be fed; the third one, /, opening towards the boiler, is the one through which the feed wat_r is to be forced by the steam. Tha injector operates upon the prin- ciple that a curient of steam at high velocity will produce by suction a vacuum which draws the air from above the water in the supply pipe B\ when the water rises it is forced through the nozzle D; the steam meeting the water from th© 214 ELECTRIC KAILWA Y HAND BOOK. supply pipe carries the water with it, on account of the energy of impact and condensation into the boiler lifting the check valve, IT, to gain admission. The watjr is t hcrcrore injected into the boiler hot. The injector is usually installed in railway plants as an alternative method for feeding the boilers where pumps or other methods are U3ed; but its economy is so poor as a method of feeding the boilers that it is not used in regu.ar service. Injectors may be placed either in a horizontal or vertical position. They work best where the suction U not over 20 ft., and should be located as near the boiler as possible. It is the usual practice to supply an injector for each boiler or pair of boilers. Injectors work more effectively at low steam pressures than at hif;h, but should be adjustable to work at varying steam pressures. Steam Pumps.— A piston speed of 100 ft. per minute is the ordinary practice for a direct acting pump, but in a boiler feeding under heavy pressures, especially where hot water has to be pumped, a slower speed is advisable. The table herewith gives the sizes and capacity of pumps from Sins, to 12 ins.; this is the theoretical maximum amount that can be pumped, but on account of slippage and the leakage of the valves, the actual amount of water pumped will be less than that given in the table : Fig. 174. injector. THEORETICAL CAPACITY OF STEAM PUMPS: SPEED PISTON OR PJLUNGER 100 FT. PER MINUTE. OF Diameter of Pump Gallons discharged Diameter of Pump Gallons or Plunger in per Minute. or Plunger in discharged per Inches. Inches. Minute. 2 16.33 5 102.0 m m m 20.67 25.52 112.0 123.0 30.88 b% 135.0 8 36.75 6 147 3M 43.13 VA 172 50.02 7 200 57.42 7% 229 4 65.34 8 261 73.76 &x 205 8;). 7 9 330 m 92.14 9^ 368 10 408 10^ 4oQ 11 4 4 12 5C7 In a duplex pump the number of gallons delivered per minute is found by multiplying the displacement of one plunger by twice the number of strokes. ^ ELECTRIC RAILWAY HAND BOOK. 215 The direct-acting steam pump is one in which the steam cylinder and watei cylinder are centrally in line with each other so that the water plunger and steam piston are connected to the same piston rod. This form of pump gives the least first cost and occupies less space, but is perhaps the most wasteful and extrava- gant form, for the reason that the steam follows at full pressure throughout the stroke, getting none of the economies due to using the steam expansively. The duplex steam pump consists of two steam pumps of equal dimensions, placed side by- side, and so arranged that the piston of each pump has a controlling movement of the slide valve of the opposite steam cylinder. This allows one piston to proceed to the end of the stroke and gradually come to a state of rest, while during the latter part of this movement, the opposite piston moves forward in its stroke and also gradually comes to a state of rest; but in moving forward and before reaching the end of the stroke, the slide valve controlling the first piston is reversed, and in consequence the first piston returns to its original position. These movements continue uniformly as long as steam is supplied to the pistons. "When the boiler pressure is from 65 lbs. to 100 lbs., a gain of from 20 to 35 per cent can be made over direct acting cylinders by compounding. But for pumps handling the amount of water necessary for railway plants of 1000 horse-power and under, the economy would not be of sufficient import to warrant the additional expense of a compound pumping apparatus, as the total amount of steam required for feeding the boilers is about rf-g of the output of the boiler. The table on the opposite page gives sizes of suction and delivery pipe for piston speeds of 100 ft. per minute. FRICTION OF WATER IN PIPES. Friction loss in pounds per square inch for each 100 feet of different-sized clean iron pipe discharging a given quantity of water per minute. Gallons Inside Diameter of Pipe. per Minute. 114 in. l^in. 2 in. ^in. 3 in. 4 in. 5 in. 6 in. 8 in. 10 in. 12 in. 20 4.07 6.40 9.15 12.4 16.1 20.2 24.9 56.1 1.66 2.62 3.75 5.05 6.52 8.15 10.0 22.4 39.0 .42 ' " '.bi' 25 30 35 40 45 50 75 .21 .10 1.60 2.44 5.33 9.46 14.9 21.2 28.1 37.5 .81 1.80 3.20 4.89 7.00 9.46 12.47 19.66 28.06 .35 .74 1.31 1.99 2.85 3.85 5.02 7.76 11.2 15.2 19.5 25.0 30.8 .09 ' ' .33 ' ' '.69 ' 1.28 1>9 2.66 3.C5 4.73 6.01 t .4J 0.03 0.06 0.10 0.16 0.23 0.32 0.42 0.H5 0.94 1.28 1.68 2.10 2.70 5.40 100 .05 li5 150 .10 175 SCO .17 .26 .37 .50 .65 .81 .96 2.21 250 .07 .09 .12 .16 .20 .25 .53 .03 .04 .05 .06 .07 .09 .18 .01 300 * 350 .02 400 450 .03 50 .04 750 .08 2l6 ELECTRIC RAILWA Y HAND BOOK. DIAMETERS SUITABLE FOR SUCTION AND DELIVERY PIPES FOR DUPLEX DIRECT- ACTING PUMPS: PISTON SPEED 100 FT, PER MINUTE. Vater-Cylinder. Suction-Pipe. Delivery-Pipe. Diam- eter. Area. Diam- eter. Area. Velocity of Flow at 100 Feet. Diam- eter. Area. Velocity of Flow at 100 ft. Inches. 4 5 6 7 8 9 10 12 14 12.57 19.04 28.27 38.48 50.27 63.62 78.54 113.09 153.93 Inches. 3 4 5 6 6 8 8 10 12 7.07 12.57 19.64 28.27 28.27 50.27 50.27 78.54 113.09 Feet. 178 156 143 136 180 126 156 144 136 Inches. 2 3 4 5 5 6 7 8 10 3.14 ,7.07 12.57 19.64 19.64 28.2? 38.48 50.27 78.54 Feet. 400 277 ; 224 196 256 225 204 224 196 THEORETICAL HORSE-POWER REQUIRED TO RAISE WATER TO DIFFERENT HEIGHTS. Gallons 60 75 90 100 12| 150 175 200 250 300 350 400 per Minute feet feet feet feet feet feet feet feet feet feet feet feet 25 .37 .47 .56 .62 .78 .94 1.09 1.25 1.56 1.87 2.19 2.50 30 .45 .56 .67 .75 .94 1.12 1.31 1.50 1.87 2.25 2.62 3.00 35 .52 .66 .79 .87 1.08 1.31 1.53 1.75 2,19 2.62 3.06 3.50 40 .60 .75 .90 1.00 1.25 1.50 1.75 2.00 2.50 3.00 3.50 4.00 c»"?ent to 12 per cent of the heating value of every pound of coal, since ihifi^u^pAir 3ir has to be drawn through the fire and heated from 60° to 500° fc'SiireiHieit *rnc oP?c£&t In a chimney is produced by the difference in weight of the JO l gases tr:^' chimney and the cold air outside, and can be considered as an On :?alanceo n verted siphon with -the heavy cold air on one leg attempting to *es"';ore ._ e equilibrium by forcing air through the grate fires, flue and chimney, ^arrught is usually measured in inches of water, that is, with a U-tube partly tilled with water, one end connected to the draught to be measured and the other end open; the difference of water level will give the draught pressure in inches. One inch difference in water level means 0.5TT ounces per sq. in. The chimney should be located so as to give the least length of flue from boiler to chimney, preferably in the middle of a battery of boilers. The foun- dations should be carefully proportioned and independent of any building foundations; and no connections of the breeching to the chimney should be made until the chimney is completely erected and settled down on its foun- dations. Construction of Brick Chimneys.— The total weight of a brick chimney must be greater than the total wind pressure against it. Every square foot exposed of a square chimney should be designed to withstand a maximum wind pressure of 56 11 «. per sq. ft. A hexagonal chimney reduces this to 42 lbs. ; an octagonal, 36 to 34 lbs. and a circular, 30.8 lbs. The circular is the best form for a chimney as it makes the best flue and is economical in material. Roughly the diameter at the base of a chimney should be ^ of its height. The following table gives the height and sizes for chimneys for different horse-power of boilers, based on an assumed evaporation of 7 lbs. of water per pound of coal, or an equivalent evaporation of 5 lbs. of coal per hp-hour. The ratio of the cross section of a chimney to the grate area is usually taken as 8 to 1. J. J. De Kinder found that 73 ft. was the best height for free burning bituminous coals, 115 for slow burning bituminous coals and 125 to 150 ft. for ne anthracite coals. 218 ELECTRIC RAILWAY HAND BOOK. CHIMNEY DIMENSIONS WITH CORRESPONDING HORSE- POWERS. Area u . G> 00 ■+-> D la ft HEIGHTS IN FEET. Square Feet. 75 80 85 £0 95 100 110 120 130 140 150 175 200 COMMERCIAL HORSE TOWER. 3.14 3.69 4.28 4.91 6.59 6.31 7.07 8.73 10.56 12.57 15.90 19.63 23.76 28.27 88 48 24 26 28 30 32 31 33 40 44 48 54 60 66 72 84 96 108 120 75 90 78 92 106 122 81 95 110 127 144 162 98 114 130 149 168 188 117 133 152 171 192 237 287 120 137 156 176 198 244 206 352 445 164 185 208 257 310 370 468 577 697 215 267 322 3*4 484 600 725 862 1173 279 337 400 507 627 758 902 1229 1584 413 526 650 784 932 1270 1660 2102 2596 672 815 969 1319 1725 2181 2693 1044 1422 1859 2352 2904 50.27 63 62 1983 2058 2511 78.54 3100 Allowance must be made where flues are longer than 50 ft., and height added to the chimney to make up for the loss in heat of the draft and friction. The table following gives the loss in effectiveness of chimney draught in percent, which is due to long flues. REDUCTION OF CHIMNEY DRAFT BY LONG FIXTES. Total length of flues in feet.. 50 100 200 400 600 800 1000 2000 Chimney draught in per cent 1C0 93 79 66 58 52 48 35 Local conditions, such as adjacent hills, atmosphere ladened with moisture, elevation above the sea level, etc., reduce the theoretical draught; and ample allowances should be made so that the chimney will be able to completely burn the combustible under the worst conditions, both of coal and weather, It is well to bear in mind that without proper draft all other boiler economies are futile. Iron Stacks.— These are formed of plate iron, lap or butt riveted, and made up in sections of convenient size. They may be made self-supporting with a flaring base, or maybe guyed two-thirds of the way up with four or more guy rods or chains. They are usually lined with fire brick part of the way up. Tireeching and Flues.— The connecting flues between the boiler and chimney are preferably round, as that form prcsent3 the easiest passage for the gases at the lowest first cost. All bends should be made with an easy curve. All joints should bo riveted and scaled with luting, means being provided for cleaning. Where two flues enter the main flue opposite each other, there should be a baffle plate interposed in the main flue between the openings to prevent back drafts. The covering of flues with insulating covering that will resist the tem- perature will increase the effectiveness of the chimney. ELECTRIC RAIL WA Y HAND BOOK. 219 Dampers.— Pivoted gates are introduced into the flue so the draft can be controlled, depending upon the steam required from the boiler, and the main damper can be regulated automatically by the steam pressure, so that when the pressure falls the dampers open to produce greater draft. Mechanical Draft.— As the chimney is limited in capacity by its dimen- sions and weather conditions, mechanical methods, of producing the flow of air through the furnace have been adopted in a number of railway plants, with economy in first cost over an equivalent chimney, and giving greater and more readily controlled draft. By this means the boiler can respond readily to ad- ditional demands, and the rate of combustion on the grate can be carried beyond that possible to be obtained from natural draft only. Induced Draft.— A steam jet may blow up the stack, inducing a draft. This method, while being the most economical in installation, consumes from 8 to 20 per cent of the stertm made by the boiler to produce the required draft. The steam introduced into the hot gases reduces their volume and effectiveness, and such an arrangement is so noisy that very few power stations are so located as to be able to use it. In mechanically induced draft systems a fan is introduced between the boilers and the stack, which draws the air through the fire and boiler, and ejects it from the chimney. The fan can be operated by an electric motor or an engine, whose speed can be controlled in order to vary the rate of combustion in the furnace to meet the required steam demand. By having a controllable air supply, complete combustion and consequently greater evaporative results from the coal can be obtained at a cost of 1 per cent to 4 per cent of steam from the boiler. Forced Draft or Plenum Method.— This may be accomplished in two ways : first, by making the ashpit practically air-tight, and forcing into it suf- ficient air for combustion; or second, a method only practicable in steamships, by making the fire room air-tight, maintaining sufficient air pressure in this room to produce the required draft. The first method, which is applicable to street railway boiler rooms, does not give the results in practice of the induced draft, and subjects the fireman, where hand firing is done, to considerable heat on opening the fire doors. The test carried out on the steamer, IT. M. S. IV.yphcmus, gives comparative results between the forced and induced method as follows: I RESULTS OF EXPERIMENTS AT PORTSMOUTH DOCK YARD WITH BOILERS OF H. M. S. POLYPHEMUS. P 1 Induced Forced CD el o £4 Temper- ature. 74.2 77.3 62° 51° 1 0Q o a 69.9' 49.8' o g o ► 12 Wg O n E-« 80,G00 777,044 94,500 759,338 1 M Lbs Water C 1> a . Evaporated per lb. Coal. *»2 o3 , < ■M < Lbs. c sumod Sq. Ft 9.64 11.13 40.4 803 9.3 47.3 Lbs. Water Evaporated per hour per Sq. Ft. Grate _ 93 . «a 4*fc< < < 389.6 450.4 381 444. 0* go; < 426. 395. aao ELECTRIC RAILWAY HAND BOOK. THE STEAM ENGINE. The steam engine, to successfully maintain potential on a street railway system at maximum economy, must possess features which ft it particularly for that work. Increasing the number of car equipments averages out the character- istics of the individual equipment, as the character of service changes with the number of equipments. The engine will be considered with regard to its eco- nomical performance, regulation and maintenance in connection with the follow- ing demands: 20-car road, 35-car road, 60-car road, 150-car road, and 300-car road and over. With 20 cars the variations are large and rapid, and an engine, which will not respond readily to an increased load, drops the potential on the system, and retards the acceleration or speed of all cars operating, thus decreasing the possible external efficiency. It will be seen that the slower the initial speed of the engine, the larger the volume of steam that is taken at each stroke in order to give the same horse-power. In the Corliss type of engine for small roads, the load can vary much more rapidly than the governor can control the steam, and this rapid vari- ation throws strains throughout the engine in the interchange of power between the flywheel and piston. The greater the inertia of the flywheel, the longer will be the period required for the governor to respond to the changing demands. The first cost of a slow-speed engine, where direct-connected to the generator, is higher than that of an equivalent high-speed engine. The economy of the Corliss type of engine has often been judged for railway work from its full load efficiency, whereas carrying efficiently loads of one-half to three-quarters full load is required for this class of power station service. Consequently the adapta- bility of engines to the different character of loads found in the various railway stations must be carefully considered, in order to determine what class of engine will give the best average economy. In determining this for a new road where the equipment is fixed, the con- stants for the equipment should be determined for grades, speeds and loads. "Where mixed equipments are used, the data should be based on the largest equipment where the traffic ever requires their use over the entire system. In determining the current and potential required and the size of the unit for a new railway, the following points have to be borne iu mind: The operation of the equipment under large line drops requires a greater current delivery, as its speed falls off on account of loos in potential on the line; allowance must be made in the size of the operating unit to make good all these losses without exceeding the allowable overload on the unit. Hence in collecting data for the purpose of accurately proportioning the engine and generator to the load, the current re- quired by the equipment under maximum and mean transmission losses should be determined. The greater the number of cars, the nearer the average load becomes a con- stant load, varying only with the number of cars in operation. Approximate results can be determined by referring to the curves in Fig. 173, which have been based on tests of roads operating equipments of two Wcstinghouoe 12- A motors, with a 28 ft. car body and K-10 controller, the total equipment weighing 22,000 lbs. Effect of Grades on Engine Loads.— The physical conditions of the road are reflected immediately on the engine demands. Curves given herewith, Fig. 175, show the relation of the maximum and mean demands on the engine for roads of from one car to fifty, and for thrco characters of roads, one possessing heavy grades, one moderate grades and one on level roa4, ELECTRIC RAIL WA Y HAND BOOK. 221 Q JO 20 30 W fiO DUMBER OT CAPS OPERATING Fre. 175. — ouuKjBjrr iter ©ah ttfder putbrewt eoNDiTiovg. 222 ELECTRIC KAILWA Y HAND BOOK The effect of moderate grades below 3 per cent is to increase the starting current on the equipment, but the effect of the stored energy in the equipment in climbing such a grade i3 not fully recovered on the drifting of the equipment returning on this grade. The effect of a grade, averaging approximately between 3 per cent and C/% per cent, will reduce the ratio between average and maximum current demand a but increases the maximum demands, due to starting and carrying the equipment up the grade. In grades above 4*£ per cent the maximum and average demands per equipment are both increased. 2 = j t\ : J Si uo 60 M/A/UTES 60 /00 /SO NORMAL LOAD AT 6 O'CLOCK Fig. 176.— curve showing variation of demand on station. #0 In order to determine the effect of the grade on the required output of the power station, the profile of the road is required, from which is determined the time from the schedule when the equipments require their maximum demand on the power station. It is very important in improvements in a power station for roads with 20 cars and under, or in a new station of the same capacity to have the average dally schedule carried by one engine, and to have this engine rated at maximum cQcicncy at about five-eighths to three-quarters of its maxi- mum load. For roads in operation, the proper unit for maximum efficiency can be determined by the main ammeter readings. For fixed schedule on a 20-car road and under there will be found on taking minute readings, a cycle of changes which arc periodic in character, depending upon the profJe of the road and the coincidence of equipments requiring the maximum demand at the same time. This is especially marked in a single track road with turnouts. Maximum Engine Efficiency.— Supposing a report from several daily observations, taken under different track and weather conditions, or estimated from the profile of the road and known equipment demands, showed a variation ELECTRIC RAILWA Y HAND BOOR*. 223 1 like that in Fig. 17G, it is easy to see that here the maximum load obtained is 420 hp, the mean average load is 205 hp. For one unit to carry this service con- tinually, the point of maximum efficiency of the engine should be 300 hp, and the maximum capacity of the engine 5u0 hp, including an overload possibility of 25 per cent, and the circuit-breaker set for this number of amperes. The ad- ditional load, caused by passengers at 7 a. m. and 6 p. m. and for days of special demand, will be much more profitably carried by an auxiliary unit than to operate two engines for th^ whole day, increasing the depreciation of the plant while neither engine would be using steam economically in the cylinder, and the friction losses would be doubled. 40 60 CO 7v 60 90 100 110 120 130 i40 U* Fig. 177.— steam consumption for different loads on engine. In one test made by the author on a 32-car road, operating two Corliss engines, the operation required 62.3 lbs. of steam per kw-hour and both engines had a mean load of % full load; with one engine carrying the whole load, the steam consumption fell to 42 lbs. per kw-hour. The commercial value of this change to a single engine effected a saving in coal and oil sufficient to pay for all the power station labor. Varying of Efficiency with Load.— When the load on an engine is reduced much below the rating, the friction per cent increases and also the cylinder condensation. The aggregate losses are shown in Tig. 177, as given by Prof. K. C. Carpenter, on a single non-condensing engine, 14 ins. x 16 ins., 120 lbs. fcteam pressure, 210 revolutions per minute. While, in the 20-car to 80-car road, the number of equipments averages out the rapid maximums, the speed of the engine should still be moderate. With from 35 to 60 cars the question of the type of engine is varied by thecharacterietics of J 224 ELECTRIC RAILWA Y HAND BOOK. the road, as both high speed and moderate speed engines have shown about the same gross economy. The gross economy includes the cost of production, as well as fixed charges against the power station property; this is treated as a factory selling its product (current) at a cost which will cover every expense of operation, maintenance, depreciation, interest, taxes and insurance, and that proportion of executive expenses that the cost of the station bears to the whole property cost. With more than CO cars the conditions of load become one w T here the engines can be worked at a slightly varying load from 80 per cent to 60 per cent of full load; and with their maximum efficiency between these points. Every means taken to increase the economy of the plant should be reflected in the operating costs per kw output in power plants of this size. Records of the performances of different types of engines under railway loads have been obtained from the average of fifty-six tests taken on the basis of a four-hour and twenty-hour run, which are the average periods of service for all day operation and overload periods, and are tabulated below. Non-Condensing Engines. Lbs. of Coal per Indicated HP-Hour. Tons (2,000 lbs.) per HP per year. Day, 20hrs. Day, 4 hrs. Simple, High Speed, Slide Valve, average. . 4k " " " «• best Simple Corliss, average 4.71 4.52 3.46 3.00 4.09 3.91 3.18 2.22 2.41 1.80 17.19 16.49 12.62 10.95 14.92 14.27 11.60 8.10 8.79 6.57 3.44 3.30 2.52 " " best 2.19 Compound Slide Valve, average 2,98 *' '« best 2.85 Condensing Engines. Compound Slide Valve, average 2.32 " " »' best 1.62 Compound Corliss, average 1.76 " '« best 1.31 The tests on the engines show the value of the condensing type to a plant, but in the comparison between the Corliss and compound-condensing engine the loads on the Corliss tested were more favorable to economy than the loads on the compound-condensing moderate-speed engine; and, if the question of interest on first cost, oil and attendance were included in the aggregate expense of operation, the outcome will be in favor of the compound-condensing moderate-speed engine. The tandem-compound, from the results of observation, gives the best results in stations under the 35-equipment class, both condensing and non-con densinc, and the cross-compound for stations of the larger classes show good economy under operating conditions. The elements of the economy of the Corliss valve gear have been introduced In moderate speed engines, giving this type of engine an additional economic value in railway work. One station has shown under mixed load 300 watt-hours output per lb. of coal under constant operation, using this type of engine. The general classification of engines in regard to use of steam in cylinders is based on the number of expansions through which the steam passes in the engine. In the simple engine there is only one expansion; this may be of the high-speed ELECTRIC RAILWA Y HAND BOOJC. ±25 or low-speed Corliss type. The compound engine uses steam expansively through two cylinders, which may be arranged in tandem, or the engine may be cross- compound, having one high and one low pressure cylinder on separate cranks 90degs. apart, or the three cylinder compound engine where the first expansion takes place in the high pressure cylinder and the second expansions in two low pressure cylinders. The double tandem compound engine consists of two tandem compound engines, coupled to the same shaft 90 degs. apart. The cost of an engine for any given power increases with the number of cylinders and expansion stages, but the economy of engines, with approximately constant load, increases with the number of expansions, providing the steam pressure is raised for the proper economy of steam through these expansions. So, in estimating the most profitable engine to procure for a given condition, the following values have to be balanced against each other; the interest on first cost, depreciation, cost of oil and attendance. The engine should be so selected that, with the road under consideration, it would bring the lowest interest charges on the cost of the boiler plant, and the least cost on coal and water consumption, the coal being the largest item, bearing about one- third to two-thirds of the total cost of power production. Only with very cheap coal and water and variable duty, will the simple high-speed engine show an economy of operation. Division of Units.— The power station requires a duplicate set of engines, and in 20-car roads and under the stand by investment is large. The advisable division is usually as follows, the sizes of engines being selected to fit the gene- rators as manufactured for railway work: 200 Horse-power Maximum .2- 200 300 ,c " " 2-30O 400 " * " 3-200 500 " " ** 3- 250 or 4-175 600 •• '• *■ 3-300 800 '• u « ' 3- 400 or 4-800 1000 " " " 3- 500 " 4-350 1200 *« •« •* 3- 600 «• 4-400 or 5-800 1500 w " " 3- 750 "4-500 2000 " " " 3-1000 " 4-750 In a road requiring more than 2000 horse-power, the division is limited by thr size of the units which the market supplies; by combining, however, several size of units the greater factor of overload can be purchased for the least cost. There are several advantages in having a system of power units all of the same type, as they operate together and the engineer can experiment on them to determine how to get the best results in operation, but in long roads with few cars a small unit can often be used during the end and beginning of the schedule of each day, with a saving that will warrant the expenditure for this smaller unit. Mechanical Strains.— The mechanical construction of an engine must give ample strength to the parts in order to withstand straining on overloads, and when the circuit-breaker opens. Engines built; for factory service, especially of the slow-speed type, do not require the strength of parts of the railway engine, and to use that type of engine for railway work has led to a high rate of depreci- ation thereof. All parts subject to reciprocating strains have to be strengthened, and the engine is classed under the heavy duty type. Where a great section Of 226 ELECTRIC RAILWAY HAND BOOK. metal is put in the frame of the engine, the piston-rod, connecting-rod, shaft and crank bearings are increased in size, and the flywheels should be constructed on different principles from those employed for factory loads. Rotary and Piston Speeds.— The table on page 208 gives the revolution per minute and pi3ton speeds of a number of types of engine, direct-connected to generators for railway and lighting work. The following table gives the average approved revolutions per minute for railway engines: 500 kw 135 r. p. m. 1000 kw 100 r. p.m. 2000 kw and up 75 to 70 r. p. m. DIMENSION OP ENGINE PARTS. Clearances allowable with high-speed engines, with valves providing relief for entrained water vary with the size of the engine. The clearance volume bears no fixed relation to the total steam volume in the different engines for railway work, varying from 14 ins « to % ins. in the different sizes and types of engines. The clearance on the crank end is greater to take up the wear on every working joint between the piston and crank-pin; this allowance is generally T ^ in. for each joint. There are several methods for cutting down clearances, which have to be filled with steam at every stroke without doing useful work. One is to have the valves raised off their seats to allow relief for entrained water. Engines of this type have operated with 3 per cent clearance without trouble. Another method is to introduce pop valves, opening into the clearance spaces to relieve the entrained water. Still another method is a diaphragm placed so that it will be broken open when the pressure reaches 100 per cent over the maximum steam pressure. Cylinder Walls.— In railway engines the bursting stress on the cast-iron cylinder walls should not exceed 2500 lbs. Cylinder Heads.— The thickness of the cylinder heads vary with the diameter of the piston.. 10 ins. diameter averages on a basis of 100 lbs. unbalanced pressure .68 ins. ; 30 ins. diameter, 1.48 in. ; 50 ins. diameter, 2.30 ins. An old rule is to make the cylinder head 1^4 times the thickness of the walls. Webbed heads should give equivalent strengths. Cylinder Head Bolts— -No bolt smaller than % in. should be used in cylinder heads. They should be spaced at a distance of about four or five times the thickness of the flange, and the strain on them should not exceed 5,000 lbs. per sq. in. for steel and 4,500 lbs. for wrought iron. The nut should engage threads to a greater depth than the diameter of the bolt under thread. Piston Head.— The general rule is that the thickness of the piston head is equal to \/ length of stroke x the diamoter of the piston. Piston packings should be made approximately 1 per cent larger than the diameter of the cylinder and sprung into place. A section of ring is usually ^ of the diameter of the cylinder plus % y and for width % in. is added to the thickness. The fit of the piston rod into piston is usually made by a combination of a straight and taper surface, the taper being about 3 ins. to a foot, which is drawn up to a shoulder by a nut. The strain on the bottom of this nut should not exceed 7,000 lbs. per sq. inch for steel and 5,500 for wrought iron. ELECTRIC KAILWA Y HAXD BOOK. 227 Diameter 0/ Piston Rods.— The average diameters of piston rods for railway en- diameter of cylinder / — — : — — Xy Maximum working pressure-f-15. gines should be at least 65 Piston Rod Guides.— The pressure on the lubricating surfaces of piston rod guides should not exceed 350 lbs. Thurston gives the following value: The product of the relative velocity of the two surfaces in feet per minute of the guide multiplied by the maximum intensity of pressure should not be greater than 60,000. Connecting Rod.— The ratio of the connecting rod length to stroke varies from 2:1 to 2*^:1. Some of the more modern engines for railway work show slightly less than a 2 to 1 ratio, but this increases rapidly the wear on cross-head guides and friction surfaces; and with small clearances railway experiences certainly dictate longer connecting rods, even at a sacrifice of floor space in horizontal, and head room in vertical engines. Crank Pin.— The pressure on a crank pin should not exceed 500 lbs. per sq. in. projected area or its length of bearing surface by diameter of pin. The crank pin is preferably made part of the crank arm or disc. In station engines the crank pin should be an integral part of the crank arm. Engine Shaft.— With direct-connected units special conditions arise which throw strains on this shaft not encountered in belt driving. As an armature gradually falls out of alignment, due to the wear on the main shaft, the magnetic field is disturbed and an unbalanced pull occurs due to the smaller clearance on the lower part of the armature; this for an % i n » difference in a 200-kw machine throws an additional pressure on the bearings of 21,400 lbs. approximately. Often when the bearings commence to heat in a direct-driven unit that has previously run smoothly, this is the place to look for the trouble. This can be found electrically by taking off the brush connecting cables with brushes down and fields excited SIZE OF STEEL SHAFTS FOR DIRECT-CONNECTED UNITS. Kw Output 575 Volts. Rev. Size of Shaft. Inches. Rev. Size of Shaft. Inches. Rev. Size of Shaft. Inches. 100 150 275 200 5^ 9 200 300 200 150 10^ 14 150 120 15 100 16 400 500 150 120 16 18 120 100 18 18 1C0 90 18 18 650 800 90 120 20 22 100 22 80 22 1000 1200 80 80 25 27 1600 2000 2400 75 75 75 27 30 30 and the generator running; if the field is distorted, due to unbalanced magnetic circuit, and the field winding is in good condition, a voltmeter will show higher voltage between those brushes bridging pole pieces which are too close to the 228 ELECTRIC RAILWAY HAND BOOK. armature. Boxes for shafts of direct-connected nnits should all be adjustable bo that the generator can be re-aligned to make up for wear thereon. The propor- tions of the shaft depend on whether the generator is overhung or provided with outboard bearings; both methods of connection have been used and given satis- faction on railway loads. The overhung armature requires less floor space than that with outboard bearing. With a belted engine the outboard bearing is usually used. There is a tendency for a shaft beyond the engine to be deflected on account of the pressure on the crank pin. The sizes of steel shafts given on page 206 are advised for generators of 575 volts. Engine Bearing?. — Engine practice shows weight of bearings for direct- driven units as high as 460 lbs. per sq. in. of effective bearing surface; the belt- driving engine, 151 to 375 lbs. per sq. in. The length of the bearings on overhung armatures is 234 times the shaft diameter and 1% times the shaft diameter with outboard bearing. Automatic, forced oil circulation has a great value in carrying away the heat from these friction surfaces ; some engine makers introduce pipes into the pillow block casting through which water can be circulated in order to reduce the temperature. The character of shaft metal and the boxes in which they run should produce a glass surface and one on which the lubricant can reduce the friction coefficient to the lowest point. Fly Wheels.— Armatures do not give sufficient centrifugal force to steady the engine aud the drag of the armature through the field tends to make it behave as a brake wheel. An additional flywheel is necessary. While there is no case known of the bursting of a solid flywheel run on a high-speed engine, those on slow-speed engines have been wrecked, due to several causes. Where governor balls have been used for regulation in railway loads, the collar is continually working up and down over a narrow band with the result that at some time, if this point is not given particular attention, the governor will stick when the load goes off, and the engine will commence to race, or the safety stop may be out of order. A slack governor belt will let the engine run ahead of its rated speed. On examinations of flywheel explosions, where proper care has been taken of the engine, the failure has been due to two causes i one, the structural weakness of the flywheel and the other its location. The structural weakness occurs where a rim speed of 5000 ft. per minute and under and no greater strain on the rim section than 6000 lbs. per sq. inch is allowed. In the segmental form of casting, used where the spokes are cast directly to the rim, the fractures found on investi- gation show a very coarse grain at the fiacture between the spoke and the rim. This would be due to shrinkages taking place between the rim and spoke, produc- ing a character of metal here which has less than the calculated tensile strength. Built-up wheels should have the spoke and rim of separate castings, if possible for slow rotative speeds, or the Tim maybe built up of sheet iron; wire-wound flywheels have been used with success. The wheel pit has often been made with small clearances between the wheel and the masonry of the foundation, and in two cases the driven pulley fractured first, the parts were drawn into the wheel pit and jammed under the flywheel and the engine wrecked. The maximum diameter of flywheels of cast-iron allowable for railway work •hould not exceed the following figures for 5000 ft. peripheral velocity; 80 Rev. per minute 83.25 ft. in circumference 26.5 ft. in diameter. 100 " " 55.5 4k M 17.7 " 150 M " 41.67 " " 13.3 " M 800 « •* 31.36 " " 9.9 " •* 850 M M 22.73 •• w .....7.3 «■ •• i ELECTRIC RAIL WA Y HAND BOOK. 229 & P 'Jl H 8 S H S 5 § © 1 ,d ■*a • fc i • o •J J* p. c £* t3 8 § = o p< P? o T3 P O Q »d O o o pi o • P4 o o Q § a o d I o Comp. Comp., ft a o a o O d a o d' S o P P Eh Pi. a o p. a o a U ^ s O © o> O O 0) S3 Eh I o F-l 2 2 00 oc O Pi Q CO CQ O t-t CD as O U O PS o p .o P o p CO CD o (-1 O CO X O 5 g 8 " A ^H 8 8 8 CO rf ^ 8 8 8 o t=> X 8 8 •dq *XT?xt 'dq po^t?a p99dg no^sid CO iO one another, and would compare properly with any theoretical expansion curve. (Prof. A. B. W. Kennedy, Froc. Inst. M. E., Oct. 1880.) Fig. 100 shows a combined diagram of a quadruple-expansion engine, drawn according to the usual method, that is the diagrams are first reduced in length to relative scales that correspond with the relative piston displacement of the three cylinders. Then the diagrams are placed at such distances from the clearance line of the proposed combined diagram as to correctly represent the clearance in each cylinder. Clearance.— The clearance of an engine may be measured by filling with vaseline, the space between the piston and cylinder head when the engine is on centers and the volume of vaseline required to fill the space measured. This yoJ- 234 Fig. 181.— friction indication simple valve engine. Fig. 183.— indication c? friction FOUR VALVE ENGINE. Pig. 185 — htghphes«tjre card for four valvic compound engine. Fig. 187. — graduated load indication. Fig. 182.— full load indication simple valve engine. Fig. 184.— full load of four valve engine. Fig. 186.— low pressure card for four valve compound engine. Fig. 188.— maximum and minimum indication. Fig. 189 — corliss condensing engine indication. J ELECTRIC RAIL WA Y HAMD BOOK ume divided by the volume swept through by the piston per stroke, equals the per cent of clearance. In case the clearance can not be measured in this way, it may be roughly drawn from an indicator card by the following process. Draw a straight line, c b ad^ across the compression curve, first having drawn O X % Fig. Fig. 190.— combining multiple expansion engine diagrams. 191, parallel to the atmospheric line and 14.7 lbs. below. Measure from a the distance, a d equal to c b, and draw Y O perpendicular to O ^through d; then will TB divided by A The the percentage of clearance. The clearance may also be found from the expansion line by constructing a rectangle efhgy and drawing X 5 J" Fig. 191.— method of obtaining clearance lines from diagram. a diagonal gfio intersect the line O X, This will give the point, c*>, and by erect- ing a perpendicular to O X we obtain a clearance line O Y, Both these methods for finding the clearance require that the expansion and compression curves be hyperbolas. Prof. Carpenter (rower, Sept. 1893) says that with good diagrams the methods are usually very accurate, and give results which check satisfactorily. 236 ELECTRIC RAILWAY HAND BOOK. HORSE-POWER PER POUND MEAN EFFECTIVE PRESSURE. Hp. per lb. M. E. P. = Area In eg. in. X piston speed 83,000 Diam. of Cylinder, Speed of Piston in Feet per Minute. ins. 100 240 300 400 450 500 550 600 650 750 4 .038 .091 .114 .152 .171 .19 .209 .228 .247 .285 *M .048 .115 .144 .192 .216 .24 .264 .288 .312 .360 5 .06 .144 .18 .24 .27 .30 .33 .36 .39 .450 5^ .072 .173 .216 .283 .324 .36 .396 .432 .468 .540 6 .086 .205 .256 .342 .385 .428 .471 .513 .555 .641 ®& .102 .245 .307 .409 .464 .512 .563 .614 .698 .800 7 .116 .279 .348 .466 .524 .583 .641 .699 .756 .874 7^ .134 .321 .401 .534 .602 .669 .735 .802 .869 1.002 8 .152 .365 .456 .608 .785 .761 .837 .912 .989 1.121 ®A .172 .413 .516 .688 .774 .86 .946 1.032 1.118 1.290 9 .192 .462 .577 .770 .866 .963 1.059 1.154 1.251 1.444 9^ .215 .515 .644 .859 .966 1.074 1.181 1.288 1.395 1.610 10 .238 .571 .714 .952 1.071 1.190 1.309 1.428 1.547 1.785 11 .288 .691 .864 1.152 1.296 1.44 1.584 1.728 1.872 2.160 12 .342 .820 1.025 1.366 1.540 1.708 1.880 2.050 2.222 2.564 13 .402 .984 1.206 1.608 1.809 2.01 2.211 2.412 2 613 3.015 14 .466 1.119 1.398 1.864 2.097 2.331 2.564 2.797 3.029 3495 15 .535 1.285 1.606 2.131 2.409 2.677 2.945 3.212 3.479 4.004 16 .609 1.461 1.827 2.436 2.741 3.045 3.349 3.654 3.958 4.567 17 .685 1.643 2.054 2.739 3.081 3.424 3.766 4.108 4.450 5.135 18 .771 1.849 2.312 3.083 3.468 3.854 4.239 4.624 5.009 5.780 19 .859 2.061 2.577 3.436 3.865 4.295 4.724 5.154 5.583 6.442 20 .952 2.292 2.855 3.807 4.285 4.759 5.234 5.731 6.186 7.138 21 1.049 2.518 3.148 4.197 4.722 5.247 5.771 6.296 6.820 7.869 22 1.152 2.764 8.455 4.607 5.183 5.759 6.334 6.911 7.486 8.638 23 1.259 3.021 3.776 5.035 5.664 6 294 6.923 7.552 8.181 9.44 24 1.370 3.289 4.111 5.482 6.167 6.853 7.538 8.223 8.908 10.279 25 1.487 3.569 4.461 5.948 6.692 7.436 8.179 8.923 9.566 11.053 26 1.609 3.861 4.826 6.435 7.239 8.044 8.848 9.652 10.456 13.065 27 1.733 4.159 5.199 6.932 7.799 8.666 9.532 10.399 11.265 12.998 28 1.865 4.477 5.596 7.462 8.395 9.328 10.261 11.193 12.125 13.991 29 2.002 4.805 6.006 8.008 9.009 10.01 11.011 12.012 13.013 15.015 30 2.142 5.141 6.426 8.568 9.639 10.71 11.781 12.852 13.923 16.065 31 2.288 5.486 6.8C5 9.144 10.287 11.43 12.573 13.716 14.866 17.145 32 2.436 5.846 7.308 9.744 1^.962 12.18 13.398 14.616 15.834 18.270 83 2.590 6.216 7.770 10.360 11.655 12.959 14.245 15.54 16.835 19.425 34 2.746 6.59 8.238 10.984 12.357 13.73 15.103 16.476 17.849 20.595 35 2.914 6.993 8.742 11.656 13.113 14.57 16.027 17.484 18.941 21 855 36 3.084 7.401 9.252 12.336 13.878 15.42 16.962 18.504 20.046 23.130 37 3.253 7.819 9.774 13.032 14.861 16.29 17.919 19.548 21.177 24.435 38 3.436 8.246 10.308 13.744 15.462 17.18 18.898 2). 616 22.334 25.770 39 3.620 8.648 10.86 14.48 10.29 18.1 19.91 21.62 23.53 27.150 40 3.808 9.139 11.424 15.232 17.136 19.04 20.944 22.848 24.752 28.560 41 4.002 9.604 13.006 16.0(8 18.009 20.00 22.011 24.012 26.013 30.015 42 4.198 10.005 12.594 16.792 18.901 20.99 23.089 25.188 27.287 81.485 43 1.40 10 56 13.20 17.6 19.8 22.00 24.2 20.4 28.6 33.00 44 4.606 11.046 13.818 18.424 20.727 23.03 25.333 27.636 29.939 34.545 45 4.818 11.563 1 1.454 19.272 21.681 24.09 20.399 2.^.908 31.317 36.135 46 5.043 12.086 15.128 20.144 22.602 25.18 27.008 30.216 32.754 37.770 47 5.256 12.014 15.768 21.024 23.652 26.28 28.908 £1.536 34.104 39.420 48 5.482 12.846 16.446 21.928 24.609 27.41 30.151 83.152 3") 033 41.115 49 5.714 12.913 17.142 22.856 25.713 28.57 31.427 34.284 37.141 42.855 50 5.950 14.28 17.85 23.8 26.775 29.75 32.725 35.7 38.675 44.625 61 6.180 14.832 18.54 24.76 27.855 30.95 34.045 37.08 40.205 46.425 52 6.432 15.437 19.296 25.723 28.944 3:3.16 35.376 3S.592 41.808 48.240 53 6.684 10.041 20.052 20.730 30.078 32.42 36.762 40.104 43.446 50.130 54 6.940 16.656 20-82 27.76 31.23 33.7 38.17 41.64 45.11 52.05 55 7.198 17.275 21.594 28.792 32.391 35.99 39.589 43.188 40 787 53.985 56 7.462 17.909 22.386 29.848 33.579 37.31 41.041 44.772 48.503 55.965 57 7.732 18.557 23.196 30.928 34.794 38.66 42.526 46.392 50.258 57.99 58 8.006 19.214 24.018 32.024 36.027 40.03 44.033 48.036 52.039 60.045 59 8.284 19.902 24.852 33.136 37.278 41.42 45.562 49.704 53.846 62.13 60 8.356 20.558 25.698 84.264 88.547 42.83 47.113 51.396 55.679 64.245 ELECTRIC RAILWAY HAND BOOK 237 TRUE RATIO OF EXPANSION AS AFFECTED BY CUT-OFF AND CLEARANCE. %4 14 II 1 PER CENT OP CLEARANCE. 1 2 3 4 5 6 7 8 9 10 PER CENT OF CUT-OFF. 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1% .789 .796 .794 .792 .790 .787 .785 .783 .781 .780 .663 .660 .657 .654 .650 .646 .643 .640 .636 .633 1* .567 .563 .559 .555 .550 .545 .541 .537 .533 .528 2 .495 .490 .485 .480 .475 .470 .465 .460 .455 .450 2U 2H .439 .433 .428 .423 .417 .411 .406 .400 .395 .3S9 .394 .388 .382 .376 .370 .364 .358 .352 .346 .340 2& .358 .351 .345 .338 .332 .325 .319 .313 .306 .300 S .327 .320 .313 .307 .300 .293 .287 .280 .273 .267 3^ .279 .271 .265 .257 .250 .243 .236 .228 .221 .214 4 .243 .235 .227 .220 .212 .205 .197 .190 .182 .175 4y 2 .215 .207 .199 .191 .183 .175 .168 .160 .152 .144 5 .192 .184 .176 .168 .160 .152 .144 .136 .128 .120 53^ .174 .165 .157 .149 .141 .132 .125 .116 .108 .100 6 .158 .150 .142 .133 .125 .116 .108 .100 .092 .083 6^ .145 .137 .128 .120 .112 .103 .095 .086 .078 .069 7 .134 .126 .117 .109 .100 .091 .083 .074 .056 .057 7V, .125 .116 .107 .099 .090 .081 .073 .064 .055 .047 8~ .116 .108 .099 .090 .081 .072 .064 .055 .046 .037 8^ .109 .100 .091 .082 .074 .065 .056 .047 .038 .029 9 .102 .093 .084 .076 .067 .058 .049 .040 .031 .022 s% .096 .087 .078 .070 .060 .052 .043 .034 .025 10 .091 .0*2 .073 .064 .055 .046 .037 .028 ...» 1034 .086 .077 .008 .059 .050 .041 .032 .023 ..., 11 .0 C 2 .073 .064 .055 .045 .036 .027 UK .078 .069 .060 .050 .041 .032 .023 ..... 12 .074 .065 .056 .047 .037 .028 ( 13 .068 .058 .049 .040 .031 .021 14 .062 .053 .044 .034 .025 15 .057 .048 .039 .029 .020 ...... ..... 16 .053 .044 .034 .025 .... 17 .049 .040 .031 .021 18 .046 .037 .027 ..... 19 .043 .034 .024 ...... ..... E0 c040 .031 .021 ...... 23$ ELECTRIC RAILWA Y HAND BOOK. COMPRESSION OF STEAM! IN THE CYLINDERS. Best Periods of Compression; Clearance 7 per cent. % TOTAL BACK PRESSURE, IN PERCENTAGES OF TOTAL INITIAL PRESSURE. OQ O •i-« a> Jt4 ntag Stro 2^ 5 10 15 20 25 30 35 s £ © £ PERIODS OF COMPRESSION, IN PARTS OP THE STROKE. 10% 15 65% 58 57% 44% 32/^ - I 52 40 20 !U% i 20 52 47 37 27 22 25 47 42 42 39 34 32 26 25 21 20 17% 16 30 14% U% 35 39 35 29 23 19 15 13 11 40 36 32 27 21 18 14 13 11 45 33 30 25 20 17 14 12 10 50 30 27 23 18 16 13 12 10 55 27 24 21 17 15 13 11 9 60 24 22 19 15 14 12 11 9 65 22 20 17 15 14 12 10 8 70 19 17 16 14 14 12 10 8 75 17 16 14 13 12 11 9 8 STEAM CONSUMPTION DISTRIBUTION AND VARIOUS EFFICIENCIES OF AVERAGE ENGINES FROM 300 TO 500 H. P. Engine Class, NON-CONDENSING. Throttling, small Simple, J ep. Valve Compound Dep. Valve. Simple lndep. Valve.... Compo'nd lndep. Valve CONDENSING. Simple Dep. Valve Compound Dep. Valve. Triple Dep. Valve Simple lndep. Valve.... Compo'nd lndep. Valve Triple lndep. Valve Comp. or Triple lndep. Valve, very large 0Q 0Q 2 . be 1 """' I CO 80 100 130 100 130 100 130 160 100 130 160 170 Dry Steam Con- sumption, Lbs. per 1 H. P. hour. 03 d 17.83 16.08 14.30 16.08 14.30 8.81 8.27 7.85 8.81 8.27 7.85 7.74 . 27.17 16.92 9.70 11.92 7.70 r.19 11.73 9.15 13.19 9.73 7.15 5.26 a 1 III d ° & 45 83 24 28 22 27 20 17 22 18 15 13 Sag Wo 14.29 15.77 17.62 15.77 17.02 26.50 28.14 29.44 26.50 28.14 29.44 29.85 oco£ bco d O 30.7 48.T5 59.6 59.5 65. 8^.65 41.35 40.2 40.05 4 .95 52.4 59.5 o a) o a> •St: H o ~y S-3 e o * I .* o o . ©.SU- 5.07 85. 7.' 5 94 10.50 92 9.05 92 11.44 90 8.65 91 11.63 90 13.61 90 10.61 88 12.91 87 15.43 87 17.77 93 4.82 7.19 9.66 8.32 9.28 15.35 Note: The condenser pressure is assumed at 2 lbs. absolute. ELECTRIC RAILWA Y HAiVD BOOK. 239 •[boo jo # q[ I3d smoq ub^i | o || ^ •H *iW *H g •H *dH t £ ® a "gjd QD 5 m ou jC (»a0 05 0>OC-0-^rH«« ,h 'saggo-x pus %* £ 4 eaiiBiiixny 2nt °p,o -pnpujpa;Bino[83 se* — ~ .0 _ 0Q^ •pa^saji [Boo jo a q[iad io;bm. jo'sqi to n § H * §6 • oos*o^o 2 :- S* CSCJi CO 00 ■^ CO tH OS rl OJ ^ 00 cc o i>©l W r-J35c55^Ha5^>Tr..o^i'^^ COJOOOWN^rs«WWNN(MNOi>tf3 •uni uj bitiotj &&&& ^OOQ^COTtt.*i>i ; -;oi~i>in30- Eh s ^ a 1: 3 „ iS " . .2 S c o a c o Eh ° ; o © 4 5 S 5 * 240 ELECTRIC RAIL WA Y HAND BOOJT. *5C&«©«0 K.!Orf«©?CO &t mia^? co co co o> io . oc . ^ 8? <*5 '0 ' -i i y-i r-i i~* i>i>i>OJ> CO iO?> tjj rfi 10 10 i> •*d* -^ ■<# co © ocoq coo 8& :« 3 I c § o- 9 w ^^^ 3 . p p 3 g ^3 05 cs 03 3 '4-3 ♦ • • Sowoa oaow gQ OOOJON nPJHCO • od : a : p . k :8 . o 2 o £ 5q 'pqpq BcraTo'i-r^ hhWCO OOOJK) COHTf© s CO — CO"tf» rfOSOiOS^t 1 OQ0OO CO OS $? £- O o? o o ■ CD O -* ( HnO' *- H — — * rlHrt«H s-is'l c §3 p o 3^3^ slsl •ctrct: o a) o a> 5 I? -a d OihN«0«OHM- 00 J> CO CO CO CO CO CC ~S »~iG0 t^co c«t»THT-jcoa3 co ■ looq rp o Tj5 "^ t> i> t> i.^ O »-" 0* 0C ; C* odod lO CO CO CO CO CO -^ ©i oi <<# ^ ^ -^ co CO o? C0i> cot- CO \fi OJ CO CO — < 0? "*• "■* ©* OJ JCOOOSOOOiOCOCO !>. Ci o o o co* co' co i> i> i.- -rr id lO^i-tCOCO^CO Tt- -o CO O CO CO CO L- O CO CO a. —; coco*coco*cocO'^'tj''«j' ©i co' t^i-i need , coo • B o II E si ; 4* C a -(-; P | +: 4 C, eg : ph • « « M ft Ph w S3 aqwt-co I co ffi'i'«t ©*©*(}* CO CO CO CO 05 I ra- ft c H 7-Tco S n pqcqMpqcc O O th iO""* P- ^Ji JfESSS * MH T-t (-? GO f>00 ooooo OO OCOO COCO-^QOO CO i> i> J> HNO ^cot-^oco^ coo? t-JT lO 00 iH (N lO OS 00 co c7* :o ^j »o" •<* t- CO OJ ^J< ^1 Oi ^ ^J* 3J ^ «oc>ooo«.>-ooo ooaosoNsOo rlOOWOOOr^CO^ co „ u u u j. u „ u m u 45 >* u >» 60 >» >J >% a> >% >* m i .2 Pi 2? .2 *3 -us Pi °5 1 Pi c3 § p OJ Pi P .2 "0 +5 P< . 03 O P .2 & . *o 83g .8.3 £§ cS § SpS C g ® ^ •p fl e § e § CD < Wo ^ k8 CO . c3 «- . ce ^ M • C D t-i . 03 a> «- . C3 i> f- . c3 O CD £ **• AM S A PC S A ft K gp* fi3 g Pt a« 09 ftH s 00*^ CD DD ""' ' OD^ G V BO*" 1 O QC iH «^ ,d ,Q .P rO ,P £J ,P rQ ,P ,Q ,P rQ b* ^ E* t-J B h3 H h3 En ^ ^ kI 75 12.51 20.45 26.8 8.60 23.6 9.98 21.6 11.07 20.1 12.00 18.9 12.90 80 12.95 19.75 27.2 8.46 24.07 9.77 22.07 10.82 20.55 11.73 19.38 12.56 80 13.40 19.05 27.55 8.35 24.44 9.62 22.5 10.60 21.00 11.47 19.8 12.27 90 13.86 18.40 27.95 8.22 24.84 9.45 22.9 10.40 21.4 11.23 20.22 12.01 95 14.29 17.83 2R.3 8.11 25.20 9.32 23.3 10.21 21.8 11.03 20.62 111. 75 10u 14.68 17.32 28.65 8.00 25.60 9.19 23.62 10.06 22.2 10.81 21.00 111. 53 l(fi 15.06 16.89 28.97 7.91 25.30 9.04 23.98 9.91 22.5 10.65 21.35 11.33 110 15.40 16.48 29.3 7.81 26.22 8.91 24.3 9.76 22.85 10.48 21.70 11.14 115 15.77 16.08 29.55 7.73 26.50 8.81 24.6 9.64 23.15 10.32 2^.00 10.98 125 16.45 15.40 30.13 7.57 27.10 8.60 25.2 9.39 23.75 10.05 22.6 10.66 135 17.05 14.81 30.61 7.45 27.6 8.44 25.8 9.16 24.3 9.82 23.18 10.38 145 17.62 14.30 31.1 7.32 28.14 8.27 26.25 8.98 24.85 9.57 23.70 10.14 155 18.14 13.89 31.5 7.20 28.6 8.10 26.7 8.82 25.3 9.40 24.?0 9 92 165 18.61 13.51 31.95 7.10 29.0 7.99 27.2 8.64 25.8 9.20 24.63 9.72 iro 19.10 13.12 32.35 7.00 29.44 7.85 27.6 8.51 26.2 9.04 25.1 9.53 185 19.55 12.82 32.7 rtQn 29.85 7.74 28.0 8.37 26.6 8.90 25.55 9.36 200 20.15 12.42 33.22 6lb0 30.35 ,6X 28.53 8.20 27.2 8.69126.1 9.15 STEAM PIPING. The proper arrangement of piping in a station is such an important matter that the relative location of boilers and engines is largely considered with regarcj to their steam connections. In general, the live steam velocity should not exceed 6CC0fc. to 8000 ft. per minute, the lower velocity being used with slow-speed en- gines; and 3 per cent drop may be allowed at the end of the steam main furthest from the boiler in a single line of pipe. The method of calculating the proper size of steam pipe is to first estimate the effective length by adding to its actual length the number of globe valves, automatic relief valves, separators and T's where the direction of steam flow is changed, and multiply the sum by 5. Then add together all the right angle elbows, and multiply their sum by 3><, and add the number of Y's and T's through which the steam passes without turning, multiplying this sum by 1.6. The sum of these products thus found multiplied by the actual internal diameter of the pipe in inches, and the result in feet added to the actual length of pipe line will 242 ELECTRIC RAILWA Y HAND BOOK. give the effective length. The diameter must be assumed and can be checked from the table, on this page. By obtaining the foot run of pipe, as above, and the pounds of steam per hour at each position required on a single header system, and the pressure losses assigned to these different parts of the piping system, we can calculate the size of pipe. The pounds of steam per hour multiplied by the square root of the dividend obtained by dividing the effective length of pipe in feet by the pounds pressure to be lost gives from the table below, under the column of the initial pressure of steam, the nearest number to that obtained by applying the formula; which is the proper diameter. CONSTANTS FOR IXOW OF STEAM IN PIPES. CD is Gage Presssure. Pounds per Square Inch. 2 100 120 140 160 180 200 2 . Cc nstants = Lbs. of £ >team per Hour X A SB / rt. Run P i Lbs. Loas of Pressure. 1 2 3 228 1,110 3,960 620 3,020 10,800 1.530 7,450 26,600 1,650 8,070 28,800 1,760 8,590 30,700 1,870 9,110 32,500 1,960 9,580 34,200 2,060 10,000 35,900 4 5 6 8,390 15,200 23,800 22,900 41,500 65,000 56,400 102,500 160,000 61,100 110,800 173,000 65,000 118,000 185,000 68,900 125,000 196,000 72,500 132,000 206,000 76,000 138,000 216,000 7 8 9 35,400 50,000 67,700 96,500 136,003 185,000 238,000 336,000 456,000 258,000 364,000 493,000 275,000 388,000 525,000 291,000 411,000 557,000 306,0^0 432,000 585,000 321,000 453,000 615,000 10 11 12 88,700 113,000 141,000 242,000 308,000 384,000 597,000 762,000 948,000 645,000 822,000 1,027,000 687,000 877,000 1,094,000 730,000 930,000 1,158,000 765,000 978,000 1,217,000 805,000 1.025.000 1,280,000 13 14 15 172,000 208,000 247,000 470,000 567,000 675,000 1,160,000 1,400,000 1,670,000 1,255,000 l,510,0i>0 1,800,000 1,340,000 1,610,000 1,920,000 1,420,000 1,710,000 2,040,000 1.490,000 1,800,000 2,140,000 1,560,000 1.890,000 2,240,000 16 17 18 290,000 339,000 392,000 793,000 925,000 1,070,000 1 960,000 2,280,000 2,640,000 2,110,000 2,470,000 2,860,000 2 260,000 2,630,000 3,040,000 2.390,000 2,790,000 3,230,000 2,510,000 2.930,000 3,390,000 2,640,000 £.080,000 3,560,000 19 20 21 449,000 512,000 579,000 1,220,000 1,390,000 1,580,000 3,020,000 3,440,000 3,900,000 3,260,000 3,720,000 4,210,000 3,480,000 3,970,000 4,490,000 3,690,000 4,210,000 4,760,000 3,880,000 4,420,000 5,010,000 4.060,000 4,640,000 5,250,000 22 23 651,000 728,000 1,770,000 1,980,000 4.370,000 4,900,000 4.730,000 5,300,000 5,040,0^ 5,650,000 5,340,000 5,990,000 5,610.000 0,300,000 5,900,000 6,610,000 A loop system of piping, Fig. 192, is installed to give two methods of feeding from the boilers. In this case one side of the loop should be able to carry two- thirds of the aggregate steam demand. If there were liability of a breakdown the loop system is very effective, as repairs on the piping plant can be made while steam is kept constantly on the mains. Statistics on steam pipe breakdowns in railway stations, show a permanent structure with such a remote liability of breakdown that the additional first cost and constant condensation cost is not ELECTRIC RAILWAY HAND BOOK. 243 compensated for. Large plants may be built up on the unit system with a single main, Fig. 103, or the combination system, Fig. 104. A system of smaller pipes shows less first cost than a single lar^c steam main. It is doubtful whether it is profitable to exceed a 24-in main for pressures over 100 lbs. per square inch. e e e Tl e mrfi iTiwa Fig. 102.— loop system. • VALVES bBOILERS e-EA/G//V£$ Fig. 193.— unit system. b r-o e b ' b vr-O Jh 1 1 b ir-O f b r-O b 1 < 0--v i 1 1 1 Jb b 1 ~1> b < i ' Jb e d ! - e 1 1 1 1 e 1 1 1 - 1 e 1 e 1 1 e \ 1 1 e 1 1 1 1 e 1 < e 1 1 e 1 1 e e e 1 1 e 1 1 c FlO. 194.— COMBINATION SYSTEM. In railway work the pounds of steam assumed for pipe calculation should bo in excess of the average hourly demand, for the reason that the overload falls on all engines working in multiple and a fall in pressure varies as the square of the rate of flow, thus affecting the regulation of the engine. The steam pipe, if short, should contain at least twenty times the volume of the engine cylinders fed from !ft ELECTRIC RAILWAY HAND BOOK STEAM PIPES. Velocity of Steam in Pipes Corresponding to a Constant Pressure Loss. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 ir 18 19 20 21 22 23 -5 s IS Gage Pressure, Pounds per Square Inch. 100 120 140 160 180 200 Velocity, Feet per Minute at 1 lb. Loss of Pressure per 100 feet. 12,110 13 0.0 23,450 27,850 32,420 35,300 38,550 41,600 44,500 47,250 49,750 52,100 54,300 56,500 58,600 60,550 62,500 64,500 66,250 68,100 69,950 71,600 73,250 4,440 6,625 8,600 1,800 2,680 3,482 10.230 11,00) 12,930 4,140 4,815 5,230 14,130 15,270 16,330 5,720 6,175 6,605 17,320 18,250 19,110 7,010 7,380 7,730 19,910 20,690 21,500 8,050 8,375 8,700 22,200 22,950 23,650 8,980 9.280 9,570 24,300 24,950 25,600 9,830 10,110 10,390 26,250 26,850 10,630 10,890 1,667 2,482 3,230 3,835 4,460 4,850 5,300 5,720 6,120 6,500 6,845 7,170 7,470 7,760 8,060 8,330 8,600 8,860 9,110 9,370 9,620 9.850 10,080 1,561 2,325 3,020 1.475 2.197 2,855 1.403 2,090 2,716 3.590 4,175 4,540 3,390 3,945 4,290 3,230 3,750 4,080 4,960 5.3)0 5,730 4,690 5,060 5,415 4,460 4,820 5,150 6,085 6,410 6,710 5,750 6,050 6,340 5,470 5,760 6,030 6,990 7,270 7,550 6,610 6,870 7,130 6,290 6,530 6,780 7,800 8,050 8,300 7,370 7,605 7,840 7.000 7,235 7,460 8,530 8,770 9,000 8,060 8,280 8,500 7 660 7.880 8,090 9,210 9,440 8,710 8,910 8,290 8,480 1,338 1,992 2,590 3,080 3,580 3,890 4,255 4,590 4,910 5,210 5,490 5,755 6,000 6,225 6,470 6,685 6,900 7,110 7,310 7,510 7,715 7,900 8,080 it. Engines that staggered badly have been cured by adding local steam storage where there was a throttling action of the steam main for instantaneous demands This is more noticeable in slow-speed than in high-speed engines. As an example of pipe design, we may take a 1000 ihp condensing engine with 120 lbs. boiler pressure. As we may under peculiar circumstances desire to run non-condensing at full load, the live steam and atmospheric exhaust pipes should be designed to carry 23,000 lbs. per hour, assuming 23 lbs. of steam per hp-hour. Overloads if they come are cared for by raising the boiler pressure, as the engine is not intended for regular non-condensing running, but the condenser exhaust should be proportioned for 25 per cent overload— that it*, 1225 ihp at about 16 lbs. per hp-hour or a total of 19,C00 lbs. per hour. Assume the total length of steam pipe to be 75 ft. and that there is one separator, one globe valve and four sharp right elbows in tho line. We thus have (boiler entrance 1, sepa- rator 1, globe valve 1 = 3) X 5 = 15. Also 8j^x4 elbows = 13>£, and the total sum is 28. Assuming D at 8 ins. we have 75 ft. -f (8 X 28) = 299 ft. effective ELECTRIC RAILWA Y HAND BOOK. 24S Velocity of Steam in Pipes Corresponding to "Weight of Steam Delivered per Hour. X « o p 1 Gage Pressure, Pounds per Square Inch. 100 120 140 160 180 200 Velocity, Feet per Minute, per Pound of Steam Delivered per Hour. 1 2 8 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 532. 133. 59.1 33.2 21.3 14.8 10.85 8.30 b.56 5.32 4.40 3.70 3.15 2.72 2.36 2.08 1.84 1.64 1.47 1.33 1.21 1.10 1.00 71.5 17.9 7.94 11.74 2.935 1.305 10.1 2.52 1.12J 8.84 2.21 .982 7.89 1.97 .877 7.13 1.78 .792 4.47 2.86 1.99 .734 .470 .327 .631 .404 .281 .553 .354 .246 .493 .316 .219 .446 .285 .198 1.46 1.12 .881 .240 .183 .145 .206 .158 .125 .180 .138 .109 .161 .123 .0972 .145 .111 .0880 .715 .590 .497 .117 .0971 .0816 .101 .0835 .0702 .0884 .0730 .0614 .0789 .0651 .0548 .0713 .0589 .0496 .423 .365 .318 ,0695 0599 0522 .0598 .0515 .0449 .0523 .0451 .0393 .0467 .0102 .0350 .0422 .0364 .0317 .280 .248 .220 .0458 .0407 0362 .0394 .0350 .0312 .0345 .0306 .0273 .0308 .0273 .0243 .0279 .0247 .0220 .198 .179 .162 .0326 .0294 .0278 .0280 .0252 .0239 .0245 .0221 .0200 .0218 .0197 .0179 .0198 .0178 .0162 .148 .135 .0243 .0222 .0209 .0:91 .0183 .0167 .0163 .0149 .0147 .0135 6.48 1.62 .72 .405 .259 .180 .132 .101 .0799 .0648 .0535 .0450 .0384 .0330 .0288 .0253 .0225 .0200 .0180 .0162 .0152 .0134 .0122 length. Three per cent pressure loss is 3.6 lbs.; and A / * vo - = 8.89 per pound x / 299 \1mF 23,000 lbs. = 204,000. We find in table, page 221, that this corresponds to a little more than a 6-in. pipe. Assuming the pipe to be 6 ins. and calculating over again we get an effective length of 242 ft. and a constant of 188.000, corresponding to a pipe just larger than 6 ins. If the globe valve was replaced by a gate valve, and long sweep elbows used, the effective length would be 95 ft. and the corresponding diameter is just 1 in. smaller: the construction cost would be 20 per cent less, and the condensa- tion would be 17 per cent less than with the other fittings and larger pipe. Material and Sizes of Steam Piping. — Wrought iron pipe and cast-iron fittings are generally used. The pipe is made in ''standard weight," "extra strong " and " double extra strong " grades. Fittings are "light weight," "stand- ard weight" and "extra heavy " grades. Ail pipe and fittings are rated by "nom- inal inside diameter ' of pipe up to and including 12 ins. There is no 13 in. size, and all material above 13 ins. is rated by the actual outside diameter, and 60 246 ELECTRIC RAILWA Y HAND BOOK. specified. Thus a pipe 13.25 ins. inside diameter is called " 14 in. O. D," and the lies t size " 15 in. O. D." * w Standard " pipe should be proved to 300 lbs. per eq. in. hydraulic pressure in sizes lip to 1*4 in.» and to 500 lbs. in larger sizes. Pipe should be good for a working pressure of half its proof pressure. Thus " stand- ard weight" pipe is generally used; but on account of insufficient thickness at the threads, for 1 in., 2J^ in. and 3 in. pipe, if a strong job is desired, especially at boilers and main line taps, "extra strong " is generally specified. Brass pipe is generally used around machinery for gages and oiling systems on account of the ease of bending, better finish and less liability to leak with oil than iron pipe. Copper pipe is used for long sweeps and expansion bends, the pipe generally being riveted and brazed to brass flanges. Soft steel pipe is used where the pipe has to be flanged over the flange ends, and is an alternative for copper pipe in long sweeps or expansion bends. Fittings.— In regard to fittings, "light weight" is good for 25 lbs. pressure and is therefore used for atmospheric exhaust work. Double, galvanized, spiral- riveted, flanged-iron pressure pipes are also used for this work. For condenser exhaust work it may also be used, but the real trouble with light weight fittings is the liability of breakages in making the flanges on the pipe and drawing them up tight if there is any strain due to poor alignment. ■• Standard weight " is very satisfactory for exhaust work and steam pressures below 100 lbs., though it is often used up to 150 lbs., the only objection being that it is difficult to get the flanges tight enough together for high pressure work without breaking the bolts. There is an objection to the extra heavy fittings, in that the number of bolt holes in most of the flanges are not even multiples of 4, and in some cases are an odd number. This arrangement makes a quarter turn impossible. Valves.— Valves are of " globe" and "gate" patterns, the seats being made of a variety of metals. Bronze seats give very good results for globe and gate valves; but exhaust gate valves, and valves seldom used, may have babbitt seats in order to reduce the cost. It is advantageous to have the seats renewable. Valves with outride screw and yoke are often made with a cone top on the stem just under the gland, the top and corresponding seat being ground. This makes it possible to pack the gland under pressure by opening the valve wide. On ac- count of the collection of dirt or scale on [the stem seat, the valve may blow too much for packing under pressure. Valve seats as well as all piping should be cleaned just before erection, and after erection steam blown through them to the air in order to remove unavoid- able dirt. To close a valve no attempt should be made to screw the gate down harder than the manual effort of the handwheel, for it generally results in mar- ring the seat or twisting the stem. A clot of water suddenly released has suffi- cient velocity given it with steam behind it to break cast fittings, therefore care must be faken in opening valves. Babbitt seat valves for condenser exhaust work should be by-passed for 20 ins. and over. Atmospheric exhaust valves do not need by-passes. Bronze seat valves should be by-passed at 12 ins. or 14 ins. and over for 100 lbs., at 8 ins. or 10 ins. for 150 lbs., and at 6 ins. or Sins, for 200 lbfi. pressure. In opening np pipe or fittings shut off from the steam supply by valves, some positive means should be taken to learn that the valves are not leaking danger- ously; therefore before risking the opening of the pipe line, drill a }£-in. hole in the pipe, which can afterwards be plugged. " Steam Pipe Joints.— Tight joints are the combined result of good design and workmanship. The ecrew threads should be perfect. Where flanges are - ELECTRIC EAIL WA Y HAND BOOK. 247 used they should be made up tight, and the pipe ends should not come flush with their faces. All threads, flange faces and gaskets, but not ground faces, should be painted before assembling. For work that is permanent Caliban's cement is very satisfactory, or graphite mixed with boiled linseed oil. If these are not convenient, a mixture of 2 of white lead to 1 of red lead in boiled linseed oil is very good. Red lead alone is liable to crack under strains. In any case the paint should not be thin and should be thoroughly and uniformly applied with a brush over the abutting surfaces. Small pipes are joined throughout with screw couplings, but large pipes should have flange joints at all fittings, and screw couplings elsewhere. Light and medium pressure flanges are screwed on the pipe, and have plain faces. High- pressure flanges are screwed or welded on to the pipe ends, or else the latter are flared over the flange faces and expanded into recesses in their hubs, the portion of the pipe flared over being usually finished for a ground joint. Where ground joints are not used it leads to better results to have the flange faces tongued and grooved. Square-head bolts and hexagon nuts are preferable, and wrenches or span- ners for them should have hardened surfaces and be extra strong so that a small pipe may be used as a lever. A hammer should not be used to turn the wrench. Where cap bolts have to be used, as in attaching to separators, etc., they should have a hexagon head; and considerable care has to be used before assembling, to see that the bolts are an easy fit, and afterwards that they do not break or strip the thread. Gaskets cut out of sheet packing may be used where they are not liable to blow out, as on exhaust pipes and inside the tongue of high pressure flanges, but corrugated copper gaskets are not expensive and are more easily applied and gen- erally better, though in some cases the superior elasticity of heavy sheet packing may stop a leak, even in a live steam pipe, where coppei fails. The hole diameter of gaskets should be between the inside and outside pipe diameters. The outside , diameter should be equal to the inside of the bolt-hole diameter for copper gaskets : and the outside flange diameter for sheet packing unless tongued and grooved I flanges are used, in which case it should equal the inside tongue diameter. Tongued and grooved flanges may also be packed at the bottom of the groove with asbestos or sheet packing. In any case gaskets or packing should have no radial cuts. If flange faces are found to leak after the steam is on, the pressure should be I entirely relieved before attempting to tighten them. If screw threads are found 1 to leak they may sometimes be caulked with soft copper wire with the pressure on or on! as desired. Steam Pipe Supports.— Pipe hanging of the best order is absolutely neces- < sary if tight joints are desired. Trouble with vibration is chiefly due to turns in 1 the pipe being reacted upon by steam puffing through the steam main. Pipe is 1 best hung from short rigid centers in such a manner that the $>ipe may move longitudinally under expansion strains due to heat, but not transversely under i any conditions. Longitudinal vibration will be prevented by the shortness of the suspension radii. Where pipe must be hung on a long radius from above, it may be successfully accomplished by a three-joint or four-joint suspension with the upper suspension ends well spread apart. The suspension rods should have turn buckles and must be provided with means to prevent their transverse vibration if they are very long. Separators, Engine Drains, Etc.— Separators for oil or water depend for their success upon a few simple conditions. On the live steam eide they should be 24 8 ELECTRIC RAILWA Y HAND BOOK. placed as close to the engine as possible. In case dry steam is expected from the boilers, they should be pretty cheap, but if the boilers are expected to prime, no desirable quality should be omitted. Of course, in any case there is no use for them if they will not separate, but the cost is mainly dependent on the capacity and naturally this need not be so large where water is but a possible contingency as where it is a probable one. Large separators, if they can be placed near the engine, undoubtedly equalize the pressure and thereby help the speed regulation. The qualifications of a good separator are that, in entering it, the steam shall immediately change its direction of flow and reduce its velocity. The water or oil having greater specific weight will not change so rapidly and may therefore be thrown to surfaces down which they may run to a receiving chamber. In doing so they should not leave the surface or be blown along it into the current of steam. After reaching the receiving chamber they should be protected from violent waves or rotary motion and preferably also from contact with the steam current. An ample chamber and drain should be provided in steam separators to take care of sudden large quantities of water which may come over by priming. One of the best precautions against water troubles is 20 degs. to 40 degs. Fahr. of superheat in the steam. Pipe coverings help, but condensation in the pipes is not a great source of danger unless it is allowed to collect. The chief trouble is due to the gradual collection of water in improperly drained pipe, and the carry- ing over of large clots of water from a boiler which is priming. In general, pipes should not rise vertically in the direction of flow. If it is necessary to do so, a separator should be placed on the horizontal run as near the riser as possible. In exhaust pipes a drain pipe may be used instead of a separator. Pipes should never rise gradually in the direction of flow, as it is impossible to drain them, though they may slant downwards without harm. Particular care has to be exercised with fittings, particularly reduction fittings, to ensure that they do not partially pocket a run of pipe. The fitting pockets, themselves, should be, and generally are, small enough to be immaterial. Drainage is usually accomplished by pipes ^ in. to 2 in. diameter. 1-in. pipes will care for a good deal of water; and it does not pay to make them too big on account of the cost and radiation. Live steam drainage pipes, which are continu- ously in use, should be covered; they should also be blown out with live steam every six months. They should be provided with valves wherever necessary, but these valves should be periodically inspected to see that they are hard open wherever they should be. In condensing engines the cylinder drain cocks must be piped to the exhaust pipe. In non-condensing engines they may be piped to the exhaust or to waste; never to the drainage system as the water contains oil. As such pipes carry more or less oil they should be of ample 6ize. Where the steam pipes descend to the engine they should have a small drain pipe, about J^in., with a valve just above the throttle to take out the condensed water in the pipe before the throttle valve is opened. The same applies to the steam chest drips. With condensing engines the exhaust pipe should drain itself into the condenser, which should be located below the lowest point in the pipe. With non-condensing engines where the exhaust pipe is not self-draining, there should be s^ in. to 1 in. drips to waste just before the rise, and at the lowest joint of the pipe if there is any other; these drips should not have valves. Automatic cylinder relief cocks should be drained by l^j-in. to 2^-in. pipes to waste so that the drain may be observed from time to time to detect undue leakage at the cocks. The cylinder steam jackets, and the receiver coils of com- pound engines, where the latter acts as a reheater should also be drained. Some arrangements must be made to take the water from the live steam pipes ELECTRIC RAILWA Y HAND BOOK. 249 without opening them to low pressures. One or two traps of large capacity, into which all pipes drain, should be connected with high water alarm whistles and hand by-pass valves. The "steam loop," especially as modified in the "Holly System," gives a positive method. The plain steam loop shown in Fig. 195 draining a separator, operates as follows: The pressure in the separator being supposed to be 95 lbs., and thj boiler pressure ICO lbs., water will rise in the "drop leg" 11^ ft. above the boiler water level so as to balance the 5 lbs. pressure difference. Steam enters the "horizontal" from the separator and is condensed by radiation, the water flowing down the drop leg. More steam rushes up the "riser 11 to fill its place and in doing so entrains water from the separator with it. The whole goes to the "horizontal" where the steam is condensed. Thus the action is continuous. It will be noticed that the condensation in the horizontal being small, the resulting action is correspondingly weak. Also that if the separator becomes filled by a sudden flow of water so that the steam cannot reach the riser, the drainage will gtop. This is the vital objection to the simple " loop." O THROTTLE HOM20NTAL » ■ *- STEAMPft % fVSEA DftOP & a 1 WTERTrnir' m BO/LER CHECK fEPAAATO* Fig. 195.— pulik steam loop. Fig. 196 shows the "Steam Loop and Ilolly Gravity Return System," which is a modified steam loop suitable to practical conditions. A shows the receiver, placed below the lowest point to be drained, into which all drainage water flows by gravity, and which is of sufficient capacity to care for sudden large quanti- ties. / is preferably but not necessarily the highest point to be drained, not much below boiler pressure, and is likely to have condensation water most of the time. This water flows through the suction T, Z, which, on the injector princi- ple, helps to draw water through the header £-2 from such points as may not be so favorably located as /. The water in A passes partly through the perforated plate, i7, and as steam rushes up the riser, C, it has to do so through the perfor- ations over the water surface, which materially assists it to entrain water and carry it through the T, <9, into the discharge chamber, B, which is in reality only the top of the drop leg, D. There is no " horizontal " in this system, i. e., no arrangement for the con- densation of steam by radiation within the system itself; but instead the pipe P takes a small continuous supply of steam to some place where it can be used. Generally the only place where it can be used continuously is the feed water heater, and usually there is but small return for live steam put in there. If the boiler feed pump pumps cold water, a little may generally be fed through the pipe P-% and the "spray " Into B % whick then acts as a condenser and draws the 250 ELECTRIC RAILWA Y HAND BOOK. water up higher in Z>, and the steam rapidly up C, without experiencing the loss previously mentioned. A is equipped with sight gage and loud alarm whistle; Z"-2and E should be furnished with atmospheric discharge valves and pipes for emergency use; the pipes E-Z should have valves close to the header. All valves ezcept checks should be gate valves. P\ is a reducing valve, /'-S, a three-way valve, F is a check valve and the starting valve shown is for blowing out air. The boilers should be interconnected by a steam pipe of ample size to equalize their pressure under all conditions. After the system is once properly started it will run indefinitely without attention. ft [ffi DETAfL OF SUCTtOX TEE I r 70 HEATER TOCONDENSEh TO RADIATORS STARTING VALVE- \OM.WAGE Tig. 196.— holly loop. Coverings.— Steam pipe losses result from friction of the pipe walls, bends and valves, and from radiating heat through the walls of the pipe ; but when either of these losses are reduced in the dimensions of the pipe, the other is increased. As the radiation can be decreased largely by insulating the pipe, these losses cannot be equalled for the least profitable investment. There is no inconsiderable loss from conduction through supports and connections in any steam main; an engine indicator connected to it will show the variations of steam pressure for different steam demands, which can be judged from main ammeter readings. If the volume of stoam contained in the header is known, the effectiveness of the insulation to ELECTRIC RAIL WA Y HAND BOOK. 351 DIMENSIONS OF STANDARD WEIGHT WROUGHT-IRON PIPE. 1J4 & n( i smaller, proved to 300 lbs. per square inch by hydraulic pressure. V/z and larger, proved to 500 lbs. per square inch by hydraulic pressure. Nominal Inside Diameter, A ctual Outside Diameter. Thickness Actual Inside Diameter. Weight per Foot. Threads per Inch. Taper of Threads Inches. H 1 Inches. 0.405 0.54 0.675 0.84 1.05 Inches. O.OGS 0.038 0.091 0.109 0.113 Inches. 0.207 0.364 0.494 0.623 0.824 Pounds. 0.243 0.242 0.561 0.845 1.126 Number. 27 18 18 14 14 1 1.315 1.66 1.90 0.134 0.140 0.145 1.048 1.380 1.611 1.670 2.258 2.694 2 o ■ 2 2^ 2.375 2.875 0.154 0.204 2.067 2.468 3.600 5.773 a a 8 3^ 3.50 4.00 0.217 0.226 3.067 3.548 7.547 9.055 8 8 o Pi 4 4.50 5.00 0.237 0.247 4.026 4.508 10.66 12.34 8 8 < 5 6 7 5.563 6.625 7.825 0.259 0.280' 0.301 5.045 6.065 7.023 14.50 18.767 23.27 8 8 8 8 9 10 8.625 9.625 10.75 0.322 0.344 0.366 7.982 9.001 10.019 28.177 33.70 40.06 8 8 8 £ 11 13 12.00 12.75 14.00 0.375 0.375 0.375 11.25 12.000 13.25 45.95 48.98 53.92 8 8 8 p m g M 0) 14 15.00 16.00 18.00 0.375 0.375 0.375 14.25 15.25 17.25 57.89 61.77 69.66 8 8 P. . ■ 20.00 22.00 24.00 0.375 0.375 0.375 19.25 21.25 23.25 77.57 85.47 93.37 .2 "fO radiation can be obtained by closing the connecting valves from the boiler and to the engine, and noting the fall in pressure, and the time that will give for the header its rate of radiation. In order that this test be reliable the valves must be tested for steam tightness. The valve losses from radiation in steam piping are considerable; in a plant with an output of 4200 hp., non-condensing, the losses i were as follows: Condensation and conduction, .36 per cent; leakage, .83 per I cent; total lbs. of steam per hour lost, 5G00 lbs., at an annual cost of production of $1120 per ycir. Another plant of 2i00 indicated horse-power, condensing, I showed .83 per cent leakage only. The condensation loss may be roughly approximated as equal to .55 B. T. L. per hour, per inch external diameter, per foot of bare pipe, per Fahr. degree 1 temperature difference between the pipe and air. The loss per square foot per 'i hour per Fahr. degree temperature difference is about 2.1 B. T. U. Actually, as 252 ELECTRIC RAILWA Y HAND BOOK. JDIMCENSIONS OF EXTRA STRONG WROUGHT-IRON PIPE. Nominal Actual Actual Kominal Inside Inside Outside Thickness. Weight per Diameter. Diameter, Diameter Foot. Inches. Inches. Inches. Inches. Pounds. V& .20 .40 .10 .29 H ,29 .54 .12 .54 % .42 .67 .12 .74 1 ,54 .84 c!4 1.09 .93 1.05 .15 1.53 1 .95 1.81 .18 2.17 1/4 1.27 1.6G .19 8.00 i^i 1.49 1.90 .20 3,64 2 1.93 2.37 .22 5.02 2Vz 2.31 2.87 .28 7.67 3 2.89 8.50 .30 10.25 3^ 3.35 4.00 .32 12.47 4 3.81 4.50 .34 14.97 4^ 4.25 5.00 .35 17.60 5 4.81 5.56 .37 30.54 6 5.75 6.62 .43 28.50 7 6.62 7.62 .50 37.60 8 7.50 8.62 .56 47.85 IRON PIPE SIZES OF.SEAMLESS DRAWN BRASS AND COPPER TUBES. "Will thread to fit iron pHe fittings. Approximate "Weight per Ft. Iron Tipe Inside Diameter, Outside Diameter. Length Feet, about. Size Brass. Copper. H ,27 11 12 .30 .31 i .36 ft 12 .43 .45 .49 n 12 .58 .61 .62 ii 12 .80 .84 % .82 i& 12 1.17 1.23 i 1.04 ia 12 1.67 1.75 a 1.38 w 12 2.42 2.54 1.61 m 12 2.92 3.07 2 2.06 *% 12 4.17 4.38 w% 2.46 *A 12 5.00 5.25 8 8.06 8^ 12 8.00 8.40 &A 3.50 4 12 10.00 10.50 4 4.02 4% 12 12.00 12.00 5 5.04 5.56 8 to 10 15.93 17.30 6 6.06 6.62 6 to 8 20.69 22.38 7 7.02 7.02 Special 26.28 27.77 8 7.98 8.62 Special 29.88 33.69 ELECTRIC RAILWA Y HAND BOOK. 253 Pounds of Steam Condensed per Jlour per Foot of Covered Pipe, Covering 1 inch thick, having conductivity of \^% Temperature of Air 80 clegs. Falm External Steam Gage Pressure, Pounds. Diam. of Pine, Inches. 80 100 120 140 160 180 200 1 .050 .054 .057 .060 .053 .065 .067 2 .101 .108 .114 .120 .125 .130 .134 3 .151 .162 .172 .180 .188 .195 .208 4 .202 .216 .229 .239 .250 .260 .269 5 .252 .269 .286 .299 .312 .325 .34 6 .302 .323 .34 .36 .38 .39 .40 7 .35 .38 .40 .42 .44 .45 .47 8 .40 .43 .46 .48 S-0 .52 .54 9 .45 .48 .51 .54 .56 .58 .61 10 .50 .54 .57 .60 .63 .65 .67 12 .60 .65 .69 .72 .75 .78 .81 14 .71 .75 .80 .84 .88 .91 .94 16 .81 .86 .91 .96 1.00 1.04 1.08 18 .91 .97 1.03 1.08 1.13 1.17 1.21 20 1.01 1.08 1.14 1.20 1.25 1.30 1.34 22 1.11 1.19 1.26 1.32 1.38 1.43 1.48 24 1.21 1.29 1.37 1.44 1.50 1.56 1.61 ECONOMY DUE TO SUPERHEATED STEAM. Amount of superheat Boiler pressure, gage Temperature of superheated steam Indicated hp Lbs. of steam per lb. of coal Lbs. of steam per 1 hp-hour I 19.75 Lbs. of coal per 1 hp-hour ! 3.147 Per cent saving in steam due to superheating ■ j 20.9 •• " " " coal " •' " j 17.6 118.3°F. 99. 455.3°F. 491. 6.024 15.63 2.593 III a m 126.9°F. 94. 4(30. 4°F. 502.3 6. 15.61 2.513 20.9 20.1 the pipe temperature increases, the loss increases a little faster. Another method of comparing the value of different coverings \z to have a cone of hoat-insulating material fitting over the pipe surface to be tested, and with the same steam tem- perature for the different samples measure the rate of rise of temperature in the air space inside the cone. The loss from covered pipes depends on the thickness, kind and quality of the covering and somewhat on the extent to which it is com- pressed. Coverings in common use are carbonate of magnesia and asbestos, the latter being sometimes combined with other materials such as hair and woolen felt. A 254 ELECTRIC RAILWAY HAND BOOK. little asbestos is usually put in magnesia, coverings to bind the material. So called "Magnesia Covering " is usually 1 in. thick up to 12-in. pipe, over which it runs 1 J4 i ns « *° 1/6 * ns - * n thickness. This covering contains practically no magnesia. The loss for two samples having a small percentage of good asbestos was .65 and .87 B. T. IT. per sq. ft. per hour per Tahr. degree temperature difference, /. e. y the loss was 31 per cent to 41^ per cent of uncovered pipe. The latter value was for the denser sample. Pure asbestos generally gives the same or a trifle greater loss than bare pipe. "Air Cell Asbestos " about equals " Magnesia Covering." The following tcst3 of commercial coverings were made by Geo. M. Brill and reported in Trans. A. S. Mo E., Yol. XVI., page 827. A length of CO ft. of 8-in. steam pipe was used in the tests, and the heat loss was determined by the condensation.. The steam pressure was from 109 lbs. to 117 lbs. at the gage, and the temperature of the air from 53 degs. to 81 degs. Fahr. The difference between the temperature of steam and air ranged from 263 degs. to 286 degs., averaging 272 degs. REPORT OF Ti*STS ON STEAM PIPE COVERINGS. (G. M. Brill.) ! 0> u & P u u o P* a) d . ft. per hour average dif- perature. 00* -1-" Is O M o ■*» S3 §8. IS. Kind of Covering. © o w 5 CD "^ o a o o v OS** CO * ftgo i_* be o o o d *-« 5 3 CS.T o > S b* « Hp§ 33 9 03 « h3 pq'p«2 Ulxi P4 H'3n-l Bare pipe , . • . • • .846 12.27 2.706 100. 2.819 Magnesia 1.25 1.60 1.30 .120 .080 .089 1.74 1.16 1.29 .384 .256 .285 .726 .766 .757 14.2 9.5 10.5 .400 Hock wool ...• .267 Mineral wool .297 Fire-felt 1.30 1.70 .157 .109 2.28 1.59 .502 .350 .689 .737 18.6 12.9 .523 Manville sectional... . .564 ! Manv. sect. & hair felt 2.40 .066 0.96 .212 .780 7.8 .221 Manville wool-cement 2.20 .108 1.56 .345 .738 12.7 .359 Champ, mineral wool. 1.44 .099 1.44 .317 .747 11.7 .330 Hair-felt .82 .75 .132 .298 1.91 4.32 .422 .953 .714 .548 15.6 85.2 .4?9 Riley cement .093 dossil-meal .75 .275 3.99 .879 .571 325 .916 Exhaust Piping*. — In long exhaust pipes, radiation is very objectionable, for the steam becomes ladened with the condensed moisture and the weight opposed to the engine exhaust is greater. Every pound of pressure lost in this way cuts off just so much from the bottom of the indicator card andcalbfora higher steam supply to do the same work. It is therefore imperative that this loss should be kept down as far as possible In exhaust pipes. In condensing engines the exhaust pipe may generally be made comparatively short, though the Mo E. P. of condensing engines is rather lower than in the non-condensing type. The loss should be a function of the length of pipe, and in either type of engine ELECTRIC RAILWA Y HAND BOOK. 255 should not exceed 1 lb. per 100 ft. actual run plus \i lb. In exhausts over build- ings the weight of the steam column must be added to the back pressure, and the size of exhaust pipes will show economy for larger sizes than the formula for live steam pipe indicates. An exhaust connected to a vacuum must be air tight in order that the condenser will not fail at full load. Exhaust Heads should be placed on atmospheric exhaust pipes where the noise or water and oil cf exhaust are objectionable. Cases have occurred where the exhaust pipe opening over the building roof and only one or two engines running, the roof and floors below have been set into periodic vibrations by the varying air pressure over the roof. The exhaust head is in principle a good deal P/TT6Bl/ftGH £CUP$£ Fig. 197. — types op exhaust heads. euhot like a separator, in so far as the separation of fluid is concerned; the reduction of noise is accomplished by the reduced velocity of the steam from the end of the cone and also by the steam chamber action similar to the air chamber principle in water pumps. Fig. 197 shows several types of modern exhaust heads. EXHAUST-STEAM CONDENSERS. The Jet Condenser.— This consists of a chamber into which the exhaust Bteam and a jet of cool water are conveyed, the exhaust steam being condensed by actual mixing with the latter. The volume of this condensing chamber is ordinarily from one-third to one-half that of the cylinder of the engine. The water of condensation acting directly upon the steam will make a given lowering of temperature of the exhaust steam with less weight of water and less bulk and weight of condenser. To condense steam requires from twenty to thirty times the weight of water in cool seasons or climates, and from thirty to thirty-five times with warm water, as shown by the table on opposite page. Where the condensed steam is to be pumped back into the boiler, the injection water goes with it; therefore it must be water that is not objectionable for use in boilers. The total heat contained in 1 lb. of steam as it leaves the low pressure cylinder of a condensing engine, is about 1138 B. T. U. above that contained in 25° ELECTRIC KAIL WA V HAND BOOK. COMPARATIVE WEIGHTS OF INJECTION WATER AND STEAM. Temperature of Hot Well. Degs. Fahr. 100 110 120 130 140 Corresponding Back-Press, in Cylinder. Lbs. per sq. inch 0.94 1.27 1.68 2.21 2.88 Temp, of Injection Water, Fahr. 40 50 70 80 £0 Ratio of weight of injection water to weight • of steam. 17.8 21.4 26.8 35.7 53.5 15.1 17.7 21.2 26.5 35.3 13.1 15 17.5 21.0 26 3 11.6 13.3 14.9 173 20.8 10.3 11.4 12.9 14.7 17.2 107.0 53.0 35.0 23.0 20.6 1 lb. of water at 32 degs. Fahr. and the weight of water required to condense this 1138-4-^-4- 7* 11^0 T* Bteam is 11"°', — = ~~, — — - where T is the temperature of the hot . * "~~~ * l ' — /, well, and / the temperature of the injection water, ris usually from 100 degs. to 120 degs. Fahr. The area of the injection pipe is approximately where Wis the weight 130 4/ ^ of injection water required per minute in lbs. and h the head of water in feet. The Surface Condenser.— This type differs from the foregoing in the fact that the exhaust steam is not mixed or brought in actual contact with the water Fig. 193.— method op securing tubes in surface condensers. which condenses it. In the snrface condenser the steam is separated from the cool water by metallic partitions or tubes, the ordinary arrangement being to pass the cool water through brass tubes around which the steam is caused to circulate, or vice versa. The condensing surface required is usually from lj^ to 3 square feet per indicated horse-power. The surface condenser, while more heavy and bulky to handle for cooling a given weight of steam cliacharged as exhaust, can be used with any kind of water. The condensed steam can be used again in the boiler, but the effect of distilled water is to increase the corrosion in the boiler, and from 10 per cent to 12 per cent of its weight of fresh water has to be added in order to reduce this effect. ELECTRIC RAIL WA Y HAND BOOK. 257 ? Oil must be separated from the steam, and oil separators should be used in the exhaust steam main before it enters the condenser. Tliis method is useful w'v the only available water contains solid matter, salts or acids which woi" TOWEH HOT WATER. COLO WATER. SVCTlON TANK Fig. 199.— self-coouno condenser. Injurious to the boilers. The same water is used over again and so the steam circuit is practically a closed one. The brass tubes of the condenser are solid drawn, and are pcnerally tinned outside and inside. They vary in diameter from ££ in. to 1 in., but generally are $£ in. outside diameter. Such tubes are about ,048 in. in thickness. 258 ELECTRIC RAILWA V HAND BOOK. The tubes are generally secured to the tube plates by screwed glands and stuffing boxes, packed with cotton cord or a ring of thick tapes, as shown in Fig. 198. They are placed zigzag, and their pitch measured from center to center, may be from 1.5 to 1.7 their diameter. The thickness of tube plates equals the diameter of the tubes iu inches, plus -in- Fig. 200.— syphon or injector condenser. ^ In the surface condensers of modern triple expansion marine engines the amount of cooling surface is from 1.1 sq. ft. to 1.5 sq. ft. per indicated horse- power. Prof. Yrhitham's rule for the amount of cooling surface is : Where 3* equals cooling surface in sq. ft. IV " weight of steam to be condensed per hour in lbs. T " temperature of steam to be condensed. / " mean temperature of circulating water which is the arithmetical mean of initial and final temps. Z, •• latent heat of steam of temperature T, ELECTRIC RAIL WA Y HAND BOOK. 259 WL Then S= „ r 180 (r— *) If T equals 135 degs. and / about 75 degs., then i* 17 IV 180 The amount of cooling water required is determined in the same way as for jet condensers, except that it must be noted that the temperature of the cooling water as it leaves the condenser is not the same as that of the condensed steam. The formula for determining this weight is as follows: Where H — total heat in 1 lb. of steam above that contained in 1 lb. of water at 32°. T — temperature of condensed steam. / = " of circulating water as it enters coudenser. t x = " of circulating water as it leaves condenser. W— weight of circulating water (in lbs.) required for each lb. of steam condensed. Then W = /f+32 Fig. 801.— injector condenser with pump. 26o ELECTRIC RAILWAY HAND BOOK. Self-Cooling Condenser.— This type, shown in Fig. 199, consists of twc parts: the condenser in which the exhaust steam of the main engine, or engines, is condensed, and the tower in which the heated discharge water from the con- denser is cooled to proper temperature, to be used again in the*condenser for the further condensation of the exhaust steam. As this process is carried on contin- uously, only a very small supply of circulating water is required. The heated water falling through the tower is cooled by three processes: first, radiation from the side of the tower; second, the contact of cool air; and third, evaporation. The cooled water falls from the grating to the subsiding tank at the bottom, and is fro'm there drawn by the condenser to be again employed in condensation. The current of air is passed through the tower by a circulating fan. The Siphon or Injector Condenser. — This condenser is shown in Fig. 200. The exhaust steam receiving a downward direction in passing through the goose neck at the top of the apparatus passes through an inner cone, surrounded by an annular cone of water. The steam is condensed in this conical space, and falls with the injection, whose velocity is so graded by the cross section of the condenser that air in the injection is entrained and has no opportunity to remain in the space where the vacuum is. The small vacuum cone being continually filled and emptied prevents the trouble from air. There is no air pump, but the injection pump is required as before. Where a height of water of 9 ft. to 12 ft. above the hot we'l is available a natural flow of water can be used instead of the supply pump. The Injector Condenser with Pump. — There are many places where the height required for the long leg or siphon of the barometric condenser is inconvenient. This has given rise to a design of condenser, Fig. 201, in which the small bulk of the injector and its efficient action are combined with a pump to maintain the vacuum by continually drawing off the water and the air. The exhaust steam enters through the inlet fi, which is controlled by an inner pipe C, that carries a deflecting nozzle D ; this throws the injection in a finely divided state into the annular exhaust steam passage F, and the air pump below continu- ously draws off the water mixed with air, to which a higher velocity is given by reducing the cross-section, so that the bubbles of air once caught in the water have no chance of rising into the vacuum space below D, STJEAM TURBINES. There are many ways of classifying steam turbines. They may be classified as axial or radial flow, as the flow of the fluid is axial or radial ; as impulse or reaction turbines according as the pressure in the space between the fixed blades and the moving blades is the same or greater than that at the exit, etc. The maximum efficiency can be obtained, only when superheated steam in connection with a condenser giving a high vacuum is used. Superheated steam prevents water hammer which is detrimental to the blades, and more over tests have shown an increase in efficiency of about 10 per cent, for each 100 degs. Fahr. superheat, through the load range. A test on a 2,000 k. w. turbine, carrying 1,800 k.w., showed that when the vacuum was raised from 26 to 27 ins. the economy increased 5.2 per cent., and when still further raised to 28 ins., the vacuum increased to 6.75 per cent. The turbine requires facilities for superheating and abundant water for condens- ing. The ratio of the weight of condensing water, to steam condensed is often as high as 70 to 1, depending upon the temperature and the adaptability of the condenser to this special class of duty. ^ ELECTRIC RAIL WA Y HAND BOOK. 261 Counter current condensers show the best results. The steam spaces between the tubes should be large, and the ratio of cooling surface to pounds of steam should be 1 to 8 or better. The condenser should be located as near as possible to the turbine and in some types it is practically a part of the turbine. Turbine speed i. e., the peripheral speed of the wheel must be high in order to produce an economical use of the steam. The blades may form an integral part of the wheel or are recessed into the rim and held in position against the impact of the steam by steel lacings passing through the blades and binding them together. The great stresses which these wheels are subjected to by centifugal force, and the reaction from steam pressure necessitate the most careful selection of material. R.PM. 1500 [ 1450 1400 40000 36000 32000 Vacuum j Lctua 28" _Vac uum V Com icted o23' aouu m s 26* 25 ^ jf f^ # 3 24000 20000 16000 12000 8000 i Cffi f > a 1 4 ffl" g 3 1 .Wate 5 O ii 3 mmmm ' — — ■ — ( lctua orrec ed (3 * I Full 1 Load ^ _ L >ad-B rake ! lorSe Powe B 8 20 18 16 12 10 200 400 600 800 1000 3000 4000 Fig. 201-A One method of construction now used in turbines for street railway work is to have the generator located directly above the turbine and the rotating parts sup- ported by a balanced thrust bearing. Some general data of these turbines is given below: Output k.w 15 500 1.500 1,500 5,000 Revolutions per minute 3,000 1,800 1,800 800 500 Weight with dynamo 1,830 36,lt0 94,800 121,250 385,800 The governing may be done in several ways, each of which has its own peculiar characteristics. Some are controlled by ordinary throttle governors, and others by electrically or hydraulically operated valve mechanisms which are designed to do away with the loss due to wire drawing. The relative floor space occupied by a turbine and a equivalent reciprocating engine, in this case of 2000 k. w. units, is as follows : A three cylinder engine 38 in. x 58 in. x 58 in. x 54 in. driving a 2000 k. w. generator, occupies 1230 sq. ft. of floor space or 0.61 sq. ft. per k. w. An equivalent Parson turbine generator unit 333 sq. ft. or 0.17 sq. ft. per k. w. This difference will be greatly in favor of paying higher prices for property, which is located near the most economical center of dis- tribution, also the head room required for crane facilities are less with the turbine. The time required to start and get on the line with a steam turbine is claimed to 262 ELECTRIC RAILWAY HAND BOOUT be less than that with an equivalent Corliss engine. Starts from rest to phasing in being made in one minute to one minute and a half. The safety in a quick start lies in the clearances between the blades and casing, these clearances can be made from y& in. to T 3 S in., it is claimed, without interfering with the steam economy. The starting of a Corliss engine of quivalent capacity from a standstill to full speed, if hot all over, requires at least five minutes. A by-pass to heat a turbine before it is required to be put in operation is a safety precaution if it can be arranged to get rid of all entrained water. There is no available data for the all day efficiency of steam turbines operating under railway loads, but test runs show an economy slightly better than Corliss engines operating under the same loads, but in order to excell the Corliss engine both superheating and higher vacuum are necessary. TESTS OF WESTINGHOUSE-P ARSONS TURBO- GENERATOR. B O .2 0° a A H 02 o o a o (-1 03 h ■ ft a; d o 0) d cog d o o P £ & .5 . o No. Date. M M o a* OQ 02 .4 1 03 d g.4 03 O ft^a & M o3 d H ft pa I— 1 d. M £ JH QQ . :8fc 3 885 S 155.5 M 24.13 M 1 Jan. 27 748 580 6 30.70 26.22 32.17 2 " 28 1657 18*0 1480 6 151.3 30.73 28*00 40.08 61.34 19.85 15.15 20.2 3 Feb. 1 1998 2185 1900 4 155.4 30.27 26.91 41.56 55.05 32.45 14.43 19.10 4 May 7 471 730 310 6 121.8 29.86 26.62 19.10 29.00 3.50 23.97 31.96 5 44 8 888 980 750 6 152.6 30.04 25.83 32.90 47.50 12.00 19.90 26.53 6 44 9 1371 1570 1110 6 151.9 29.81 26.26 32.10 38.60 12.50 16.46 21.94 7 44 12 834 940 660 6 153.2 30.26 27.26 35.40 45.10 20.10 18.50 24.60 8 44 13 364 520 150 6 153.1 30.06 27.40 29.00 45.00 2.50 25.10 33.47 Fig. 201a shows curves plotted from the efficiency test of a 1500 k. w. turbine working with steam at 150 lbs. pressure per square inch. GAS ENGINES. These are used successfully in several railway plants, and, where the gas is used directly from the holder without distribution expense, they show a high thermal efficiency, and their cost of operation per kw-hour compares favorably with a steam plant. A multi-cylinder engine is necessary in order to give constant voltage, the flywheel has to be large, and the capacity of the generators should be considerably under the capacity of the gas engine. In gas engines the con- sumption of coal gas may be taken at 20 cu. ft. per 1 hp-hour, 24 cu. ft. per brake hp-hour. With engines of 100 hp the equivalent coal consumption was 1.1 lbs. of coal per hp-hour, and the mechanical efficiency, 85 per cent; this improves with larger engines. WATER POWER. The value of water power for railway work depends upon the supply being ample at all times and seasons to operate the whole load. Small water powers, u6 a vaiiable supplementary, do not often show an economy sufficient to encour- ELECTRIC RAIL WA Y HAND BOOK. 263 age their development, where the location is not adapted to directly supply the railway system. The water flow should be measured at that season of the year, which has the lowest flow for the water shed drained by the stream. In order to determine the flow of water in an open stream, where the channel has a fairly uniform depth and width, twelve to twenty equidistant measurements across the channel should be taken from the bottom of the channel to the surface of the water; their sum divided by their number will give the average depth. The velocity of the flow of water can be measured by the time required for a float to pass between two parts located 100 ft. apart. As the surface of a stream at the center moves approximately 83 per cent faster than the sides and bottom, certain allowaiice has to be made. The cross section in feet multiplied by the BESUI/TS OF TESTS ON A "CYLINDER GATE" VICTOR TURBINE. Horse Bead Involutions Cubic Feet Power developed by Wheel. Percentage Size op Wheel. in of Wheel Water Useful feet. per minute. per minute. Effect. 30-inch Full Gate 3751 168 4440 119.56 81.35 Vs " .... 37.62 163 3892 104.93 80.03 % M .... 17.95 163 3392 88.24 76. G6 % " .... 18.10 155 2 93 70.97 71.28 x " .... 18.20 159 2265 51.42 63.46 36-inch Full Gate.... 16.78 135 6106 158.18 81.80 >8 "" « • • • 17.14 135 5J22 141.58 80.71 H " •• 17-35 140 4708 118.22 76. C8 % 4fr ••-■ 17.05 129 39C2 91.62 71.50 \k " .... 17.48 134 3-02 66.87 63.30 39-inch Full Gate .... 14.66 116 6873 152.66 80.37 "i :: :::: ft -:::: 14.53 118 5920 129.41 79.80 16.84 125 5517 135.56 77.40 17.06 123 4005 10S.22 71.67 17.39 124 3856 81.00 64.07 48-inch Full Grt3 . . . . 13.23 91 10072 201.71 80.11 Vs *' •••■ 14X6 89 9042 192.41 78.42 % '* .... 14.75 89 78G9 165.23 75.34 % " 14.87 85 6744 132.76 70.06 - i 15.28 87 5526 100.66 63.09 velocity in feet per minute, will be the discharge in cubic feet pjr minute. By taking levels to obtain the height of fall that can be secured, and multiplying this height in feet by the cubic feet per minute, multiplied by C2.C6 lbs. (the weight of 1 cu. ft. of water at 60 degs. F.) and dividing by 33,000 the gross horse-power of the water power can be obtained. In estimating the recoverable water power, allowances have to be made for the turbines of from 75 per cent to 85 per cent efliciency at full gate. There is also loss of head due to weirs, and the necessary drop to produce flow in the flumes and raceways, and in addition the suction effect on the turbines in a penstock is sometimes reduced by the presence of air decreasing the draft tube effect. All these losses combine to reduce the possible recovery of power and should be given, ample allowances in estimating water powers for railway work. 26a electric railway hand book. Turbines, etc.— The power obtained by turbine wheels is due to the impact of the water against the curved buckets attached to the rotating shaft. The form and angle of these buckets and their spacing varies with the different types of turbines. For heads above 100 ft., the Pelton wheel can be used effectively, and the regulation by a deflected nozzle responds much more readily to the load changes than in the case of turbines, where the regulator opens and closes the gate and a£ccts the flow of water through the turbine. Automatic regulators, however, have been made for turbines, which give very satisfactory results for railway work, if close attention is given to their adjustment. Overshot wheels, due to their large inertia value, show some points in favor of their use in railway work for small plants, but their efficiency is so poor as to prohibit their use except where there is an abundant surplus of water. The table given herewith shows the results of tests on a Victor Turbine, made by the Stilwell-Bierce & Smith-Vaile Co., carried out at the testing flume of the Holyoke Water Power Co., Holyoke, Mass. THE RAILWAY GENERATOR. The generator it, the most economical transformer of energy in the station, and consists essentially, in its simplest form, of two parts : the armature, which in revolving induces a potential in the copper conductors wound on its surface, when these conductors pass through or across a magnetic field; and the field magnet, whose function is to produce a flux or flow of magnetic lines through the revolving armature. The successful design of a generator is the happy com- promise of many conflicting losses, and it is not within the scope of this hand book to discuss these complex relations, which can be found fully treated in " Dynamo Electric Machinery " by Sylvanus P. Thompson, and similar books. Efficiency. — As all losses in the generator appear in the form of heat, the temperature is the criterion accepted as a gage of efficiency. The field magnets require a certain amount of energy developed by the armature; this varies from .75 per cent for generators above 500 kw to 1.8 per cent for 150 kw generators. The temperature of the field should not rise more than SO degs. Cent, above the air, the temperature of the air being 20 degs. Cent, by the thermometer; or show by resistance measurements a resistance greater than that corresponding to 45 degs. Cent.; 1 watt per square inch of external surface of field gives a rise approximately of 62 degs. cent, by the thermometer. For field surfaces the ap- proximate requirement is 15 sq. ins. per kw output. The Field. — A railway generator, is usually compound-wound. In this type there are two separate systems of field coils, one is the shunt, which is connected across the full potential, and in series with this circuit there is a rheostat for varying the current through the shunt coils, in this way changing the magnetic field through which the armature rotates. The resistance of this rheostat should be sufficient to bring the potential of the armature, run on open circuit at full speed, 20 per cent below the normal bus voltage; this requires in different typos of generators a rheostat resistance of from ^ to 2 times the field resistance. The other, the series fields, increases the field intensity due to the current from the generator passing around the field, and tends to maintain the potential of the generator. The generator is over compounded for increasing loads so that the degree of compounding required depends on the drop of potential on the distri- bution system which the generator supplies. Twenty per cent over compounding ELECTRIC FATLWA Y HAND BOO FT. 265 is the usual amount employed for railway work, but the compounding coils can be shunted by a resistance to reduce this effect to any desired per cent of com- pounding. This is the usual way that manufacturers adjust machines for com- pounding requirements under 20 per cent. There is a condition arising in railway plants extending over considerable territoiy, where the line drops are considerable but not sufficient to warrant a booster, w r hich can be met by having the generator wound to give two percentages of compounding. This can be done by opening the shunt coil around the series compounding, and separating the outlying feeders from the short feeders on the switchboard. It also requires two equalizing busses, and the generator can then be operated for a large percentage of over-compounding to make up for the line drop and operated independently on the long feeders, which will produce better potential delivery at their ends. The watts lost in each field should be equal to the product of the drop across each field and the current flowing through it. If the fields vary, it is due to short-circuited turns on the field or poor connections; the former is usually found in overheated shunt fields, and the latter in the series field connections. The Armature. — This is composed of discs of thin sheet iron or steel as- sembled on the shaft. The modern armature has slots on the periphery of the armature body, through which the windings pass. The current density varies from 300 circ. mils pe7 ampere to 800 circ. mils per ampere in the ventilated types; 800 circ. mils per ampere to HOC circ. mils per ampere in the unventilated types. The ventilation in the armature is effected by separating the different groups of discs by an open spider, which allows the air to pass from the interior of the armature body to the exterior, it being thrown out by centrifugal force due to the rotation of the armature. The energy lost in the armature is due to the resist- ance of the armature windings, the internal drop varying from 7 per cent for 50-kw to 2 per cent for 2000-kw machines. The other loss in the armature is due to hysteresis and eddy currents set up in its iron which acts as a conductor cutting the magnetic field. These losses aggregate from 6 per cent for a 5j)-kw generator to 2.6 per cent foral200-kw generator. The radiating surface per kw output in armatures should be from 20 sq. ins. to 18 sq. ins. in a 1200-kw generator. The peripheral speed varies from 2000 ft. to 3200 ft. per minute. The insulation resistance should be at least from 1 to 3 megohms, cold, and 1 megohm to 750,000 ohms, hot, and should be subjected to an alternating pressure of 2500 volts for 5 miuutes without break down. The latter test is much more severe when made while the armature is hot. The Commutator. — This is built up of a number of segments of copper insulated from each other and from their mechanical support by mica. The purpose of the commutator is to rectify the alternating currents which are induced in the armature coil when passing from one pole face to the next. The conditions of the commutator require that as few turns as possible be connected to the brush at the same instant, for, when in this position, the coils are short-circuited and a local current circulates through them and the brush bridging the coils. The electromotive force, due to the flux distribution in the field and the number of bars connected to the coils between pole centers, limits the possible local current that can circulate in the coils under the commutator. The sparking, which the opening of this circuit produces, is shown in the character of wear on the commutator. A dark bronze uniform color of the commutator is the desired surface. Sparking produces bright metallic lines around the commutator, or pits the commutator next the mica segment towards 266 ELECTRIC RAIL WA Y HAND BOOK. the brush in the direction of rotation. The terms of undue and excessive sparking, as applied to commutation in the generator, are relative and the proper commutation should be defined by the character of the surface and wear on the commutator. The radiating surface on a commutator varies between 5% to 3££ sq. ins. per kw output. Brashes and Brush-Holders.— The brushes for railway work are usually of carbon. The curvature of brush should be fitted to the commutator so as to give uniform wear to the segments and mica insulation. It should be hard enough to produce a gloss without cutting, and not soft enough to give a black film on the commutator. The brushes must be of uniform density and should show a bright contact surface which has the appearance of graphite. The loss in potential between the commutator bar and the brush should not exceed 1.4 volts at full load, and the current density on the contact surface should not exceed 40 amps, per square inch. »In early generators there was from % volt to 1 volt loss between the carbon brush and its holder ; this has been reduced by con- necting a flexible lead between the brush and its holder, looping out this contact resistance and reducing the drop to % of a volt or less. The brush-holder should be sufficiently flexible to yield to any inequality in the rotating commutator. and yet produce a pressure of 1.2 to 1.7 lbs. per square inch of contact surface. The movement of the brush-holder to accommodate wear should be such as to keep the brush parallel to the wearing arc. The best position of the brushes relative to the commutator is generally marked on the rocker, or its position can be found on loading the generator, and finding the point of least sparking for the normal load. For polishing a commutator use only sandpaper. Have a concave wooden segment covering about J4 °f the commutator, between the brushes, to which is secured the sandpaper, and which is pressed against the commutator surface with the armature revolving at moderate speed. Bearings. — These are about 3.8 time's tr^ir diameter in length for 100-kw to 300-kw, and from 3.2 to 3 above 300-kw direct-connected generators. The engine manufacturer supplies the shaft and engine bearing, and the dynamo manufact- urer supplies the outboard bearing, when one is required. The armature is built up on a keyed sleeve which is forced on the engine shaft. The engine manufact- urer usually supplies the dynamo manufacturer with a gage from which the arma- ture sleeve is bored, so that it will take the desired pressure in tons to force the armature home on the shaft. The bearings are always self-oiling either by rings or a chain, revolving over the shaft, the lower extremity moving in the oil well. Where the oil circulating system is used, oil is kept flowing into the oil well and drained from an overflow pipe, keeping the oil at a constant level. This has the additional effect of cool- ing the bearings. As the air, passing through the armature, draws any oil spray thrown by the reciprocating parts, the armature should be screened from it, for manufacturers justly demand, that in order to maintain their guaran- tees, the armature must be free from oil. Oil carbonizes at the temperatures attained, and acts as an adhering surface to which carbon dust from the brushes can cling, thus tending to break down the insulation of the armature windings. The armature shaft bearings should be scraped to a fit and be adjustable for wear, to maintain the alignment of the armature in its field. In belted machines, above 250 kw, it is considered good practice to provide an outboard bearing to carry the belt strains. In direct-connected generators the bearing friction is ELECTRIC RAILWAY HAND BOOK 267 about 0.32 per cent of delivered horse-power, and in belted, from .62 per cent to 1.25 per cent of delivered horse-power, depending upon the size and bearing design. SIZES AND APPROXIMATE DIMENSIONS OF WESTINGHOUSE STANDARD ENGINE-TYPE RAILWAY GENERATORS. £ So 0Q O S A B C D E F M ci O Pi < £ Ins. Ins. Ins.' Ins. Ins. Ins. 36 36 36 150 150 200 270 270 364 8 8 8 200-225 170-185 200-220 18,700 24,000 24,400 15H 15** 13 27** 30 30 46*4 51*4 51*4 82% 94*4 94*4 120 120 • 120 250 250 250 455 455 455 8 8 10 150-170 120-125 90-100 30,600 43.200 50,000 17 16 18 30^ 36 32 54% 62** 67*4 101 114** 124*4 120 144 144 36 39 39 300 825 400 546 590 725 8 8 10 145-160 90 90-100 43,400 61,500 78,000 17 19 22 39 38 39 62** 71*1 73% 114** 131 ^ 138^ 144 144 176 39 39 47** 500 500 800 910 910 1455 8 10 10 150 90-100 80-90 62.000 100,000 135,000 18 21 24 38 39 48 71** 80*4 88*4 131% 152% 166% 144 171 204 39 42 52** 1050 1200 1500 1910 2180 2730 12 12 14 80 75-80 75-80 147,000 185,000 200,000 22 22 23 50 52 54 95% 10094 109M 180** 191% 209% 252 252 276 53** 53** 55 G Ins. 37*4 42*4 42*4 45% 52** 57 52** 60% 63** 60% 70 76 83 97** H Ins, 84 96 96 102 10S 114 108 120 126 120 144 144 168 192 204 Note. — In the above table thelnachines which are indicated as having either of two speeds may be operated at 550 volts at either speed, or at 575 volts at the higher speed. SIZES AND APPROXIMATE DIMENSIONS OF SPECIAL ENGINE- TYPE RAILWAY GENERATORS. CO CO -4-> CO* P4 Appro x. Weight. Lbs. A Ins. B Ins. C In8. D Ins. E Ins. P Ins. G Ins. H Ins. 800 1230 650 10 80 135,000 24 46 8^14 166% 180** 204 52U 76 144 1050 1830 575 12 80 147,000 22 50 95% 252 53** 83 168 1500 2300 650 14 80 200,000 19 47% 1092£ 2095^ 276 55 97^ 204 1800 4400 410 20 75 247,900 24 56 129*4 2449* 312 62 111 204 2700 4700 575 24 75 312,500 24 58 144£ 277& 336 62 125% 216 268 ELECTRIC RAIL WA Y HAND BOOK. grtt.t*JWur-« KM £ ^ r-t *" 1 -J r-l «"* "St tH . H* »* DQ a §8 00 i—i •pi K O . -:* Ph a QO OS CO ©* 1— t 1-1 fc CO CO . r*o S a 00 00 1— ( rM TH, ft — H CO . r-W ■"** ft 1—1 CO . H a oc • la 1 * •-W O 3 CO OS CO M n !£'» r-S* tt a iO CO B r*# «■•* a TT 00 -< ^ 4-3 *? ** "W 'J *H Ci CO I'S •gajadray ^ LHl § s^ . R S S bf OS Tf cs rc tt 83 ^_, TRfioa 5 r, X r.x •e £3 H c a "TOUi 1J t-pa rH c^ «s< "" ' Pi •qn-jjo X O -O rA HJ3 u^ f I a ^^ " Ci w 3 ft •80T?£ CO co a CO CO •UIBTd a CO CI CO 10 CO . -W! r4* hi H Eh a Hi GO CO CO CO a •saiadray 00 li CO *.vi H 1 8 § th- T*. CN 2/0 ELECTRIC RAIL WA Y HAND BOOK Fig. 204.— side view op westinghouse belted generator. SIZES AND APPROXIMATE DIMENSIONS OF GENERAL ELEC- TRIC DIRECT-CONNECTED RAILWAY GENERATORS. MP, FORM H, 575 VOLTS. CLASSIFICATION. APPROX. WEIGHT. A B C K L M do •d 8 $ -ry- -----1 J _ 2T - !6 01 — P^-^-b" f! AMMETER^ \or£ALMd \ StV/TCH 60, .Y0L7 , , /Here*- SMTCH \\ & OfHAMO" ///Mr 3 Mo^/^aw stY/rcj* a Fig. 206a.— switchboard connections of storage battery to railway without eooster. TRotter 22S C£LLS poft soo yoirs I mm SHI//VT MA/rt G£rt£/t/lTaft WOSTZX 0/?/f£/V Fig. 206b.— diagram showing connections of differential booster and storage battery. 274 ELECTRIC RAILWAY HAND BOOK. In regard to the switchboard connections Fig.206A shows the method of con- necting the storage battery across the railway generator. In order to regulate properly for the variations of load that occur on the railway generator a booster is connected as shown in Fig. 206b. Here the booster is so adjusted that it charges when the generator is above a fixed potential, and discharges when the potential falls below it. The booster has a series field, through which the main battery current flows; opposed to this field is a shunt field which works differ- entially against the series field, so that this booster compensates for the high potential required in charging and the low potential of discharge. 600 ?550 500 ©200 1 150 Sioo £ 50 * $-0 I 50 1 100 # 150 glOO I 50 5-Second ^^A/^!l^^ Readings FlG. 206cr*BFPEOT ON railway loads op storage battery. ! ELECTRIC RAILWAY HAND BOOK. 275 J Fig. 206d. — bwitohboabd connections eor storage battbby, boosts* GENERATOR. J. 276 ELECTRIC RAILWA Y HAND BOOK. The curve, Fig. 206c, shows the effect on the generator operating in multiple with a booster and storage battery, the compensation for the fluctuation o f current demand and also the fluctuations on the engine and generator alone. Fig. 206d shows the switchboard connections an 1 arrangement of instruments and circuit breakers; it has cell regulation which is not now used in street rail- way work, the differential booster automatically regulating the charging and discharging of the battery. The booster is preferably driven by a 500-volt motor coupled direct, but may be belted or direct-connected to an engine. TYPES OF BATTERIES There are several different methods employed in the construction of a storage battery. One is the paste method, which is used by the Chloride Accumulator Company. Here a sheet of lead has on its surface small rectangular cells into which is forced the active material. This battery by charging and discharging ia readily " formed " for use. Another type of storage battery which is used in railway work is manufac tured by the Gould Storage Battery Company. In it a plain sheet of lead is grooved by rotary knives so that the lead is forced up between the knives and form ribs and corresponding grooves. These ribs vary in width from .008 to .024 of an inch. There are about 450 sq. ins. active surface per pound of plate, and 186 sq. ins. per cubic inch, while the ratio of the contact surfaces to superficial area is as great as 17^ to 1. The active material is formed in the interstices of these ribs by electrolytic processes. From tests on these plates it is shown that they are capable of main- taining a high electro-motive force with discharges at high rates, which is an advantage to be considered in railway work. The plates for railway work are 15J^ ins. x 15^ ins. divided into 16 ribbed plates which are formed out of a homogeneous sheet of lead. DATA ON STORAGE BATTERY INSTALLATIONS. The Buffalo Railway Company, Buffalo, N. Y.: Capacity of battery, 1200 hp-hours, 250 amps.; consists of 270 cells all in series, with 41 plates to each cell, the size of the plates being 15^ ins. x 15^ ins. The positive plate weighs 24 lbs., the negative, 16 lbs. each. Outside dimension of tank, 59% ins. x 21^ x 24% ins., with room in tanks to increase the capacity two-thirds. The Lansing Street Railway Company, Lansing, Mich. : 240 cells of 9 plates each, 10^ ins. x 10^ ins., and room in jars for 13 plates. Capacity, 320 amp.- hours at 8 hour discharge rate. This battery was used at the end of the line, and will run from about 25 to 50 amps, on average charge, and maximum discharge of 200 amps, with a machine variation of about 25 amps, maximum. The battery is located in a power house about a mile and a half from the generating station. Battery Installation and Attendance. The acid when put into the cells should have a specific gravity of 1180 degs. to 1190 degs. The charge should then at once be commenced at about half the normal rate. After charging at this rate for a short time and it is determined that all connec- tions are well made, the rate should be raised to the normal and continuted for about 20 consecutive hours or until the potential of each cell reaches 2.5 volts and all the cells are gassing freely from both positive and negative plates. The specific gravity of the electrolyte which fell shortly after the cells were filled, should now have reached at least 1200 degs. At this point the charging rate should ELECTRIC RAIL WA Y HAND BOOK. 277 be reduced to one-half the normal and continued until the electromotive force of each cell has again reached 2.5 volts. The regular service of the battery may now be commenced. On the subse quent charges to the number of five, the cells should be brought up to 2.6 volts per cell at the normal rate or preferably 2.5 at half that. When the battery is in use as a regulator enough geuerator capacity must be carried to meet a little more than the average demand of the load, and the bus voltage must be kept up to the average. That is, the battery must charge a little more than it discharges. The battery will regulate best when about 75 per cent full. „The individual cell voltage will then be about 2.08 volts. The specific gravity should be be tween 1190 degs. and 1200 degs. In its work as a regulator the battery should not discharge at a higher rate than specified by the manufacturers, nor should any individual cell at any time read lower than 1.8 volts when discharging at the normal rate. The battery must never stand discharged, but must be thoroughly charged on reaching the above point. A full charge should be given the battery once a week, when all the cells should be individually tested with low reading voltmeter and hydrometer. No cell at the end of this charge should read less than 2.5 volts when charging at the normal rate. At the end of this charge the specific gravity should not be below 1200 degs. Pure water, distilled if necessary, must be added to make up for electrolyte lost by evaporation. This water should not be added in large enough quantities to reduce the specific gravity to any considerable extent. It should be added at the bottom of the cell through a rubber hose or glass tube to insure its thoroughly mixing with the electrolyte. The plates should always be covered by the electrolyte. The positive plates should have a dark brown velvety appearance. Any lightness in color indicates insufficient charging. No attention need be paid to I a whitish precipitate that sometimes appears on the plates. The negatives ! should have a clear bluish lead or light slate color. If there occurs a time during which it is not convenient or possible to carry , on the generators the entire average load on the plant the discharges of the bat- , tery may be allowed to exceed the charges up to the capacity of the battery. About the only form of trouble that is likely to occur in a cell is a short cir- l cuit complete or partial between the positive and negative plates. This will be 1 indicated by low voitage and low specific gravity and should be at once removed. • Its most probable cause is the lodging between the plates of some foreign article ! or a loosened part of the plates themselves. It may also be due to the depth of J the sediment in the bottom of the cells reaching the bottom of the plat. s. If the 1 short circuit is due to a foreign body, it should be removed; if to a loosened por- i tion of the plates, it may be forced to the bottom of the cell; if to sediment, the 1 cell should be cleaned out. When Chloride Accumulators are in use, it will be found that there is a constant slight loss of solution. This is principally due to the evaporation of the : water from the mixture of water and sulphuric acid, of which the solution is com- < posed. Use pure water to replace that lost by evaporation. The water should I be absolutely free from chlorine (common salt), and contain not more than a ] trace of iron and other metals. Always use distilled water when it can be ob- tained; fresh rain water is also suitable. The solution should always entirely \ cover the plates in every cell. The proper density for the solution in a charged I cell is 1200 degs. The specific gravity of the solution should be tested with a 278 ELECTRIC RAIL WA Y HAND BOOK. hydrometer at least once a week. The test should be made just after the cell has been fully charged. A decrease in the density of the solution in a fully charged cell is not due to evaporation, as the acid does not evaporate. Some of the acid in the solution may be lost, however, by the spraying which occurs dur- ing the latter part of the charge. By the violent evolution of gases at that time small particles of dilute acid are thrown upward and prevented by air currents in the room from falling back into the cells. As this is replaced by water in the regular filling up of the cells, the specific gravity may be lowered from this cause. For this reason, it is not safe to always replace evaporation only with clear water on the assumption that no acid has been lost. Loss of water tends to increase the strength of the solution. When hydro- meter readings, taken at the end of charge, indicate that the density of the solution is low, a mixture of pure sulphuric acid and water of a specific gravity of 1400 degs. (one part sulphuric acid and one part water, by volume— not by weight) should be prepared and when cool, a sufficient quantity of the dilute acid should be thoroughly mixed with the solution in the jar to raise the specific gravity to 1200 degs. as shown by hydrometer readings. The density of the solution will vary with the condition of the cell, the density in a discharged cell being lower than in a charged cell. During the dis- charge, the acid is drawn from the solution into the plates; and during charge, this acid is again released. A low density of the solution, when the cell is ap- parently charged, does not, therefore, necessarily mean a lack of acid, as the low density may be caused by insufficient charging, that is, the acid may be in the plates instead of in the solution. Before adding the mixture of acid and wateivto the solution in the jar, it shouid be known that the cell is fully charged. A ceL may be considered fully charged when with the normal charging current flowing, voltmeter readings show the cell to have an e. m. f . of 2.5 volts. If the cell be charged at three- quarters of the normal charging rate, the charge should be continued until the cell shows an e. m. f. of 2.45 volts; if the cell be charged at one-half the normal THE CHLORIDE ACCUMULATOR.— TYPE " G." Sizes op Plates, 15^ Ins. x 15^ Ins. Number of Plates For Discharge in I ^ hours Amperes j g » Normal charge rate Weight of each element, lbs... . ^ . •* ir 4. (Width.. Outside MeasurmentJ Length. of Tank in inches, j Height.. Weight of acid, lbs Weight of cell complete, with I acid in lead-lined tank, lbs. f Height of cell over all, inches.. 15 140 196 280 140 300 18* 3 19| 22£ 197 621 26 21 200 280 400 200 422 23| 19| 255 829 26 27 364 520 260 544 291 20| 23i 312 1066 28 35 340 476 340 707 36 20f 23| 388 1351 28 41 400 560 800 400 829 40£ 20| 23g 444 1563 28 49 480 672 960 480 991 m 23g 520 1848 61 600 840 1200 600 1235 581 21* 24| 635 2277 71 700 980 1400 700 1439 66* 21* 24| 729 2633 29 920 1288 1840 920 1886 84| 21 £ 24f 938 3418 105 1040 1456 2080 1040 2131 21* 1051 3845 29 125 1240 1736 2480 1240 2538 1114 21* 24| 1242 4560 29 ELECTRIC RAILWA V HAND BOOK. 279 THE GOUU3 STORAGE BATTERY.— TYPE "S. Dimensions op Plate, 15J^ Ins. x 15^ Ins. Element Number Number of Plates Normal Charging Rate Discharge in j For 8 hours.. Amperes, 1 u 3 u " Capacity in j At8hrs. discharge Ampere, -< " 5 hrs. discharge Hours. ( M 3 hrs. discharge Weight of element, lbs Outside dimensions ( Width, of heavily glazed \ Length I earthen tank in ins. ( Height.. Outside dimensions (Width., of lead-lined tank •< Length. in inches ( Height.. Height of cell over all in ins. . . Weight of acid in tank, lbs Weight of cell, complete, lbs.. . 605 5 40 40 56 80 320 280 240 160 11 21 24 11 20 23 26 100 275 607 7 60 60 84 120 480 420 360 180 12* 21 24 18* 20 23 26 120 345 609 9 80 80 112 160 640 560 480 200 13* 21 24 13| 20 23 26 140 415 611 11 100 100 140 200 800 700 600 220 15* 21 24 15 20 23 26 160 485 613 13 120 120 168 240 960 840 720 17 21 24 17 20 23 180 555 615 15 140 140 196 280 1120 980 840 300 18* 21 18* 20 23 26 197 625 617 17 160 160 224 320 1280 1120 960 340 20 21 24 20 20 23 26 216 619 19 180 180 252 360 1440 1310 1080 380 21 24 22 20 23 26 235 768 rate, the cell should have an e. m. f. of 2.4 volts, and if the charge be at one- quarter of the normal rate, the cell should have an e, m. f. of 2.35 volts. If a voltmeter is not available, a cell, generally speaking, may be considered fully charged, when both the positive and negative plates have been gassing freely for fifteen minutes. To prepare dilute sulphuric acid, always pour the acid into the water, never the water into the acid. It is advisable to prepare the solution at least twelve hours before using, in order that it may thoroughly cool. Solution of specific gravity of 1200 degs. is composed of one part sulphuric acid having a density of 66 degs. Beaurae, and three parts of water. THE BOOSTER. The function of the booster in a railway plant is to assist the long feeders to maintain their potential at the ends of the line. Usually a series wound booster is used, whose function is to increase the line potential as the current flow through the booster increases. The rise in potential in the booster can be made equivalent to the drop in potential on the feeder. This may be obtained from a booster either by having its series winding compensate for the line drop or by shunting part of the series turns by a shunt winding whose magnetic effect can be opposed to or in the same direction as the series winding. It is usually the custom to group the distant feeders together on one booster. This booster may be operated by a motor or by an engine, but it is preferable to use the latter. In a number of stations there will be found the older type generators, kept for reserve capacity, and these can be changed into boosters by making the con- 28o ELECTRIC RAIL WA Y HAND BOOK. nections shown in Fig. 206b. The machines can still be nsed as generators when required. The feeders to be boosted are grouped together on a bus-bar, which is con- nected to the main bus-bar by switch, A. Feeding this bus is a generator having a double-throw double-pole switch. When switch A is open and switches /?, /?, from the generator are thrown in the lower position, the current then passes through generator F from the bus to the boosted feeders in a direction to increase the potential of this current. When the switches on generator F are thrown in the position C, C, the gene- rator is used simply as a generator and A can be closed, and the feeders will not be boosted. This condition is especially useful for parks and special outlying STRA/GHT F ££0£ff$ BOO<$T£D f££D£X§ IV IV E E IV § E E ri D i — [ID — (0 kvv-^o)J *?^ wr(dj hi f [) 3 Fig. 206e.— diagram op connections for changing generator to booster. attractions, where the traffic is large for a short time; and it saves the invest- ment for a booster. The capacity of generator F should be equal to the maxi- mum demand on the feeders, or a number of generators can be supplied with double-throw switches and connected as shown for the single generator, and used as multiple boosters. The shunt field can be excited or not if the copper is light for the maximum demand, and in this way cars can be moved more economically for short periods than by the investment in copper or boosters, and gives a double use for the same generator. The amount of boosting will depend upon the series turns on the generator, but if too much, it can be shunted and if too little, can be increased^jby the shunt winding on the generator. ELECTRIC RAILWA Y HAND BOOK, 281 ROTARY CONVERTERS AND DOUBLE-CURRENT GENERATORS. Direct-current dynamos generate alternating current in their armature wind- ings which are rectified by the commutator and delivered to the external circuit as direct current. If collector rings are connected to the windings at suitable Fig. 207.— alternating direct current generator. TWO W/A£$. THREE MASS. &'* &'* Figs, 208 to 211. —types op rotary converters. points, see Fig. 207, alternating current can be delivered ( externally. If both (commutator and collector rings are used the machine can deliver both direct (current and alternating current at the same time, in which case it is a double current generator. If inotead of being driven by external power, it is driven as a direct-current or synchronous alternating-current motor, and at the same w 282 ELECTRIC RAILWAY HAND BOOK. yv/WWWWW Ww WVWW r/W /V/AST SECONDARY TO ROTARY CONVERTER. A B. MAAAAAAAAAAA /WWVIWVWVV THREE PHASE PRIMARf T HANS MISSION LINE; 6-t. 8-2. f f rf-2. C-t. A-t W^AAAA^^A/^^"^sAAAAAAAAAA/ TftftT /90Wf SECONDARY TO NOTARY CONVERTER, A-B. C'A. THREE PHASE PR/MARY TRANSMISSION LINE. A-2. B-2> *& a-/. <>/. a-* Ws/WWVW WwWWVV WWVWVW 7X//?££ /5BM£ SECONDARY TO ROTARY CON/ERTER A-B. 6-C. C-A. Y^/MWWMAA^ ^WVWWVWWW \ ^ WWWA/WVVWW^ W/ttT /WAB" PRIMARY TRANSMISSION LINE. A-2. C-2. B-2. B-1. A z h C-l 1 I f 7^/Tff PHASE SECONDARY TO ROTARY CONVERTER. A. B. O. 7W?££ /V/>45£ PRIMARY TRANSMISSION l/NE. Figs. 212 to 215.— transformer connections tor rotary converter*. ELECTRIC RAILWA Y HAND BOOK. 283 A-2. 6-2. c-r. C-2. A-7. a-7. £-1 frt a-2. *e-2. k 4 www Www Www VWVAv d/X PHASE SECONDARY 70 POTAPY CONYERTER. a-b. a-&. C-A, *e-a. /W\^VVWWA/WWVVVVVV \ y w\AAAAMAAA/W\/W\/\M 7HPES PHASE PP/AfAPY 7PANSMJSSW/Y l/N£. 3-/. 6-2. c-r. b-2. k I -*. C-A. ^-a yWWWWWWWWWL^^ THPEE PHASE PP/MAR7 TRANSMISSION l//Y£. \ A-r- Ik *c-r 4-2 A-Z* S-/ A B-2 3* wwvwh Vwwww vw^ r^ r T^> r^ 5 ** *^ +Y-*-r* «-i*+i*-* SIX PHASE SECONDARY 70 P07APY C0NYEP7EP. A-C. a-Xi. B-A-C. &-a-£ in. to 5^ ins. long. Laps should abso- lutely have no filling strips. Single belts should have an ultimate tensile strength of 3C00 lbs. per sq. in., aud double belts should have 4C00 lbs. per sq. in. If tests are made, the average of three pieces selected at random, should be taken. HORSE POWER TRANSMITTED RY DOUBLE LEATHER BELTS. (1 INCH WIDE, 550 PT. PER MINUTE = 1 HP). Speed in Feet per Minute. Width of Belt in Inches. 4 6 8 10 12 14 16 18 20 22 H.P. 24 H. P. H. P. H. P. h. p. H. P. H. P. H. P. H. P. H. P. H. P. 400 600 800 2| 4i 5| 41 61 * 81 5| 81 HI Ti- ll 141 81 13 10 15 201 "1 171 23 13 191 26 141 22 29 16 24 32 17| 26 341 1000 1200 1500 71 81 10f 11 13 16J m 21| 22 27J 211 26 321 251 301 38 29 341 431 321 39 49 36 44 541 40 48 60 43$ 52$ 65$ 1800 2000 2400 13 141 17j in 21| 26 26 29 842 32| 361 44 39 431 521 451 501 601 52 58 691 59 6-1 781 651 721 88 72 80 96 78J 87 105 2800 3000 3500 201 211 251 30| m 38 ■ 401 50| 51 541 631 61 651 76 71 76 89 81 871 101 911 98 114 102 108 127 112 120 140 122 131 153 4000 4500 5000 29 321 361 481 49 64J 581 65 721 72| 82 91 87 98 109 101 114 127 116 131 145 131 147 163 145 163 182 160 180 200 174 196 218 Note. — The belts are not supposed to be unduly strained. For single belts estimate only two-thirds as much as for double belts. Pulley Dimensions to Avoid Abnormal Belt Bending Strains. No. of Ply. Min. Pulley Diameter. Min. Ratio of Diam. to Width. Single 3 ins. Immaterial Double 6 ins. 3:4 Triple 10 ins. 1 : 1 The power transmitted is dependent on the arc of contact. If the belts are very oily, the power may be reduced one-half. Paper pulley coverings or paper pulleys generally increase the power obtainable by 10 per cent to 20 per cent. The power that can be transmitted by belting is greatly increased by the use of Cling Surface dressing which increases the life of belts by decreasing the wearing off of the surface of belts due to slippage. Belts which are not horizontal will generally tend to slip on the lower pulley if worked at their max- imum power. The arc of contact is increased by having the elack side of the ELECTRIC KAILWA Y HAXD BOOK. 287 belt on top and this method is, therefore, preferable. All pulleys should have crown faces unless it is intended to shift the belts over them. Belts, particularly on long drives, sometimes give trouble by wobbling from side to side. This may be due to vibration or movement in the shafts or pulleys to which they run, or it may be caused by the belt being stiff and requiring too much force to fit it over the crown. In the latter case washing once a week on the side next the pulley with one part of beef tallow to two of castor oil mixed warm with a little pulverized rosin, until the leather is pliable, will generally remedy the trouble. Belts may also run badly if the pulleys are not properly aligned, are untrue or are out of balance. Trouble will also occur if the belt is not made or fastened truly. Waves in the belt are often caused by irregular power or untrue pulleys, but most heavy belts tend to wave slightly. The length of drive, i. e., the distance between centers of driving and driven shafts, may be widely varied according to necessity. If a belt is too short it has no elasticity; if too long, it has a tendency to wobble or wave; but the limits are quite broad. Center Distances for Belt Drives. WIDTH OP BELT (INCHES). 3 6 12 18 24 96 4S 60 CENTER DISTANCE (FEET). Minimum Preferable Maximum 4 8 25 6 12 30 9 17 32 11 20 34 12 22 37 15 25 40 17 30 45 20 35 50 SHAFT KEYS AND BEARING CENTERS. Proper Distance Proper Distance between Shaft Diameter. Size of Keys for between Ins. Couplings and Pulleys. Bearing Centers, Bearing Centers, Line Shafting. Jack Shafts. Feet. Feet. !?« % X £ $h i\ ■ It 7 * T ? S X , 7 6 6| 5; ■ m T 7 3 X ft 7 5; ■ iii i 7 e X x 7 s 7| 6] i 2t\ 1% X ft 8i s 2& ft x ft 9 7 m u x n 9| 2* m HXH 10 8 •a lixH 11 8* 3*1 *t x ?| 12§ 13} 10 4i 7 B *§ x n 10| ll| 4ji ii x *i 14* 5| ii x n 15 12 5* iixii 16 12| 6| 17 13j H 18 14* 8 20 16 288 ELECTRIC RAILWAY HAND BOOK. DIMENSIONS OF HEAD OR JACK SHAFTS. Shaft Revolutions per Minute. Diameter. Ins. 100 125 150 175 200 2:5 250 300 350 400 Horse Power op Jack Shafts. h% 1.7 2.1 2.5 3. 3.4 3.8 4.2 5. 5.9 6.7 3.1 3.8 4.5 5.3 6. 6.9 7.6 9.1 10.6 12. 4.5 5.6 6.8 7.8 9.1 10.? 11.2 13.6 15.7 18.2 liS 7 8.7 10.5 12.2 14. 15.7 17.5 21. 24.5 28. *A 9.3 12.5 14.7 17. 19.5 22. 2\ 29.5 34. 39. V* 14. 17.5 21. 24.5 23. 31.-3 35. 42. 49. 56. m 1S.5 23. 28. 31. 37. . 41. 4". 5 5i. 62. 74. *tt 24. 30. 3d. 41.5 47.5 53.5 59.5 71. 83.5 95. 3/s 38. 47. 56.5 66. 75.5 85.5 94.5 113. 132. 151. 311 56. 70. 84. 98. 112. 126. 140. 168. 1^6. 224. 4/ 8 80. 100. 120. 140. 160. 180. 200. 240. 280. 320. 4|| 109. 136. 164. 191. 218. 246. 273. 328. 382. 437. 5f 146. 182. 218. 255. 291. 328. 364. 436. 510. 582. 5$ 189. 236. 283. 330. 378. 425. 472. 567. 661. 755. 6f 238. 297. 357. 416. 475. 535. 595. 713. 832. 950. 6£ 294. 367. 440. 514. 587. 661. 735. 880. 1030. 1177. 8 448. 560. 671. 783. 895. 1010. 1120. 1345. 1570. 1790. Two or three belts may be run tandem, but where the pulley diameters are small the belts should not touch each other at points on the drive. Where belts are run tandem, the maximum power of each belt is not reduced, but rather increased. PUIXEYS. Pulleys are usually made in two weights for single or heavy belts. Crowns of pulleys should be from $, of the face for small or slow-running pulleys, to 5 £ of the face for pulleys of 21-in. face and over. The latter figure is quite usual in large dynamo pulleys and gives entire satisfaction. The increase in diameter at the crown is twice the amount given. Where the crown is too high with a fast running belt, the latter is liable to leave the edges of the pulley and thus concen- trate all the strain and wear on the belt center. Pulleys and clutches should be balanced. Set screws should have cup ends. The bore should be just large enough to fit closely on the shaft. Pulley centers and not edges should be aligned. Of course it is necessary to align from the edges, but allowance should be made where width differs, so as to bring the centers in line. Contrary to general ideas, belts do not " tend to climb to the high side of pulleys' 1 ; but where two shafts are not parallel, the belts will run on both pulleys toward the low side, *. ^., toward the point where the shafts are nearest together. .Shafting. — Counter shafts are usually made in lengths of 24ft. or less. They ■hould be straight before erection, and should be properly supported so that their hangers can not shake or vibrate. Journals are pieferably made " self -oiling. M ROPE DRIVES. The limit of belt transmission for railway work is in the neighborhood of 500 hp. Beyond this power rope drives should be resorted to for transmission. Curves herewith, Fig. 226, give the power that can be transmitted by manila rope. ELECTRIC RAILWA Y HAND BOOK. 289 The English method of independent ropes driving in multiple is not suc- cessful in railway work, for the reason that unequal tension causes undue strains to fall on the ropes having the greatest tension. For the variable railway load, American practice is to have a continuous rope wound around the grooves of driving or driven pulley grooves, and a slack loop taken from one side of the drive, which is held in tension by passing around a tension pulley and kept taut by weights. As the diameter of pulleys decreases, the wear on the rope increases. The table herewith gives the smallest pulley that should be used; larger diameters than those given should be employed where possible. 10 20 30 40 50 60 70 80 90 VELOCITY OF DRIVING ROPE IN FEET 100 W 120 130 MO PER SECOND. Fig. 226.— curves for power transmission by rope driving. DIAMETER OF PUIXEYS AND WEIGHT OF ROPE. Diameter of Hope in Inches. % 1 ¥ Smallest Diam. of Pulleys, in Iuches. 20 24 30 36 42 54 60 72 84 Length of Rope to Allow for Splicing in Feet. 6 7 8 9 10 12 13 14 Appox. Weight in Lbs. per foot of Rope. .12 .18 .24 .49 .60 .S3 1.10 1.40 2go ELECTRIC RAIL WA V HAND BOOK. POWER STATION SWITCHBOARDS. The location of the switchboard should be central with respect to the units it controls. In stations of over 8000-hp output or twenty-five feeders a switch, board attendant is generally required. Here the elevation of the switchboard in a gallery saves floor space and gives the attendant view of the generators he con- trols. The " unit system " where the generator panel is located adjacent to the generating unit, cutting down the internal conductor cost and giving the engi- neer electrical as well as steam control of the unit will prove useful where large units are installed. In this case, the feeder circuit-breakers can be operated by pneumatic control for distant parts of the station, and the feeder panel board located conveniently near the distribution wire tower or underground ducts. In the usual methods of construction the generator panels consist of an am- meter, circuit-breaker, quick break main switch, equalizer switch, voltmeter for throwing in the machines, a regulating rheostat and a field switch. Figs. 227 to 230 show the forms adopted by various companies. The switchboard surface may be selected of slate, enameled or marbleized, or of marble. The thickness should not be less than 1% ins. for a 20-in. panel and 2 ins. for a 26-in. panel, as the circuit-breaker in flying open is liable to fracture thinner slabs. They should be secured to iron framing made of T's or L's using asbestos washers as a bedding between the slate and its support. Slate slabs should not be secured to wooden verticals, as warping of the supports will event- ually crack the panels The edges of the panels, have usually %-m. bevel, and are fastened by finished hexigon bolts passing through the panel and iron backing. Bus-bars should be figured for the current density given by table herewith. They should be rolled medium hard and insulated from a cast iron supporting bracket on the back of the board by slate or porcelain. COPPER BAH DATA. (Bus-Bars.) Size. 1 X Y\\r\. vX x H " in x y 4 " iA x n u M x H " 2 X %" 2K X % " 2U X \{ " 23/a x V, " 2 xKa u No. 0000 B. A S. \A in. Round Amperes. 433 530 626 725 676 798 916 1,035 1,154 1,500 1,715 1,222 257 305 426 560 861 Circular Mils. 318.310 397,290 4:7,465 556,400 596,830 716, 200 835.600 954,930 1,074,300 1.591.550 1.989,440 1,273,210 211.600 250.000 390,625 P62,500 1,000,000 Square Mils. 250,000 312,000 375,000 437,000 468,750 562,500 656,250 750,000 843,750 1.250.000 1,562.500 1,000,000 Ohms per Foot. .00003*6 .0000269 .0000223 .0000192 .0000179 .0000149 .0000128 .0000112 .000009P5 .00000672 .00000537 .00000840 .0000505 .0000428 .0000273 .0000190 .0000107 Weight per Foot. .97 1.21 1.45 1.70 1.82 2.18 2.54 2.92 3.27 4.86 6.07 3.89 .C4 .76 1.18 1.71 3.05 ELECTRIC RAILWAY HAND BOOK. 2g t I i 1 \ r 292 ELECTRIC EAILWA Y HAND BOOK. © ,f=i n ^> ®® r-~i ^^ ■0 b^ } ._., ,- - "~s H ?4 ELECTRIC RAIL WA Y HAA'D BOOK. 293 They should be connected together by lapping, and iron bolts used for bring- ing these surfaces together, which can be figured safely with smooth bus bars to carry 190 amps, per square inch of contact surface. The connections between the bus-bars and switches, etc., are preferably made by a copper link clamped be- tween the bus-bar, and between nuts threaded on the stud projecting behind from the switches, circuit breakers or ammeters. The current density should not exceed 60 amps, to 100 amps, per square inch on thread and nut surfaces. Railway switches should have at least 4 ins. between breaking terminals, and should be provided with an auxiliary snap or carbon break to shunt the current carried by the contact surface, and so reduce injurious arcing effects. The copper contact area should be 60 amps, to 80 amps, per square inch of switch contact surface with parallel, well-adjusted switch surfaces. A composition machine- finished lug held down by an iron bolt should not be expected to carry more than 120 amps, per sq. in. of contact surface. INSTRUMENTS. Ammeters.— -The ammeter should be dead beat if possible, especially for the generator, and have a full scale reading 35 per cent to 60 per cent greater than the maximum output of the generator, in order to prevent overloads damaging the instrument. It is useful to have the manufacturers mark a red line on the dial of the instrument for full load amperes on the generator. Where shunts are used, they are connected in the bus behind the board and the leads and shunts marked with the number of the instrument. Care should be taken that the corresponding instrument should only be used with a given shunt. For main ammeters the shunt-type instrument is generally used and the shunt inserted in the main bus as it passes from the generator panel boards to the feeder panel boards. Ammeters are not essential on feeder boards, where the circuit-breaker is reliable and their expense can be saved by introduc- ing a plug device, that can be plugged into the bus side of any feeder switch, on which readings are to be taken, and the current diverted through the common shunt to the ammeter when the feeder switch is opened. This saves space on switchboard as well as reducing the cost. Voltmeters. There are generally two of these, reading exactly alike on each board. They can be mounted on a swinging bracket so that they can be seen from any part of the board. One is used to maintain the station voltage, and the other to bring the machines to the bus-bar potential. Before throwing ma- chines together thore is usually provided a receptacle into which a voltmeter plug is inserted, when the machine is to be adjusted in voltage so that one voltmeter is sufficient for all the generators. For throwing in, where two pressure boards are used, three voltmeters are required. Wattmeters. Wattmeters should be installed where records of output are to be kept. Periodic readings of the ammeter and voltmeter multiplied together invariably give a result from 20 per cent to GO per cent above the average watts; and some remarkable station performances have been accounted for from this cause. The natural tendency is to read the ammeter at its maximum swing. The wattmeter should be carefully screened from magnetic effects from the bus-bars which may throw it out of calibration. With large units it is the modern railway station practice to have a wattmeter on each unit so that the depreciation on the units can be averaged. Circuit' Breakers. To prevent violent overloads on the generator, or to take Current off a ground on the line, the automatic circuit-breaker U ordinarily ne$- 294 ELECTRIC RAILWAY HAND BOOR'. essary, although fuses are sometimes used instead of circuit-breakers on feeders. The circuit-breaker should respond to a rise in current above its set value in amperes, and open the circuit which it protects; the arc formed on breaking should be taken care of by auxiliary contacts of carbon or a magnetic blow-out. The contacts should be kept in good order, so that the friction between their surfaces will not increase the time constant and so strain the generator, or on feeders throw the generators instead of the feeder breaker. A circuit-breaker should be capable of breaking a circuit before a Weston dead beat ammeter will record 100 per cent over the circuit-breaker's set current value. Switchboard Connections. — In modern railway practice the positive side of the railway generator armature should be connected to the trolley, through the series winding of the generator. The equalizing connection is taken from the middle point of the switch to the equalizing bus, Fig. 231, but the present practice in power stations is to equalize at the dynamo with the equaliz- ing switch either mounted on the frame of the dynamo itself, or on a pedestal by the side of the dynamo. In other cases again, the equalizer is tied together permanently between all the dynamos. The disadvantage of having the equalizer opened is that there is a danger of the machine being thrown in circuit before it is equalized. In order to provide against this accident, several suggestions have been made. One is to make the switch at the dynamo double pole, carrying both the equalizer and positive connections, throw the generator in first; another method has been used where the throttle of the engine is connected to the equalizer switch, so that before the throttle is open, the switch is closed; and the generator cannot be thrown in before it is equalized. For balancing and adjusting compound generators see page 48. The field of the railway generator may be connected up in two ways: the one shown in full lines in Fig. 231 is the bus-bar excited method, and the one shown in dotted lines is the self-exciting method. When a large number of generators are to be handled, a dynamo galvanometer or voltmeter is connected across the dynamo terminals of the dynamo switch, instead of plugging in the voltmeter in order to show when the generator is of the right potential to be thrown in. It is also usual to allow for a panel between the dynamo and feeder panels, on which to. mount the main ammeter, integrating wattmeter, voltmeters and pressure switches. The positive bus-bar only is taken to the feeder board, and the feeders are provided with a single-pole switch, ammeter, circuit-breaker and reactance coil to choke back any lightning discharges, and force the arrester to operate. The dynamo panel should be provided with a small double pole, lighting switch where the station is lighted from the power generators, so that any generator can light the station independently of the power bus-bar. This lighting circuit should be looped inside of the circuit-breakers. The present practice indicates that the best results are obtained when the lightning arresters are located as near the point where the feeders enter the station as possible. Behind the switch- board is not. the proper place for the lightning arresters as a rule. The panel form of construction is now universally adopted, the apparatus being mounted on an upper panel, with a foot plate, about 20 ins. below it. These panels are made interchangeable for the different units and feeders, and the exten- sion of a switchboard, only requires that the bus-bar and iron frame be extended, giving a very flexible method, and amply providing for the future growth of the system. It is useful in some cases to be able to separate the feeder systems, so that they can be supplded by independent generators, where extra demands of frame require a higher potential to be obtained on the congested part of the sys- ELECTRIC RAIL WA Y HAND BOOK, 295 j tern To effect this result, the dynamo phould be provided with a double-throw switch, and the equalizing system should also be double, with a double-throw equalizer switch. If double-throw feeder switches are also provided, the feeders can be operated on independent generators when required. It is the usual pract- FlG. 231.— DIAGRAM OP RAILWAY SWITCHBOARD CONNECTIONS. ice to tie all rail and return grounds to a common negative bus-bar, but to reduce electrolysis, in some cases the ground returns which are tapped directly to the water or gas pipe systems, are brought to one ground bus-bar, and the rail or return feeders are connected to a separate one. * 296 ELECTRIC RAILWAY HAND BOOK. The conductors behind the board are supported on porcelain insulators, or threaded through porcelain blocks. All conductors over No should be stranded and the field wires should in all cases be a stranded conductor. In some cases asbestos or lead covered leads are used, Where bare rubber is employed for the insulation great care should be exercised to prevent oil from reaching these con- ductors, as fire has originated in several railway stations from this cause. Ex- posed terminals of different potential adjacent to one another, especially where the line is subject to lightning storms, should be taped and insulated, or so shielded that no spaik can jump between them, for when the circuit-breaker opens on overloads, there may be quite a rise in potential on the dynamo, which sometimes starts a damaging flaring arc between exposed adjacent surfaces. Lightning Protection. — There are several principles employed to protect the generators and apparatus from lightning discharges. The leak arrester con- sists essentially of a water resistance connected between the positive trolley and + BuS5 Fig. 232.— tank lightning arrester. ground, so that the potential between the feeder system is maintained at 500 volts difference. Any tendency for static discharge to select the feeder wire as its path to ground, is neutralized before an abnormal difference of potential can exist. The tank arrester, shown in Fig. 232, is plugged into the circuit at the approach of a thunder storm. Three tanks are usually employed, so that if a discharge passes one, it is dissipated through the other and leaks to giound. The magnetic forms of arresters have an air gap about T V in. over which the static sparks jump to ground. The main current passes around an electromagnet, whose field of force is so arranged as to blow out of this gap the arc formed by the line current following the static spark. Every obstruction offered to the flow of this discharge b^ the ground wire subjects the station apparatus to an electro* ELECTRIC RAILWA Y HAND BOOK. 297 static stress, tending to break it down at its weakest point, and every means should be used so that the lightning discharge can jump the spark gap of the arrester and pass to ground. The high frequency which a discharge possesses N I ?""* I Station Transformer 1 Main Station Swttob CpOB. O.B.C~> XX J— L, fm D 0. P—4f D C Feeder ] Shunt Fi»W D.C. Feeder DC Feeds r Fig. 233. — connection between main and stjb-station, two-phase system. gives it a tendency to travel on the surface, rather than on the interior of the wire, and in this way choke its own passage. This e£ect increaoes as the diameter of the wire increases, and consequently only a small wire is used for the ground con- 298 ELECTRIC RAIL WA V HAND BOOK. ductor, No. 6 being the usual size. A bend in a conductor greatly increases its self induction, consequently the wire should be as straight as possible from the point of connection at the lightning arrester to the ground connections. Carrying this wire parallel or near masses of iron will also tend to retard by self-induction the passage of the discharge to earth. To use a water pipe system for earth is not the best practice, but where it is necessary an iron lug can be clamped to the water j)ipe and the contact surfaces amalgamated, the ground wire being soldered into this lug. After the connection is made, it should be painted over with two coats of air-drying asphalt varnish. No ground connections that are used for any other purpose should be used for the lightning arrester ground. No part of an iron structure or piping through the building should be used for this purpose. The ground conductor should be connected to the water system as near its entrance to the earth as possible. A ground near running water or naturally moist earth will give the best results, but in all cases it must be below the frost line. If these conditions cannot be secured, a hole can be sunk in the ground until water is reached. A copper plate two by two feet, with the conductor firmly soldered to it, will in ordinary cases be adequate for lightning grounds. Loose waste metal does not materially increase the actual contact area of the earth plate. If such material is'used for the earth plate, each piece should be connected with the ground conductor itself. The best material to use to get a low resistance ground is broken coke, which should be tamped well in the bottom of the hole to the depth of about 2 ins. The copper plate should then be laid on, and about 4 ins. more coke tamped well over it. Earth can then be thrown in and tamped lightly. POLYPHASE SUB-STATIONS. Fig. 233 shows the main station and sub-station switchboard connections for a two-phase transmission system, with an alternating-direct-current generator at the generating station and a rotary converter at the sub-station. POWER STATION COSTS— AVERAGED. Land Buildings Foundations Stack flues and breeching Flooring and material structures Crane Boilers and settings Stokers Fuel economizers Ash and coal conveyors Coal storage Steam piping Engines Generators, DC Lubrication Pumps Feed water heaters Switchboard Interior wiring conduits and lighting protection Totals , Maximum Dollars per kw. 7. 17.20 3.90 2.35 2.10 .70 18.40 3.00 4.20 1.60 4.30 9.80 33.20 21.00 .60 1.90 2.16 3.80 4.60 $141.81 Minimum Dollars per kw. 3.50 10.50 1.65 1.20 1.00 .00 9.50 .00 .00 .00 2.40 6.40 21.00 18.25 .30 .90 .60 2.10 3.80 $83.10 Eateof Depreciation per cent. 1.2 1.4 2.5 2. 5. 8.5 9.1 2.3 4.5 3.2 4.5 9.6 10.6 6.7 5.2 2.5 3.5 ELECTRIC RAILWAY HAND BOOK. 299 The engineer is required to find for each case what investments in the construc- tion of a plant will prove profitable to install where the price of coal and the avail- able supply, water supply and the price of labor are all taken into consideration. The capital employed in building an economical station will reflect largely in the possible earnings of the railway system and the nicety in selecting investments in power plant for economical appliances and their earning power which will cut down operating expenses differentiates the good from the poor engineer. The engineer has to rely largely on his own common sense and keep before him the physical con- ditions surrounding the power plant under construction and not base his determina- tions on what the same auxiliary economical methods have accomplished in other plants where all the conditions are not understood. We generally find the most economical power plants in plain but substantial structures and low labor costs per k. w. Where the designer has comprehended ease in maintaining cleanliness, and an absence of polished brass, and the coal cost per k. w. are less, where coal and ash handling facilities had been carefully thought out in the original design always bearing in mind that a $1.50 a day man, in operating, costs the same as the use of $10,000 added to the capitalization and by designing labor saving appliances to cut down the operating force by this amount for an investment of $2500 in the station a return of 20 per cent, can be made on investment. It is the additional salary of the auxiliary help oilers, cleaners, coal passers increasing the cost per unit that can be saved by the proper application of labor saving devices. SECTION V.-THE LINE. Direct current distribution should be employed on roads not exceeding six miles radius with moderately condensed trade and eight miles radius with grades and light traffic. For roads which reach further from the power station than this, the question of the most economical method of distribution will have to be solved for each individual case, as there are too many variables entering into the problem to make a general solution possible. One question in the design of the system for larger territories when the principal part of the road lies within the six mile radius is whether to use more copper to expand the area to, say, nine and one-half miles, [or use boosters with less copper For distant distribution, the general method of solution is given below and also data, from which the copper line can be figured. The cost of the different methods should be compared with that of the direct feeding method as a criterion. The area to which the estimates for alternating current distribution should be applied is certainly beyond the five mile radius, and the capital investment and cost of copper for feeding the out- lying territory alone should be considered. The elements involved in the consideration of what would be the best system of transmission to use are the fixed and operating charges. The former include the cost of line copper for permissible line drop, additional cost on pole line to carry copper and the cost of bonding for the return circuit; the latter, the interest on capital, the depreciation on the line and the cost of line losses per annum. To make the substation profitable the cost of these two charges should be greater than the sum of the cost of the substation building, the boiler, generator, switchboard, cost of line, bonding, and the interest charges on the substation investment, depreciation charges, cost of supplies, labor charges and reduction in main station efficiency due to loss of load. In comparing a direct transmission of 550 volts against a rotary converter substation, the same principle holds good. Compare the original cost of the550-volt distribution system and the operating charges, as given above, against the cost of building the rotary converter substation, the additional cost of generators in the power station, cost of rotary converters and static transformers, cost of trans* mission line and insulation and switchboard, and the interest, depreciation, attendance and supplies, as well as the annual cost of transformation losses. The booster overcomes the line drop, and the economy of using a booster to produce proper potential at distant points depends upon the relation between the copper cost, depreciation and transmission losses for direct current distribu- tion, and the booster cost, depreciation and transmission losses. It is not usual that boosters operating continually for any considerable load will show a letter investment than copper; but for transient loads they do make an econom- ELECTRIC RAILWAY HAND BOOK. 301 ical showing. The distribution can be carried by copper alone on roads between six and ten miles radius, where the equipments are operated by feeders only under normal loads. Data will be given in this chapter to determine the copper line costs; and equivalent rotary or substation construction could be estimated from the manu- facturers' quotations for specific performances. THE MOST ECONOMICAL ARRANGEMENT OF FEEDERS. The next matter to be decided in line construction is the proper amount of copper to use and its most economical disposition. VTe will first view this question from commercial considerations. Returns from the investment in copper may come from several sources. The first and direct loss caused by a de- ficiency of copper in a feeder system is in the loss of energy ; second, in the in- creased depreciation of the car equipment due to the higher temperatures attained by motors operated by low voltages ; third, in the added expense of operating more equipments where a given headway between them is maintained, due to the lower maximum speeds and slower acceleration under low voltages. In connection with the energy it can be seen that the smaller we make the feeder for a given load, the greater the loss and the less will be the fixed capital charges against this feeder per annnum. On the other hand, the cross section may be increased until the interest charges are largely in excess of the energy saved. Lord Kelvin determined that the most economical sized feeder to use was the one in which the annual interest charges were equal to the annual cost of the energy lost, and this is accepted^ as a general rule for the determination of the proper capital investment in the feeder. To the cost of feeder should be added the cost of its insulation and pole line or of conduit, and the interest charge can be fixed by local conditions. The price p :r unit of energy generated in the station should be based on that charge for which a power station could sell all its output without profit or loss. The method for computing this charge is given on page 267. Taking this prime value, the cost for the losses on the line will be some amount less than this cost, per unit lost, depending upon cost of this additional production of energy, and the cost varies on each plant for this loss, but the cost of increasing a load 10 per cent on a station will make little difference in its consumption of coal, oil and water, except where an extra unit has to be operated to maintain the usual 25 percent overload margin allowed in operating capacity; and under these con- ditions the losses in this added unit are chargeable to the line losses. Strictly speaking, the fixed charges belong to that portion of the energy of the station which produces a revenue, and again the increased loading of a unit brings up its efficiency, and this line loss is reflected in decreased cost of the total output. If these coGts and current deliveries are determined for any road, it is very easy to construct a table in which the cost of the energy lost is compared with the capital charges, and this determines what size of wire can be most econom- ically used. The next and most di£icult question is the fixing of the current required to propel the car or cars which are fed by the copper to be supplied. Assume the energy consumption a3 1.2 few-hour per car mile for level track. 28-ft. car body weighing 1S,C00 lbs., single truck, two G. E. £00 motors with K-2 con- troller, and speed 10 miles per hour. Thi3, with 530 volts, gives the average of 24 amps, per car for current delivery. On the first step the car would require 60 amps, at 500 volts; but, with this flow of current, a drop will occur over the copper conductors. Say the voltage fell to 450 volts with the controller on the 302 ELECTRIC RAILWA Y HAND BOOK. first notch, then the current will be 54 amps. The second notch of the controller cuts out 4££ ohms., and this should largely be taken up by the counter e. m. f. of the motors, which have started and commenced accelerating. The speed gained on the first notch reduces the current obtained when the second notch is reached, and the greater the feeder drop the slower the accclera* tion. The greater the amount of energy required to bring the equipment to speed, the higher will be the temperature attained by the motors, and the greater the rate of equipment depreciation and the maximum demand on the power sta- tion. These losses and station investment can be reduced by greater line copper investment. In this problem, both the copper and ground return system have to be considered. Allowing 20 per cent drop, which means an effective voltage delivery of 400 volts to the equipment, the maximum current at starting, which will be on the second notch, can be assumed as about 90 amps. Say that 8 per cent drop is allowed for the ground returns, and 12 per cent for overhead copper; this gives a maximum feeding resistance of % ohm per equipment. The average demand then is 24 amps, per car, and the maximum demand 90 amps, for the case under discussion. The effect of grades is to increase both the starting current and also the running current value. The chart, Fig. 234, shows the relation which exists between the traction coefficient, which is plotted as ordinates on the chart, and the watts at 1 mile per ton per hour, which are plotted as abscissae on the chart. Diagonal lines are drawn across the chart corresponding with grades from level up to 19 per cent, to assist in giving what is really required, i. e., the amperes flowing for different speeds and grades. On the right hand of the chart is given a scale in amperes, assuming the current delivery to the equipment is at 500volts, for speed of 1 mile per hour. While the current delivery is not a rectilinear function of grade and speed, approximately proportional results can be obtained by multiplying the speed on grades by the wegiht of the car and then by the current given in this chart at the required grade, which will give the total amperes required. As an example, suppose we have a car weighing 18,000 lbs. climbing a 4 per cent grade at 10 miles per hour, with a track coefficient of 25 lbs. Pass up the vertical line 25 until the diagonal line indicating 4 per cent is reached, then pass horizontally until the 1 mile per hour scale of amperes is reached; this gives 0.42 amps, per ton. For 9 tons this would be 0.42 (amps.) X 9 (tons) X 10 (miles) = 37.8 amps. The two other vertical columns at the extreme right of the chart, one 6-mile speed and the other 12-mile speed, give the current required, including 20 per cent loss on the line. For example, ,assume an 8 per cent grade, 30 track coefficient, car weighing 20,000 lbs., running at 8 miles per hour., at 20 per cent transmission loss. This will give 5.7 amps, per ton, adding J^ more for increased speed and multiplying by 10 for weight in tons would give 5.7 X 4 X 3 X 10 = 76 amps. This, of course, is without any rheostat in circuit with the motors. The same problem can be worked out by the table given on page 271. The following example will illustrate the use of this table: Example— Given a car weighing 12,000 lbs. loaded; the grade at the point where we wish to know the current is 4 per cent; speed required is 7 miles per hour; traction coefficient on this track is, say, 20 lbs.; motor efficiency is 80 per cent; current delivered at 500 volts. The current taken at this point will be equal to watts in table, shown at intersection of 4 per cent grade and 20 coefficient, multiplied as follows: • 1 98.9 X 7 (miles per hour) X 6 (tons weight) ^gO (efficiency) X 500 (volts pressure) ~ " ^ amp8# ELECTRIC RAIL WA Y HAND BOOK. *>o% Fig. 234.— chakt showing relation between traction coefficient and watts at 1 mile per ton per hour for different grades. It is extremely difficult to give the exact speed at which a car will ascend a given grade, for each equipment will fall in speed in mounting a grade until the counter e. m, f. of the motors has been reduced to such a point that the current is sufficient to propel the car up the grade. This point is vaiiable, depending upon the equipment. 304 ELECTRIC RAILWA Y HAND BOOK. In considering the copper service on grades, the cars coming down grade require less current than tliose ascending, and generally above a 3" per cent grade a car will float with open controller after being started up to speed. It is'very important to maintain potential at grades in order that the car can climb the grade at good speed, and to reduce the heating of the motors; and if the schedule can be maintained up grades the motorman will not be called on to make up time by coasting too fast down grade. The usual practice of feeder taps from trolley to feeder every ten or eight poles on levels should, on grades, be reduced to six or four poles, in order to maintain the feeder pressure at trolley wire. The headway of the cars will have to be known in order to get the average current demand, but the average demand is taken care of when provision is made for the maximum demand. In the operation of a railway it is necessary at times to operate more cars over a section of track than are required by schedule, and fixing the maxi- mum current demand is purely a local problem. The maximum demand is usu- ally figured for a total drop of 140 volts where 500 volts are used at the station, and 150 volts with a 550 station voltage. There are a number of ways in which the copper investment can be reduced to handle this maximum load, some of which are given on page 276. By estimating the possible number of cars that could be massed together and using the constant given for each individual car, the maximum demand can be determined. Theoretical Watts Per Ton of 2,000 [Xbs. and Per Mile Per Hour, with Various Grades and Traction Coefficients. Grade Coefficient, in Pounds Draw Bar Pull Per Ton. Percent. 12 13.5 15 18 20 25 30 35 40 50 60 1 2 3 23.9 63.7 103.4 143.2 26.9 66.6 106.4 146.2 29.8 69.6 109.4 149.2 35.8 75.6 115.4 155.2 39.8 79.6 119.4 159.1 49.7 89.4 129.3 169.1 59.7 99.5 139.2 179.0 69.6 109.4 149.2 189.0 79.6 119.4 159.1 198.9 99.5 139.2 179.0 218.8 119.4 159.1 198.9 238.7 4 5 6 7 183.0 222.8 262.6 302.4 186.0 2^5.8 265.6 305.4 189.0 228.8 268.5 308.3 194.9 234.7 274.5 314.3 198.9 238.7 278.5 318.3 208.9 248.7 288.4 328.2 218.8 258.6 298.4 338.1 228.5 268.3 308.3 348.1 238.7 278.5 318.3 358.1 258.6 298.4 338. 1 378 278.5 318.3 358.1 397.9 8 9 10 11 342.4 381.9 421.7 461.5 345.1 384.9 424.7 464.5 348.1 387.9 427.7 467.5 354.1 393.9 433.7 473.5 358.1 397.9 437.6 477.4 368.0 407.8 447.6 487.1 378.0 417.8 457.5 497.3 387.9 427.7 467.5 507.2 397.9 437.6 477.4 517.2 417.8 457.5 497.3 437.1 437.6 477.4 517.2 557.0 12 13 14 15 501.3 541.1 5^0.9 620.7 504.3 544.1 .'83.9 623.6 507.2 547.1 586.8 626.6 513.2 553.0 592.8 632.6 517.2 557.0 596.8 636.6 527.2 567.0 606.7 046.5 537.1 576.9 616.7 656.5 547.1 586.8 626.6 666.4 557.0 596.8 636.6 676.4 576.9 616.7 656.5 696.3 596.8 636.6 676.4 716.2 Multiple feeder Table I (Fig. 235a) gives resistance for any distance up to 6JJJ ft. for all ordinary feeders and trolley wires, together with their usual combinations. Multiple feeder Table II (Fig.235B) gives resistance for any distance up to 48,000 f t,f or all ordinary feeders and trolley wires, together with their usual combinations- Theie are practical feeder charts based on resistance and circular mild* ELECTRIC RAILWAY HAND BOOK. 305 1 VERTICAL LINES "SHOW THE LENGTH OF SECT I ON % FED BY ANY SIZE OF_ FEEDER. ONLY SCHEDULE LOAD CONSIDERED Fig. 235.— diagram showing the distribution op feeders on a road "with cars 500 feet apart. 3 o6 ELECTRIC RAILWAY HAND BOOK. 4000 ,2000 3000 4000 50.Q0 Distance. In Feet. Fig. 235a.— multiple feeder table i. 6000 ELECTRIC RAIL WA Y HAND BOOK. 307 4000 8000 12000 16000 20000 24000 28000 32000 3G000 4Q0OO 44000 48000 Distance in Feet, Fig. 235b.— multiple feeder table ii. 308 ELECTRIC RAILWA Y HAND BOOK. PROPORTIONING FEEDERS, After the current for the feeders has been determined, the next question is the location of the feeding sections and the proper disposition of the copper to get maximum potential delivery. This copper may be in one or several feeders. Where it is combined into one feeder, the cost of copper, per volt drop, is least. The cost of supporting this feeder is less, and the strains which it im- poses on the pole line and the surface it presents to wind pressure, are all in favor of the single feeder. The sub-division of the feeders and the connection of these separate divisions to different circuit breakers in the station is for safety. The sub-division of the feeders may be said to have been originally due to the em- ployment of fuses as safety devices. The action of the fuse required the dividing of the feeding systems up into a number of independently fed sections, but the modern circuit breaker, being much more prompt in its action, provides ample safety for the electrical machinery. Fuses, if placed between separate feeders on the line, will open when any section is grounded, so that in rewiring or recon- structing old distribution systems it is desirable to inter-connect the neighboring feeders by fuses, and thus get the most effective use of the copper. The diagram on page 272 shows the application of this principle of locating feeders to a road 16,500 ft. long, on which the cars are 500 ft. apart with a 12 per cent drop; ordinates are drawn every 500 ft. or for each car, and their length represents the ampere feet required at each point. To apply this diagram to the case of a road with cars a greater or less distance apart the ampere feet required will be inversely proportional to the car spacing; thus with cars 3000 ft. apart the values in ampere feet will be one-sixth of the values given in the diagram. The ampere feet are given in column 15 in the table on page 274, from which the different conductors can readily be selected and applied to find the least feeder cost f orthis distribution. It will be seen from this diagram that the feeding sections grow smaller as the distance increases from the station, so also this method of laying out feeders gives each feeder uniform service. The limiting distances for No. 0, No. 00 and No. 000 feeders, applied to the problem worked out, are shown in the diagram. The ordinates below the datum line show the current consumed by the cars at each point. This current multiplied by the distance from the station, gives the ordinates above the datum line, which are the ampere feet. With a road in which the feeders traverse short cuts, i. e. do not follow the trolley line, the problem would have to be arranged so that the distribution takes place from the inter- section of the feeder and trolley, and the distance in feet to the station from this point, would be the feeder length. Wiring Diagram for Simple Transmission.— The diagram on page 275 will give the correct size of wire to use in power transmission, from a distance of 1000 ft. to 25,000 ft., and from no drop to 200 volts drop. On the lower margin of this diagram will be found current in amperes. On the right vertical edge will be found distance in feet. On the left vertical edge will be found volts drop. On the top of the sheet will be found sizes of wire, with a heavy line vertically through the diagram for each siue of wire. Any wiring problem within the values given on the diagram can be solved as follows: Suppose we had 200 amps, to carry 3 miles with 50 volts drop. Start at the bottom of the diagram at 200 amps., follow this vertical line up until it intersects the horizontal line from 50 volts on the left hand vertical scale, pass along the radial line from this point until the horizontal line from 3 miles is met. The vertical line passing through at this point will lead upwards until the scale of ELECTRIC KAILWA Y HAND BOOK. 309 •^93^ 9.TTmbg m m?ds £ '1& 001 J9d 9jiav' "qotjg- •eucdg -^ 00X n39A^8a diij& jo ^qSio^ owinxo«oxomowooco?|inooqoxo' ^CONOMJ>iOK«OXNiOWr-Ol-W'*1"iONN' w -doici %Z\ WAV. ja P ao J ^ jo ^pedBO ;aoj ojadiny OOiO»On''V^COTOM«(MW«i-imrt S -Japnfl .&noBdBO Sin -jCjibq 'sduiy uinraixtij^ -OOOOOOOOiOOiftOOO^NWOiOtOnOI' oNx^oc^xccotoicoKr-icwxiowria. •e;ioA00S doia^I IR!« ■fa.iBQ iiiAi 9-iiAi 90Ug;eiq; W» \*» •>*>S 4. ~— -»- *-* %J' <#*•; \#^ 1 1 V 1 ^> V •*■•' NT Ni ■ '■^ *.^- •*-•* '*'•' * ■ **-*■ v **-' ••■' 16 00 ojotfToic'o 00 ^^■^■'tf rf-rfco 00 co cocoof oi •q^ 009 £3 ?B sraqo S iBiioiiBtiiaiui 'Q[\iSL wd ooub;9isojj C5^Cifiiir)"NXCJ;CC5(NXTrX»NOC>.WM'-DTt! csco :oic ; ;«o • -of 00 •ir.xo^x CO t^ • iO Tf TT WO< O* O* Tf uC ocaci>Ci 00 /*»t3 .SPH 32 X s niBTa 3 © flS C£ ic o • i— < 4 ;qST9jVi jojonpnoo paoj, 0% ubi'iBinsuj jo 11193 J3 CO CO CO < •13 co w N • CO O ' w ® NOi'.aoCrCHLlXHXO-'OOC!'" •^jooo't-^a^a OXifi OOOQO 10 as tji cjs -«- 00 OXNO'tN CC«I^NW in c* oxokcx »o 01 *c cc -.:: c* o :c o i- cc cc cc i^ cj t- o* c - c ^ - o :; ..-> o c c» i-H cs^cc x »eoc«o cr. i> -j o ■v w w w « r- ^ 30 0^0 i- I- C4t-i«C«J ooir.oocc. o 00 c? o re 1-- »c w « o^oo fi «o eo h o^i- a tt ^ w w w t^ cc" co co co o? 0? of !>f '^f of i-T r-T i-T -^-T i-T — T qOQT UB JO pu£g ;89JB9]^ Tt< *p9piBjg: 9iddux JOOJJ J9q;B9 i W 'tnBTQ[ apisjno vC?is^n?»\ W «e* 1 n» ■^:S" •jo;onpnoo papnB.qg JO J9;dXUBI(J 9pT9^tio KOOtMiocdooi ir: cr. o x cr. o o o o o o o 1-1 t- 1— 10 oj 05 iD 00 ci cc c? cs ;.c — 1- oj i- re Ci CC I- ?* J> CO Ci ^r JO r-TiOOOcic-.c.xxxi-i.-oceir.TrrrrtwwwM C« «8 •— od o OOOOOOOOOOOOOOOOOO1C1O COOOOOOOOOOOOOOCCOOC^Oi OOOOO OOOOOOOOOOOOr-Ci^ Ol ^"* — O OTf CD CJOXXNt'aOOiO^^WMNNi-iT-iTi i 00000000000000000000 WCO^ 3io ELECTRIC RAILWAY HAND BOOK 'J.33J Nl 30MV/1SIO 'S1T0A Nl sscn ELECTRIC RAILWAY HAND BOOK. 311 8 a H «3 . S3 1-^ 0) .P > r p >>•" ■5 £ S P »«> bfii» P P «3 03 «** ^."S ** 50 C3 «-> 03 -J7-*-/8-+-I7-*r ^ !~^-„ •*/-• *-- 22— •-- 21-* *• BT2--! — *to^ 2A v--— — [ — -jo'o-—- — -s I^J FlO. 250.— MISCELLANEOUS TYPES OF CROSS ARMS. at which the cross arm has to bear the greatest strain. In order to avoid piercing the wood at this point with wooden poles and also where wooden cross arms are used on iron poles, straps and plates are employed, as fth'dwn in Fig. 350. The pole is gained in the regular way when wooden, but ELECTRIC RAILWAY HAND BOO/?. 327 ^= 1 1 « 1 s 1 [ 1 N 1 7 1 ' 1 1 TopFicw Fig. 251.— feeder wire supports. 328 ELECTRIC RAILWAY HAND BOOK. where the pole is iron, «a saddle is cut into the -cross arm to make it fit the pole and thus increase the bearing area of the cross arm. Fig. 250 shows a cast-iron fitting to attach a wooden cross arm to an iron pole. 7, Fig. 251, shows a split cross arm construction where the insulator pin is also the spacing piece between the two halves. For securing cross arms to poles in railroad feeder work, use ^-in. galvanized bolts, and lag screws instead of J^-in. Cross arms should be planed straight grained and painted with two coats of Prince's metallic paint, made up in the proportion of 7 lbs. to 1 gal. of pure linseed oil. Three designs of split cast-iron brackets are shown in Fig. 249. The two halves clamp the pole by two bolts which pull them together; the threads of the top bolt catch the whole strain in this method of fastening. The diameter of the pole at point of connection with cross arms can only vary 1 in. for this style of fastening. Instead of a cast-iron cross arm, a sheet-steel punched cross arm made in two duplicate halves, is shown in Fig. 251. The elasticity of the two halves allows for considerable variation in the size of poles. Four bolts hold the halves together and either iron or wooden insulator pins can be used. Side arms are very useful in dodging trees; where the poles .have to be set at a fixed distance from the curb, the arm may be swung either side of the pole, and the feeders in this way cleared of tree contacts. 8 and 9 show the two, three and one pin arm method of fastening. In some instances the feeders have to be supported on structures. 10 shows them carried under an elevated railroad. At A is shown the general ar- rangement, at B the suspension irons, at C the porcelain insulator, at D the spacing washer between insulators. A %-in. iron pipe threads these insulators, which are placed every 12 ft. to 25 ft. 11 shows how a wire support is made where house connection can be made, but no pole can be set. The feeders are tied in the insulators in this case, and the span wire is also insu- lated by strain insulators. ERECTION OF SPAN AND TROLLEY WIRES. After the poles have been set long enough to have the concrete hardened, or the ground settle, the span wires are strung in position. In some cases guard wires are required, but as a general rule guard wire construction is falling into disfavor, because it has been found to add a greater hazard to the overhead line construction. The span wire usually employed is of galvanized steel, or Swedish iron,consist- ing of seven strands, and with general dimensions as follows : Diameter of each wire in inches 07 .11 .12 .135 Outside approximate diameter in inches J T ^ | jl Weight per 100 ft. in pounds 10 21 29 36 Yards per 100 lbs 307 209 130 88 Breaking strains in pounds average 2500 3950 4600 6100 In some cases No. 1 B. & S. galvanized wiped Swedish iron is used, with No. trolley in short spans, and two No. 1 wires are twisted together for long spans and anchor guys. Stranded span wires give considerable less line maintenance cost, after several years of use. Guard wire spans are No. 8 B. & S. galvanized iron, and for longitudinal guards. No. 10 B. & S. is usually specified. Before erecting the pole it should be provided with the proper gains and holes drilled in them to receive the pole fixtures. With wooden poles the span wire may be attached to the pole and adjusted by means of an eye bolt, shown at 1 in Fig. 252. or the fork bolt 2, or ratchets 0, ELECTRIC RAILWAY HAND BOOK, 3^9 1 Pie. 252.— POLE SPAN WIRE FIXTURES, J 33o ELECTRIC RAILWAY HAND BOOK. 7 and 8, For iron poles, clamps, as shown at 3, 4 or 5, can be used. Where attach- ment is made to a lattice pole below the pole top, or to a lattice column, the device shown at 9 may be used. Where a wall is used for the support of the span, a rosette and expansion eye bolt, shown at 10, can be employed. Where iron poles are used, turn buckles, shown at 11, are required for taking the slack out of the span wire; sometimes one eye of the turn buckle is insulated to reinforce the trolley insulator. Insulated turn buckles are shown at 12 and 14. At 13 is shown a wooden strain insulator. At 17, 18 and 19 are shown different methods of splicing in the span wire to the eye of the pole fixtures. Nos. 15 and 16 show span wire insulators. The general methods of disposition of labor in active construction of trolley line work varies in nearly every case, depending on the conditions, material and labor on hand. The work is generally carried out in the following order; placing of pole fixtures, stringing up of the span wires, erection of guard wire, erection of trolley wires, erection of feed wires. When a dead trolley wire is put up it is generally unreeled under the span wires in the middle of the track, and tied with a xJVMPSff I •sp/i/f mnt -as o 7XP^M#£ 9 F££D£/fTJP y $/w/y jr/zif, /tfSC/LA70X /mi/lATOR \mtjul \ *»**»"* 1 <$P/I/V Wfi£ Fig. 253.— methods of connecting feeder to trolley. □ temporary tie wires to the span wire, starting the trolley wire from two strain guys; the tension is then brought upon it. To erect a live trolley wire the reel is mounted on a flat car with brake levers and is pushed ahead of the construction car with the trolley under tension, and immediately attached by means of suspension insulators to the span wires. In this way some 6 miles of trolley can be put up in a day. Where the feeder taps to the trolley wire, two methods are employed: One is to let the feeder tap act also as a span wire, and connect it by a jumper into the feeder ear, as shown in Fig. 253; the other method is to put the feeder ear on a regular iron span wire, and above it stretch the feeder tap, and connect by a pig tail from the feeder tap to the trolley ear. This latter method of construction has several advantages; one is, if the trolley falls, it may break the pig tail, and disconnect itself from the feeder. It also gives a ready method of disconnecting a grounded feeder from the trolley wire. It is found in practice, that a feeder for a Bpan wire has neither the strength nor the durability of the iron span wire, and this weak part of the overhead system can be strengthened by making the feeder tap only an electrical connection, without imposing on it further the carrying of heavy line strains. ELECTRIC RAILWAY HAND BOOK. 33 , 1 CAP AND CON£ $TfiAJGHTUAt£ /AtfVLATOP D0i/3L£ CU/?y£ ^aeP£Afs/o/f CAP AND COAf£ S/Af&L£ cv#y£ 0OVBL£ C(/BV£ SUSPBrtS/QAT JWLAT£D C*OSS/m C APMO COA% S/ "f*iif"*!K BOC/BLB C(//?y£ s/mL£ a/py£ CAPA/VO COM£ C0ff$TMC7/0Al M$VLAT£0 BOL7<$(/$P£AL&°At *££/U $USP£ff$tON BPACK£Ti '^gSermpe**' " STRAtAt£AR. &WCK?TSV$P£A'&0# ^\j SPl/C/Af& SAB f££0£P BAR Fig. 254.— trolley wire insulators and fittings. 332 ELECTRIC RAILWAY HAND BOOK. Fig. 265 —types of side-arm construction. ELECTRIC RAILWAY HAND BOOK, 333 The following strain table will give an approximate idea of the strain on wires when erected for span wires. This table gives the distance of span in feet and the dip of the span wire in inches, with a single trolley and a double trolley: TABLE SHOWING SAG TENSIONS OF SPAN IN INCHES FOR DIFFERENT AND LENGTHS OF SPANS. Strain on Poles in Pounds. Span in Feet. D. T. S. T. 500 lbs. D. T. S. T. 800 lbs. D. T. S. T. 1000 lbs. D. T. S. T. 1500 lbs. D. T. S. T. 2000 lbs. 40 15.4 20.8 26.3 31.9 37.6 43.5 49.5 55.6 61.9 10.6 13.6 16.7 19.9 23.2 26.7 30.3 34.0 37.9 9.6 13.0 16.4 19.9 23.5 27.2 30.9 34.7 38.7 6.5 8.5 10.4 12.4 14.5 16.7 18.9 21.3 23.7 7.7 10.4 13.1 15.9 18 8 21.8 24.8 27.8 30.9 5.3 6.8 8.3 9.9 11.6 13.4 15.2 17.0 18.9 51 6.9 8.8 10.6 12.5 14.5 16.5 18.2 20.6 3.5 4.5 5.6 6.6 7.7 8.9 10.1 11.3 12.6 3 9 5.2 6.6 8 94 10 9 12 4 13.9 15.5 2.7 50 3.4 60 4.2 70 4.9 80 5.6 90 6.6 100..,.....,. no 7.6 8.5 120 9.5 In pulling up span wires, the temperature of the air in which the work is done must be considered, for a span pulled up to 1500 lbs. at 10 degs. below zero Fahr., will only give a strain of 350 lbs. at 90 degs. Fahr., yet a T V in - span wire pulled up with 850 lbs. at 50 degs. Fahr. will reach the breaking strain at 8 degs. Fahr. below zero, providing the pole does not yield. But the result of construct- ing high tension line constiuction in hot weather is to throw the poles out of alignment. With a trolley wire more attention has to be paid to strains, because the distance over which these strains can be transmitted being greater, they fall di- rectly on the overhead line construction, and the effort to displace the overhead line fixtures by this tension must be taken care of by strain guys. In order to keep these strains from distorting curve work, strain guys must be placed at the tangent of the curve, adjacent to the curve, in order to relieve the pull offs from any line strains. Copper trolley wire changes its length in each mile, 4J4 ft. for 90 degs. variation of temperature. The right tension to put on a trolley wire is such that the rise and fall are taken up in the dip, and do not move the line fixtures longitudinally. If, for instance, the poles move with a change of temperature, it may be assumed that the line strains are being trans- mitted unduly. The way to determine whether this is taking place is to throw a long plumb bob line over the span wire and mark on the pavement the position of the span wire in the cool morning, then again at noonday when the sun has thoroughly warmed up the overhead construction. If there is no change in these two positions of the plumb bob with a taut trolley wire it is safe to assume that each span is automatically adjusting itself to temperature changes without caus- ing undue strains on the overhead construction. Overhead line construction is put up with all degrees of tension, "but it does not require very much observation to see that the high tension construction leads to less trolley wheel wear, and the wheel does not leave the trolley at high speeds as readily as with slack overhead construction. Appearances also certainly favor taut lines, 334 ELECTRIC RAILWAY HAND BOOK Fig. 255a,— trolley wire insulators. ELECTRIC RAILWAY HAND BOOK. 335 For approximating the strain pnt on lines by the block and fall, the distance moved by the pulling rope, divided by the distance this pull moves the line under tension, multiplied by the weight applied to the pull, gives the line tension. The different standard forms of trolley wire insulators are shown in Fig, 254. These fixtures shown hold a round trolley wire, but the ear may be so arranged as to hold a figure 8, or grooved trolley wire. Side- Arm Construction.— To decrease line cost and pole obstructions, side-arm construction is often resorted to. Fig. 255 shows some of the types used. The trolley wire fixtures for bracket suspension shown in Fig. 254 are used for securing the trolley wire to the side-arm or bracket and insulating it therefrom. CURVE CONSTRUCTION. Overhead work on curves should be so designed that the wheel will not leave the trolley wire in going around the curves. This is practically accomplished when the following precautions are observed: First. All line tensions should be taken off the trolley wire at the end of both tangents to the curve by running strain wires to take up this tension, (2, Fig. 256.) Second. The location of the curve of the trolley wire should not be directly over the center of the track except at points of tangency, but should depart from these points toward the center of the curve, the departure being greatest at the last named point. (1, Fig. 256.) The amount of departure at the center increases as the radius of the curve decreases. The table on page 300 shows what this should be for different radii. Third. Radii of the curves should not be less than 40 ft. Where curves as small as this must be used, an improvement can be made by making the switches at the ends of the curves of a greater radius than the main part of the curve such as using 70 ft. radius at switches on 40 ft. curves; this eases the curve for about 10 ft. at each end depending upon the position of track and center of trolley stand. The proper position of trolley wire around curves can be found by marking the center of stand on outline of celluloid car body in the same way as given on pnge 124 for locating curves. Fourth. There should be a sufficient number of pull-off s around the curve, so that the deflection of the trolley wire at any one point will not be greater than 10 degs. This is accomplished by properly spacing the pull-off s for curves of different radii and also by having long ears bent to radius of curve. The table shows the distance apart the pull-offs should be in order that the angle be- tween the trolley wire and the pull-offs should not tend to throw the trolley wheel from the wire. In the construction of turnouts no additional poles are necessary as the poles for span wires are sufficient with the large radius of curvature. The pull-offs are connected with poles midway between the two overhead switches and the poles on the line with overhead switches. There are a number of different methods of attaching the pull-offs and of locating the pull-off poles, In Fig. 256, 2 shows one method. All pull-offs, independent of their arrangement, must be provided with turn-buckles, so that they can be varied in length, thus enabling the trolley to be adjusted to the proper curve. Another method requiring short pull-offs is shown in 4, Fig. 256. The poles are set on opposite sides of the curves and heavy span wires run between them, the pull-offs being adjusted to this wire. In another method, known as the flexible method, the trolley wire is connected 336 ELECTRIC RAILWAY HAND BOOK. to heavy span wires by means of pull-off s, 5, Fig. 256. Here all the pull-off s are arranged at right angles to the trolley so that this method has the advant- age of equalizing all the strains on the different pull-offs, thus tending natuially to hold the trolley wire to a curve. Fig. 256.—curve construction. When there is a switch in the line, where it branches in the shape of a Y, a general method is shown in 3, Fig. 256, where the pull-off pole is located iu line with the switch, ELECTRIC RAILWAY HAND BOOK. 337 n TROLLEY WIRE SUSPENSION ON CURVES, Distance be- Distance be- Distance be- Distance be- tween Center tween Pull- Radius in tween Center tween Pull- Feet. of Track and Offs on Feet. of Track and Offson Center of Curves. Center of Curves. Trolley wire. Trolley wire. Ins. Ft. Ins. Ft. 35 14 6 100 4 12 40 12 7 120 4 14 45 10 8 140 4 14 50 8 9 160 3 14 55 7 10 180 3 14 60 7 11 200 3 14 70 6 12 250 3 16 80 6 12 300 2 10 90 5 12 350 2 16 TROLLEY WIRE. This should be hard drawn and in as long length as possible. Several sections besides the round are used as shown in Fig. 257, which leave the lower surface unobstructed for the trolley wheel to roll over. On high speed roads the flashing of the trolley at points of support gives trouble, and equipment breakdowns occur, due to the rise in induced potential caused by this sudden partial rupture of the circuit. The trolley wheel should not strike the insulator bell when worn down to l^-ins. diameter. The tensile strength of round wire, soft and hard, is given in the following table. PROPERTIES OF SOLID COPPER WIRE. (B. & S. GAGE. ENGLISH AND METRIC SYSTEMS.) DIAM. AREA. WEIGHT. Breaking Strain. o 00 8 00 f-i +a a ■H 0> 00 C3 r*. O ^0 00 is <» OQfH So 22 00 j«-5^ FlG.279c.-CROSS SECTIONS OF UNDERGROUND CONDUIT CONSTRUCTION, NEW YORK. Fig. 279d.— side elevation and section op insulator por underground CONDUIT. ELECTRIC RAILWAY HAND BOOK. 35' It was found both in conduit rails and exposed rails that the positive rail retains its insulation when current leaks over the surfaces, whereas it is much harder to hold the insulation with the negative rail on account of the tendency of moisture to increase the negative leakage. AXLE FRAME H TOP OF TRAM RAIL Fig. 279-e— method of attaching plow to car. On the third-rail the current is collected by a shoe on a projecting arm be- yond the car sliding over the rail and flexibly connected to the arm by two laced joints, but the electrical connection is made directly from the shoe to the car wiring circuit. Both the third rail and the underground conduit rail have been successfully operated when under water, especially where the water was pure. Slush is the most difficult thing to contend with in exposed conductors. ^ 358 ELECTRIC RAILWAY HAND BOOK. ELECTROLYSIS. The decomposition of the iron of the rails or subterranean piping systems due to the flow of current from the metallic surfaces into the adjacent moist soil is known in street railway work as electrolysis. The constituents of the soil may cause corrosion, due to the chemical affinity between soluble matters in the soil and iron, forming an iron rust of several molecular combinations of oxygen and iron. This action may be accentuated by the flow of current from the iron sur- faces into the surrounding moisture; but the results of this action are not distin- guishable by any visual or microscopical examination over that caused by natural oxidation of these buried structures. Soils high in chlorides will carry this rust TROLLEY (T^r:r^^::;;;^^-:a~-::^^^3cr: ^s f. X >-_- »-::i^ •-r.s-'-'-^T IPE SYSTEM o: - x ^- DYNAMO S> Fig. 280 — general distribution or current. through adjacent soil. There is, however, a chemical test that can be made im- mediately after uncovering a badly oxidized pipe, which will in some cases determine whether the cause of depreciation was due to natural oxidation or electrolysis from escaping currents. In undertaking to determine whether subterranean metallic structures are affected by the rail return current, it is necessary to find the current flow in this structure, and not to base the deductions on potential measurements between the subterranean structures and the rail. Tests given on pages 43, 44 and 45 give data from which the current flow can be determined, and by carefully tracing this current flow along the pipe system its entrance and exit from the pipes can be determined. It has been found in some railways that the current is carried directly into the water pipe system by a portion of the subterranean metallic •tructure being brought into actuai contact with the rails. Such connections ELECTRIC RAILWA Y HAND BOOK. 359 should be removed before considering any of the methods for testing. The rela- tions of the water-pipe system to the railway system in regard to the pumping station and main arteries of the water-pipe system and the location of the rails and power station are most important features of the problem and should be carefully plotted out. as well as the current flow in the pipe system laid out by means of ordinates on different parts of the piping system, the ordinates corre- sponding in length to the current intensity in the piping system. From this data can be ascertained at what point of the piping system the leakage current can be drained away so as to produce the least flow in the piping system, and to reduce the electrolysis of the rails and pipes to the least possible value. The custom of connecting the pipe system at the station directly to the nega- tive bus-bar rarely leads to satisfactory results. There will be points in the TROLL EY „, rMTIAt . O N RAlk « OROP JN POTEMT IAIO-— = '*(/ RAIL IE PIPE E M& T=r D mi ir~-4— ii ii II ii i i— =ff= POINT OF MAXIMUM CURRENT FLOW IN PIPI POTENTIAL POfENTiALOF WATER PIPE Fig. 281.— distribution op potential: no connection between pipe and RAIL. TROLLEY imr* £&L ^5L h"*l — / / HI — II Fig. 282.— changes in distribution of potential when pipe and rail are connected. piping system where the flow of current is maximum, and there is a varying tendency for the current to flow to and from the rails and pipes. If the piping system is tapped at this point and current led away from the railway system and earth resistance interposed between the rails and piping system, this flow will not be found to be large. It should not exceed 8 per cent of the total rail return in any one section. Fig. 280 shows the general distribution of current between the railway system and the subterranean piping system, and Fig. 281 shows the distribution of poten- tial between these two systems. Between A and B, Fig. 281, is shown the drop of potential along the rail; between C and Z>, the drop of potential on the water. 36o ELECTRIC RAILWAY HAND BOOK. j >0 / 1 / / / / / / /' eJ r / if 7 */ / 1 J / 7 1 / / V 2: / / / 3= O 1 / / in / D 1 / / / / / / / / / / 5 v / A A / <^ / 1 MOURS 10 15 Fig. 283.— curves showing increasing resistance op different soils with 20 amperes per square yard current density. ELECTRIC RAIL WA Y HAND BOOK. 361 pipe. Fig. 282 shows the change in potential when a connection Is made between the water-pipe and rail and ground return feeder. Tests on pages 44 to 46 will show whether the bond is faulty, and deflecting the current into the water-pipe system. Other methods for more complicated systems have to be used for the reduction of this current flow in the water-pipe systems. Connecting a separate set of feeders to the water pipe at their maxi mum current flow points, and maintaining its potential at this point above the rail potential, by a separate booster, will tend to dam back the flow of current from the rails. This booster sometimes is compounded by the flow of the main circuit around its fields, so the difference of potential will automatically vary with the output required on the system. Where gate boxes are brought in actual contact with the rails, the iron vertical portion of the gate boxes should be substituted by wood or terra cotta pipe, and no metal portion of the water- pipe system should be allowed to come within 36 ins. of the rails, unless in- sulated. The problem of deducing the electrolytic effect on the water-pipe systems, where it is of any consequence, is very complicated, on account of the inter- relations between the two systems and the complex current distribution. It is always possible, however, to find simple means for reducing the current flow in the water pipes to a negligible quantity. Electrical Resistance of Earths. Clay, 6 per cent moisture 50 ohms per cu. yd. Sand, 8^ per cent moisture 35 " " " Loam saturated with salt water 20 " " " Gravel and loam, 3^£ per cent moisture 87 " H ** Dry Sand 23,000 " '« " Cement, Portland 1,800 " " " Asphalt, Barber paving 37,800 " " •« Wooden ties, oak rail 3 in. from earth, damp. 8,600 " " sq. ft. rail contact Wooden ties, dry 18,700 " " " contact The resistance between a pipe surface and earth rises as the rust accumulates on the pipe. Different soils show considerable variation in these changes, de- pending upon the soluble constituents in the soil. Fig. 283 shows these varia- tions in a few cases with a current density of 20 amps, per sq. yd.: A, soil from Broadway, New York City; B, soil from Brooklyn, N. Y.; C, soil from Paterson, N. J.; D, soil from Peoria, 111. Increasing the current density causes the resistance to rise and tends to cut down the current flow. Service pipes passing under the rails in the vicinity of the power station often show the effects of electrolysis, on account of the high current density which the adjacent rail surface focuses on these points. Insulated service pipes made for this purpose should be used in these locations; but if iron pipes are used they should be enclosed in wooden troughs and surrounded by hot asphalt under the rails and for a distance extending ft. beyond the rails on both sides of the track. 362 ELECTRIC RAILWAY HAND BOOK. METHOD- OF DETERMINING WHETHER ELECTROLYSIS DOES TAKE PLACE ON UNDERGROUND METALLIC SURFACES. As the natural corrosion of pipe is identical in every respect with that corrosion caused by electrolysis it is important to make a physical test to determine whether this action is due to current flow from the pipe. It can be done in the following manner:— Have a split sheath, which approximately surrounds the pipe under test, cast from the same sort of iron as the pipe is made of in the case of cast iron pipes, and a wrought iron sheath, in the case of wrought iron pipes. Two sheaths are re- quired at each test point. They are carefully cleaned of all dirt and the two halves which go together are numbered and accurately weighed. The length of the sheath should be at least eight times the diameter of the pipe. One sheath is put around the pipe, after they have been carefully cleaned and amalgamated, both on the inside of the sheath and on the outside of the pipe which it covers, and an amalgam alloy put between them, so they will be in electrical contact when clamped around the pipe. The other sheath, after being weighed, is placed around the pipe in the same vicin- ity but insulated from the pipe, by sheet rubber wrapped around the pipe. The pipe with its two sheaths should then be covered up and filled in, in the regular way, and left for a year, when it is opened up again, and the sheaths removed and cleaned by the application of crude oil, and scrubbed with a stiff scratch brush until all deposit is removed. They are then weighed again and the difference in weight between the connected sheath before and after exposure will give those losses in weight, due to both corrosion and electrolytic action ; while the loss in the insulated sheath will be that due to corrosion only, and the difference between the loss on the connected sheath and that on the insulated sheath will be due to electrolytic action. SECTION VI.-CAR HOUSE. In order to design the maximum available storage capacity for a certain area, cut out a plan of a car to a convenient scale allowing the clearance space required on the sides of the car. Where internal pillars are used for supporting trusses, greater distance is required between tracks for clearance. The greatest dis- tance is generally found with open cars with the running boards down, where there should be at least 8 ins. between the running board and pillar. A number of such templets covering the space and clearances required by the type of car to be housed can be arranged in various ways over a plan of the property on the same scale in order to determine the maximum storage capacity with the least special track work. The width of the property available will determine the economical length of span and support of roof. Wooden roof trusses are advised as decreasing the fire hazard, as the structural iron trusses collapse and prevent fighting the fire from outside. The distance between the car house front and the main track should be sufficient to make the entrance curves outside the car house of moder- ate radius. For layout of car house tracks, see Page 157. Transfer Tables.— For handling the cars within the car house, transfer tables are used for which there are several general methods of construction. In one a shallow transverse pit is constructed, wide enough to accomodate,with 18 ins. to spare on each side, the largest single or double truck equipment. The transfer table carries on it a track, upon which, when aligned with any fixed track, a car can be moved. The transfer table can then convey the car to any other desired track. Where the transfer table runs through the middle of the car house, it also forms a bridge to pass over the transfer pit. Another method is to have the transfer table roll on the surface on tracks at right angles to the main tracks. In this case, there are spring sloping tongues which are depressed to the head of the rail, and up which the car rises until it is on the transfer table. This method of construction makes the main track continuous, and the transfer table in this case is only used to transfer the cars at right angles to their length. In case of fire, the latter method has advantages as the cars can be taken out of the car house more expeditiously. The transfer table can be used in small car houses, and light cars moved by a geared hand winch; or a street railway motor can be geared to the wheels of the transfer truck. The current can be carried to this motor by a trolley wire over- head, or a protected third rail which is under the floor, a shoe being used for collecting the current for the motor on the transfer table. The maximum speed is generally 4^ miles per hour, and the motor geared accordingly. Overhead Construction.— The overhead trolley may be strung up on span wires as in outdoor construction or under the beam, using a fixture like that ^64 ELECTRIC RAIL WA V HAA r D BOOK. shown in Fig. 284. A light T-iron is also sometimes used, insulated from the roof work by blocks of hard wood dipped in insulating paint while the wood is warm. It is important in fixing the height of the trolley wire in the car house to have it so high that the trolley pole tension springs are not under considerable tension, for this weakens them, and makes the trolley liable to leave the wire in service, where the height of the wire in most cases is 22 ft. above the rail head. Fl fixture for ET Doors.— The car house doors may be either of the indoor use. swinging or sliding type. The swinging door is most gen- erally used. This door does not require a break in the trolley wire over the sill, if the door is swung both ways from the center. The doors should be well framed and at least 3 ins. thick with double diagonal panelling of yellow pine, and the fastening should be such as to force the door against its jam. Heavy doors should be locked open as well as closed. Floors.— Car house floors are usually of wood. Wooden block paving set on end has been used successfully. There should in any case be an air space be- tween the floor in the car house and the earth, in order that the hot motors will not sweat badly on cooling over a moist floor. For this reason earth and cement floors, especially where the location of the car house does not afford excellent drainage, should not be used under the equipment where stored. The author has found two cases where high rates of motor depreciation were clearly traced to this cause. Where there is a low track with water standing over it, through which the equipment has to run, several companies place a line of steam pipe be- tween the rails above the floor but clearing the motor, on which steam is kept in wet weather and the equipment thoroughly dried at night. Equipments having low insulation, can be baked in this way. General Heating.— Where there are inspection pits it is customary to arrange steam piping around the sides of the pits for heating purposes. But for general heating, the indirect heating methods give the best results for a given weight of steam used. Here flues are carried to the different parts of the car house, and heated air distributed. Flues are also located between the tracks with registers so that this warm air can be blown up under the equipment. The air is heated by first passing through a bank of steam pipes, when by means of a blower it is forced through the ducts of the distributing system. General righting. — The lighting of the car house can be best effected by a group of incandescent lamps or enclosed arc lamps which are arranged along the aisles or passage ways. Light for the night inspector should be especially pro- vided, as without proper light his work is only half done, and as no other man can render such valuable assistance in the maintenance of the equipmeut, every convenience should be placed at his hand. In respect to the proper light to work by there have been several satisfactory methods used. One is to have a flexible cord with a 32 c.p. lamp on one end, and a plug at the other end, plug receptacles being placed around the building. A short cord can then reach from the recept- acle to the interior or underneath the car, four other lamps being banked together at some convenient location in series so that the trolley circuit can be used. The inspector's lamp should be connected on the " ground M end of the series. A bicycle lamp arranged with a handle to be carried in the hand also makes a good source of light. For day work windows for side light, and, where the car ELECTRIC RAILWAY HAND BOOK 365 house is wide, roof light should be provided. Where short roof spans are em- ployed, mill construction with a saw tooth roof having glass on the perpendicular side as shown in Pig. 285 can be employed and gives abundant interior illumination. The front of the car house can be finished in any shape desired to hide the serrated roof. This method makes one of the cheapest forms of car house structures, where A ELECTRIC RAIL WA Y HAND BOOK. 366 wood is employed. The result of dark car houses is dirty cars and greasy floors with scrap heaps at every available corner. CARE OF CARS. Washing.— Where this is done in the car house, a track is generally desig- nated for the work; it should have a cement or asphalt floor pitched % in. to the foot so as to be well drained. The car is first washed down with a stream of water from a hose. For cleansing the inside floors and seats and sweeping out, com- pressed air is used with success, as well as for blowing out the interior of the motor and controller. There is no better method to quickly cool a hot motor or boxes. Lubricating Methods.— Grease is generally put into the journal boxes by means of a paddle, by which it is gouged out of a bucket and smeared into the box. The man who does this, if he is not unusually careful, soon has the floor smeared with grease, and this is the begin- ning of a dirty car house. It is then transferred to tools, workmen and trucks, and the efficiency of a work- man is decreased when he has to look like a coal heaver on account of the grease and dirt surrounding him. The best method is to use a bucket with a short ^-in. spout soldered to the lower side' and a tightly fitting plunger which when screwed down forces the grease out of the spout. This spout can be shoved into the journal box, a turn or two given to the screw and the proper amount of grease injected into the journal box or gear, causing a saving of waste and dirt. Sand. — This is generally sup- plied to the cars at the car house. For charging the sanding boxes dry sand only should used and a con- venient drying arrangement can be provided as follows : Over the boiler room a sealed loft can be built that is used to heat the house and repair shops. Fig. 286 shows the side elevation of the loft and boiler room. It will be seen that the iron stack from the boiler down stairs passes through this room, and around it is a wrought-iron funnel with a circular opening about 2 ins. wide around the stack at the bottom, and with a flare about 2 ft. wide at the top. The wet sand is introduced into this hopper by a motor-operated conveyor from the sand pile. The sand when dried by the heat of the stack spreads over the flue of the sand bin, and is perfectly dry by the time it reaches the spout whence it is delivered into the sand car. In this way the sand is dried practically automatically, and by heat that would otherwise be wasted. Fig. -METHOD OF DRYING SAND. SECTION VII -THE REPAIR SHOP. General Arrangements.— There are several methods in the design of the repair shop by which the equipment can be readily dismantled. In some cases the tracks are elevated from 4 ft. to 6 ft. above the general floor level as they enter by being built upon trestle work with bents about 10 ft. apart. Another method is to locate the repair shop on sloping ground so that the repair shop proper is at a lower level than the car house where the disabled cars are stored. Here the motor and trucks can be lowered to the working floor by some conveuient hoisting arrangement. The usual method is to have a pit the width of the tracks or even wider by supporting the tracks by an occasional I-beam, and this pit is generally laid in brick or concrete and made water-tight. In order to handle the motor and take it from underneath the car several devices are used. One is to have a portable horse which can be placed inside the carbody and to which tackle can be rigged to lower the motor into the pit or onto the repair shop floor in the case of a two-story construction. The more modern method is not to place any rigging work inside the carbody but to use a hydraulic jack mounted on a carriage. On the end of the plunger is arranged a cradle which will engage the motor and lower it below the truck parts; this carriage can be run either along a track or the floor and the motors in this way moved where desired. The tendency in larger repair shops is to raise the carbody off the truck. Fig. 287 shows a rig which will raise the whole carbody well above the truck, so the work can be carried on from above without the use of a pit. 1^ he other method is to jack the carbody off the truck by means of four jacks and two cross- beams; after being disconnected the truck is run from under the carbody and carried into the repair shop. In designing or improving methods of handling car parts in the repair shop, every effort should be made to reduce the time and facilitate the dismantling of a car, so as to remove any of the parts with the least delay possible. With the motor, wheels or any other portions of the equipment many arrangements have been used to carry these parts to the repair department. The sloping track from the pit to the floor of the room can be curved through several pits, or all the pits can lead into one general area-way, so the truck can be hauled up the incline to the repairing floor. Another method is to have an overhead crane over the equipment, which can carry the parts to the repair shop. Where the overhead trolley is in the way, a long insulated flexible cable is used which can be hooked over the trolley wheel and used to bring in or take out the equipments; this gives clear overhead space for crane work. Another method largely used is an overhead track, made up of two I-beams which are supported by brackets, and on the lower flange between the I-beams rolls the carriage on which the equipment parts are transferred from one part of the shop to the other. Either a differential tackle or air lifts attached to an overhead carrier can be used for raising or lowering the equipment parts. J ELECTRIC RAIL WA V HAND BOOK. L ELECTRIC RAILWAY HAND BOOK. 369 The best method to adopt in any repair shop depends entirely upon the rela- tions of the machine and repair shop, and the character of the structures in which this work is carried on. There is still another method for dismantling practiced by some companies which consists of bolstering up the car body and depressing the track, in this way separating the equipment from the body; then the truck is rolled along the tracking pit at right angles to the car body, and is again raised and carried away from the equipment. Lathes.— In small shops a screw-cutting lathe with centers high enough to clear the largest motor armature and long enough bed to take a car axle for turning is required. One such lathe for roads operating 30 cars, two for 60 cars, three for 150 cars, should take care of all commutator turning and axle work, even where repairs are heavy. Where an overhead track is used, it should pass over this lathe, so that the work for the lathe can be brought directly to it. It is found in railway repair shops that the heavier types of lathes are required for this work. A grinding attachment is very useful for bringing the bearings to correct dimensions and polishing. With the harder variety of sand stones, commu- tators could be ground down with less loss than turning. Never use emery wheels on commutators. Tools with individual motors show economy over shafting in machine shop work, and leave room overhead for the crane. In a small shop a double-gap lathe can be made a universal tool, which will take in the gaps the car wheels and the grinding arrangements fitted to true th<5 wheels. The expense of one tool, where the repair work is light, is less than two lathes, one for axle and armature work, and one for wheel grinding. Where the wheel grinding lathe is used instead of the double-gap lathe for truing car wheels, this should, if possible, be placed away from the other machinery, as the dust and dirt arising from this machine injure the bearings on all the other machinery. Drills.— An upright drill, with 14 ins. between the center of the bed and post, is a convenient size for the repair shop. Truck frames, structural work, special work and rails can be handled on this drill, as well as line material, con- trollers, brake rigging and station repairs. In small shops the electric-driven track drill, with flexible shaft, can be attached to an overhead structure and used in the shop for general drilling, when not required on track work. Planers and Profilers.— Where special work is made up, a planer is required with a 24 in. bed. IJST OF MACHINERY AND TOOLS REQUIRED IN REPAIR SHOP EQUIPMENT. MACHINE SHOP. 30-car. Speed lathes, 6 ins : • • • 1 Lathes,swingl4ins. Double-gap swing, 36 ins., screw-cutting 1 Axle lathe Wheel grinding lathe Wheel boring machine Hydraulic press for wheels and armature pinions * Automatic hack saw J Drill 24 ins. vertical * Planer, 24 ins Shaper and Blotter, 16 ins • • Profiler, 14 ins J Automatic tool grinder 1 -SIZE OP ROAD.— 60-car. 300-car. 1 2 2 4 1 2 1 1 • • . 1 1 2 1 2 2 4 • • • 1 . • • 1 1 a 1 1 37o ELECTRIC RAILWA V HAND BOOK. ARMATURE AND FIELD REPAIRS. , SIZE OF ROAD 1 30-car. 60-car. 300-car. Armature stands 4 6 12 Fieid winding machines 12 3 Baking ovens 1 1 2 BLACKSMITH SHOP. Forges 2 4 6 Drop hammers ... ... 1 Shears, cut ^ in. x 6 ins ... ... 1 Punches, hole 1 in. x % in ... ... l WOOD WORKING TOOLS. Planer's surface Ill Splitting saw 112 Moulders Ill Joiners 1 1 1 Sandpaper machine ... 1 1 Vertical mortising machine 1 1 l Tenoning machine ... 1 1 Boring machine 1 1 2 * Where there is much brass work done in the shop, such as overhead line material and equipment parts, the profiler certainly gives the greatest and most direct method of finishing these castings, and practically can do any work that a small milling machine can Other Tools.— The hydraulic press for pressing wheels on and off should also be provided with a device for pressing off the armature pinion and com muta- tor. The armature pinion is often started by a blow from a heavy hammer which in many cases has bent the shaft. There is no small tool more needed in the repair shop than an automatic hack saw which will save considerable labor. One large enough to take the standard rails should be secured. Armature Stands.— There are a number of methods of designing arma- ture stands, the most primitive being two horses in which there is a V cut on the top of the back, in which the armature is rotated. A deviation from this plan is to make two A-shaped supports braced together, on the cross-bar of the A being a platform for tools and coils. Instead of a plain V two rolls are often used for the shaft to rest in so that it "can be readily turned around. The more modern method of making armature stands is to provide two pedestals secured to the floor, which can be elevated or lowered by means of screws and the armature rotated in a fork on the top of the pedestals. This allows the winder to get right over his armature. At his side is a stand for tools and coils. For winding bands on armatures a handle can be clamped over the pinion or shaft, and the binding wire wound under tension. Field Winding.— This is often done in some shops on a lathe, speeded at low speed. If this is a screw-cutting lathe with a capacity to give the proper turns per inch for the wire, by clamping the wire in a tension device secured in the tool post, the wire can be laid on automatically by the feed of the lathe, and at the end of the turns, with a little practice, the winder can soon reverse the feed so there will be no lap over. Where field winding machines are built especially for the purpose, a screw working into a worm gear gives the best results as this locks against any slack, and an automatic arrangement can be attached to the wire feed, consisting of a right and left hand screw which are opposite the side of the field. These screws Will rotate one turn for every turn of the field, and the threads per inch are equal ELECTRIC RAILWAY HAND BOOK. 371 to the turns per inch required on the field. The wire is held on the right hand screw in going forward and depressed on the left hand screw in going back. Bake Ovens.— Small bake ovens for armatures and fields can'be made in the form of a box, into which the armature is slid or the fields hung. This box should have an iron bottom with a number of perforations, and be raised about G ins. from the floor. Underneath this perforated iron floor are arranged a series of five lamps, which are enclosed by a false box in the bottom of which are drilled a number of holes. This should clear the floor about 2 ins. On the top of the box there should be two rows of J^-in. holes about 4 ins. apart, and a slot sliding over these holes. There are holes in the slide so placed that they will register with the holes on the box; and in this way sufficient ventilation can be produced to take away the moisture or fumes which are driven from the armatures or fields. In a number of cases where bake ovens have not given satisfaction it has been found that they did not have sufficient ventilation, the air in them becoming so charged that it would take up no further moisture. The temperature of a bake oven should not be carried over 180 degs. Fahr., especially if linseed oil is used as an insulating medium. General Remarks.— Where compressed air is used for cleaning the arma- tures, a pipe can be run to the blacksmith shop. If a small nozzle is introduced into the center of the supply pipe, a small jet of compressed air will induce a flow of air which can be easily regulated and controlled for the forge fires. The repair shop should be lighted and well ventilated, and where over 30 ft. wide, this should be done from the top.as well as the sides. To paint the inside of the repair shop with cold water white paint gives a neat appearance, and increases greatly the interior illumination. The floors can be concrete or wood, and where pits are in the repair shop the floor space can be increased by putting sliding doors, with handles to slide on battens between the rails. These doors can be slid underneath the truck when only the truck is being worked on. The pit should be lighted by side lights which are covered with wire shields. The pits should be slightly sloped towards one end or the middle where a depression can be made in the cement for collecting any seepage or moisture, which may accumulate in the pit. It is best to have the pit floored with 2-in. hard pine flooring, laid over the concrete, which gives a better surface for rolling the trucks and handling the material. The usual practice in locating machinery in a repair shop is to so place it, that the work passes from one part of the shop to the other in its repair, without passing the same point twice. It is great economy to have all tools belonging to the company marked, and have a tool-room and a workmen's check system, in order that the tools can be replaced, or, if mislaid, the person responsible can be known. In small shops the foreman is generally tool-room keeper, and keeps the tools in order, and in many cases an automatic tool grinder is located in the tool-room. It expedites work for each repair man to have his own set of tools, consisting of large and small wrenches, hammers, cold chisels, pliers, calipers and rules, which are supplied him by the company and charged to his account, credit being given on their return. In repair shops, where the cost of repairs is large, it has almost invariably been found that a large amount of time is lost in looking for tools which can be left anywhere on the work-room floor. Poor or damaged tools do not lead to the best work and discourage the workmen. The numerous labor saving tools and wrinkles for the repair of the different parts of the equipment can be found in their proper place under " Equipment," SECTION VIII.-THE EQUIPMENT. THE CAR BODY. Modern practice in car body construction for heavy high speed work is indi- cated by the specifications given below, which are abstracted from those used by the South Western Missouri Electric Kailway, and are for a car body 30 ft. 6 ins. long, with smoking compartment. This type of car body has been selected for standard specifications as it covers all the details required in the most extensive systems. Whereas the structuraf woods, which have been selected in this specification, conform to the general nomenclature regarding the kinds of wood to be used, the specific species of wood is not often included in the car body specifications, except that for interior finish. The woods forming the modern structure of the car are specified in a general manner since a specific kind of wood is very often difficult to procure. Furthermore the lumber market to-day does not furnish, in the prop- er quality and lengths, certain varieties of oak which some years ago were readily procured. The lengthening of the car body has made these long timbers, partic- ularly the sills, difficult to obtain, especially so when a particular species of oak is specified. The tendency for car body construction is to abandon timber for the longitudal beams on the car and to introduce angle iron around the sill, the whole length of the car, to produce the necessary strength. This method also decreases the depth of the sill. The details are fully designated in the specifications, which form a guide to the general make up and dimensions of the parts entering into car-body construc- tion. The size of car bodies varies. The dimensions of a number of car bodies, which have been recently ordered, are given for reference in the table on page 324. Car Body Specifications.— General Dimensions. Feet. Ini. Length of car body over corner posts 30 6 " " " over vestibules 40 2 41 •' •* •' bumpers 41 2 Width of car body at sill, over plates 8 3^» 44 " 44 " atbeltrails 8 3^6 4i " " 44 over water drip rails 8 4^| Height of car body from under side of sills to top of roof, not includ- ing trolley board 9 Height inside, from floor to top plate 6 3 44 44 t4 top of floor to underside of upper deck ceiling. . . 8 2 44 from track to underside of sills when car body is mounted on trucks S 1 Bottom Framing.— Side sills of very best quality long leaf yellow pine, size 4^ ins. x 7% ins. finished, reinforced on the outer sides with steel plates ; plates to be % in. thick, 7 ins. wide, extending full length of car body and around corners ELECTRIC RAIL WA Y HAND BOOK, 373 a SO C40>> «- a a 03"* i| www* P* OiOO TjiCOO «3 p ® p O w «H • o o Kind of Seats. .fl CJC 33 „ i „ : ."2 I *0Q : 2~ ' 3 : 2 tH 4> .O « B*i CM CMCMCM • ao«>© • r-i fc m gor-o «o o: w »o ■*H & j£ *• 3 . S *" i> ®i t~ -*-3 • ^h on 2 - - j>fl» s ^ : : • ibi 0D 3D flop "- 1 -«-s "^ -• •* - " % 22 8" " ' CO CO ■♦J «o o b£^_ *» .«*-■ '53 o"^ ^ , • ««-•« a 00 5« ^T ^ 1 ^^ ^* CM ,P Lengt over Posts P o QO 00 £ CM J3 00 eo p «OOS ^£ * a p >— m £ n www OS «o«o CM CMOJCO (H I u T3 fl a> OQ £ o -a o s ^ C3 A J I 3 09 £ s 2 - ^3 3D 09 §!§.= - W o ~ * tf'tf- «a p 3 p- 'O Q) 1 ** a fl fl p a> a»«. 5 o o aa- c CJ © C O P cocoooooooo -J.n«ooo*>«oos;o o r ft„- &V2 P oo cc r-f Tt< i> OS o ,53 c3 Pi Choosxosn P P eU p .S^OOOSOt-h-i ZZ P « 5 w o o x> o ~ O c3^-e3"S 53-P j, v~ u ^ >«- S^^^o °°o"cf^Ot A — _ —-^ *^" c O »- v 09 J 374 ELECTRIC RAILWAY HAND BOOK. to door posts on ends in one continuous piece and securely bolted to sills with oval head carriage bolts. End sills of oak, size 7% ins. x 4% ins. finished. Cross sills of oak, size 6% ins. x 2% ins. finished. Diagonal braces of oak, size 4 ins. x V/± ins. finished. Trap door framing to be arranged to meet the requirements of the otor, for two ...motors to each car, of type No and of hp. sach. Bottom framing to be also arranged to meet the requirements of the double trucks, with wheels 33 ins. in diameter, wheel base of trucks , dist- ance between centres of truck bearings , bolsters to be set back from end of car body, gauge of track 4 ft. 8}4 ins. Bottom framing to be tied together with % in. round refined-iron rods, one for each through cross sill, extending across the whole width of bottom frame and through side sills and plates, with thread and nut at each end. Short fram- ing and trap door framing to be also tied together with %-in. round refined-iron rods and plated where necessary. Body bolsters to be made of 9 ins. x % in. and 9 ins. x % in. iron filled with oak, made to suit the requirements of the trucks and securely bolted to the car body. Inside truss rods to be made of 2 ins. x 1^ ins. refined iron with 1 in. round ends with thread and nut. These trusses to be placed as high as possible and out into posts and securely fastened to posts and sills. Also truss rods of 1 in. round iron to be placed underneath side sills to support centre of car, with turnbuckle in the centre, and securely bolted to sills at ends. Floor to be of best quality quarter-sawed yellow pine, % in. thick with 2%-m. face, tongued and grooved, closely laid longitudinally with car and securely nailed to bottom framing. Floor to be of two thicknesses, the lower layers being laid crosswise and the top floor to be laid longitudinally with car. Trap doors to be fitted with large wrought-iron handles, counter sunk into trap doors and bolted through. All parts of bottom framing and under side of floor to be painted with brown oxide of iron paint, and all mortises and tenons to be thoroughly white leaded when being put together. Body Framing.— All posts, rails and other parts of body framing to be of best quality tough ash. Ventilator rails to be faced with oak on the inside and molded. Side of car to be framed for 11 windows on each side with straight siding below windows. Ends of car to be framed for double doors and one window at each side of door opening. Posts to be well braced. Diagonal cross bracing between posts, made of oak % in. x 3% ins. and tightly fitted. Outside of car below win- dow to be sheathed with matched poplar sheathing vertically and well nailed to body framing, sheathing to be % in. thick and 2 ins. wide on face. Inner side of sheathing to be backed with best qualily linen scrim well glued to sheathing, posts and bracing, and when the glue has become hardened the scrim to be painted with brown oxide of iron paint. The joint between sill plate and sheathing to be covered with 1*4 ins. x X in. half oval iron molding extending full length of body at sides, also on ends. The joint between arm and rail and top edge of sheathing to be covered with 1J4 in. x y^ in. half oval iron molding extend- ing full length of body and around corners to door posts on ends In one continuous piece. Posts to be secured at bottom with strap bolts made of 1*4 in. x & in. iron, extending through sill with thread and nut at bottom. Strap bolts to have heel bent on top and cut carefullv into posts and to be securely screwed to posts. ELECTRIC RAILWAY HAND BOOK. 375 Posts to be tenoned, and sills and top plates to be mortised, drawbored and pinned and all pinning to be done in such a manner that pins will draw tenons into mor- tises. Tops of posts to be provided with wrought-iron T-plates securely fastened to posts and top rails. Side top rails to be of best quality yellow pine, size 4% ice. x 3 ins., finished, strengthened by a heavy letter panel made of ash \% ins. thick and 7 ins. wide, which is gained into top rail as well as posts. Under drip rail to be 234 ins. x ^34 i ns - x 34 in. angle iron, extending from corner post to corner post in one continuous piece and securely screwed to posts and letter panel. Arm rail of ash, size IX ins. x 4% ins., finished. Side posts of ash, 234 ins. thick, cut to pattern. Corner posts of ash, 4 ins. thick, cut to pattern. Diagonal bracing be- tween posts of oak, size % in. x S% ins. Truss plank at bottom, inside, of ash, size 1J4 ins. x 9 ins. Door posts of ash, size 2% ins. x 2^ ins. Each door post shall be provided with one % in. round iron rod, extending from top of headpiece through to bottom of end sill, screwed to headpiece and let into door posts, and to have thread and nut at bottom. Roof.— To be of the monitor-deck pattern full length of car. All parts of roof framing to be of best quality tough ash. Ventilator rails to be faced with oak inside and molded. Lower ventilator rail of ash faced with oak, 2^ ins. x 4)4, ins. Upper ventilator rail of ash, size 234 ins. x 3^ ins. Ventilator mullions of ash or oak, size 1% ins. x 234 ins. Side of lower deck carlines of ash, size 1% ins. x 2 ins., cut to pattern, glued and tenoned into lower ventilator rail, and shouldered, glued and screwed to side top rail. Centre of upper deck carlines of ash, 1% ius. x 1% ins., cut to pattern, glued and screwed to top ventilator rails. Roof to be further supported by ten steel carlines, one over each intermediate side post, made of 134 in. x % in. steel, forged to shape of roof in one continuous piece and extending from side top rail to side top rail, with a foot at each end, which is securely screwed to side top rails. Steel carlines to be securely bolted to wood carlines. Roof framing to be covered with matched poplar sheathing ^ in. thick and 3 ins. wide, closely laid and securely nailed to roof framing. Outside of roof sheathing to be painted with brown oxide of iron paint, and then covered with No. 8 cotton duck, and outside of duck covering to be painted three coats of lead and oil paint before trolley board is placed in position. Inner side of roof sheathing to be painted with brown oxide of iron paint. The entire upper deck of roof to be covered with roof mats made of ash slats, placed on ash cleats laid in white lead, and securely screwed to roof framing. Trolley Board.— To be of best quality white pine, made of two boards 1% ins. thick, 6 in. wide and 11 ft. long, placed 4 ins. apart on centre of roof, placed on ash cleats laid in w T hite lead and securely screwed to roof. All parts of trolley board and roof mats to be well painted with lead and oil paint. Bolts to be pro- vided for securing trolley pole base stand, and to be located to suit the base. Roof mats to be placed on lower decks at diagonally opposite corners of car, also black steps on corner posts and grab handles on roof at same corners, of black. Platforms and Vestibules.— Platform outside knees of oak, 3^ ins. thick, 8 ins. wide. Platform inside knees of oak, %% ins. thick, 8 ins. wide. All platform knees to be re-enforced with 5 ins. x % in. iron plates extending full length of knees and securely bolted to same. Platform floor to be best quality quarter-sawed yellow pine % in. thick with 234 in. face, laid crosswise and securely nailed to platform framing. Distance from top of car floor to top of 376 ELECTRIC RAILWAY HAND BOOLT. platform floor 6% ins. Platforms to be 4 ft. 10 ins. long from end of car body to outside of vestibule front at centre of car, enclosed at front by stationary vestibule, with folding doors at both sides hung to vestibule corner posts, also folding gates for summer use. Vestibule to have three sash at front, the sash in centre opening to drop, and all other sash to be stationary. All framing parts of vestibule to be of best quality tough ash. Lower part of vestibule fronts to be sheathed with matched poplar sheathing, placed vertically same as on sides of cars, sheathing to be % in. thick and 2 ins. wide on face and securely nailed to vestibule framing. Inner side of sheathing to be backed with linen scrim and painted same as on sides of car. Inner side of vestibules to be finished in oak throughout, finish at bottom to be panel work, all finished in the natural color of the wood and varnish. Hoods.— To be of the street car pattern. The bow to be of oak 1% ins. x 1% ins. steamed and bent to shape. The carlines to be of ash % ins. x 1J4 ins. steamed and bent to shape, shouldered, glued and securely screwed to bow. Hood carlines against end of car to be of ash, \% ins. x 1)4 i ns - Hoods to be covered with matched poplar sheathing % in. thick and 2}4 i ns - wide, bent to shape and securely nailed to bow and carlines. Outside of sheathing to be painted and covered with cotton duck, and outside of duck covering to be painted three coats of lead and oil paints. Inner side of hoods to be painted same color as outside of car body. Outer edge of hoods to be provided with an iron guard to prevent trolley pole from wearing out canvas covering. Smoking Compartment.— Each car to have a smoking compartment at one end, 7 ft. 8*4 ins. long between end linings, with longitudinal side seats made of oak and varnished. Partition between smoking compartment and balance of car to have a single sliding door, 25-in. opening. Windows at each side of door open- ing in partition to be glazed with clear glass; also glass in door to be clear. Windows.— Eleven on each side of car, and two at each end. Each window opening to have two sash, the upper one to be stationary and the lower one to drop flush with arm rail. All window openings at sides of car to be provided with a hinged casing covering space between sash and inside lining, which will close the opening both when sash are up and down. The lower outside end sash to be made to drop, and the inside end sash' to be hinged and fitted with brass wire cloth. Interior Finish.— Interior finish of car body and vestibules to be of quar- tered oak throughout, of modern design and secured in place with solid black oval-head screws. End and side linings of oak, % in. thick, with raised panels T 9 g in. thick. Doors: Center rails of oak, 6^j ins. x 1 T 5 B ins. Lower rails of oak, 8J4 i" s - x l^g ins. Upper rails of oak, 4^ ins. x 1 T 6 S ins. Side door stiles of oak, 4^j ins. x 1 T B B ins. Center door stiles of oak, 4 ins. x 1 T 6 6 in. Panel mutins of oak 2 ins. x 1 T 6 B ins. Raised panels of oak, T 7 6 ins. thick both sides. Sash: Bottom rail of oak, 3^ x % in. Top rail of oak, 3% ins. x % ins. Side stile of oak, 2^ ins. x % in. Deck sash: Lower rail of oak, 1% ins. x % in. Upper rail of oak, 1/fe ins. x 2£ in. Side stiles of oak, l T 6 g ins. x % in. Transom sash in ends: Lower rail of oak, cut to pattern, 1J4 ins. a % in. Upper rail of oak, 1 in. x % in. Side stiles of oak, 1% in. x % in. Ceiling. — Of three-ply birch veneer, plainly decorated and varnished, ceiling and ceiling moldings to be secured in place with screws, and ceiling moldings to ELECTRIC RAILWAY HAND BOOK. 377 be grooved on back to receive the lamp wires. Back of ceiling to be painted with brown oxide of iron paint. Doors.— Automatic double doors at each end of car, made of oak with oak panels, hung at top with contra-twist door fixtures. Door in partition to be hung with hangers and track. Door openings at ends of car to be 40 ins. wide and 6 ft. 3 ins. high, and door openings in partition to be 25 ins. wide and 6 ft. 3 ins. high. All aoors to have stationary glass. Sash.— All sash to be % in. thick and made of oak. Deck Sash.— Eleven on each side of car pivoted and made of oak. The ends of ventilator or monitor deck to be divided into three spaces, with pivoted sash in center opening and stationary glass in side openings. Glass— The glass in all windows and doors to be first quality double thick American window glass, imbedded in molded rubber on all edges to prevent rat- tling. Glass in deck sash to be double thick white chipped, with 1 in. clear edge and imitation bevel. Curtains.— All side and end windows, also outside of end door openings to be provided with curtains. Curtains to be made of material, pattern color , mounted on 1-in. spring rollers and fitted with the . . .- fixture at the bottom. Curtains on outside of end doors to be made up in same manner as the other curtains, to be placed in a neat oak box over door opening with side pieces extending down to arm rail to form guides for curtain fixtures. Seats.— There are to be eight (8) on each side of car in large compartment, six of which are to be of Walkover pattern, and four stationary seats; the two stationary seats at end next to vestibule to be placed longitudinal with car, and the two stationary seats next to partition to be placed crosswise. Cushions of all reversible and the two longitudinal stationary seats to be 33 ins. long and the two stationary seats next to partition to be 31 ins. long. Backs of all cross seats to be 22 ins. high, and those of the longitudinal seats to be of same height as side lining (13 ins.). All reversible cross seats to have grab handles on coiner of back at aisle end, also movable foot rest and thumb latch on seat back levers. All seat cushions to be 18 ins. wide. All seat and back cushions to be covered with canvas-lined rattan with hard enamel finish. Space between center of stationary cross seat and the first reversible seat to be 37 ins. Reversible seats to be spaced 17 ins. between edges of seat cushions when passengers are facing each other, and 16^ ins. between edges of seat cushions when backs are all turned same way. Aisle through center of car to be 20 ins. wide. Hand Rails.— To be placed in smoking compartment only, supported by black ornamental brackets. Rails to be made of oak 1% ins. diameter, with pol- ished black ends. Each rail to be supplied with six (6) padded hand straps made of fancy leather and fitted with black buckle. Floor Mats.— Everett pattern, made of ash slats % in. thick, % in. wide, placed % in. apart, extending longitudinally in aisle full length of car and sunk flush with the floor. Trimmings.— Of very best quality black metal, dead finish and secured m place with solid black screws. Grab Handles.— Long vertical grab handles to be placed on posts at each Bide of each vestibule entrance, made of 1-in. steel-lined black tubing fitted into 378 ELECTRIC RAILWAY HAND BOOK. black end sockets, 36 ins. long, and securely screwed to posts with solid black oval -head screws. Window Guards.— Three bar window guards to be placed on outside of all end windows, made of ^ in. heavy black tubing filled with hard wood and secured in place with solid black screws. Signal Bells.— Two 6-in. steel conductor's signal bells to be supplied with each car, with necessary cords of ig-in. round* oak tan leather extending through center of car suspended from ceiling with suitable black hangers with 13-in. drop. Register.— One fare register of latest pattern to be furnished with each car, with necessary cords of r 5 e -in. round oak-tan leather extending along ventilator rail at each side full length of car and onto both platforms, sus- pended by suitable brackets or guides. Wiring.— Car bodies to be wired for light and trolley circuits. Light circuits to be arranged for four single lights on each ventilator rail in large compartment and two single lights in upper deck of smoking compartment. All wiring mate- rial, sockets and switches to be furnished by the railway company. Headlight.— Each car to be equipped with one headlight complete, arranged to hang on front of vestibule. Heaters.— Each car to be equipped with sixteen electric heaters, type to be selected by the railway company. Sand Boxes.— Each car to be equipped with two (2) sand boxes— one at each end at diagonally opposite comers of car, placed under seats, arranged to operate by foot lever, and supplied with removable hose. Gongs.— Two 14-in steel foot alarm gongs to be supplied with each car one under each platform. Brake Staffs.— One on each platform, 1% ins. round at the bottom, forged tapering to 1 in. round at the top, well braced, fitted with 12 L 2-in. black ratchet brake handle and %-m.. twist-link Norway-iron brake chain. Platform Steps.— Double steps at each platform entrance, sides of steps to be made of steel plate *4 in. thick, with treads of ash % in. thick and 8^ ins. wide. Distance between end of car body and inside of platform crownpiece 38 ins. Distance from outside of car side to outside of platform knee 13 ins. Dis- tance from top of platform to first step 11 ins., and from top of upper step to top of lower step \0% ins. Distance from top of platform floor to underside of bot- tom step 24 ins. Distance from top of platform floor to top of car floor 6% ins. Edge of bottom step to project 1% ins. beyond side of car. Outer edge of step treads to be covered with iron molding. Draw Bars.— Extra heavy radiating spring draw-bar at each end of car, with necessary slides, all securely bolted to car body. Height from track to center of draw-bars when car body is mounted on trucks to be 22^ ins. Bumpers.— Angle iron bumpers to be placed on front of vestibules, made of 6 ins. x zy% ins. x % in. angle iron bent to same shape as vestibule front, and extending full width of same, projecting 6 ins. beyond front of vestibule and securely bolted to platform knees, which project out for that purpose. Height from track to center of bumpers to be 33^ ins. Material ani> Workmanship.— All material entering into the construction ELECTRIC RAILWAY HAND BOOK, 379 and finish of these car bodies to be of the very best quality ; all sills to be full length without splicing; mortises and tenons to fit each other tightly without Zalse filling, and to be thoroughly white leaded when being put together; all lumber to be of the very best quality and thoroughly well seasoned and dried; and all work to be done on a strictly first class workmanlike manner. Corner Posts and Headpiece.— To be securely tied together with an ang.e iron brace let into top of headpiece, corner post and side top plates. Outside of corner post, where it joins plate and headpiece, to be protected by a heavy iron plate. Corner iron to be placed in corners of hoods where the bow joins the rear carline. Painting.— These car bodies to be painted in the best possible manner, let- tered, ornamented and striped as desired by the railway company, and varnished throughout with railway varnish. Outside of car to be painted in the following manner: Two coats of pure white lead and linseed oil. After sec- ond coat car to be puttied and plastered where necessary. Car to be rubbed with pumice stone and water until a perfectly smooth surface is obtained. Two coats of flat body color, and if necessary an additional coat will be put on, depending on color used. One coat color and varnish, on which striping and lettering will be done. Two coats rubbing varnish, second coat to be rubbed with pulverized pumice stone and water. Two coats railway body varnish. All striping to be done in gold. Lettering and numbering to be of same size and 1 style as shown by photograph to be sent by railway company. Roof of cars to be painted three coats of lead and oil paint, each coat to be allowed to dry before the i succeeding coat is applied. Inside work to be finished in the following manner: All parts of inside woodwork of car and vestibule to receive one coat of oil filler, one coat of pure linseed oil, and three coats of rubbing varnish, last coat of varnish to be rubbed to a cabinet finish. No shellac to be used on any part of the cars. Inspection.— The railway company shall have the privilege of sending a re- presentative to the shops of the car builders to inspect and examine the cars while : being built. Trucks, Mounting Motors and Installing Electrical Equipment.— The ! company to furnish and deliver free of any expense at the works of the car build- | ers all necessary trucks, motors, wire, switches, sockets and all other electrical 1 equipment for these cars, and the car builders to install same without any extra charge. Time op Delivery.— These car bodies to be delivered complete, f . o. b. cars, on or before 19...., subject to delays caused by fire, labor trouble or any other cause beyond the control of the car builders. NOMENCLATURE OF CAR PARTS. (See Figs. 288, 289, 290.) Abreviations. WOODWORK. METAL WORK. O Oak C. I Cast Iron A Ash W. I , Wrought Iron Y. P Yellow Pine M. I Malleable Iron P Poplar W. S Wrought Steel 3 8o ELECTRIC RAILWAY HAND BOOK. lOj ELECTRIC RAIL WA Y HAND BOOK. 38l ^ 8. Sill, Y. P., 0. 56. Upper deck. 9. End Sill, 0. 57. Deck bottom rail. 10. Transverse floor beams, 0. 58, Deck post. 11. Cross tie rod, W. I. 59. Deck window. 15. Side post, A. 61. Deck end ventilator. 17. Corner post, A. 64. Window. 18. Door post, A. 66. Window stile. 19. Belt rail, A., Y. P. 67. Sash lift. 20. Belt rail band, W. I. ( half oval. 68. Sash stop bead. f»»m»mm^ '4-5^ Fig. 289.— end view op car body. 21. Fender rail, A., Y. P. 69. Window blind. 22. Fender gnard, W. I., half oval. 72. Window blind mullion. 23. Body truss rod, W. S. 73. Window blind lift. 24. Body queen post, W.S. ,M.L, C.I. 77. Window guards. 25. Truss rod plate, W. I., C.I. 78. Door stile. 27. Outside panel, Convex P. 79. Door mullion. 1 28. Lower outside panel, Concave P. 81. Middle door rail. 29. Upper end panel. 82. Top door rail. ! 30. Lower end panel. 85. Mirror. 31. Inside frieze panel. 86. Door case sash. J 32. Panel strip. 89. Fare wicket. | 33, Panel furring. 91. Sliding door handle. 34. Seat bottom. 102. Platform timber clamp. 35. Seat leg. 103. Platform end timber or crown i 36. Front seat rail. piece. J 382 ELECTRIC RAILWAY HAND BOOK. 38. Back seat bottom rail. 39. Back seat rail. 40. Lower seat back rail. 42. Seat back board. 43. End seat panel, 44. Upper belt rail. 45. Window ledge. 47. Plate. 48. Eaves moulding. 49. Window blind rest. 50. Window sash rest. 51. Outside window stop. 52. Inside window stop. 53. Carline. 108. Platform post. 109. Platform post boss washer. 110. Platform rail. 112. Dash guard straps. 113. Body hand rail. 114. Side step. 115. Hood. 116. Hood bow. 117. Hood carline. 119. Hood moulding. 120. Brake shaft crank. 122. Brake shaft. 123. Upper brake shaft bearing. 125. Brake ratchet wheel. Fig. 290.— sectional view of car body. ELECTRIC RAILWAY HAND BOOK 383 The modern tendency is to lengthen car bodies, especially where there are many short distance riders, since it leads to increased comfort and the probability of the passengers always obtaining seats, and this will attract more traffic al- though the cost for operating is very slightly increased. Where double trucks are substituted for single trucks, the car body can be lengthened 4 ft. and in- creased in weight proportionately without perceptibly affecting the demand on the power station, or increasing the power supply per equipment; and, the labor item being a constant, the cost per passenger will be less. SPLICING CAR BODIES. A number of roads have increased the length of their car body by cutting in two and splicing. The Union Traction Co., Philadelphia, has followed the following practice : Taking its short 18 ft. car bodies, which had six windows, Floor Line %" Fine Sill of old Car #*6" ' Qa/t #z~*/2" Georqta Pine Tur/fBucAle /?' AuFs/f Fig, 291.— method or splicing car bodies with wooden beam the body is sawed in two. The side sills are cut dovetail, and the inserted length- ening sill is dovetailed into the side sill. The car is lengthened so as to just receive two more window frames, which gives a car body length of 24 ft. when complete. The inserted sills are only used to hoid the uprights. For strength a steel angle is fitted to each side of the car for the length of the body. The sides of the angle are steel plate, L-shaped, 6 ins. x 4 ins. and % in. thick, the 4 in. side being placed under the sill. In addition to this angle iron, a truss rod is placed behind the sills to reinforce the angle, and two maximum traction trucks are placed under the body. Another method is to splice two small cars together. This has been carried out by one company in the following way: The two bodies are set on horses; the back end is removed from one of these bodies and the front end of the other. The corner posts and end sills are ripped in half. Then the two bodies are butted together and fastened with %-m. carriage bo'.ts, placed so as to be out of sight. The sills are reinforced by ^a-in. angle iron plates, extending 4 ft. each way from the splice. The old panels and water rails are removed, and a piece of Southern pine, 4^£ ins. x 12 ins. x 32 ft. is set into it flush with the bottom of the sill, 384 ELECTRIC RAILWAY HAND BOOK. and bolted throngh the old sill every 22 ins., using ^-in. bolts with cast washers on the pine side. On. each side of the splice two rods %-in round went clear through the body. The 4^ in. x 12 in. sill was framed to fit the top curvature of the post. Fig. 291 shows a section of this car body and the method of trussing to support it. To give sufficient strength to the roof, two pieces of Southern pine, 2 ins. x 8 ins., extend the length of the roof and fit into an angle of J^-in. iron at each end. Two tie rods, % ins. x 32 ft. x 5 ins. pass through these angles with double nuts on each end. By using these tie rods it is possible to give the roof any de- sired cambre to hold it there. The 2 in. x 8 in. pine pieces are 14 ins. from the roof at center of the car. On top of this truss is mounted the trolley stand, and a double truck is placed underneath the car. MODERN INTERURBAN CAR CONSTRUCTION. The size and weight of interurban cars has been greatly increased during the last few years. The bottom frame is now frequently built with six longitudinal sills -two center, two intermediate and two side sills. The side sill is usually made of one piece of long leaf yellow pine 5 in. by 8 in. and one piece of the same 2 in. by 6 in. enclosing a 6 in. by y% in. steel plate between them, all being securely bolted together. The center and intermediate sills are 6 in. steel I-beams with filling pieces on each side, all bolted together, forming square sills into which the bridging or cross sills can be framed. The four center and intermediate sills extend, in many cases, from buffer to buffer, thus making the platforms level with the car floor and capable of support- ing a heavy weight without sagging. This construction gives great protection in case of collisions. A number of cars have been built in which the entire bottom frame was made of steel I-beams with wooden strips on top for fastening the floor. The latter is in two layers, one or both laid diagonally. In a modern body frame the side frame from the sill to the belt rail have been greatly strengthened. Fig. 291b shows how the side frame is trussed and braced which aids greatly in preventing the breaking up of the car in an accident. The roof is of the steam railway coach type with hoods covering the vestibules at both ends and built in the same manner, with steel carlines at every side post. The usual length is from 50 to 52 ft. over all, and the total weight, including trucks, motors and all equipment, is from 50,000 to 70,000 lbs. Owing to their solid construction and the costly interior finish now used, which is usually mahogany, these cars, such as is shown in Fig. 291c, cost from $6,000 to $9,000, com- plete with all equipment. FIRE PROOF CARS. Since the beginning of underground operation of heavy electric cars, several improvements have been made along the line of fire protection. Some of these cars are built of wood much the same in design as heavy interurban cars, but the entire space under the car is covered with a fire proof material similar to asbestos board, % in. to % in. thick. On this is placed the car wiring. The sides of the cars from the under side to the belt rail are sheathed with sheet copper, this will aid in fire protection, and it never requires painting. Recently there has been put in service a number of the Gibbs all steel cars which are practically fire proof. ELECTRIC RAIL WA Y HAND BOOK. 385 »Ki5S f Jl? <. fa*ra ssn'jT ST. 386 ELECTRIC RAILWAY HAND BOOK. ft 1 1 =&r-| AL — I- □J '■ W ' - ELECTRIC RAILWAY HAND BOOK. 387 TRUCKS. Test on Peckham Truck.— All castings except center bearing and side poles were malleable iron; side pieces were formed of 4 bars of flat iron, riveted to pedestals, placed in pairs, to take compression and tensile stresses. The truck was the same as a 14-A, except heavy enough to carry loads of 30,000 lbs., with a factor of safety of 6. The test was made by the Robert A. Hunt Co. on a wheel press. Load, Tons. Total Deflection, Ins. Permanent Set, Ins 5 .00 .00 10 .00 .00 15 .02 .00 20 " .03 .00 25 .04 .00 30 .07 .03 35 .16 .10 40 .19 .13 45 .26 .17 At 50]^ tons, the lower tension members broke through the first rivet hole and the malleable casting at one end. Figs. 292-294 show the No. 14 Peckham truck. The length of the car body, upon which the trucks were mounted, was 24 ft. The length over all 33 ft. The distance from rail to bottom of car sill, 28 ins. The distance from rail to car steps 16 ins. The width of car body, 7 ft. 7 ins. The diameter of axle, 4 ins. The wheel base on trucks, 5 ft. 7 ins. Diameter of wheels, 30 ins. Method of Increasing Traction. — It has been urged against the double truck that the weight on the driving wheels was not sufficient to mount grades. This has been successfully overcome in one instance, where a 14^£ per cent grade had to be mounted, by placing double trucks under the car, but both motors under the rear truck. It must also be borne in mind that the hanging of the motor for- ward of the front axle of the car increases the weight on the driving wheels, on account of the additional weight of the motor; in this way greater tractive effect can be given for maximum traction or double truck cars. In descending grades there is twice the coefficient of adhesion between the eight wheels that there is with the single truck four wheels. As a result of care- ful tests between the double and single truck, it is found that the double truck requires less power, for the same car body, than the single truck, especially where the single truck allows considerable teetering of the car. In one case a car body, weighing 3 tons more with a double truck and the same motors, showed 1.21 kw J ^ 388 ELECTRIC RAILWAY HAND BOOK, ELECTRIC RAIL WA Y HAND BOOJC. 389 per car mile, whereas a single truck, with the same motors, showed 1.37 kw per car mile over the same track. The only change was made in the truck. In looking for the difference in efficiency of these two types of trucks, a cyclometer was put upon each driven axle, and it was found that the actual slip- page ranged between 10 per cent and 18 per cent between the front and lear wheels when the car was on a level and when climbing a 5^ per cent grade, the front wheel having the slippage. It must be remembered that, in case of wheels slipping, the heating of the track and wheel is lost energy, and produces no useful result. SELECTION OF TRUCKS. It is probable that, even with the cantilever extension truck, a 22-ft. closed car body (being about 30 ft. over all), is approximating the limit of a size of body which can be successfully carried on a 7 ft. wheel base. Fig. 294 shows the cantilever extension truck, which is the size used under a 22-ft. car body success- fully. The spring support to the body is generally differential, that is, an elliptical double or single takes the weight of the car body light. The best method sets the car body over an elastic support, which will not be too light at low loads, allowing the car body to oscillate or pitch, and not coming down too hard under heavy loads. The springs should be so arranged among themselves as not to repeat the move- ment over bad joints, but their interaction should tend to damp out any oscillat- ing effect. As the load comes on the car body, the elliptical springs are depressed. When they are compressed to a point where they lose their resilience the weight is taken up by spiral springs. The arrangement of springs under the truck to support the weight varies with different truck manufacturers. The points, however, which should be obtained in truck construction are : rigidity of frame to withstand the stress tending to throw the axles out of alignment on rounding curves; the power to resist the iongitudinal strains thrown on the truck frame by sudden changes in track contour, and reduction of uncushioned weight on the wheels. The truck must be so constructed as to allow easy access to motors, wheels and journal boxes. The attachment to the car body should be made so as to be readily remov- able. The method of constructing the truck frame can be either riveted bridge construction or solid side frame. CAR AXLES. Cold rolled steel and wrought iron are both used for axles, wrought iron having the preference. The diameter varies from $% ins. to 4 ins. ; 3J4 i ns - i 8 the most common diameter for the journals. The gear keyway is generally made for a key % in. wide and cut ^ in. deep in the axle. Axles are found to break where square corners are present for the fracture to start. All corners should, therefore, be turned with fillets. If the keyway is cut with a milling machine so that it has sloping sides, the axle will be less liable to break at this point than if the keyway is drilled at each end and slotted out. The size that the axle should be turned before forcing on the wheels can only be determined by experience and depends upon the density of the wheel hub and the axle. The length of the axle for standard gage varies with the dif- ferent trucks. The Taylor takes an axle 6 ft. 3 ins. long ; Peckham, 6 ft. 4% ins. and also 6 ft. 6% ins.; Brill, 6 ft. 5 ins.; McGuire, 6 ft. 5 ins. and 6 ft. 6 ins.; Bemis and Baltimore, 6 ft. 5*4 ins. j Diamond, 6 ft. % ins* J* 390 ELECTRIC RAIL WA Y HAND BOOK. ELECTRIC RAILWA Y HAND BOOK. 391 Y APPROXIMATE WEIGHT OF MOTOR TRUCKS. Gage, 4 ft. 8}4 in. MAKE OP TRUCK. Weight of Wheel. Lbs. Diameter of Wheel. Ins. Weight of Truck. (Bare) Lbs. Weight of truck equipped with G. E. 800 motors. 1 Motor. Lbs. 2 Motors. Lbs. Bemis, four-wheel 250 300 300 280 300 J300 1200 280 • (300 |200 300 300 30 30 30 30 30 30) 22) 30 30) 24 f 30 30 3,123 3,500 3,000 3,600 6,&40 5,400 6,400 5,000 4,000 5,000 5,300 4,800 5.400 4,900 4,500 5,000 6,800 6,200 6 800 Brill, four-wheel 7.100 McGuire, four-wheel 6,600 Tripp, four-wheel 6,900 Bemis, eight-wheel Brill, (maximum traction) . . . Tripp, eight-wheel Robinson, radial 8,600 Peckham, four-wheel " eight-wheel 8,000 STANDARD DIMENSIONS FOR BRIIX NO. 31-E. TRUCKS. (See Figs. 295, 296, 297.) Width Centres Width GAGE. over top of top over jour- Length of Wheel Total Spring plates. plates. nal boxes. axle. base. length. base. A B D t, E F G Ft. Ins. Ft. Ins. Ft. Ins. Ft. Ins. Ft. Ins. Ft. Ins. Ft. Ins. Ft. Ins. 4 5 ioy 8 5 SYs 6 10% 6 3% 6 14 3 13 2 4 8% 6 5 m 7 6 5 6 6 14 9 13 8 5 6 2% 5 11% 7 2% 6 7H 7 15 3 14 2 5 2y 2 6 5% 6 2% 7 5% 6 10)4 7 6 15 9 14 8 5 3 6 5% 6 2% 7 5% 6 1034 8 16 3 15 2 8 6 9 16 9 17 3 15 8 16 2 Height of Diameter truck. with of wheel. weight of body. M N O P R S T H I Ins. Ins. Ins. Ins. Ins. Ins. Ins. Ins. Ins. 30 25)4 16 18% 18% 10 15 28 K 33 26% 392 ELECTRIC RAILWAY HAND BOOK, INTERURBAN MOTOR TRUCKS. The necessity for heavier and stronger trucks has led to the general adoption of an all steel swing bolster, equalized truck almost identical in design with the Master Car Builders standard for steam roads. A partial exception to the above is the equalized truck shown in Fig. 297a. In order to secure side play, instead of hanging the bolster on the ordinary swing links, it is hung by the equalizer springs from the side frame of the truck, In this form of truck the blow from the car wheel is transmitted to the car body through the journal box springs, the equalizer springs and the bolster springs in series. Improvements in the design of the motors are constantly being brought out. Many of the later designs are especially arranged so as to be accessible from above i. e.t in repair work the trucks are removed from beneath the car and the work done on the floor instead of in the pit. Insulation has been improved some, as well as ventilation of the motor frame. In mechanical details also an advance has been made ; rarioua new methods of lubricating the armature and axle bearings have been devised. A decided tendency to abandon grease as a lubricant and substitute oil is evident. Several patented devices for feeding oil at the proper rate are in use. A compromise wheel design has been widely adopted in which the tread is 8 in. wide and the flange y± in. to % in. deep. This makes a reasonably safe wheel but when made of chilled cast iron, the pavement rapidly chips off the outer edge of the tread and the shallow grooves in city special work break off flanges. This wheel will not run with safety through the frogs and switches of a steam road on account of its narrow tread. Most of the city and interurban cars have 2}£ in. tread and y± in. flange. A 3 in. tread overhangs the head of the rail and breaks up the pavement. CAR AXL.ES, BEARINGS AND OILING METHODS. The diameter and weight of car axles for interurban service have been greatly increased to keep pace with the increasing weight and speed of the cars. The diameter of axles through motor bearings has been made Q% i n - * n the heaviest type of car, and this diameter is increased to 7}4 i n - or 8 in. at the gear fit, in order that the key way may not cut into and reduce the strength of the driving axle. The main journal bearing for this axle is 5 in. diameter by 9 in. long. The motor axle bearings are 6^ in. x 11^ in. The labrication of the axle bearings of the motors is now usually accomplished by wool waste saturated with ordinary journal oil in an oil well under the bearing which is part of the motor frame. The lower half of the axle lining is cut away considerably, allowing the waste to come in contact with the axle. This method is that in use on nearly all cars for lubricating the main journals Other methods for lubricating main journals are in use. One of these uses no waste in the journal box and substitutes two or nine light wheels supported in a frame which fits in the bottom of the box. The lower part of their periphery dips in oil held in the box and in rolling on the axle carries the oil to it. In the turning of new axle bearings or in trueing up old ones a good plan is to roll the surface after the finishing cut has been made. This is done by means of a steel roller held in a tool which fits the tool post of the lathe. This roller is forced against the bearing surface and rolls or burnishes it until it is smooth. This method is considered by some to be better than grinding and is in use by most steam roads. ELECTRIC RAILWAY HAND BOO A". 393 6 H ft 394 ELECTRIC RAILWAY HAND BOOK P H 2 PQ 6 LENGTH OF AXLE Lt 1— ( b CO b OS b WIDTH OVER JOURNAL BOXES Dt L O CO TH b OS 00 00 RADIUS OF RUB PLATES V 1—1 CO CO b b T-l b b CENTERS OF FRAME Bt ©J b lb 2r o ib b b OS b b b H ao CO b CO b 2r 00 b b CO b LENGTH OF AXLE Lt It b NOD OS lb CO b b St b o T-t b o T-l b WIDTH OVER JOURNAL BOXES Dt M H K O S5 lb 2r TH b NX o b CO b OS b §5 o 3? o o 02 M o ft \00 b Jo o b b \Q0 etfs lO effs RADIUS OF RUB PLATES V CO CO 00 b b T-l b b CENTERS OF FRAMB Bt 1 lb 2fe 00 ib OS ib b b b GAUGE CO CO Jo CO b 00 b b b CO b 5 ^ cc (M CO CO CC 1— < «c CO CC os . -t- 3 • fl . o : Tf. PQ • a ^ a >ya 4= d _ &5 P I : a. u z o. c W o £.0. w w 8 ^ § a X3 >> w oq be a I o 3& •§5 : o « |n OS'S o o ■PhD o S 52 «n £°i-2 9^ g^ -^ PQ-f? « d a "S .0 §lf 2 § 35g QQ O < OD Cj »-> ^ a> -111 7 Sl^ ELECTRIC RAILWA Y HAND ROOK 395 Under no circumstances should a file be used on a finished bearing. The old form of split axle gears for interurban cars has never given iatisf action owing to bolts working loose. Solid cast steel gears with tangential spokes are now coming into use. They are forced on over a key by hydraulic pressure as are wheels. Reamers are kept in stock so that when it is necessary to force new gears over the same axle to replace those worn out, the bore of the new gear is reamed to give clearance to go on with the same pressure as did the first one, which is forced on at between 25 and 30 tons pressure. The same method applies to wheels, one of which has to be renewed whenever a new gear is applied. Such gears seldom give trouble and run until worn out. The diameter of the axle is usually made larger under the gear than at the wheel fit, so that the kewway does not extend deep enough to weaken the axle. The latest type of solid gear has a flange on the side towards the wheel and the latter is so designed that the gear can be bolted to the wheel. This gear is pressed on over a key also and the above gives much additional strength and reduces the axle breakage which most usually occurs between the wheel and gear on gear side, due to the short radius of strain set up between the gear and wheel, w hich causes a crystalization. CAR WHEELS. The following table gives the composition of car wheels which showed a long life and stood thermal and blow tests. Graphite Combined carbon Silicon Manganese Sulphur Phosphorus ANALYSIS OP CAR WHEELS Which Stood Thermal Test for 60 Mins. Max. 3.28 .95 .75 .53 .088 .48 Min. 2.65 .32 .50 .20 .055 .35 Which Stood 40 or More Blows Drop Test. Max. 3.31 .90 .70 .46 .086 .52 Min. 2.65 .55 .50 .24 .040 Which Gave 5 or More Years of Service. Max. 3.18 1.24 .94 .34 .085 .49 Min. 2.23 .56 .58 .13 .047 .25 It will be seen that these limits are rather wide, but below are given what are considered to be the desirable limits for the chemical constituents of wheels: Desirable Wheel Analysis. Graphite 2.75 per cent to 3.00 per cent. Combined carbon 50 " " .75 " Silicon 50 " " .70 " Manganese 30 '« " .50 " Sulphur.... 05 M " .07 " Phosphorus.... 35 " " .45 " The proper amount of manganese is an important element, for upon it de- pends the capability of the wheel to stand the preliminary test and take a good deep chill. _ J 39^ ELECTRIC RAILWAY HAND BOOK, There are a great variety of methods and variation of mixtures used by the different wheel manufacturers, on which they base their mileage guarantees, but the following are the elements of general specifications for car wheels. They must all be cast in true metallic chills of the same internal diameter and uniform cross-section. The body of the wheel to be of clean, soft, grey iron smooth and free from blow holes. The hubs to be solid and free from drawing. The tread and throat of the wheel must be smooth and free from deep and irreg- ular wrinkles, slag or sand wash, and practically free from chill cracks and sweat. The depth of clean white iron should not exceed % in. at throat and 1 in. at middle of tread, nor be less than % in. at the throat or ^ in. at middle of tread; nor should there be more than J4 in. in variation of the depth of chill throughout the same wheel. The blending of the grey with the white iron must be without distinct line of demarcation. See Figs. 298-300. In each wheel, when a true metallic ring is placed so as to bear on the cone no part of its circumference will stand more than T X B in. from the tread of the wheel. No wheel made in a solid chill will be passed whose circumference differs from 1% in. or less than % in. from the circumference of the chill in which it is made. Wheels cast in contracting chills should not differ more than 2 ins. from the circumference of the chill. All wheels during inspection must stand three heavy blows of a 6-lb. sledge under the flange and between the brackets, and must withstand a pressure of 50 tons when being forced on the axle. (mm Fig. 298.— chill for car wheel and method of testing. With each pouring of 100 wheels two additional ones must be furnished for the following tests. One wheel is placed, flange downward, on an anvil block weighing not less than 1700 lbs., set on 2 ft. of rubble masonry and having three supports for the wheel to rest on, not less than 5 ins. wide. The wheel is then struck centrally upon the hub with a weight of 140 lbs. falling from a height of 12 ft. The wheel should stand fifteen blows without breaking. If it breaks in only two places and the depth of chill is uniform, the wheels may be accepted providing they stand boring and mounting with 50 tons pressure. The thermal test is carried out as follows : The test wheel is laid, flange down, in the sand, and a channel way 1^ ins. wide and 4 ins. deep, must be molded with green sand around the wheel, the clean tread forming one side of the channel way. This is then filled with molten cast iron, which must be hot . enough when poured to form a ring when the metal is cold that shall bo solid and free from wrinkles or layers. ELECTRIC RAILWAY HAND BOOK, 397 The weight of car wheels has gradually increased from 250-260 lbs. to 325-350 lbs. for 30-in. wheel, and to 370-380 lbs. for 33-in. wheel. For interurban high- speed service the 38-in., 400-lb. wheel is now coming into favor. Sections of car wheels are shown in Figs. 299-300. The mounting of the wheels on the axle is done as follows: The a^les are pressed in y± in. less than the gage line between the center of the fillet between SO* oiamctck SECTION AT A -B 7 A*MS Fig. 299.— section op new york car wheel, 30 inches. the flange and the tread, where the road is not in good alignment and where 60-lb. rails are used. This is to allow of lateral play and avoid cramping the flanges and wearing them unduly. The surface of the flange presented to the special work at frogs and switches is becoming more of a flat f urface than form- erly to avoid wearing and cutting these parts of the track. With a grooved rail the flange end should present as much of a cutting surface as possible, in order to Fig. 800.— section op new tork car wheel, 33 inches clear the groove of dirt and not pack in at the bottom of the groove and increase the power necessary for operating the car. The ou'ftid^ flange should have a slight slope to prevent cutting into guard rails, and the tread should not over- hang the head of the rail so as to come in contact with paving blocks or similar obstructions. 398 ELECTRIC RAILWAY HAND BOOK. Flat Wheels.— Mat wheels are primarily caused by sliding and grinding a flat on the wheel. There are a number of causes assigned for this trouble. One is that the wheels which become flat were not perfectly true with respect to the axle so that as the brake shoe was drawn up to the wheel it locked the wheel when the largest diameter rolled against the shoe and tended to stop the wheel always at one point, thus focusing the wear at one place and producing flats. In new and old wheels, where the chill first wears through, a soft metal will be pre- sented to the attrition between the wheel and rail in braking the car. The retarda- tion is caused by the difference between the length over which the car passes, and the distance through which the wheel rolls. The maximum retardation is approximately when this difference is 22 per cent., and falls when the difference passes this point until the car slides. The thermal test on car wheels is an important one, for the foot tons in the moving equipment appear as energy dissipated in the rim of the wheel, at the brake and under the brake shoe. AIR BRAKES. Heavy motor cars are now generally equipped with some form of power brakes, the power usually being furnished by compressed air. There are two methods in use for supplying compressed air. One in which each car carries a small air com- pressor driven by an electric motor, which automatically maintains the air pressure necessary to operate the brake, in a small reservoir. A governor actuated by the reservoir pressure cuts the motor driving the air compressor in and out of the cir- cuit, as the pressure rises and falls within predetermined limits. In the other method, air compressor stations are established at convenient points, where large electrically driven compressors supply the storage reservoirs carried by the cars with air at a pressure af about 300 lbs. per square inch. This supply is sufficient to last the car some hours, dependent upon the service, and is supplied from the storage tank to the regular reservoir, at the proper pressure, through a reducing valve. The remainder of the brake equipment is the same as the first method employs. The two systems in general use are the so called straight air and automatic air. The former is used on individual motor cars, and sometimes on short trains. The automatic air brake is the same as that used on steam trains, in which the brake is automatically applied to both sections of a train, if the latter should part. The essential difference between the two systems is that, in the former the train pipe is empty when the brake is not in use, and air is allowed to pass through it from the main reservoir to the brake cylinder when the brake is applied; while in the latter the train pipe is always under pressure, and maintains the proper air pressure in the auxiliary reservoirs under each car in the train. Any reduction in this train line pressure by opening it to the atmosphere, whether caused by the motorman using his brake valve, by the use of emergency valves in any of the cars, or by the bursting of a hose between cars, applies the brake on the whole train; each auxiliary reservoir supplying air to the brake cylinder on its car through the medium of the triple valve, which is operated automatically by this reduction in train line pressure. To release the brake this pressure must be restored by the motorman, who admits air to the train pipe from the main reservoir. This operates the triple valves which allow the air in the brake cylinder to escape, and the auxil- iary reservoirs to recharge to their proper pressure. To release the straight air brake, the motorman open the train line to the at- mosphere and the air in the brake cylinders passes out through it. ELECTRIC RAILWAY HAND BOOK. 399 vv 400 ELECTRIC RAILWAY HAND BOOK cf£&<^t ELECTRIC RAILWAY HAND BOOK 401 1 402 ELECTRIC RAIL WA Y HAND BOOK. Wx b a Wxb or „ Wx I F F + W b — Fxa or b _ Fxl ~ W F+W |f- fXO, b F -Wxb a a ^Wxb ov a _ Wx d F W F b Fx a or b Fxd ■ w ~W F Fxg b F Wxb ~ a Wxb or a== Wx d F-W h _ Fxa or /, _ Fx d -~W~ F-W Fig. 300-d [ELECTRIC RAILWA Y HAND BOOK. 403 BRAKE LEVERAGE. The maximum braking power applied to a car at the brake shoes should be per cent of the light weight of a four motor car, 90 per cent for a trailer car, and 70 per cent for a freight car. In calculating the dimensions of the various levels, lengths should always be in inches taken from center to center of pins, and pressures in pounds. The usual sizes of brake cylinders to give 100 per cent braking power on differ- ent cars is as follows : Cylinder Diameter. 6 Inches 8 " 10 " 12 " 14 " Light Weight of Car. Up to 20,000 lbs. 20,000 to 36,000 " 36,000 to 52,000 " 52,000 to 72,000 " Above 72,000 " Fig. 300b shows an example graphically worked out. A four motor car weigh- ing 40,000 lbs. is taken. It is desired to secure a braking power of 100 per cent or 40,000 lbs. pressure on the shoes, which is equal to 10,000 lbs. per brake beam or 5000 lbs. per shoe. to fiftkc Staff 4- ? J 3 i - -*a*I iE»tbooo* ft»«06* It is desired to secure a pull of 10,000 lbs. at the middle hole of the truck dead lever, which corresponds to the brake beam. The live lever then must also exert a pressure of 10,000 lbs., at its middle hole, which it will do, for the two levers must always be in the same proportion, but not necessarily the same dimensions, in order to equalize the pressure on the truck. In order to get 10,000 lbs. at the middle pin of the live lever the pull at the upper pin, from the formula, would be F = where B, the distance from the a bottom pin or fulcrum, to the middle pin, is 5 ins. and a the distance from the ful- crum to the top pin, or the entire length of the lever, which is 20 ins. 10000X5 Therefore F = 20 = 2500 lbs. 404 ELECTRIC RAILWAY HAND BOOK, As the two trucks are usually similar, it is evident that to secure the required braking pressure there must be exerted a pull of 2500 lbs. on each live lever by the brake rods. It now remains to exert this pull of 2500 lbs. at the end of each cylin- der lever to which the brake rods are attached. Assuming an 8 in. diameter brake cylinder, and air pressure of about 50 lbs. per square inch will exert a force of about 3000 lbs. on the push rod. This force is exerted on the end of the floating cylinder lever whose fulcrum is the tie rod pin. To find the position of this pin it is first assumed that the total length of the cylinder lever is 27.5 ins. on account of the available space; then from the formula 2500 X 27.5 'i . "= 3000 + 2500 = 12 - 5in8 - b = 3000X27.5 = 16in8> 3000 -f 2500 The sum of a and b must of course equal the total length of 27.5 ins. The cylinder dead lever must be divided in exactly the same proportion as this lever, though it may differ in length. The force of the cylinder push rod is then trans- mitted through the tie rod to the cylinder dead lever, which gives the same pull to its brake rod, 2500 lbs. The tension on the tie rod is the sum of the forces on the ends of either cylinder lever. ELECTRIC BRAKES. The Westinghouse electric brake (Fig. 300-p) consists of a track shoe which is magnetized by a winding energized by current produced by the car motors acting as generators, and is powerfully attracted to the rail by its magnetism. The downward pull, and horizontal drag of this shoe resulting from its friction on the track, is transmitted through suitable rods and levers, to the ordinary brake shoes acting on the wheels in the usual manner. The essential points in the rest of the equipment are merely that the car con- trollers be provided with the necessary braking notches to properly connect the motors to act as generators, and conduct their current to the track shoe magnets ; also that a suitable resistance be provided to dissipate the energy generated over and above that required by the brake. The Price-Darling brake (Fig. 300-g) consists essentially of two brake controll- ers mounted on the platforms, an electromagnetic brake cylinder, an automatic controller and a transfer switch under the car, The brake controller contains the necessary mechanism for cutting off the trolley current from the car motors, and applying it to the brake cylinders, and for converting the motors into generators. The brake cylinder consists of a solenoid, the core of which is movable, and when energized, first by the trolley current, and afterwards by that generated by the car motors, exerts the necessary pull on an equalized system of brake levers which transfer it to the shoes. The automatic controller regulates the current supplied to the brake cylinder by the car motors, automatically graduating the braking pressure as the speed of the car is reduced, thus preventing skidding of wheels. The transfer switch automatically cuts off the trolley current from the brake cylinder when sufficient current is generated by the motors. After applying the brake a locking device on the brake cylinder holds the brakes on without using any power, as the car approaches a stop, until released by the motorman. This renders it possible to hold the car on a grade without using the hand brake. ELECTRIC RAILWAY HAND BOOK. 405 Fig. 300-f Fig. 300-g w 406 ELECTRIC RAILWAY HAND BOOK. BRAKES. The leverage in hand brakes varies between 40:1 and 72:1, depending upon the weight on wheels, grades and conditions of track. The amount of power a motorman can exert on a brake wheel is given in the following table. POWER OBTAINED BY DIFFERENT MOTORMEN ON BRAKE WHEEL,. Weight of Gradual pull with Jerk with both Emergency jerk motorman. one hand. hands on hand with both hands wheel. on hand wheel. Lbs. Lbs. Lbs. Lbs. 140 112 135 275 200 135 275 385 287 145 235 312 175 125 212 285 153 125 245 310 185 150 200 300 170 150 275 350 155 135 210 325 135 110 175 325 135 125 250 350 160 125 250 405 176 100 200 400 185 276 250 375 Av. 131.7 Av. 224 Av. 338.23 The pressure on the brake shoe should not exceed the pressure between the wheel and rail. The effect of speed, brake pressure and traction coefficient is given in the following table for hand brakes. TABLE SHOWING RELATIONS OF SPEED, BRAKE PRESSURE AND TRACTION COEFFICIENT. Speed revolutions Brake pressure. Traction Coefficient. per minute 33-in. wheel. Lbs. Lbs. Per Cent. Varying, 150 900 87.4 9.7 125 900 91.7 10.2 100 900 99.8 11.1 78 900 118. 13.2 56 900 133. 14.8 38 900 150.4 16.6 20 900 154. 17.1 4 900 174.6 19.4 Constant, 105 300 29.4 9.8 500 50.5 10.1 750 91. 12. 100 1,150 125. 11.2 1.500 178. 12. 2,200 305. 14.4 94 8,780 488. 13.2 The brake rigging takes a number of forms, the inital effort being given by the motorman through a brake handle on the wheel, which wraps a chain around the brake staff. In some cases the brake staff has a pinion which engages in a s. ELECTRIC RAILWAY HAND BOOK 4o; 408 ELECTRIC RAILWAY HAND BOOK. RECORD OF CAR Materials. Weight of Cars. Mileage of Wheels. Mileage of Shoes. Stops per mile. Gradients 8 9 10 11 12 Xo. Records at all 12 to 22000 Abt. 225000 Abt. 75000 As hign as 10% & plenty of them 6% heaviest 13 14 About 12000 15000 4 to 6% tons 12000 Cabc 7500 E. 14500 6 to 8000 4 to 5 tons About 6)4 tons 25000 to 30000 35000 20000 to 60000 55244 40000 Abt. 40000 15 16 17 18 19 20 21 3000 4to7000 4864 7000 Abt. 12000 Not over 7% 214 to 5^% 3 to 6% Level to 12% Highest 2% Abt. 10 per m. 5 per m. 8 perm. 30000 to 32000 22 Very f re- One of 10% 8 to 9% in placet 6% About level 0to9% 0to6^% Abt. 2^% 23 24 25 26 7 7 to 8 tons 12 to 15000 5 tons 4 to 6000 Cab. 7000 E. 16000 11000 20000 6^ tons 10000 Mo. 16000 Trail 6000 8 to 9 tons 8 tons One year 30000 to 40000 Not worn out but flat in 5 One year 33000 27000 45389 30000 20000 to 24000 Abt. 36000 9000 5 to 7000 6000 2 to 8000 Very f re- Not fre- quent 5 to mile Every 300' As usual in cities 28 29 5% heaviest 30 4500 10 to 14000 1500 81 82 88 Every 500' 20 per m. Not over 2% 4 to 11% 3 to 7% 84 35 4000 to 30000 35000 Comp. 20000 C. 1.8 to 10000 45000 Usual in cities 7 torn. Max. 3% 7% ELECTRIC RAIL WA Y HAND BOOK. 409 WHEELS AND BRAKE SHOES. No. of Truck Patterns. Shoe Patterns. No. Hangers. Shoe Patterns Separate Hangers. Stand Shoe Wanted Stand Shoe Hanger Wanted. 4 Brill 1 Yes Present 1 practice 1 5 or 6 Yes 3 6 Not possible Yes Yes 3 2 2 3 1 Not possible Yes 1 Yes 2 Brill Yes 2 2 2 1 2 Yes 4 2 3 3 Yes 6 Yes 4 Most em- 4 2 2 Loop bolted to brake bar 1 1 1 phatically Yes 2 each Durable but railway % • hard to keep good fit Yes 4 Yes 1 Yes 1 2 Yes 1 All 1 2 Yes 2 1 Yes 3 Yes Yes Remarks. Shoes wear down to *4 in. thickness or less before giving out. Shoe interior and fit stand- ard hanger. Shoes from Bemis Co. only. Use shoe as made by truck manufacturers. 60% chilled iron, 20* soft Lappin. Chilled iron in shoes. Hard iron shoe to brake on tread only. G0$ soft I; C. I. with wood, also with steel plugs. Medium C. I. Congdon shoe (cast steel plugs in C. I.) Soft iron and wood, ill-fit- ting hangers. Ordinary C. I. shoe. Chilled 1. shoes, 2 patterns. Soft I, with wood plugs. Soft iron shoe. Have used soft I. & hard I. and iron and wood plugs. Have used soft I. & hard I. with wrot. plugs & wood. Same as 25 above. Prefers hard I. Thinks soft I. wears wheels faster titan hard I. Impossible for one shoe to suit all Ky. men. Congdon shoes. C. I. with wrought I. plugs. McGuire type shoe, chilled I. wheels. Hard C. I., 4 steel segments 3 ins. apart. Soft C. I. and same with wrought I. plugs. Soft C. I. and comp. shoe. Soft C. I.; hard C. L, C. I. with wood plugs. Wood. 4io ELECTRIC RAILWAY HAND BOOK. ELECTRIC RAILWAY HAND BOOK, 4ii gear, to whiob is attached a sprocket through which the chain is wound ; this pulls the brake rod attached to the end of a brake lever which is connected to the brake beam by which the shoes are forced against the wheel. There are a number of adjustments for the stretching of the brake rod and the wear of shoes. The brake rigging has to be so aligned that there will be no cramping of the brakes when the equipment passes around curves. In the Price Momentum Brake, instead of the brake staff directly transmit ting the power necessary to draw the brake shoe against the wheel, the brake staff is connected to a clutch. This clutch actuates a drum which winds the brake chain around the car axle and pulls the brake shoe against the wheels. Fig. 301 shows the general construction of this arrangement, the end of the brake chain being Fig. 304.— ^standard air brake for double truck cars. attached to the drum sleeve on one of the axles. This drum is not keyed to the axle and does not turn with it except when a stop is to be made. By a series of levers the edge of the drum, which is in the form of a disc, is then pressed against a corresponding disc on the inside of the car wheel. Between the two discs is a leather washer to take up the wear. The friction caused by pressing the drum against the car wheel causes the former to revolve, winding up the chain and setting the brake. Auxiliary power brakes may be actuated by compressed air, or the current generated by the motors. In air brakes the compressed air may be stored or pro- duced by an axle drum or motor driven compressor. The Magann storage system is shown in Fig. 302. The compressed air is produced by one or m^re steam or electric air compressor plants located at the power stations or car hoases, where there are large storage tanks aDd drip tanks connected therewith to eliminate moisture. The pressure carried is generally 300 lbs. per square inch. From the storage tank the supply pipe is taken to some locality convenient to charge the reservoirs carried on the cars, which have an aggregate capacity of 20 to J85 cu. ft. The air is first reduced to a pressure of 20 to 50 lbs. per square inch, 412 ELECTRIC RAIL WA V HAND BOOK. by a reducing valve according to tonnage and operating conditions, before it reaches the auxiliary reservoir, from which the brake cylinder is supplied through the controlling valve under the hands of the motorman. From 400 to 600 stops can be made without recharging. The axle driven compressor has the compressor pump geared to one of the car axles, and an automatically controlled valve, by which it keeps the reservoirs charged. The electrically driven compressor has an independent motor with an automatic switch, actuated by the initial air pressure so as to throw the motor in and out as the pressure rises and falls between fixed limits. The general arrange- FlG. 305.— METHOD OF APPLYING STANDARD AIR BRAKES TO MAXIMUM TRACTION TRUCKS. ments on the car of the Christensen electrically-driven compressor and brake are shown in Fig. 303. Fig. 304 shows the Standard Company's method of applying the air cylinder to double truck cars. Fig. 305 shows methods of applying the air cylinder to maximum traction trucks. THE MOTOR EQUIPMENT. As long as the insulation is maintained the current through the motor follows the proper paths and the motor can be operated. Temperature, oil, moisture, as well as time, all tend to depreciate the insulations on these conductors. A prac- tical limit to heating is the ability of the various materials used for insulation to endure the high temperature without perishing or losing their insulating quali- ties, and in order to obtain a long life from a motor its temperature should not rise 40 degs. Cent, or 70 degs. Fahr. above the air. This brings approximately the ELECTRIC RAILWAY HAND BOOK. 413 ; ultimate temperature of the motor to 62 deg. Cent, or 143 degs. Fahr. in summer, and to 54 degs. Cent, or 129 degs. Fahr. in the winter, under working conditions. A motor raised above these temperatures will gradually carbonize the insulating material between the coils on the armature and the body of the armature, as well as between the field coils and their cores. The cotton insulation and covering on the windings will become charred, and the stress to which these windings are submitted on opening the motor circuit by the controller, will tend eventually to pierce them and break down the insul- ation. "With the exception of the mechanical wear on the commutator and the bearings, the whole problem of motor repairs is one of successful insulation. CARE AND REPAIR OF MOTORS. The Fields.— The fields of all railway motors are wound with double cotton covered wire. There is a new wire insulated first with asbestos and then cotton over this; the advantage of this double insulation is that if the cotton becomes charred the asbestos will still offer sufficient resistance to prevent adjacent layers of wire from short circuiting out the turns around the field. The field windings for each of the railway motors generally used will be found under data for eaeh motor. The general precautions to be used in winding are common to all of them. Some motors require the field to be wound on forms, and other methods of design have a field spool on which the wire is wound directly. In both cases it is advisable to varnish each layer of wire. Shellac has been advised for this purpose, but it is nearly impossible to dry out a coil thoroughly which is filled with shellac, and the oxidizing of the alcohol tends to carbonize the cellulose in a cotton winding, and in this way it neutralizes the good effect that might result from the lacquer. Never use wood alcohol for this purpose, it is deliquesent, whereas grain alcohol will maintain the insulation resistance. In some cases it is advisable to wind the field coils dry and dip into an insulating varnish and hang up to drain and dry out. If the field coil is warmed before it is dipped into the varnish, the varnish will soak completely in and fill all interstices with a good insulator which will prevent the entrance of moisture into the coil. After this treatment the coil is insulated and different manufacturers advise different methods, but to cover the coil first with mica or micanite and overlap this medium so that there are no seams left, and then tape ! this over with two layers of adhesive tape, and over this cover with canvas, and finally paint with some good air-drying asphaltic insulating varnish, is one of the approved methods of insulation. This treatment requires that the field coils be I baked before they are used, so as to thoroughly dry out all the solvents used in 1 the insulation. Where the field coil is wound directly upon a core it should be thoroughly j insulated with mica, canvas and duck, and the edges of this insulation should project beyond the field coil so that they can be lapped over after the field coil is 1 completed, and over the overlapped insulation should be wound adhesive tape in the same direction as the winding of the wire. This should then be painted over I with asphaltic varnish. The field terminals or lugs which project from the surface of the core should be specially insulated, since this is the point where moisture enters the field coil. The tape should be brought up close to these lugs and the I wire leading from them should be well taped before the insulation is put on the field coil. This point should be well treated with varnish and paint. In a few types of motors the field coils are further protected by being encased in sheet copper. . ._ - 414 ELECTRIC RAIL W A Y HAND BOOK A well insulated field coil should stand the following test: It should lie in an inch of water for ten hours without its insulation falling below 400,000 ohms. If placed on the pole piece, a difference of 2,000 volts between the field winding and the pole piece, or the field winding and its spool, should not break down this insulation. The field spool should be maintained at 140 degs. Fahr. for 10 hours without the insulation resistance of the coil falling when again cooled. It is well to apply these tests, when using any new material for insula- tion, on several field coils before it is adopted, as there are many proposed insu- lating compounds which are seriously impaired in their usefulness when main- tained for a length of time at an elevated temperature. This condition of temperature arises in practice when the motor is subjected to an overload or an improperly handled controller. The Armature. — The break-downs of a street railway armature are caused by heating, flashing at the brushes, and rubbing of the armature body on the field. Oil and water form the principle external troubles, and crosses, grounds, open coils, short circuits and grounding of commutator are the principal internal troubles. As the heating of an armature tends to carbonize and destroy the vitality of the insulation, as soon as this insulation has fallen, a treatment should be given the armature which will again restore the insulating qualities, and the following methods can be used in order to accomplish this result. The armature should be first put in a bake oven, the temperature not being over 120 degs. Fahr. Passing a current through the armature in order to dry it out does not give good results in practice. If there are any leaks or moisture in the armature, it tends to set up internal electrolysis, which again impairs the vitality of the insulation. After the armature has dried in the oven for two or more hours, and while still warm, paint with thin air-drying asphalt compound, P. & B. paint. M. I. C. paint, Sterling varnish or shellac. Trials have to be made on these different compounds to find out which gives the best results with the special insulation used for insu- lating the armature body and coils. Where there is much paper or cellulose insu- lation, an asphaltic varnish generally gives the best results; with mica and cloth, varnish or shellac gives the best results. Never use anything but alcohol to mix the shellac; woodnapthaor wood alcohol gives very poor results. When using shellac do not have the armature above 90 degs. Fahr. and treat several times. After the final treatment bake the armature for 4 hours at 150 degs. Fahr., bring- ing the temperature gradually up to this point. The armature should be turned several times during this baking, in order to allow any free solvent to run and dry out. The heads or any covering over the winding should be taken off for this treatment. Of course splines should not be removed. The above treatment is useless unless the armature is first heated, since, if cold, the treatment will only be superficial, whereas if the armature is hot and then allowed to cool, the insulating compounds will be drawn in and impregnate the whole insulation. A carbonized insulator if impregnated with insulating com- pounds will break up as a partial conductor. Flashing at the brushes, due to short, broken or weak brush springs, or short- circuited rheostats, will cause sufficient arcing to burn through the heading of the armature. These heads as a rule do not afford sufficient protection from this source of trouble. A layer of asbestos paper between two layers of canvass on the head will greatly strengthen this weak point. This flashing across from the brush to the winding, causes a bucking of the motor, which, if not cut out immediately, means considerable damage to the armature. Fields are often broken down from this same cause, and a rigorous brush inspection results in reduced motor repairs where the breakdowns are caused by flashing. ELECTRIC RAILWAY HAND BOOK. 415 The striking of the armature against the field in a toothed armature bends the teeth over and they pierce the insulation on the armature coil. Several types of motors have been designed in which the armature bearing has been too small and improperly proportioned, and it is only by very close inspection that the arm- ature can be kept in proper alignment. The position of the armature should be midway between pole pieces, for if there is any looseness in the bearing, the stronger magnet will pull it towards that pole piece and the direction of the movement of the car will tend to throw it against the rotation of the pinion. The internal troubles of an armature are generally evident upon external examination, after they have crippled the armature. Flashing at the commutator may be caused by carbon dust on the commutator bead ring or carbonized oil on the commutator surface. There is also a commutator bead ring used which is carbonized \vhen an arc passes over its surface; it has a low resistance and forms a partial conductor. This material should not be used directly in contact or slipped directly over the commutator bars. Several strips of mica should be bound around the commutator bars, and the ring then slipped over these. Where an armature tests low to ground, this bead ring or commutator ring should be first wiped clean of all carbon dust and oil, and the ground, if caused by leaking over this surface, will sometimes be removed in this way. In cases of emergency where an equipment has to run and still there is a ground on the armature which cannot be removed, and the coils are burnt out, a motor can be made to operate (where there are not more than three coils affected), by cutting these coils out entirely and plugging these commutator bars together. This is best done by soldering the ears together or putting in a jumper; sometimes a hole is drilled between the bars and a brass plug driven in. It should always be remembered that where one end of a coil is cut out from a commutator bar, the other end of the coil should also be cut where it connects to the symmetrically located commutator bar. The distance apart of these two commutator bars will depend upon the number of armature coils and the method of winding the arma- ture. The armature should not be allowed to run in this way any longer than is absolutely necessary, since this armature will take more than its share of the load, if worked in parallel with another motor on the equipment, and will lead to exces- sive heating which will surely destroy the insulation after a short service. Machine wound coils are largely used, and they are furnished already insu- lated with two layers of cotton on the wire, and taped with a double layer of oil silk and from 1% to 2 layers of insulating tape, so that their insulation is amply reinforced at the weak points in windings. The weakest point in an armature body is where the turns leave the slot; if there is any movement at all in the coil, it will wear through or abrade at this point and break down. A combination of mica and fibre paper formed into a strip at least as wide as the band or heading, should be forced in at this point and project outside of the slot at least % in; the coils should be driven down into the slot and protected from the iron tooth at these points by this strip of insulation. Complete troughs of micanite are some- times used with good results. Paraffin can be used where these coils slip in hard. After the first layer of winding is on the armature, the insulation between the first and second layers on the commutator end should be several thicknesses of canvas and mica. DATA ON WESTINGHOUSE MOTORS. Westinghouse Motor No. 3.— Speed at full load, 300-350 rev. Rated horse-power, 20-25-30. Reduction ratio, 3.45. Gear, 18 teeth on pinion, 62 teeth on gear; 4 poles; 4 field coils; 732.4 turns, total of 4 coils. Size wire 150x150 416 ELECTRIC RAILWAY HAND BOOK mils square. Armature has 95 slots, 8 conductors per slot; number of bands, 22. Commutator bars, 95. Armature bearings : commutator bearing, 1% ins. x 4^ ins.; pinion bearing, 2% ins, x 5^ ins. Weight, 471 lbs. Diameter, ll^j ins. ; length, 13% ins. See Figs. 306 to 308. Westinghouse Motor No. 12A.— Speed at full load, 505 to 700 rev. Rated horse-power, 25 to 30. Reduction ratio, 4.86. Gear, 14 teeth on pinion, 68 teeth on gear; 4 poles, 4 field coils; 636.4 turns, total of 4 coils. Armature has 47 slots, 20 conductors per slot; number of bands, 14. Commutator bars, 93. Figs. 306, 307, 308.— methods of latino on coils on westinghouse no. 3 armature. Armature bearing, 2^» ins. x 6 ins. Weight, 360 lbs. Diameter, 11^ ins. ; length, 7J£ ins. See Figs. 309 to 311 for armature windings. For dimensions of this motor see Figs. 312, 313. Westinghouse Motor No. 38B.— Speed at full load, 500-525 rev. Rated horse-power, 50. Reduction ratio, 4.86. Gear, 14-24 teeth on pinion, 68-58 teeth on gear; 4 poles; 4 field coils; 380 turns, total of 4 coils. Armature has 45. slots, 12 conductors per slot; number of bands, 8. Commutator bars, 135. Armature bearing, 2% ins. x 6 ins. Weight, 525 lbs. Diameter, 13%ins; length, Pigs. 309, 310, 311.— armature windings of westinghouse no. 12a motor. 8 ins. Armature coils are slung 11 slots. To connect draw a line through arma- ture core slot over bottom end of coil and note point on commutator. Including this bar count 21 bars to the right and call that bar No. 1. Bottom lead goes in No. 1. Top leads goes 69 bars to the left facing commutator from No. 1 counting No. 1 the first bar. Bottom leads are connected as winding progresses. ELECTRIC RAILWAY HAND BOOK. 417 W~7^ : ^^gtr _^^£_._1 ^, \ > £fiO VtEW -**' Hose suspension* Eno View firra//e/ tor suspense*} Figs. 312.— dimensions op westinghouse no. 12a motob. A 4iS ELECTRIC RAILWAY HAND BOOK. Para/tef bar suspension. foff Host Jt/S/WS/OV Cress chenrtet bar to support 2500 /Os at each postmarked "S." fcRpAMuii Bar Sv$P£N$ioii Cross chattrtef tar to support /SOO /4$\ &/ sac 6 point marked *C * Fig. 313.— dimensions op westinghousb no. 12a motor. ELECTRIC RAILWAY HA YD BOOK. 419 i /free* 6 AT 4/#fr#M&m S 4-6f 4*g' *h- 7* 4-Mf J*g" 4h '& J'-*" szA'' 4f in S-2t~ Sig" ^l' mX S-4f S6n 7h /// Ssr' s /■ 4? ji'Ji' 3i %-*W * ff #;,*" /V/V UL jk Figs. 314, 315, 316.— dimensions op westinghouse no. 49 motor. 420 ELECTRIC RAILWAY HAND BOOJC: ^^^ to suit truck" End View Credit Suspension Dimensions of Hey I'x r*$~long. for J fto S'shofU,, Notes All parts shown in 'dot and dosh( )*rt to be furnished by -true* builders. Tho motor is adopted to receive, on o/ele of ony d/emoter from J§' to S'i tho sizes most frequently used ore d'ond 4f ond th e re f ore geors ond or/e heonnos for these dimeters oro cbh»: side red as stoe* J/mo$ Pigs. 817, 318, 319.— dimensions or westinghousb no. 56 motor. ELECTRIC RAILWAY HAND BOOK. 421 Westinghouse Motor No, 49. — Reduction ratio 4.86. Gear, 14 teeth on pinion, C8 teeth on gear; 4 poles; 4 field coils. Armature bearings : commutator bearing, 2% ins. x6 ins.; pinion bearing, 2% ins. x7% ins. Weight, 438 lbs. Diameter 13% ins. ; length 6^ ins. See Figs. 314 to 316 for sizes of this motor. Westinghouse Motor No. 56.— Reduction ratio, 4.86 to 3.56. Gear, 14-18 teeth on pinion, 68-64 teeth on gear; 4 poles; 4 field coils. Armature bearings: commutator bearing, 3 ins. x 7^ ins. ; pinion bearing, 3*4 ins. x %% ins. Weight 720 lbs. Diameter, 14 ins.; length, 12 ins. See Figs. 317 to 319 for sizes of this motor. DATA ON NEW WESTINGHOUSE RAILWAY MOTORS. Westinghouse No. 63.— Rated horse-power 40; gear ratio 4.85; pinion has 14 teeth; gear has 68. Four poles; 4 field coils. Armature diameter 14 in ; 55 slots; 110 coils; 109 commutator segments; armature bearings— commutator end 2% in. x 6M in., pinion end, 3 in. x 7% in. ; maximum diameter of axle, 4% in ; weight of armature 505 ibs. ; weight of motor complete, 1,950 lbs. Westinghouse No. 69.— Rated horse-power 30; gear ratio 4.85; pinion 14 teeth; gear 68. Four poles; 4 field coils; armature diameter 13 in.; 35 slots; 105 coils; 105 commutator segments; armature bearings— commutator end 2^ in. x6 in., pinion end 2% in. x 7 in. ; maximum diameter of axle, 4^£ in. ; weight of arma- ture 385 lbs., weight of motor complete 1,620 lbs. Westinghouse No. 76.— Rated horse-power 75; gear ratio 2.41; pinion 24 teeth: gear 58. Four poles; 4 field coils; armature diameter 16^ in.; 39 slots; 117 coils; 117 commutator segments, armature bearings— commutator end 3X*n. x 8 in., pinion end 3\& in. x 9 in.; maximum diameter, axle 6 in.; weight of armature 850 lbs ; weight of motor complete, 3,335 lbs. Westinghouse No. 85.— Rated horse-power 75; gear ratio 2.36; pinion 22 teeth; gear 52. Four poles; 4 field coils; armature diameter 15% in. ; commutator has 117 segments; armature bearings— commutator end, 3% in. x 7% in., pinion end 3% in. x 9 in. ; maximum diameter of axle 6^ in. ; weight of armature 995 lbs. ; weight of motor complete, 4,000 lbs. Westinghouse No. 89.— Rated horse-power 50; gear ratio 2.30; pinion 26 teeth; gear 60. Four poles; 4 field coils; diameter of amature 16^ in.; com- mutor has 135 segments; armature bearings— commutator end, 3 in. x 6^ in., pinion end 3% in. x 7]4 in. ; maximum diameter axle 5 in. ; weight of armature 650 lbs.; weight of motor complete, 2,560 lbs. This motor can be used on a minimum gauge of 3 ft. 6 in. Westinghouse No. 101.— Rated horse-power 40; gear ratio 2.81; pinion 22 teeth; pear 62. Four poles; 4 field coils; diameter of armature 14 in. ; commutator has 111 segments; armature bearings— commutator end 3J4 in. x r T s in., pinion end 3*4 in. x 8% in.; maximum diameter of axle 5 in.; weight of armature 585 lbs.; weight of motor complete, 2,400 lbs. Westinghouse No. 93.— Rated horse-power 35; gear ratio 3.66; pinion 18 teeth; gear 66. Four poles; 4 field coils ; diameter of armature 13 in. commutator has 123 segments ; armature bearings— commutator end 3 in x 6^ in., pinion end 3 in. x 7% in.; maximum diameter of axle 5 in.; weight of armature 475 lbs.; weight of motor, 1,940 lbs. Westinghouse No. 93.— Rated horse-power 50; gear ratio 3.57; pinion has 19 teeth, gear 68. Four poles, 4 field coils; commutator has 135 segments; arma- ture bearings— commutator end, 3% in. x 6% in., pinion end 3% in, x 8/3 in.} 1 422 ELECTRIC RAILWA Y HAND BOOK. maximum diameter of axle 5>£ in.; weight of armature 778 lbs; weight of motor 2,995 lbs. Westinghouse No. 91 A. C— Rated horse-power 75; gear ratio 3.1 ; pinion, 20 teeth; gear 6 -J. Four poles; 5 field coils; the laminated field consists of circular Dunchings with inwardly projecting poles, support in an outer frame of cast Btoel. The armature is of the slotted drum type with a commutator, and is wound like a direct current armature for 225 volts, and a frequency of 25 cycles. Speed at full load approximately 700 r. p. m. Motors are connected as straight series motors. DATA ON GENERAL ELECTRIC MOTORS. G. E. 54 Motor.— Rated horse-power, 25. Speed at full load, 690 rev. Reduction ratio 4.78. Gear, 14 teeth on pinion, 67 teeth on gear; 4 poles; 4 field coils; 128 turns, total of 4 coils. Armature has 29 slots, 24 conductors per slot; number of bands, 6. Commutator bars, 115 Armature bearings : commutator bearing, 6 ins. x 2% ins.; pinion bearing, 7M i ns « x2 % ins. Weight, 380 lbs. Diameter of armature 11.5 in$; length, 9 ins. Armature coils are slung 7 slots. Short leads go to the right. Begin to connect where center of bar (No. 1) lines up with center of a slot (No. 1). Connect lead from third coil from bottom of this slot into bar No. 16 to the right. The long lead of same coil should be connected into bar No. 43 to the left. One of the middle coils in slot No. 16 to the right should not be connected. G. E. 67 Motor.— Rated horse-power 38. Speed at full load, 525 rev. Reduction ratio, 3.94. Gear, 17 teeth on pinion, 67 teeth on gear; 4 poles; 4 field coils; 110.5 turns, total of 4 coils. Armature has 37 slots, 18 conductors per slot; number of bands, 6. Commutator bars, 111. Armature bearings: commutator bearing, 2% ins. xQ% ins.; pinion bearing, 3 ins. x8 ins. Weight, 575 lbs. Diameter, 14^ ins.; length, 83^ ins. Armature coils are slung 9 slots. To connect draw a line through armature core (slot) over the top end of coil, and note where the line strikes commutator. Including this bar count 15 bars to the right and designate that bar as No. 1. Connect short lead of middle coil to bar No, 1. Long leads go to 56 bar to the left facing commutator counting from No. 1 to the left and counting No. 1 bar as the first bar. G. E. 73 Motor.— Rated horse-power, 75. Speed at full load, 485 rev. Reduction ratio, 4.23. Gear, 17 teeth on pinion, 72 teeth on gear; 4 poles; 2 field coils 80 turns, 2 field coils 40 turns Armature has 39 slots, 12 conductors per slot. Commutator bars 117. Armature bearings: commutator bearing, 3% ins. x 7{% ins.; pinion bearing 3}4 ins. x 10 ins. Weight of armature, 1,150 lbs. Diameter of armature 18 ins.; length, 10J4 i ns « G. E. 66 Motor.— Rated horse-power, 125. Speed at full load, 540 rev. Reduction ratio 4.23. Gear 17 teeth on pinion, 72 teeth on gear; 4 poles; 2 field coils 56 turns, 2 field coils 29 turns. Armature has 39 slots, 10 conductors per slot; number of bands, 11. Commutator has 195 bars. Armature bearings: com- mutator bearing, 3% ins. x6/ s ins. ; pinion bearing, 4 ins. x 10 ins. Weight of armature 1,300 lbs. Diameter of armature 18 ins. ; length, 12}^ ins. G. E. 55 Motor.— Rated horse-power, 160. Speed at full load, 580 rev. Reduction ratio, 3.29. Gear, 17 teeth on pinion, 56 teeth on gear; 4 poles; 2 field •oils 54 turns, 2 field coils 26 turns. Armature has 47 slots, 6 conductors per slot; ELECTRIC RAILWAY HAND BOOK. 423 number of bands, 11. Commutator bars, 141. Armature bearings: commutator bearing 3J4 ins. x 7^ ins; pinion bearing, 3% ins. x 11 ins. Weight of armature, 1,525 lbs. Diameter of armature, 17^ ins.; length, 15 ins. G. E. 53 Motor.— Rated horse-power, 45. Speed at full load, 480 rev. Reduction ratio, 4.6. Gear, 15 teeth on pinion, 69 teeth on gear; 4 poles; 4 field ^ r . ; — sT|* — 7 , l i L i kM 1 u- •--.2 S £— j 23a?~ «• --«^-<*©* DeCw©er» jPhrv»»->ed r>ubs-i- -/»_—_- DeCaitaof Ke> Rad.f— 63-H . Figs. 320, 321.— dimensions of general electric no. 57 motor. coils, 120 turns, total of 4 coils. Armature has 37 slots, 18 conductors per slot; number of bands, 5. Commutator bars, 111. Armature bearings: commutator bearing, 3% ins. x 5% ins. ; pinioD bearing, 3 ins. x 7% ins. Weight of armature, 650 lbt. Diameter, 16 ins.; length, 7\& ins. 424 ELECTRIC RAILWAY HAND BOOHT. G. E. 57 Motor.— Rated horse-power, 50. Speed at full load, 640 rev. Reduction ratio, 4.31 Gear, 16 teeth on pinion, 69 teeth on gear; 4 poles; 4 field coils, 90 turns, total of 4 coils. Armature has 37 slots, 12 conductors per slot; number of bands, 6. Commutator bars, 111. Armature bearings : commutator bearing, 2% ins. x 6% ins. ; pinion bearing, 334 ins - x 8% ins « Weight of arma- ture 690 lbs. Diameter of armature, 14 ins. ; length, 12 ins. Armature coils are slung 9 slots. In connecting call any slot in which the bottom of a coil is assembled slot No. 1 and bar in line with center of this slot bar No. 1. The short center lead of coil No. 1 should be connected into bar No. Fig. 322.— general electric no. 57 motor. Diameter of Axle A 4 ins. 43^ ins. \\i ins. Dimensions of unfinished parts are subject to a small variation. * Weights of pinion and gear change with ratio of gearing. Weight of motor complete without axle gear and case 2632 lbs. •■ " armxture and pinion (16 teeth) 704 ' " ll axle gear (69 teeth) 200 " k ' gear case 140 " 15 to the right. The long load of coil No. 1 should be connected into commu- tator bar No, 42 to the left, counting No. 1 bar same as before, L ELECTRIC RAIL WA Y HAND BOOK'. 425 NEW GENERAL ELECTRIC RAILWAY MOTORS. G. E. 58 Motor.— Rated horse-power 37; gear reductions 3.94; pinion 17 teeth; gear 67. Four poles; 4 field coils; armature has 33 slots; \A% in. diameter, 3 coils per slot connected to 99 commutator bars. Armature bearings- commutator end 2% in. xG^in., pinion end 3 in. x 7Jg in.; weight of armature 486 lbs. ; weight of motor complete 1,865 lbs. ; diameter of axles 3% in., 3% in. and 4 in. This motor can be used on 39>4 * n - ( one meter) gauge. G. E. 53 Motor.— Rated horse-power at 550 volts, 45; gear reduction 3.91; pinion 17 teeth; gear 67. Four poles; 4 field coils; armature has series drum winding; 111 coils in. the 3 and 6 turn machines and 99 in the 4 turn. Armature bearing— commutator end 2% in. x h% in., pirion end 3 in. x 1% in.; weight of armature 669 lbs. ; weight of motor complete 2,440 lbs. ; diameter of axles 3% in., 4 in., 4*4 i n « This motor can be used on 35^ in. gauge and above. G. E. 74 Motor.— Rated horse-power 65. Gear Ratio. Pinion. Gear. 4.56 16 teeth 73 teeth 3.68 19 " 70 M 3.04 22 " 67 " 2.56 25 " 64 '• 2.13 28 " 61 " The motor has 4 poles and 4 field coils. The armature has 3 coils per slot. Armature bearings— commutator end 3% in. x 6% in., pinion end 3% in. x 8% in. ; weight of armature 845 lbs.; weight of motor complete 3,119 lbs.; diameter of axle 5^ in. G. E. 70 Motor.— Rated horse-power 40; gear ratios 4.73 to 2.91; teeth pinion 15 to 22; gear 71 to 64. Four poles; 4 field coils; armature has 3 coils per slot ; armature bearings— commutator end 2% in. x 7^ in., pinion end 3% in. x 8^ in. ; weight of armature 614 lbs. ; weight of motor complete 2,349 lbs.; diameter of axle 4}4 i Q - and 5 in. G. E. 60 Motor.— Rated horsepower 27 ; gear ratio 4.78; pinions 14 teeth; gear 67. Four poles: 4 field coils; armature hr.s37 slots and a series drum winding of 111 coils, either four or six turns per coil. Armature bearings— commutator end 2]4 in. x b% in., and 2% in. x 7^£ in. at pinion end; weight of armature 410 lbs.; weight of motor complete 1,400 lbs.; diameter of axles 3% in., 3% in., or 4 in.; this motor is designed for 35^ in. gauge and over. G. E. A. 60 4 Single-phase A. C. Compensated Motor.— Rated horse-power 75; field winding so distributed as to compensate for amature re- action; motor wound for 200 volts A. C, 25 cycles and 300 volts D. C. Two motors always connected in series. Gear ratio 8.74; pinion 19 teeth; gear 71. THE CONTROLLER. The controller used with a single motor gradually places resistances in series with the motor until the latter is directly connected to the trolley wire, when the equipment is up to speed. To change the direction of rotation of the armature a reversing switch is used, which reverses the connections of either the field or the armature, but not of both. To blow out the arc formed on breaking A 426 ELECTRIC RAILWAY HAND BOOK. the circuit at the controller, a magnetic field is used with such polarity that it tends to deflect away from the two points on the circuit the arcs thus formed, thus sniffing them out. This apparatus is known as a magnetic blow-out. With two motors the controller has a more extended combination. Some of these are shown in Fig. 323 and 324. It first places the motors and resist- ances in series and gradually cuts out the resistance. The next few steps on the controller are known as transition steps, during which the motors are placed in multiple with the resistances again in series with them. The last steps of the con- troller cut the resistances out leaving the two motors in multiple across the line. In some forms of controller a further connection is made when the motors are in series and multiple only which consists of looping around their fields a resistance, thus reducing the current flow through the field coils, weakening the field in which the armature rotates and in this way increaseingthe speed of the equipment. Figs. 323 and 324 show the motor combinations made by some well-known K and L controllers. The sizes and dimensions of the G. E. railway controllers are given on page 364 and 365. The nomenclature for the different parts of the controller is given on page 363. 9 Series Iwg ConUnoller Multiple Res. motor I. motor 2. -Q — O — WAA— O — WAA ■H — O — WW — O — WW — 9ue^ (0 iy md U "ST uuna TT aoue*y?« a ti{| o? ft ^ T35 3 =" J 8* uo»^t«uej£ I gjuiuunfci Jlaovm-ietsata ^ 10 ^7«* 14978 -* / 14623- [14629 •>•-— [14641 *••** 14638 Fig. 325.— sebies parallel controller. The figures refer to parts shown on accompanying cut. 14,642, cap for top of controller; 14,977, star wheel, with pin, for controlling cylinder (must be fitted); 14,682, contact finger; 14,683, single connection clip; 14,684, double connection clip ; 14,983, wood bar for controlling contact board : 14,922, contact tip for controlling cylinder; 14,681, contact base; 14,972, frame, fitted with bearing caps and cap screws for controlling and reversing cylinder shafts; 14,974, insulation disc for controlling cylinder; 14,701, wood base; 14,630, double switch contact; 14,646, hinge bolt, with pin and nut fastening cover to frame; 14,633, fulcrum pillar, with pivot, for two wire connection; 14,629, single switch contact; 14,641, triple switch handle; 14,638, outer switch blade; 14,632, fulcrum pillar, with pivot, for one-wire connection; 14,961, water cap and pointer for controlling cylinder shaft, with set screw ; 14,648, lock bolt with pin (used in connection with check lever for reversing cylinder) ; 14,978, check lever, with roller for conti oiling cylinder; 14,623, check lever, with roller for reversing cylinder; 14,678. water cap for reversing cylinder shaft, with set screw; 14,635, fulcrum for 14,634; 14,639, inner switch blade; 14,695, single switch contact and binding post; 14,921, safety stop nut, with pin, for controlling cylinder; 14.698, binding post, marked ( k 'T"); 14,697, binding post (except that marked "T"); 14,640. double switch handle; 14,696, single switch contact and. Ions: binding post; 14,636, bent cam lever for 14,634; 14,981, safety stop pin for controlling cylinder; 14,963, bracket fastening controller to dasher; 14,647, wire guard; 14,939, cap screw, with wrench attached, fastening pole piece to magnet core; 14,998, star wheel, with pin, for reversing cylinder; 14,692, short contact for reversing cylinder; 14,693, long con- tact for reversing cylinder; 14,687, contact base; 14,688, contact finger; 14,686, wood bar for reversing contact board; 13,804, double washer for 13,369; 14,690, wood body for reversing cylinder; 14,992, wide strip for arc deflector; 14,993, nar- row strip for arc deflector; 14,994, division plate for arc deflector; 14,990, hinge pole piece; 14,938, hinge joint foi pole piece; 14,700, magnet spool, with terminal and flexible lead (Form 2) ; 14,644, hinge joint for cover, with pin and rivets; 14,645, slotted lug for cover, with rivets; 14,699, sheet iron cover (Form 2), complete; 16,922, reversing handle; 16,921, controlling handle. ELECTRIC RAILWA Y HAND BOOK, 429 T Hcet»x «w H* Hx * HmWx icix h4 rf OO TT 1-1 t- 9 10 ©* ©* CO OSl-. QC TH T" H 00 Hs« -I «-* H» •HO * -I Cm r^ O 50 ?D CO 95 os« - CO^f K OS Tf T-H 10 O* IT. tH tH Eh T t-XMX «W H* HX nta -«£ t«XI« -*« * ~f«H> 8 .HO Hno** "K h* ^^2 -5 M COi> «T1 tH « c?oc Tf CO ©*T-t i>ir: 00 00 ir t-X "*• CO ^c« T-t 1-H «^ W « w| » H*- jo HN«W "h ; ^^ t>5^f «»!2 W cot- CO ©irt T-t« 385 ) Tf CO t-iO 0000 ( in i>^ ^ O «*J» t-wieK Hi 10* ^ —He-* .to -1 ''MlO'X He»«T» "S i^w ^^£ "R fci CO N- CO CNJrf tHCC 33 ! £2 t>o 00 00 tH ( »o 1>?C T»« OS 1 «2sa° t-Xlfi C Hi 3 JO •V-IOSO * H2^2 "S « eg*. ©ftf th OC woe Tf CO ©*rH s>m oooo •O t-?D rf M W °9 CJ^iS 10 (,*0 t H •a JO h HhioMO Hfe,^ "5 r 1 Eh M COfc- co N-^ *H« ) Tfa t (MrH *>m 0000 •irt t-O Tf ■r • «R«* Hccisk » w| «i .10 - *V,iett H)S«W ">"■ : ^ t.»^2 w - * COt- CO ©irj TH« ' 33 > Nth i>m 0000 • >Q J> O "^ CD rt •*-«<* «W"M » H » ^H2 3U ^MHI ^2 • — — — * HO w GOOD CO^ • ts« 1 *>cr coco NiO 00O» CO «GV ) (Mr-t T-t 7 •5«S t-OC* N H N HMNI 2 rtixiow H0»«h* *TH CO ©*Tj ■ TH C>- 5 $g 5 ©4rH j>m 0000 • O i.» O Tj» HS-B Nt- St H N -MS X r;xiax -»«Wi "S : ^ t-^^2 "S 00 ©*Tj * tHC 5 MO » 1C CO f ©*tH N«5 00 00 • fc-«o r>« w £ X 1 43° ELECTRIC RAILWAY HAND BOOK, M Hhfnte MWUSX CO co OS 00 mo nMnM" COO COrH M|x COX i-tXMr* 00 O n|X OS rH MIXMW TPI> Mlrf lOIX COCO MtoflW 0000 S 1 MXHX 3£ M|X TP O 8! MW .JS0 .HO OSOO CD O OS'* J*. w|« *n*S -S I CO "TPO COOS CBO TP CO O* tH rH nn nlc^fn -W^n 00 COOft CD TP -HOIX h|n "1h"m Hh iC-XMW" h- n|N""n hWh CD CO CO TP rH CO TP O 'DOS OS ID OS CO CDOS CD-* CO TP CO CM rH rH — >/yy s/W — ^Ar-lj^AW^-J^y-^KK^^ Rr. R, 5T — n/vv — ^vw — MM/w0-nrffTL^^ R2 Rr *t» i vAAAAi — t R 5 R 2 *^7 to Fig. 325- a. ELECTRIC RAILWA Y HAND BOOK. 433 434 ELECTRIC RAILWAY HAND BOOK. CAR WIRING. The roof wiring includes the running of the main circuit wire from the trolley through both main motor switches, and through a concealed groove in the corner post of the car. The size of this wire is giren by the different companies for their motors. The wire is run to a suitable location for connection to the lightning ar- rester and fuse box, also a lamp circuit of No. 16 B. & S. stranded is tapped to wire the trolley connection on the roof and carried to the fixture outlets and the end left to attach to the ground. Where the Mires lie on top of the roof, they should be covered by molding which is well painted to exclude any moisture, especially so where the wires pass through the roof. Additional protection of a piece of canvas under the molding, which has been thoroughly painted with white lead, is neces- sary at this point. The molding should be firmly screwed down to the roof and well painted. The above wiring should be done, if possible, while the cars are being built. The floor wiring may be done after the car is completed without injuring the finish when the body and truck dimensions of the car are given. The different motor companies will furnish a made-up cable having the proper number of wires taped together or covered with hose, and of proper length to reach from the con- trollers to the motors. A hole should be made in the platform under the controllers for the cable to pass through, and in the case of closed cars the cable should pass up again underneath the car seats. In the case of open cars all the wiring is done under the car body. The cables that pass under the platforms should be supported by leather straps attached to the floor or sills. The ground wiring should run under the car floor rather than under the seats. After the cable is in position the motor taps should project through the sills for attachment to the flexible motor leads just far enough to permit an easy con- nection, and with as little chance for vibration as possible. The cables should never be bent at a sharp angle. The joints should be well soldered, and in the case of connecting stranded wires together, the strands should be interwoven before soldering; first tape with insulating tape and then put a rubber tape over it to secure the first tape in position. Wherever the wires of the car cross each other a piece of wood should be secured between the wires, and special protection and additional covering should be given the wires where they pass over iron work or are exposed to mud and water. Wires entering fuse boxes or lightning arresters should be looped down before entering, in order to prevent water running along the wire and into the box. All wires subject to vibration, sucn as those between car bodies and motors, should be of flexible cable, and sufficient slack should be left so that under no condition will any strain be thrown on these wires. In the case of swiveling trucks, more slack will be necessary. As slack gives great opportunity for abrasion, care should be taken to leave only what is absolutely necessary. THE MOTORMAN. The old saying, that u Trifles make perfection, but perfection is no trifle," applies to the street railway service as to every other undertaking. It is only by giving the best attention possible to every detail in the complicated system that perfect results are obtainable, both in economy and efficiency of operation. From the general manager to the switchboard tender and the man who fires the boilers, each employee has an important part to play, and it is only by a thorough under- standing of his duties that he can render effective service. It is for the purpose of giving a few practical suggestions to the motormen that this section has been ELECTRIC RAILWA Y HAND BOOK. 435 compiled, in which no technical terms have been nsed but such as any person of average intelligence can easily grasp. The object of chief concern to the motorman is the controller, for, if he can perfectly manage that, he may be said to understand a large part of his work. When the car is standing still, the controlling handle should be at " off " position. If the car is to be taken out of the car house, where it has been standing with the trolley off, put on the trolley wheel and place the handles on the controlling stand. To start the car, see that the brakes are off, the canopy switches closed ; then move the controller handle to the first notch. After the car is well started, move to the second notch, and after a short time to the third, and so on to the last. Don't stop the handle between notches, and don't move it too slowly. On the other hand, do not move too rapidly from the first notch to the second. Always wait for the car to get up to the speed corresponding to the notch the controller handle is on before going to the next notch, otherwise more current will be used than is necessary. In shutting off the current the handle may be moved around as rapidly as de- sired to •• off " from whatever position it may happen to be on. When stopping at any point, the reverse lever is sometimes used to make the car go backwards. Never reverse while the car is running, unless to avoid an accident. But if it is absolutely necessary to stop the car quickly, pull the brake on with the right hand and shut off the current with the left at the same time; then, with the right hand free, throw the reverse lever and turn on a very little current. If too much current is turned on, the wheels will lose their adhesion to the rails and spin back- wards, which will increase the minmum distance in which the car may be possibly stopped. Sometimes a very violent stop must be made, when the brakes fail, possibly, or the trolley comes off, in which case reverse and put the controller handle on the highest point of the controller. This causes an interaction between the motors which brings them to a standstill. It may damage the apparatus, however, and should only be used in rare emergencies ; this method is only available when two motors are on the car. When approaching curves or turnouts the power should be turned off, apply- ing such power upon reaching the curves as may be necessary to carry the car easily around. The conductor should be on the rear platform with the trolley rope in his hand, ready to give the signal in case the trolley jumps the wire, in which case the motorman should move the controller handle to " off '» until he is notified to go ahead. The motorman should never stop on curves unless absolutely necessary. In running down grades, always have the trolley on the wire, the controller handle at "off ", and the brake arranged so that it can be applied instantly. Be- fore going down a steep grade slow up the car, and set the brakes gradually. If the wheels slide, loosen the brakes to allow them to get hold of the track; then apply the brakes again. If the brakes then fail, reverse the motors. If, in the meantime, the trolley has left the wire, so that there is no power, reverse and throw the controller handle to the last notch, which will make the car come to a standstill. In running up heavy grades, get the car up to speed, if possible, before reach- ing the grade so that it will not require so much current to climb up. If the car is started while on a heavy grade, it will require a very large amount of current. Whether to climb these grades in series or parallel positions is a question on which instructions are given in each individual case. If the wheels slip on the rails, the sand box can often be used to advantage; but always be sure, especially 436 ELECTRIC RAILWAY HAND BOOK. in wet weather, that the sand is dry. Do not use the sand too freely, as you may run short just when it is needed most. If the power gives out, notice if the other cars experience the same trouble, as it may be due to an open circuit on the line; if so, throw the controller handle to "off, 1 ' close the lamp circuit and wait until the lamps light up. If, when the lamps light up, the equipment will not move with the controller handle on the first point, the motorman should first look to see whether his fuse has blown or burnt out; if so, open the head switch, or, better yet, tie down the trolley pole and replace the fuse. If the fuse has not blown, the rails may be dirty and the car insulated from the rails. In this case have the conductor jam the switch-bar between the wheel and rail, while the motorman starts the car. In rare instances there is a case of dead rail. A length of wire should be kept in the car where possible, and one end placed on the rail back of the car towards the power station, and one on any exposed part of the iron truck. Always place the end on the rail first, otherwise a shock will be received. In case, as the car goes along, a peculiar jumping action occurs, known as the bucking of motors, the motor affected should be cut out by means of the motor switches in the controller. Instructions are given the motorman how the motors are cut out on each different type of controller. For remedies for more trouble- some accidents see "The Inspector" below. After bringing the car into the car house have the controller at " off," take off the controller handles, pull down the trolley and tie it a few inches below the trolley wire. THE INSPECTOR. Sparking at Commutator.— Natural sparking will be caused by overload- ing of motor, or by burnt-out fields ; by the shifting of brush-holder in street railway motors, by not having the brush-holder yoke so as to be at equal distance from the commutator on the two sides, or where there are several brushes on one arm or holder by their not being in alignment. The most prevalent causes of sparking in street railway motors are weak brush springs, or a brush worn too short to receive pressure from the spring. Since biushes vary in size they some- times fit tightly in the holder and will produce sparking or flashing when worn away from contact with the commutator. The number of bars apart brushes should be set on any commutator is the total number of commutator bars divided by the number of poles. Subtract from this the number of bars covered by one brush which gives the commutator's bars between brushes. A commutator in proper order should have a dark bronze color, without any biting away of copper at the mica insulation. Where two brushes are used, both should wear down uni- formly. No two brushes should be used in the same brush-holder without being separated by a solid dividing piece between them, and with separate springs to each brush, or they will wear a hollow in the center of the commutator surface. In street car motors with a roughened commutator the brushes are taken out in some cases, and with a two motor equipment this motor is cut out at the con- troller. A piece of wood provided with a handle (see Fig. 326), and having a curved surface, forms a useful device for smoothing. It is as wide as the commutator, and across the top a clamp to hold sand-paper is screwed down by a screw in the handle. The ends are turned over and securely held. No. 1 or 00 is used as required. While holding this in contact with the commutator the car is run up and down the track until the commutator shows a polish. If the ELECTRIC RAIL WA Y HAND BOOK. 437 commutator polishing block is made shorter, the commutator can be polished with the motor operating the car. Some companies use a hollow stone made from a medium hard grindstone, hollowed out to fit the commutator surface, instead of sandpaper. Emery is objected to for polishing commutators for the reason that it is so sharp that it buries into the copper as a matrix, and in turn grinds the brushes. Commutators that are out of line, or have high or low bars or bad flats, are best repaired by turning in a lathe on centers, taking as light a cut as possible in order to bring the commutator concentric again. Use a diamond pointed tool, and where the cut is rough lubricate the tool with a thick solution of soap and water. [323 FlO. 326.— COMMUTATOR POLISHER. It is the practice where the brush wear does not come to the end of the commuta- tor, to leave a small ridge around the commutator at the end next the bearing to further prevent flashing to commutator by oil and carbon dust adhering to the insulating ring at the end of the commutator. A bluish oxide on a commutator shows excessive heating, and the cause should be located. The commutator will show a bar burnt lower than the adjacent commutator bars when there is a short circuit between adjacent windings connected to that bar, the biting into the commutator will continue back in the direction of rota- tion. Where this condition is allowed to continue the commutator will come in hot, and the contact surfaces of the brushes will be black and scarred when they should have a bright plumbago appearance. For locating trouble see armature tests. Commutators, with every other bar blackened, are found in certain types of winding used on railway motors where there are practically two separate windings side by side connecting to alternate commutator bars, and the blackening of every alternate bar is caused generally by a greater difference of potential between ad- jacent bars under commutation. An open lead to a commutator bar causes flashing when that bar passes under the brushes. An open coil on an armature will show a bad bar even where this coil is connected in around the armature. The equipment continues to operate until the flashing becomes so bad as to break the commutator to ground, or burn through the head of the armature. For this reason armatures are now headed with several layers of asbestos paper under the canvas cover to prevent flashing to ground before the inspector discovers the trouble. The grounding of the armature turns produces a short circuit be- tween the trolley wire and rail, which, under this condition, has a sudden braking action, commonly known as " bucking." When this happens the motor is cut out at the controller and the equipment operated on one motor until it gets to the repair shop. The efficiency of the inspector can be readily determined by an examination of the commutators on the equipments he has in charge, allowance being made fftr some motors which require great care to keep the commutators in good ehape. 438 ELECTRIC RAILWAY HAND BOOK. Poor potential delivery and dirty tracks also increase the current flow through the motors, heating and burning the commutator surface, which should be glazed. There is a class of dull, steady sparking which leaves dirty black commutators, which is generally attributed to too soft or poor brushes. In new equipments where the mica segments of the commutator have been built up with too much shellac, the heating of the commutator works it out, also causing it to carbonize on the surfaces of the commutator. This will continue until all surplus shellac has worked out. Commutators afflicted with this trouble will show ridges of shellac forced out over the mica insulation between bars. In the morning when the commutator has cooled down, only two or three bars will occasionally show it, especially where the commutator has been repaired and shellac used too liber- ally in the mica insulation between the new bars. (For proper construction see repair shop practices). Too soft a brush will produce the same effect on the commutator surface but on feeling the surface with the thumb nail slight ridges of mica will be felt between the bars. Try the brush with a knifeblade ; a brush should not be shaved nor penetrated, if the brush is of the proper hardness and the carbon should come off in small granular pieces. This is also true of a brush that is too hard. A hard brush will give with regular brush tension a bright msmzzm Fig. 327.— examples op commutator wear. metal and splintered torn surface to the commutator ; and if the brush is uni- formly too hard, signs of copper dust will be seen around the interior of the motor case. If rings of bright copper show around the commutator the brush may have hard spots ; in this case take out the brushes and see if the contact sur- face shows a corresponding bright spot ; if this spot is found to be harder than the adjacent carbon by penetrating with point of a knife, the trouble is located. A commutator without lubrication will show the same surface as that produced by <>o hard a brush. Use the best quality of vaselene on a small rag, and use spar- ihrly. Oil is used, and, in some cases, axle grease, for labor saving reasons but not to the best interests of the commutator. A brush that is too hard but of uni- form density, as well as brushes that squeak, can be improved by dropping them in hot paraffin ; but be sure and heat the brush just before dipping, otherwise the treatment will be superficial. Hot, heavy lubricating mineral oil is also used instead of the paraffin. It is the custom on some roads for the inspector to change brushes every night on all motors, which insures the inspection of every motor. Brush inspection will show a number of things. A brush that shows pitting on its side (see Fig. 327) indicates bad contact with a weak brush spring, requiring the brush to receive cur- rent from the sides of the holder and chattering and arcing between the holder sides and the brush ; a place broken out of the brush where the brush spring rests shows a weakness and a variation here, causing arcing and wearing away of the brush. The commutator contact surfaces of the brush which show a ridge down ELECTRIC RAIL WA Y HAND BOOK. 439 their length, on both sides of which is a commutator surface, indicates a brush too narrow for the brush holder, as the commutator has worn it to one surface in going in one direction and made another surface in going in the opposite direc- tion due to the change in position of the brush with the movement of the commu- tator. A brush with burnt corners indicates sparking at the commutator ; a brush tapering toward the contact surface, as shown, indicates exces- sive heating of the commutator or brush due to poor contacts. Brush springs may- give good results until they become heated from some cause, when they lose their elasticity. Phosphor bronze springs show the least effect from this cause. THE GENERAL ELECTRIC SINGLE PHASE RAILWAY SYSTEM. This system as developed operates the single phase motors at about 200 volts per motor. The four motor equipments have the motors in pairs connected per- manently two in series. The static transformers on the car supply 400 volts to the motors and receive 2000 volts from the trolley wire. The motors are controlled by standard series parallel controllers with resistance and are adapted to run the car when the trolley supply changes from 2000 volts A. C. to 600 D. C. It is claimed that there is only a slight gain in efficiency when the alternating current is controlled by the variable potential method, and that there is not available space on the usual car to place the necessary apparatus for both control systems. On the Ballston Line equipped by this company, the A. C. trolley wire is sus- pended from a % in. steel stranded catenary. The latter is hung over porcelain insulators at the ends of the cross arms and the trolley wire is supported by the catenary between poles. The position of the A. C. trolley wire is but a few feet above and at the side of the car roof. The car is equipped with two direct current trolley poles in the usual position, and two more at the sides of the roof for the A. C. trolley. Tests on this line indicate an apparent trolley resistance of 1.3 times the ohmic resis- tance and a rail resistance of 6.55 times the ohmic resistance due to the inductive drop in the rail. THE WESTINGHOUSE SINGLE-PHASE RAILWAY SYSTEM. In this system, single-phase railway motors are operated by a current of about 250 volts at 25 cycles, obtained through a static transformer carried on the car. The trolley current at a voltage of 3300 or more is conducted to the primary of this transformer. The motors are controlled by the variation of the voltage at the motor term- inals. This varying voltage may be obtained by taps from the winding of the transformer, or by the use of a controller of th induction type. The latter is practically a transformer with a secondary coil which is movable with respect to the primary, so that the secondary voltage is varied by a change in the angular position of the coil and may either be added to or subtracted from that of the main transformer. The motors will also operate on the standard 550 volt direct current with the necessary control system added. 440 ELECTRIC RAILWAY HAND BOOK. © © 1 © w o d o o (A 2 rV\MA/S 5| -0*0i f ri S/VWVV WWW rAAMA/S pAAMAAi' rAAMAA 1 t 7vwVV * r t-CV^Oi ~ Kaawvv AAMAA vvwvv rW-AAA- rAAMAA AAMAA- rW-W-° . I — O^-^-O — I • I — O jv ^ jv KT3 — i ft vw-wv o •AVWv- I rW^AA/^ ITV^M/W s A vWVv I -YM/V^ rAAM/V^ © w 1 3 8 2 2 w Pi © © 1 ELECTRIC RAILWAY HAND BOOK. 441 The high trolley potential necessitates a more substantial line construction than has been customary, chiefly on account of insulation and mechanical strength. The catenary form consists of a stranded galvanized steel messenger 01 supporting cable from which the trolley wire is suspended at intervals of a few fee , the whole 9^ 11 M 13' 15% 17 JAIessenger "" Cable !O^^lG i O I: 4-lO 1 O I: >M0 i 0^i*-lO i O i: 4-5- -lOO^CF ^Top of Rail Trolley Wire iOO FOOT SPAN ' iri5j£'' 12k" jffiessenger Cajple i G^10 1 oJ*-i0 1 : W--l- 6&" 6'' 6" 6M" f 9% 12k 15k "17 Trolley Wire^. *5^10 1 O 1 H^-<} :: >^1W 180^0 ^Top of Rail 120 FOOT SPAN Fig. 327-b MOTOR CUT OUT. 1. Motors all in. 2. Motors 1 & 2 out. 3. Motors 3 & 4 out. 4. Balancing Coil out. 1 Fig. 327-0 being hung from heavy porcelain insulators. The trolley wire is grooved and is suspended from the mesbenger cable by means of rigid galvanized malleable iron hangers of varying length, so that the trolley wire is maintained at a nearly uniform height above the track. The messenger cable is connected electrically as well as mechanically to the trolley wire so it acts as an auxiliary feeder. The insulators are of corrugated porcelain and the system has been developed for pressures up to 6,600 volts. A SECTION IX —THE OPERATION. Schedule and Speeds.— It is usual to lay out schedules on cross-section paper, taking the longitudinal ordinates f or the distances, and the verticals for the time, as shown in Fig. 328. In this way the routes for cars can be obtained, and their crossing points determined. This method is also used for locating the posi- tion of switches in single track construction. In the matter of speeds, the grades and the time consumed on grades is an im- portant element in laying out switching points, or intersecting points, for the cars. Data regarding this can be found under -'Line Construction," or the grades Number of Revolutions Per Mile for Driving Wheels of Different Diameters. Diameter of Revolutions Diameter of Revolutions Wheel. per Mile. Wheel per Mile. 18 ins 1,116 36 ins 558 20 tt 1,005 38 tt .. 529 22 t< 914 40 tt .. 502 23 874 837 42 44 tt . 480 24 .. 457 26 u 773 46 tt . 437 28 it it tt 718 ♦ 48 50 60 tt tt ii . 420 30 672 . 402 32 628 . 336 83 609 72 " of Electric Cars. . 279 Speed 1 mile per hour. 88 feet per minute. 1.466 feet per second. 2 tt tt ti 176 .t tt it 2.933 tt ti it 3 it it ti 264 tt tt ii 4.4 it ii ti 4 tt it tt 352 tt tt ii 5.866 K it it 5 tt tt it 440 it u ii 7.333 it ii ti 6 tt tt tt 528 it it tt 8.8 tt it tt 7 tt tt it 616 tt it (t 10.266 it it ti 8 «i tt tt 704 it tt ii 11.733 ti «t ii 9 tt tt «t 792 ii tt 13.2 it ti it 10 tt it it . 880 tt it 14.666 (< it it 11 tt •t it 968 tt tt 16.133 li ii ii 12 «t it tt 1056 tt i< 17.6 ii tt it 13 tt it tt 1144 it « it 19 066 tt ii it 14 tt it ii 1232 tt • 20.533 " it it 15 it a tt 1320 tl 22. it tt 41 20 it it it 1760 ti ii if 29.333 ii it II 25 it it ii 2200 it it it 36.666 it ti ti 30 tt tt »i 2640 it ti li 43.998 II it ti 85 tt tt tt 8080 ti ti It 51.331 ii *t It 40 41 tt it 3520 it it ii 58.666 it n it 45 II tt ii 3960 tt ti ti 65.997 •« *« t I 50 tt tt it 4400 tt it tt 73.332 li it li 65 II a it 4840 it ii it 80.663 II tt ii 60 tt it tt 5280 ti ii tt 87.996 it «i II ELECTRIC RAIL WA Y HAND BOOK. 443 (AVERAGE SPEEO 10 M/LE5 PER HOUR) DISTANCE IN MILES 2 3 4 5 So'CLOdK Fig. 328.— schedule diagram. *444 ELECTRIC RAlLWA Y HAND BOOK. and speeds maintained by equipments ; or tests can be made on equivalent equip- ments in order to And the grade constants. It will be found by testing a car on several different grades that a relation can be established between the speed obtained and the square of rise in feet per second of the car body which will be approximately a constant and can be applied to determine speeds on any other grades. - The switching points will not be altered by increasing or decreasing the speed of all equipments, if their grade constants are the same. Signal Systems.- -In signaling on single track roads, it is important, in order not to delay the schedule, that a car arriving at a turn-out can maintain the block ahead clear, and clear the block behind it. A number of methods have been used for this purpose, both manual and mechanical. The principal manual method is known as the Ramsey System. This consists of a signal box at each turn-out, and a single line wire between the turn-outs, with two lamps in one box and three lamps in the other. Each signal box is provided with two handles, one for the block ahead and one for the block behind, which throws the lamps either to ground or to line. It can be readily seen that a motorman, on arriving at the turn-out, can cut the lamps out behind it by throwing the switch to ground, or to the same polarity as the switch at the turn-out back of it, and block the line ahead by throwing the switch so as to light the lamps in the signa; box. If the lamps are already lighted in the signal box, it shows that the section ahead is already blocked. To introduce a signal system of this character on a railroad is a safeguard in operation, and also has great legal weight in case of accident, as it shows an inten- tion on the part of the operator to maintain a system of safety devices for the pre- vention of collisions. Several law suits for damages occurring where this system has been in vogue have been decided against the plaintiff, since it was shown that the motorman ran against his signals and took chances, the plaintiff being the employee in these cases. The merit of a manual system is that its operation is always inspected. The rules for operating the road with a manual system should include a clause requir- ing the reporting of any inoperative signal on the road, so that they will be main- tained by a rigid system of inspection. There, are a number of automatic signals which are operated by the trolley throwing the switch, or auxiliary contacts operated by the trolley wheel, to block the road ahead and unblock the road behind the car. These are now being tried extensively on several roads, but in steam railroad practice it is found that in order to' ensure reliable results signal systems should be under manual control. Steam railroad experience does not point out the possibility of an automatic device being always reliable, and it is subject to the criticism of all automatic devices that their failure to operate is not observed until after an accident. Telephone systems have been used in connection with the single track road. The telephone system can be installed, with telephone boxes along the route, gen- erally at the point of turn-out, and the selective system used. Another method is to run two parallel telephone wires along the road connecting the telephone to these wires by means of double hooks, either one hook above the other or a double-pronged hook introduced between the two telephone wires. In these sys- tems the double return should always be used in order to cut out induction ; and where there is any trouble from this cause, the wire should be transposed every thousandfeet. Hard-drawn copper wire makes the best wire for this construc- tion, and, if .covered, it should be weather-proof, double-braided. ELECTRIC RAILWAY HaND BOOK. 445 Examination of the Motormen.— In examining the motormen for pro- ficiency, special questions should be asked to draw out the emergency methods in use on the road, especially the use of the motors as brakes and when to reverse them. Questions should be asked bearing on points of the road where social attention is required at crossings or where there are special grades to be descended. The following list of questions give those generally used for exam- ination of the motormen : Having been assigned a car by the foreman of your division, what should be your first duty before taking the same out of the shed? Who is supposed to have charge of the car? What are your duties as motorman from the time you take charge of the car until the time you turn the same in, or deliver the car to your relief man? What are your duties with reference to running over railroad crossings, frogs and switches? How would you cross railroad crossings, cross-overs, frogs and switches (with the brake set or released)? What are your duties with reference to handling your car on a down grade? In running through water what would be the most advantageous method in which to operate the motors? What are your duties in case your ear gets beyond your control in going down a grade? In case your car wheels slip in making a grade, what method should you apply to obviate same? What are your duties respecting starting up in case power is shut off ? Under what circumstances are you permitted to reverse your motors ? In case it becomes necessary to reverse the motors, what is your first duty? In what manner would you replace a fuse? In case a second fuse blows on being put in, what is your duty? If a controller acts badly, or other electrical troubles present themselves and either motor becomes uncontrollable, what means would you take to ascertain or locate same? What are your duties in respect to occupation of your time while the car is ou the stand? v What would you do in case your controller becomes unmanageable with the current on and set, and you are not able to turn cylinder to a backward or for- ward position? 1 Where are the contact switches located for the purpose of cutting out motors on various types of controllers ? To what extent is the motorman responsible for the operation of the car? Under what circumstances are you permitted to pass persons desiring to feoard your car. In passing persons desiring to board your car, what is your duty? When approaching a car on opposite track that has been brought to a stop, what is your duty ? Why should you reduce the speed of car on approaching a switch point? Why should the car clear the cross street before bringing the same to a stop? ". Should your car be derailed or from any cause blockade the crossing of a steam railroad, what would be your first duty? Why shourd you ring the gong when a vehicle is ahead of your car and along ittl* of the track t What do you contidtr tht most economical method of operating the controller fc— ill I 446 ELECTRIC XAI^iVAY HAND BOOK. Explain the path of the current from the time of leaving the generator at the power house to its return thereto. Why should the trolley never be pulled down whilst the current is applied? Under what circumstances would you operate your car faster than time points named on time table? In what condition must your car be left in the car shed? What is your duty should you find the trolley wire down? Do you consider it more important to get away as quickly as possible in the event of accidents in order to maintain your car on time, or to remain and render all assistance possible? Before bringing the car to a stop on an up grade with a slippery rail, when would you begin dropping sand? Before making a stop on slippery rail, how should sand be used to prevent flat wheels? Should sand be used on a dry rail ? Should sand be used on a clean, wet rail ? Can a car be brought to a stop in the same distance undei all conditions of the rail ? In what distance would you bring your car to a stop on a level, or slightly down grade, car being operated at a rate of 10 miles per hour, condition of track dry, and rail clean ? What is your duty with respect to the rail ahdad of your car ? In case a car does not start after stopping on a dirty rail, what means would you take in overcoming same ? In what position should your controller handle be with respect to the motors, running down grade ? If any electrical trouble presents itself with the motors and then cannot be controlled by the controller, what effort would you put forth in checking same ? In what manner should you handle your controller in building up the motors to full speed ? What are your duties with reference to brakes before bringing up the motors to full speed ? 4 What is your duty to avoid further destruction when a ring of fire passing around a commutator presents itself ? Name the two chief requirements of motormen ? Why should a sharp lookout be maintained at all times on the rail when the car is in motion ? What tools and appliances should motormen have on the car at all times ? What are the bell signals ? Why are motormen and conductors not allowed to enter a car in the car shed other than the car assigned to them ? m The Handling of the Controller.— The question of the proper handling of the controller is one in which grades, the weight of equipment, motors and controller, all enter. It is the usual practice to instruct motormen to handle their controller, so as to get the equipment up to full speed in a certain time ; but they should be fully instructed to realize the difference between the time when thsy are operating near the power station, or at the end of the line, where the voltage drop is greater. In this case the acceleration is slower, and to turn the controller on too fast will increase the drop on the line and decrease the acceleration of the motor. ELECTRIC RAIL WA Y HAND BOOK. 447 In climbing grades the question arises whether the motor should be in multi- ple or in series. This depends largely on the location of the car with respect to the potential delivery to the trolley at this point. If the voltage drop is consider- able, with the motors in multiple, the series position will be found more economi- cal, and the available energy for the equipment greater. It has been proven beyond a doubt that the proper handling of the controller will save as much as 20 per cent, in the coal bill. The curves (Fig. 329) herewith show some data obtained FlG. 829.— CURVES SHOWING ADVANTAGE OP U8ING CONTROLLER CORRECTLY. from the Chicago Street Railway, showing the difference in power consumption between a rapid start and a slow start. Repair Shop Operation.— On many roads the labor in the repair shop has been put on a piece-work basis, and improvements in the cost of maintenance have been obtained by this method of working. The following division has been used by a large repair shop, a price being fixed for each operation. These price? are, for obvious reasons, omitted. Piece Work Price List for Motor Shop. Controllers. General overhauling includes Taking out and replacing drum. Taking out and replacing wipers (II). Taking out and replacing springs (II). Taking out and replacing caps. Straightening bent cover. R#pHein* wora o*t handles. 448 ELECTRIC RATLWA Y HAND BOOK. Blowing out and inspection of connections in controller, canopy-switch, fuse box and cut-out box. Exchange drum. Exchange reverse. Exchange top (cast iron). Exchange pawl (in addition to price of exchange drum). Exchange foot Cm addition to price of exchange back). Exchange back. .„ ...,.„ Exchange blow magnet coil. Exchange and fitting broken cover. Replacing and adjusting Wiper. j Replacing and adjusting back-spring. Armatures and Fields. i j- Replacing armature. Replacing field coils (each). Replacing and adjusting brush holder and brushes (included in replacing &nd inspecting armature clearance). , Replacing and adjusting brush holder yoke. | Replacing brush spring. Replacing connecting board. Replacing dust pan or cover. Inspection. i Inspecting wheels (each). Inspecting trolley. Inspection of armature clearance, blowing out and painting. Bearings. ' " c <.y.« Replacing armature bearings. > ! Fitting bearings on exchanged armature (each). ■ Fitting axle bearings (new, each). , Axle bearing wick. , ;, Trolley. ,.. Replacing worn out wheel and spindle. Replacing worn out rope. Replacing pole. . * v V : ' . . Straightening pole. • . . ; - Replacing base. . . , , ... . , Replacing canopy switch. , . ....... ,. Replacing canopy switch handle. Replacing fuse box. Replacing fuse box plug. •.'.,..,," Replacing cut-out box. Replacing cut-out box plug. . . .- . Replacing three light cluster and lamps. Replacing single socket and lamp. Replacing lamp switch and plug. Replacing pinion (arm in place). Replacing gear (under oar) One-half gear case taken down and replaced (inolnded in inspecting tnaature elearance.) ■;,-.■ : . : • . . :: : ; . ,< ? Replacing gear pan (whole). ^. -j s ^£ ; . «... . •. ^.., ^ v , , ;. w .;;- tattlag g**r ^i axle wbwi removing from ear, — " ™ - XT ELECTRIC RAILWAY HAND BOOK. 449 Miscellaneous Electric. Replacing motor. Replacing motor frame. Making screw connection. Making soldered connection. Replacing diverter. Replacing diverter spool. Cables (under car). Stripping frame. Assembling frame. Cleaning and painting. Replacing motor with motor lift (as distinguished from same operation, re- placing motor, performed with crane). Equipment Records.— There are a number of methods of keeping equip- ment records. Some roads have a card catalogue, each card representing a car between certain dates, on which are printed the different car parts, with blanks left for remarks for the date and character of repairs. From these dates can be computed the life of the car wheels, trolley wheels, controllers and motors. This card can be made as large as 6 ins. x 8 ins., and on the bottom of it are remarks with room for dates when the car came into the repair shop. The mileage of the car is also entered. In this way a complete record of the equipment is obtained, from which can be computed the cost per year for repairs on the car. From these records the cost of maintenance for the different types of equipments, including trucks, motors, trolley stands and wheels, and controllers, can also be determined. Power Station Records. — If careful data is kept on the power station oper- ation, which includes the item of coal burnt per day and (if it is found economical to divide this to shifts) of coal burnt on each shift, the watts produced for each shift and the water evaporated can be determined. For the purpose of finding the water evaporated, it is well to have an individual water meter for each set of boilers as well as the main meter, to be read for each shift. From this data of the individual meters, and from the effective heating surface the equivalent evaporative efficiency of the boilers can be found. If their cleaning has been neglected it can be ascertained from these individual water meter records. In regard to the generators it has been customary in large units to have a wattmeter for each machine, in order that the output for each unit per day can be determined. Where there is any difference in the character of the units, by com- paring tests (covering a period of a number of days) the efficiency of these differ- ent units can be discovered, with respect to the pounds of water and watts output. By carefully noting these records and laying them out carefully in curve diagrams, very often leaky valves or undue frictions can be discovered. A more extensive record form is given on page 84, under " Boiler Room Tests. " Cost of Power.— In the purchase of power by a railway, there are two methods employed: one, by the car mile, and one by the kilowatt hour. The car mile basis is generally figured on an equivalent of 1.2 to 1.8 kilowatt hours to a car mile. The cost of power production for small roads is generally estimated on tne basis of a car day, as the rate per car mile to cover the fixed charges against a small power production of this kind would be too high. In heatinfc, 'uie heating currant is taken from 15 per cent to 80 per cent in •xcees of the operating current. The ear basis is fixed on a single truck, standard aai bo £ cj, V^ v*«T '"> - x * • A : 1 ^ ^ * -^ ^ V* \° ^ - * *** ^ ; "+<&* > %jP f 'C ^ V V T . / * „ . tf W* -,*^ \ v : ^ v ^ a> Xa o - V A A V' ' r ^ ^ .^ ^* oV m K nn