\J/ v}/ \J/ v{/ v'/ v'/ v}/ v'/ vj/
NORTHWESTERN
UNIVERSITY
LIBRARY
EVANSTON
ILLINOIS
wAA/J\A»t\AAAAA
AMERICAN RAILROAD BRIDGES.
by theodore cooper.
In all things, but proverbially bo in mechanics, the supreme excellence is simplicity.
James Watt.
Engineering News Publishing Company,.
Tribune Building,
New York.
C.XrORTATICN UKTC7
AUG 1 19H6
NORTHWESTERN UNIVERSITY
mm il4
1 c
4 .
[From Transactions of American Society of Civil Engineers.] W \ i
AMERICAN RAILROAD BRIDGES.
by
THEODORE COOPER, M. Am. Soc. C. E.
The existing and the accepted types of bridges in use to-day on
American railroads being the results of a true evolution, no attempt to
present them intelligently would be complete without a brief sketch of
the past history of bridges in America.
The rapid development of the new world and the enormous number
of bridges that has been built within the limits of the nineteenth
century, have furnished us with a wide experience, from which we have
been able to select the good and reject much that was bad or undesirable.
The pioneer life, not only of the earlier settlers, but of each genera¬
tion to the present day, has developed to a high degree the energies,
ingenuity and self-reliance of the American people. These pioneers were
compelled to be men of all trades. Their limited resources and the lack
of time or opportunity to seek for past precedents, impelled them to
solve each problem anew. They, "thought with vigor and were not
fettered with the trammels of science, before they were capable of ex¬
erting their mental faculties to advantage," as Sir Joseph Banks wrote
to Thomas Paine in 1788.
4
Having no educated "lines of least resistance," they were better
able to solve the many problems before them by new and novel
methods.
The bridging of small streams was a part of the pioneers' labor.
The crossing of the larger rivers developed specially gifted men, like
Timothy Palmer, Theodore Burr, Lewis Wernwag, and others less well
known, who built timber bridges that are looked upon as wonderful
structures, even to the present day. The records of the early bridges
of America are very incomplete, but enough remains to show what ad¬
mirable work these early bridge builders could do.
1. Wooden Bridges.
The earliest bridges, where single timbers were not sufficient to
stretch from bank to bank, were short spans supported on piles, or,
where these could not be used, on timber cribs filled with stone. Where
the conditions would not allow of structures of this character, arch spans
were usually adopted.
In 1660, "The Great Bridge," as it was then called, was built across
Charles River, between Old Cambridge and Brighton. It was a pile
bridge.
In 1761 Samuel Sewall planned and built a bridge over York River,
Maine, 270 feet long, supported on thirteen piers. Rebuilt in 1793.
In 1786 Mr. Sewall built a bridge over the Charles River, at Boston,
1 503 feet long, supported on seventy-five piers. A year or so later
bridges on the same plan were built at Maiden and Beverly, Mass.
In 1792 Colonel William P. Riddle built the Amoskeag Bridge at
Manchester, N. H. It was 556 feet long, and supported on five piers and
two abutments. It was commenced on the 3d of August, "at which
time the timber was growing, and the rocks dispersed in the river," and
completed on September 29th.
Between 1785-92 Colonel Enoch Hale built over the Connecticut
River, at Bellows Falls, a bridge 368 feet long, in two spans, taking
advantage of a rock in the middle of the river for his center pier. The
West Boston Bridge over the Charles River, 3 583 feet long, and
supported on one hundred and eighty pile bents, was finished in 1793.
Near the end of the eighteenth century a bridge was built over Cayuga
Lake, N. Y. It was a pile bridge in 25 foot spans, one mile in
length.
5
In 1795 a bridge was built over the Mohawk River, 960 feet long,
supported on thirteen piers.
In 1792 Timothy Palmer built the Essex-Merrimack Bridge over the
Merrimack River at Deer Island, about 3 miles above Newburyport,
Mass. It consists of two bridges resting on Deer Island in the midst of
the river (Plate I). "An arch of 160 feet span and 40 feet above the
level of high-water connects this island with the mainland on one side;
the channel on the other side is wider, but the center arch is but 113
feet." That part of the bridge on the Newbury side, the 160-foot span,
was removed in 1810 and replaced by a chain suspension bridge. The
part on the Salisbury side remained until 1883. The chain bridge was
built by John Templeman, of the District of Columbia. "It was the
first chain bridge built in New England." Its span is 244 feet between
bearings on towers; the towers are timber frames covered with boards
and shingles. February 6th, 1827, one of the chains broke under a
heavily loaded wagon drawn by four oxen and one horse. The bridge
was again rebuilt, and is still in use. The bridge, as it now exists,
consists of two independent roadways, each 15 feet 6 inches wide.
Each roadway is supported on two sets of chain cables, each set being
composed of three chains. These chains seem to have been repaired
in many places with different sized links. The chains generally are
formed of links about 2 feet long, made of 1-inch square iron. For
about 6 feet over the bearings on the towers, each chain is spliced or
replaced by three smaller chains with links about 1 foot long, and
of about ^-inch square iron. The floor is hung from these chains
every 7 feet by suspenders, formed indifferently of bars 1 inch square,
straps 2 x J inch, or pieces of chains (Plates II and III).
In 1793 Timothy Palmer built another bridge over the same river
at Andover. Rebuilt in 1803.
i In 1794 he built the Piscatauqua Bridge, 7 miles above Portsmouth,
N. H. It was 2 362 feet long. The greater part consisted of pile
work.
"But that part which engages the attention of travelers, is an arc
nearly in the center of the river, uniting two islands over water 46 feet
deep. This stupendous arc of 244 feet chord is allowed to be a masterly
piece of architecture, planned and built by the ingenious Timothy
Palmer, of Newburyport, Mass."
"The chord of this arch is 244 feet 6 inches. The versine of the
arch was 27 feet and 4 inches, and depth of the frame-work of the arch
18 feet and 3 inches. There are three concentric ribs, the middle one
carrying the floor of the bridge; they were selected from crooked tim¬
bers, so that the fiber might run nearly in the direction of the curves,
and are connected together by pieces of hard and incompressible wood,
with wedges driven between, the ribs being mortised to receive them;
thus the ribs are kept at a regular and parallel distance from each other.
Each rib is formed of two pieces laid side by side about 15 feet in
length; they are ad disposed in such a manner as to break joints, the
end of one timber coming in the middle of the length of the other
which is near it; their ends all abut with a square joint against each
other and are neither scarfed nor mortised, the two pieces of timber
beiDg held together by transverse dovetail keys and joints; all the tim¬
bers are admirably jointed and freely exposed to the action of the air;
any piece may also be removed in case of its requiring separation with¬
out injury to the rest of the structure."
The bridge was 38 feet wide and had three arched trusses. Another
description states that the second rib carrying the floor of the bridge
was of a larger radius than the lower to facilitate the traveling; the
upper rib served for a railing.
In 1794 the bridge at Haverhill was built by Timothy Palmer,
" contemporaneous with the first stage coach and the first newspaper."
It consisted of three arches of 180 feet; each is supported on three hand¬
some piers 40 feet square; it had as many defensive piers or sterlings
extending 50 feet above and a draw 30 feet over the channel.
In 1796 he built a bridge over the Potomac River, at Georgetown.
In 1796 Eufus Graves built a bridge over the Connecticut River, at
Hanover, N. H., consisting of a single arch of 236 feet. He patterned
his bridge after the Piscatauqua Bridge built by Palmer. The roadway
followed the line of the arch and was some twenty feet higher at the
center than at the abutments. The bridge was formed of the largest
selected white pine, hewed to 18 inches square, some of them 60 feet
long. This bridge fell in 1804, without warning, and by its own weight;
its destruction being hastened by the undermining of one of the abut¬
ments, through deficient waterway. The builder of this bridge studied
divinity, graduated at Dartmouth College, became a merchant, then a
bridge builder, then an officer of the United States Army, and finally a
physician.
In 1795 a bridge was erected at Holt's Rock, between Newbury and
7
Haverhill. It was 1000 feet in length and consisted of four arches and
one draw-span. It was carried away by the ice in 1818.
In 1796 a bridge was built between Harlem and Morrisania, over the
Harlem Eiver.
About the end of the last century a bridge was built at Windsor,
Vt., with two spans of 144 feet, and one at Romley, Mass., consisting
of 8 arches of a total length of 870 feet; one at Rowland's Ferry, R. I.,
900 feet long, with a sliding draw, supported on 42 pile bents, and the
Weybasset bridge at Providence, R. I., " 160 feet long, supported by
two wooden trestles and two stone pillars."
There were also two bridges over the Lehigh River, one at Bethlehem
and one at Easton.
These are the records as I have been able to gather them, of bridges
of any importance built in the eighteenth century.
From 1804 to 1806 "The Permanent Bridge" over the Schuylkill
River at Philadelphia was built. It consisted of two arches of 150 feet
clear and one of 195 feet clear.
"The plan was furnished by Mr. Timothy Palmer, of Newburyport,
Mass., a self-taught architect. He brought with him Mr. Carr as his
second and four other workmen from New England. They at once
evinced superior intelligence and adroitness in a business which was
found to be a peculiar art, acquired by habits not promptly gained by
even good workmen in other branches of framing in wood." "The
frame is a masterly piece of workmanship, combining in its principles
that of king post and braces or trusses, with those of a stone arch."
There were three truss and arch frames. The timber was of the best
white pine. Width of the bridge, 42 feet. The bridge was covered and
closed in from the weather. Mr. Palmer stated that from his experi¬
ence, wooden bridges uncovered would become unsafe in ten to twelve
years.
"I am an advocate for weather boarding and roofing, although there
are some who say it argues much against my own interests; notwith¬
standing I am determined to give my opinion as appeal's to be right. It
is sincerely my opinion that the Schuylkill Bridge will last thirty and
perhaps forty years if well covered. You will excuse me in saying, that
I think it would be sporting with property, to suffer this beautiful piece
of architecture (as you are sometimes pleased to call it), which has been
built at so great expense and danger, to fall into ruin in ten or twelve
years."
Plate IY is copied from an old engraving. The superstructure was
entirely renewed on a more modern plan of a wooden arch bridge in
1850, and widened for also carrying a railroad track. The foundations
8
for this bridge were very difficult of execution, owing to the lack of
experience at that day in work of this kind. Depth of water at east pier
21 feet, and at west pier 11 feet, rook bottom with slight covering. The
foundations were laid in coffer-dams. Mr. William Weston, an eminent
English engineer, who was in Philadelphia at that time, furnished the
plan for the coffer-dam; but "declared that he should hesitate to risk
his professional reputation on the event." Afterwards, in a congratu¬
latory letter to the President, he says :
"I most sincerely rejoice at the final success that has crowned your
persevering efforts in the erection of the west pier ; it will afford you
matter of well founded triumph when I tell you that you have accom¬
plished an undertaking unrivaled by anything of the kind that Europe
can boast of." "I have never, in the course of my experience or read¬
ing, heard of a pier founded in such a depth of water on an irregular
rock affording little or no support to the piles."
This pier, however, was not carried to the rock, for the weakness and
leakage of the coffer-dam became so threatening, the masonry was started
three or four feet above the rock.
In 1805 Timothy Palmer completed the bridge over the Delaware
River at Easton, Pa. It consists of three spans, 163 feet clear. This
bridge, after eighty-four years' service, is still in use and about five-
sixths of the original timber is in good condition. It has always been
covered in from the weather. (See Plate V.)
In 1801, Theodore Burr built the Waterford Bridge over the Hudson
River, consisting of four arch spans, one 154 feet, one 161 feet, one 176
feet, and the other 180 feet, clear spans. The timber is hewn yellow pine.
This bridge was not covered until 1814, but is still in use, and in reason¬
ably good condition. (See Plate VI.) The footings of some of the arches
required renewal a couple of years ago. The drawings of the Easton
and Waterford bridges are made from sketches taken from the exist¬
ing structures. Burr also built another bridge over the Hudson at
Fort Miller.
A bridge over the Connecticut River at Springfield was built about this
same date by Mr. Walcott.
From 1804-06, Theodore Burr built the bridge over the Delaware
River at Trenton (Plate VII) consisting of five arch spans, two of 203 feet,
one of 198 feet, one of 186 feet, and one of 161 feet in the clear. Each span
had five arched ribs, formed of white pine plank, from thirty-five to
fifty feet in length and 4 inches thick, repeated one over the other,
breaking joints, until they formed a depth of 32 inches. The lower
9
chord was composed of two sticks Gl x 131 inches. The roadway was
suspended from the arch ribs by vertical chains. The arch was counter-
braced by diagonal braces, formed of two sticks 6 x 10 inches spiked to
the lower chord and secured to the arch above by iron straps. Outside
of the exterior ribs wing arches 50 feet long, splayed out so as to widen
the bearing on the piers and abutments. The bridge had no other
wind bracing. This bridge is erroneously credited to Lewis Wernwag.
The arch footings required renewal in 1832, owing to decay of the
timber. In 1848 the bridge was remodeled by removing the wing
arches and adding a new and stronger arch rib on the south side, and at
the same time strengthening the adjacent old arch rib, by increasing its
depth to 4 feet, to carry a railroad track. In 1869 it was again
strengthened. It was finally removed in 1875, and replaced by an
iron structure.
This bridge may, therefore, be considered as a connecting link
between the early wooden long-span bridges for highway traffic and
those for railroad purposes.
In 1808, Theodore Burr built the bridge shown on Plate VIII, at
Schenectady, N. Y. This was the second bridge built at that place by
Burr, the first falling before or soon after completion. The traditions
in regard to the first bridge are not very reliable. The contract for the
one shown, however, states that it is to be "exactly on the site of the
bridge built by the said Theodore at the same place," and apparently
with different spans, as new piers are provided for, by the contract.
There may, therefore, be some truth in the tradition, that the first
bridge was to consist of two spans of 450 feet; that one span was
•completed and the other partially, when it fell or was swept away
by floods.
The bridge shown is a curiosity in bridge construction. It is a sus¬
pension bridge of wood. The curved ribs are formed of eight 4 x 14-
inch planks spiked or bolted together. The planks are cut to long
bevels at the ends for splicing. The ribs are three in number, spaced
about 13 feet apart. The timber was white pine. The plan shown is
copied from Burr's original drawing attached to the contract.
The bridge was used for twenty years in this condition. It then
became necessary to build additional piers under the middle of each
span to stop the excessive sagging of the bridge. In this condition it
.stood till 1873, when it was replaced by an iron bridge.
10
From 1812-16, Theodore Burr built the bridge at Harrisburgli, Pa.,
over the Susquehanna River, consisting of twelve spans of about 210
feet. The half of this bridge which is south of the Island still remains
in use.
In 1812 Lewis Wernwag built his " Colossus " bridge over the Schuyl¬
kill at Fairmount, Philadelphia. (See Plate IX.) It had a clear span of
310 feet 3i inches. The eastern abutment was founded on the rock,
and the western one on piles driven to the rock. This bridge was
destroyed 1 y fire, September 1st, 1838.
In 1811, he built the New Hope Bridge, Plate IX, over the Delaware
River, consisting of six arch spans .of 175 feet; versed sine of arch, 13
feet. Each span consisted of three arch ribs, formed of four sticks 6
x 15 inches, dressed to the curve and confined by iron clamps and bolts.
Wernwag's practice was to saw all his timbers through the heart, to
detect unsound wood and permit good seasoning of the timbers. He
used no timbers of a greater thickness than 6 inches, and separated all
the sticks of his arches by cast washers, to allow free circulation of the air.
In 1810 he built a wooden cantilever bridge over the Nashammony
River, Pa. He called it his "Economy" bridge, and claimed that it
could be used to advantage up to 150 feet spans.
From this date to 1836 he built many bridges. In 1830 he built his
first railroad bridge at Manoguay, on the Baltimore and Ohio Railroad.
His last bridge was the railroad bridge over the Canal and Potomac
River, at Harper's Ferry, in 1836.
The previously mentioned bridges with spans of any magnitude
appear to have been arch bridges.
The first marked step toward bridges of the modern truss form was
the wooden lattice bridge, patented by Itliiel Town, January, 1820.
Plate X.
The horizontal members or chords of this bridge were composed of
two or more parallel sticks, spaced so that the diagonal web mem¬
bers or lattice work would pass between them. The timbers were usu¬
ally 2 to 4 inches in thickness, and 10 to 12 inches wide. The web mem¬
bers were placed closer together for the longer spans. The chords
were also made in two sets for the longer spans, one placed above the
other.
The web members at every crossing and at their junction with the
chord pieces were fastened with wooden treenails.
The timbers being of one size and of reasonable dimensions, were
readily obtainable. The absence of all bolts, straps or rods of iron, and
the simplicity of the mechanical operations required to connect together
the parts of this bridge, made it a Cheap and popular structure.
A great number of bridges of this class, up to spans of 220 feet, were
built throughout the United States, both for highway and railroad pur¬
poses. There are many of them still in existence.
These bridges were made of uniform section throughout. In order
to get sufficient support at the ends, they were usually extended over the
abutment a distance about equal to their depth, and were made continu¬
ous over all the piers.
From the thinness of the web system the trusses were apt to warp,
and as they aged they became very flexible, owing to the want of rigidity
of the treenail connections.
Mr. Town claimed, at least in his later pamphlets (1831), that these
trusses could be made of wood, wrought or cast-iron.
The next step toward simplicity and concentration of parts appears
to have been the truss known as the Long truss, patented by Brevet-Lieu¬
tenant-Colonel Long of the United States Engineers, in March, 1830, and
November, 1839.
This also was a form of truss in which iron did not enter as a neces¬
sary part, the connections being made by framing the parts together, or
by use of wooden keys or treenails. Many bridges were built of this
form, but it never became widely popular.
Both the Lattice and the Long bridges were combined with the arch
in many cases, especially for the longer spans.
No further important advance in the styles of the trusses was made
until 1840, when William Howe patented the truss which was the basis
for the present Howe truss bridge. Plate XI.
This form of truss grew rapidly into favor, from its simplicity of con¬
struction, perfection of detail and satisfactory action under service. For
some years it has been the standard form of wooden bridge in use upon
our railroads.
The chords and braces are made of timber, and the vertical web mem¬
bers (tension) of round iron, with screw ends, usually upset to give
a stronger section through the screw threads than in the body of the
rods. The chords are made of uniform section throughout, and both
the top and bottom chords of the same number of timbers, and timbers
12
of tlie same width and spacing. The sticks of the lower chord (for
through bridges) are made deeper than those of the top chord, on ac¬
count of the loss of section from splicing, the lower strains used in
tension and to resist the bending strains due to the cross floor timbers,
which rest directly on the lower chord. The chord sticks are spaced two
or more inches apart, to allow the passage of the vertical rods without
cutting away any of the timber, to give room for the packing and splice
blocks, and to admit free circulation of air between the timbers. The
number and length of the chord timbers are arranged so that no more
than one stick need be spliced in any one panel of the truss. The
chord sticks may be spliced either with oak or iron splices.
The braces and counterbraces are square-ended, and rest upon the
inclined faces of the angle blocks. They are held from displacement by
lugs or dowels. The angle blocks were formerly made of oak, but now
are made of cast-iron. They are provided with square faced projections,
which are let into the chord timbers, to take up the horizontal shear of
the braces by direct fiber compression.
The " gib plates," as the plates on which the nuts of the web rods
bear are called, are usually made of thick, flat plates of wrought-iron.
The space between the chord timbers below the angle blocks contains
cast tube blocks, made a little short to allow for the shrinkage of the
timber.
The vertical shear of the braces is transferred through the angle
blocks and these tube blocks to the gib plates and web rods, thus
avoiding any crushing action across the fiber of the wooden chord
pieces.
The timber in a properly designed Howe truss is only strained longi¬
tudinally with the fiber.
The very wide experience in this class of bridges has enabled the
American bridge builders to so proportion them in all their details that
the best results can be obtained from the material at a moderate cost.
But for the two objections, their liability to destruction from fire or
from decay, no better railroad bridge up to 150-foot spans could be
desired. For most of our country the cost of good timber makes iron
bridges the cheaper, at first cost, for spans over 150 feet.
Very few Howe truss bridges are now being built upon the older
and financially better roads, all wooden bridges being replaced with
iron structures.
4
13
In April, 1844, Thomas W. and Caleb Pratt patented the truss known
as the Pratt truss. It differed from the Howe in making the diagonal
members of the web system of iron (tension) and the vertical members
(compression) of wood, the reverse of the Howe principle. While many
bridges were built upon this form, it never succeeded in attaining an
equal popularity with the Howe; as it required a greater quantity of the
more expensive material, wrought-iron, and was not so well suited to
the joint use of the two materials, wood and iron.
It became, however, the favored form, afterwards adopted, for iron
bridges, and is therefore one of the steps in the development of Ameri¬
can bridges.
In addition to the general forms of structures heretofore mentioned,
the variations of these broad types and other special forms were num¬
erous. Each individual builder had his own favorite at some time. It
would be impossible and useless to take up those numerous individual
cases, though they undoubtedly served a good purpose in the general
evolution.
The development of wooden bridges was entirely empirical. They
were generally of uniform section, and their proportions determined either
from the result of previous failures or from the study of models. The
advancement was, however, toward simplicity of form and construction,
and greater perfection of the details of the connection of the parts.
2. Iron Bridges.
As early as 1786 the need of a material for long span bridges, more
suitable than wood, engaged the attention of thinking men.
Thomas Paine, in 1803, wrote as follows, giving an account of his
advocacy and efforts toward the construction of iron bridges:
"As America abounds in rivers, * * * I turned my attention, after the
Revolution was over, to find a method of constructing an arch that might,
without rendering the height inconvenient or ascent difficult, extend
at once from shore to shore over rivers of three, four or five hundred
feet and probably more. The principle I took to begin with and work
upon, was that the small segment of a large circle was preferable to
the great segment of a small circle. The architects I conversed with
in England denied the principle; but it was generally supported by
mathematicians, and experiment has now established the fact."
It should be borne in mind that the only iron bridge then in existence
in England was the cast-iron arch over the Severn near Coalbrook Dale;
which arch was nearly a semicircle, having a span of 100 feet and a ver-
sine or rise of 45 feet.
14
Between 1786-87 Paine made three models, " one in wood, one in cast-
iron and one in wronght-iron connected with blocks of wood representing
cast-iron blocks." This is Paine's wording; but as the Wearmouth
Bridge was copied after Paine's model or bridge afterwards built from
this model, it would be clearer to say, the last model consisted of
wooden blocks rejrresenting cast-iron voussoirs, spliced together by
wrought-iron bands fitting into recesses in the blocks and secured by
screws.
This last model he took to Paris in 1787 and presented it as a model
for an arch bridge of 100 feet span to the Academy of Sciences for its
opinion. Its committee, consisting of M. Le Roy, Abbe Bosson and
M. Borde, the last two celebrated mathematicians, reported that an
arch upon the principle and construction of the model might be ex¬
tended to the span proposed—400 feet. Paine then had a couple of arch
ribs of 90 feet span and 5 feet rise cast at the foundry of Messrs. "Wal¬
ker, at Rotheram, Yorkshire, and tested with double their own weight.
On the success of this experiment Paine had a complete bridge of five
arch ribs of 110 feet span and 5 feet rise made by the same founders
and sent it to London, " as a specimen for establishing a manufactory
of bridges to be sent to any part of the world." Soon after this his
interest in bridges was obliterated by his interest in the French Revolu¬
tion. His bridge was sold to satisfy his creditors.
In the memoir from which the above information is taken, dated 1803,
he proposes that the Congress of the United States erect an experimental
arch of 400 feet span " to remain exposed to public view, that the method
of constructing such arches may be generally known." He offers to
furnish the proportions of the several parts and to superintend its con¬
struction without compensation.
From this time no further consideration seems to have been given to
the use of iron for bridges until about 1830, when Long and Town both
suggested that their bridges could be made in iron or wood.
In 1833 August Canfield took out the first patent for an iron truss
bridge.
The first iron truss bridge built is believed to be the one erected
over the Erie Canal at Frankford, N. Y., in 1840, by Mr. Earl
Trumbull.
" The truss was a combination of the truss and suspension principles,
and was formed of—first, S3ven cast-iron sections or panels of about 11
15
feet in length and 7 in depth, cast solid, each segment consisting of an
upper chord, a pair of diagonal braces and half of a hollow cylindrical
post at each end (of the segment), except that the end segments had full
cylindrical posts at the abutments. These semi-cylinders being bolted
or clamped together in series, formed full cylindrical posts, which were
flanged at the bottom, and through which were passed vertical bolts
securing them to wooden transverse floor beams. Second, two wrought-
iron suspension rods (1J inches diameter) attached to the top end of the
end posts, and sagging in a parabolic curve, so as to pass under and
support the two centermost floor beams, and under lugs cast at proper
elevations upon the posts intermediate between the centermost and the
end posts, whereby such intermediates were supported. Cross-sections
of chords and diagonal braces of the + formed section."
In the same year (1810) Mr. Squire Whipple built his first iron
bridge. It was a bow-string truss bridge in which the compression
members were of cast-iron and the tension members of wrought-iron.
He took out Letters Patent for this style of bridge April 24th, 1841. A
large number of bridges on this plan with spans from 50 to 100 feet
have been built, and some few with spans as long as 180 feet.
In 1846-47 James Millholland built a boiler plate tubular girder 55
feet long for the Baltimore and Ohio Railroad at Bolton depot.
In 1846 Frederick Harbach patented an iron Howe truss; the top
chord and braces were of cast-iron and the lower chord and vertical
rods were of wrought-iron. Each chord and the main braces were
hollow cylinders in pairs, the lower chord was of boiler iron riveted
together as continuous tubes. The braces and chords were connected
by cast-iron saddles shaped to fit the chords.
A bridge of this style, 30 feet clear span, 6 feet depth and 4 feet
panels was built in 1846-7 on the North Adams branch of the Boston
and Albany Railroad near Pittsfield, Mass.
About 1847-50 Nathaniel Rider built a number of bridges for the New
York and Harlem, Erie and other railroads. The Rider bridge was com¬
posed of parallel top and bottom chords and multiple systems of vertical
posts and diagonal ties. The top chord and posts were of cast-iron;
the lower chord and ties of flat bars of wrought-iron. A wedge was
inserted at the top of each post to tighten up the systems. The mem¬
bers were bolted together. The failure of one of these Rider bridges on
the Erie Railroad in 1850, following the failure in England of the
16
bridge over tlie River Dee, influenced the officials of that road to the
decision that iron bridges were untrustworthy, and to direct that all
iron bridges, consisting of several Rider bridges and some which had
been built by Squire Whipple, be removed and replaced with wooden
structures.
"The first impulse to the general adoption of iron for railroad
bridges was given by Mr. Benjamin H. Latrobe, chief engineer of the
Baltimore and Ohio Railroad. When the extension of this road from
Cumberland to Wheeling was begun he decided to use this material in
all the new bridges. Mr. Latrobe had previously much experience in
the construction of wooden bridges in which iron was extensively used;,
he had also designed and used the fish-bellied girder constructed of cast
and wrought-iron; he had adopted on the older portion of that road the
Bollmau plan of bridge for short spans. For the bridges west of
Cumberland he adopted the plans submitted by Mr. Albert Fink, his
assistant."
In 1852 a span of 124 feet upon the Bollman plan was comqdeted
at Harper's Ferry. Bollman's truss is shown on Plate XII.
• In 1851-52 three spans of 205 feet each were built over the Monon-
gahela River upon the Fink plan. Plate XIII.
Both of these styles of bridge were of the suspension truss form,
being different developments of the trussed beam.
The chords and posts were of cast-iron and the tension members
of wrought-iron.
The wrought-iron tension bars had eyes at the lower ends and were
connected to the feet of the posts by pins. At the top these bars had
screw ends for adjustment. Plates XIY and XY show other forms of the
Fink truss.
In 1852-53 Squire Whipple built upon the Rensselaer and Saratoga
Railroad, 7 miles north of Troy, N. Y., an iron bridge of 146 feet clear
span. (See Plate XVI.)
It was a double intersection Pratt truss with inclined end posts, a
form which was known afterwards as the Whipple truss. The top
chord and posts were of cast-iron. The lower chord was composed of
links of wrought-iron. The web rods were round rods of wrought-iron
with eyes at the upper end and screw ends below. They connected
to a pin through the top chords and pass through holes in the cast-iron
shoe at the foot of the posts. This cast-iron shoe had oblong trunnions
at each side to receive the lower chord links. Although this bridge was
proportioned for a rolling load of one ton per lineal foot only, it con¬
tinued to do good service until 1883, when it was taken down, as it was
17
deemed prudent to replace it with a bridge of wrought-iron adapted to
the requirements of the then existing traffic (the engines in use upon the
road having increased in weight from engines of 35 to 40 tons to engines
of 50 to 68 tons).
A few months before this bridge was taken down one of the pair of
main web rods 11 inches diameter broke with a "suspicious looking
fracture;" the remaining rod, however, had sufficient strength to prevent
a collapse of the truss.
From 1851-61 many iron bridges from 65 to 110 feet spans were built
on the western and mountain divisions of the Pennsylvania Railroad.
There were Pratt trusses stiffened with arches, the top chords, posts
and arches being of cast-iron. They were built at the company's shops.
In 1857 Mr. F. C. Lowthorp built his first railroad bridge on the
Catasauqua and Foglesville Railroad, Pennsylvania. It consisted of
11 spans, total length 1122 feet, supported on iron trestle towers 89 feet
high. The chords and posts of this bridge were of cast-iron and the
tension members of wrought-iron with adjustable screw ends.
About 1856 John W. Murphy built a Whipple bridge over the Sau-
con Creek for the North Pennsylvania Railroad, of 152 feet clear span.
The road was never completed on this particular line, and the bridge
still stands as an isolated specimen of one of the earliest forms of iron
railroad bridges of the Whipple type.
In 1858-59 Murphy built for the Lehigh Valley Railroad a Whipple-
Murphy bridge of 165 feet span, over the Canal at Phillipsburg, N. J.
In this bridge he substituted for the cast trunnions on the post feet of
the Whipple bridge pins of wrought-iron, unturned. The lower chord
was formed of wrought-iron elongated links similar to the Whipple form.
The main web bars were wrought-iron bars with looped eyes at each
end. The counter bars were also bars with looped eyes, but the lower
eye was elongated and fitted with gib castings and keys for tightening
the bars.
This is the first truss bridge, as far as the author has been able to
discover, which was pin connected throughout.
In 1869 this bridge was taken down, and put up as the middle span
in a long wooden bridge of nine spans, at Towanda, Pa., to reduce the
liability of destruction of the whole bridge by a fire.
In 1879 it was again removed, and rebuilt by substituting wrought-
iron compression members for the cast-iron, turning the pins, reboring
IS
the links, etc., and put up at Shepherd's Creek, on the Southern Central
Railroad.
In 1861 Mr. J. H. Linville built a bridge on the Delaware extension
of the Pennsylvania Railroad over the Schuylkill River, in which were
used for the first time wide forged eye-bars and posts formed of
wrought-iron sections.
In 1863 John W. Murphy built a bridge over the Lehigh River at
Mauch Chunk for the Lehigh Valley Railroad, in which he used
wrought-iron for both the posts and top chord sections. This is prob¬
ably the first American truss bridge in which the tension and compres¬
sion members were of wrought-iron. He used, however, cast-iron joint
blocks and pedestals.
In 1865 Mr. S. S. Post built the first iron bridge of the style since
known as the Post Truss, for the Erie Railroad, at Washingtonville, on
the Newburgh branch.
Between 1865-80 a large number of bridges on this style were built.
From its peculiar form, the counter system of web members consisted
of a single system, which met at the center a double sytem of direct
web members, thus rendering the strains ambiguous. Much of its early
popularity was due to the apparent stiffness of this form of truss under
moving loads—the truss being put under an initial strain by the counter
system which extends the whole length of the truss.
In 1859 Mr. Howard Carroll commenced building riveted lattice
bridges for the New York Central Railroad'and its connections. The
bridges built by him were mostly short spans, varying from 10 to 90
feet. They were well proportioned both in strength of parts and in the
detail of the riveted connections, for the service then considered as
sufficient.
This early work of an excellent character gave the preference to this
class of bridges upon this road and other neighboring systems.
Mr. Charles Hilton, a pupil of Mr. Carroll, continued his work and
extended the system to longer spans, the longest being the bridge built
by him over the Connecticut River at Springfield, Mass., in 1871. It
consisted of seven spans of 180 feet.
In 1861 Mr. Felician Slataper designed and built for the Pittsburgh,
Fort "Wayne & Chicago Railroad a bridge over the Allegheny River at
Pittsburgh. It consists of five riveted lattice spans, 178 feet each.
In 1865-66 William Fairbairn & Co., of England, built the iron lattice
19
bridge over the Connecticut River on the Hartford and New Haven Rail¬
road, near Windsor Locks, from designs by Mr. James Laurie (Past
Prest. Am. Soc. C. E.), the longest span being 177| feet.
Long Span Bkidges.
The era of long span truss bridges in America may be considered as
dating from the building of the first bridge over the Ohio River at Steu-
benville, between 1863-64, by Mr. J. H. Linville. The channel span
was 320 feet long and 28 feet deep. The top chord and posts were made
of cast-iron. It was proportioned for a rolling load of 3 000 pounds per
foot of track, a notable increase in the load heretofore in use. (See
Plate XVII.) This bridge is now being removed.
This was followed by the bridge over the Ohio River at Louisville,
built by Mr. Albert Fink in 1868-70, with two main spans of 360 and
390 feet. (See Plate XVIII.)
In 1870 Mr. Linville built the bridges over the same river, at Parkers-
burg and Bellaire, with channel spans of 340 feet.
And in 1872 he built the Newport and Cincinnati Bridge, with a chan¬
nel span of 420 feet.
During 1868-74 James B. Eads built the bridge at St. Louis over
the Mississippi River, consisting of two arches of 502 and one of 520
feet clear span.
In 1876 the Cincinnati Southern Railroad bridge over the Ohio River
was built by the Keystone Bridge Company under specifications pre¬
pared by Mr. G. Bouscaren, the design being made by Mr. J. H. Lin¬
ville. The channel span was 519 feet, the longest truss span ever built
up to that time.
The channel spans of the Henderson Bridge, 522 feet, and of the
Ohio River Bridge, Kentucky Central Railway, at Cincinnati, 550 feet,
are the only independent truss bridges up to the present time with
greater spans.
There are now in existence on American railroads over five miles of
bridges with spans from 300 to 400 feet, nearly four miles with spans
from 400 to 500 feet, and two and a half miles of bridges with spans ex¬
ceeding 500 feet, estimated as single track bridges, all exclusive of wire
suspension bridges.
The first iron cantilever bridge of importance was the Kentucky
River Bridge designed and built by C. Shaler Smith in 1876-77. It is
20
1125 feet long, and consists of three equal spans 375 feet. The second,
also by the same engineer, was the Minnehaha Bridge over the
Mississippi, near St. Paul. It was built during the winter of
1879-80, and consists of one center span 324 feet, and two shore spans
270 feet.
The third and fourth were the Niagara and Frazer River Bridges,
designed by Mr. C. C. Schneider, the first being completed in 1883 and
the second in 1885.
The St. John River cantilever was built in 1885. Since that time
many have been built, the one of greatest magnitude being the Pough-
keepsie Bridge, built by the Union Bridge Company and already de¬
scribed in our Transactions (Vol. XVIII, June, 1888).
3. Combination Bridges.
This kind of bridge is one in which all the tension members of the
web and of the lower chord are made of wrought-iron and all or most of
the compression members of wood.
In the South and West a great number of these combination bridges
have been used. They are cheaper than bridges of all iron parts and are
more permanent than wooden bridges, the wooden members being less
liable to destruction from fire and decay, where the timber is only used in
compact forms and under compressive strains. Tliey are generally de¬
signed so that the greater part, at least, of the iron members can be used
again, when it becomes desirable to renew them as wholly iron bridges.
While more permanent than wooden, they are less so than iron bridges.
The question of the adoption of this class of structure instead of all
wood or all iron bridges, must be determined for each locality by the
relative economy as determined by their first cost and the financial
ability and policy of the railroad.
Their forms are varied according to the preference of different
engineers. We illustrate (Plates XIV and XIX) forms designed by
Mr. Alfred Fink, and which have been in general use in past years
upon many Southern railroads, especially the Louisville and Nashville
Railroad.
This road has been rapidly replacing them of late with wrought-iron
bridges of modern type. While excellent bridges can be made of this
kind, they should only be considered as an advance from wooden bridges
towards the future all-iron bridges.
21
4. Designing and Pkopoktioning.
The early wooden bridges and some of the early iron bridges were
not designed or proportioned upon any very correct theory. They were
largely empirical structures, generally of uniform section throughout their
length. Practical experience had led to improvements in methods of
counterbracing against partial loads, and in providing for better detail,
to obviate the crushing of the timber or slipping of the braces and to
take up the looseness arising from the shrinkage of the timber.
In the early days of our railroads so able an engineer as Mr. Kirkwood
writes:
'' The heavy locomotives used in the freight traffic of railroads, and
the great speed of express passenger trains, render massive strength in
all railroad bridge structures essential. The many bridges which have
been strengthened since they were built and the many which have had
to be removed and replaced by stronger structures, sufficiently attest the
mistaken economy of erecting light and cheap frame bridges. The blind
confidence which the best of carpenters are apt to have in the raw
material of their trade and in the rude geometry of its usual combina¬
tions, has rendered them unsafe guides to engineers, in the planning of
bridge trusses involving irregular and shifting strains, with which their
general practice cannot make them familiar. Hence it is that, even now,
the engineer will always find it not merely a safe but a necessary rule, to
make his bridge trusses heavier and stronger than the most intelligent
professional carpenter of his acquaintance is likely to consider necessary."
The heavy engines of that day would be very light ones at the
present time.
The science of bridge proportioning was yet undeveloped. The best
that the engineer could do, was to make the bridges stronger than
heretofore, solely on the facts brought out by past experience.
The first attempt to analyze the strains in skeleton bridge trusses
appears to be the work of Mr. Squire Whipple, the inventor of the
Whipple bridge—the prototype of the present most generally adopted
form of truss now in use in America. His book was published at Utica,
N. Y., in 1847. It does not seem to have been widely known till a much
later date. For Mr. Herman Haupt, who had an extensive experience
in building wooden railroad bridges, and whose treatise on bridge con¬
struction, published in 1851, was long an authority, states in his preface:
"If any work exists containing an exposition of a theory sufficient to
account generally for the various phenomena observed in the mutual
action of the parts of trussed combinations of wood or metals, the
author has neither seen nor heard of it." " The best works on the sub¬
ject of construction that have fallen into the hands of the author contain
but little that will furnish the means of calculating the strains upon the
timbers of a bridge truss or of determining their relative sizes."
22
While the principal material of construction was wood, its cheapness
and the practical reasons for making the principal members of the
trusses of uniform section, made the necessity of proper proportioning
less apparent. The advent of iron construction, involving the use of a
material of a more expensive character, compelled a better consideration
of the relation of the principal parts of a truss to the duties to be per¬
formed.
Besides the defective knowledge as to the calculation of the strains,
there was a like deficiency in regard to the strength of materials in the
forms adapted for skeleton structures, and as to the effects of the con¬
centrated engine loads.
The use of a uniform load of one ton per foot of track for design¬
ing all parts of a bridge was in general use some years after 18G0 and on
many bridges up to even a later date. It was not till about 1870 that a
heavier uniform load was adopted for the floor system than was used for
the main trusses.
The rapid growth of the traffic on our roads and the constantly in¬
creasing loads upon our cars and engines were not fully appreciated until
very recent times.
Up to about 187-4 the designing and the construction of bridges
were, almost exclusively, in the hands of the several bridge companies.
Each of these companies had its own peculiar style of bridge ; consist¬
ing either of a special form of truss, as the Bollman, Fink, Linville,
Post, etc., or in the use of patented forms, as the Phoenix, Linville &
Piper and American Companies' posts, and other detail parts.
Each company also had its own special geographical field, or lines of
railroad, giving it the preference.
Even at points where they did meet as competitors, it was rather as
advocates for their special trusses or forms of parts, than as competi¬
tors upon any very definite specification.
Though many excellent bridges, considering the state of the art,
were built under this system, there were also many very inadequate
structures made.
The construction of the St. Louis Bridge marked a most important
advance in the development of American bridges. The investigations
made, during the building of the St. Louis Bridge, into the strength
and other properties of the materials of construction, and especially the
testing of full-sized members and their detail connections, not only ad-
23
vanced very much our knowledge of these matters, but also gave an
impulse to such investigations which has continued to the present. The
erection of the St. Louis arches, by building out from the piers, was
the first practical solution of the cantilever principle on a large scale.
The erection of two balanced cantilevers, each over '250 feet, with ease,
safety and economy, made clear to the mind of engineers that the canti¬
lever was the economic method of erecting long spans over deep
gorges or rivers, where ordinary methods of scaffolding would be too
expensive, or subject to great risks, or where navigation forbids the
obstruction of the waterway.
The construction of the Cincinnati Southern Eailroad also marked an
important step in the improvement of bridges and their construction.
The bridges upon this road, including the large span at Cincinnati
over the Ohio River, and the crossing of the Kentucky River gorge,
comprised the first work of magnitude that was offered for competition
upon specifications drawn by an engineer acting exclusively in the
interest of the Railroad Company.
Two points in these specifications of Mr. Bouscaren were especially
important ones : 1st. The use of an engine and train diagram for the
live load, instead of the prevalent method of a fixed uniform load per
lineal foot of bridge or floor. 2d. The requirement that full-sized com¬
pression members of the different competitors be submitted to test, to
determine their ultimate strength and elastic limit.
About 1874, the bridge engineer, acting in the interest of the rail¬
roads, began to exert a beneficial influence. New competitors as bridge
contractors were also entering the field.
These influences all worked together to make a great improvement
in the general character of the structures built after that date.
The failure of the Ashtabula Bridge in December, 1876, not only
alarmed the general public, but also shook the blind confidence of the
railroad companies in their existing bridges.
Upon the principal roads a general examination was made into the
character and condition of their bridges by competent engineers with
very beneficial results. Not only were many defective structures rebuilt
or strengthened, but the defective methods heretofore in use, both in
the proportioning of parts and in constructive detail, became apparent.
Floor systems designed for a uniform rolling load, regardless of the
length of the panels, were found to be inadequate for the concentrated
24
engine loads in use. For the same reason the counterbracing was also
deficient. Broken castings emphasized the unreliability of this material
for bridge purposes. A careful examination and analysis of the details,
upon which the capacity of the structures depended as much as upon
the sectional areas of the several members, showed the need of greater
attention to such matters.
The Erie specifications drawn by the writer in 1878, embodying the
results of his experience in the designing of bridges, their shop con¬
struction, the testing of material and the study of existing bridges and
their defects, was the first general specification covering the designing,
proportioning and detail of construction with that completeness neces¬
sary to give the railroad company the full advantage of the competitive
method, with a certainty that the resulting structure would in all ways
be up to the advanced state of the art.
It was the first paper on bridge construction in which that relic of
ignorance, "the factor of safety," was entirely omitted.
It definitely specified the working strains to be allowed on the differ¬
ent parts of the structure according to the service they were to per¬
form.
While aiming to obtain the best results for the railroad company, by
clearly specifying for what loads and under what working strains each
part should be proportioned, and the perfection of workmanship which
must be adhered to, it left to the contractor perfect liberty of selecting
his own plan of structure, as to form of truss, relative proportions of
depths and panel lengths and methods of detailing.
These specifications were adopted very widely. They have been modi¬
fied and extended from time to time, by the author and other engineers,
as advancing knowledge and experience justified.
The competitive struggle of the bridge contractors has still been
allowed free scope in certain directions, but has been modified by the
new factor, the bridge engineer acting in the interest of the railroad
companies.
The bridge engineer acting for the contractor has for his principal
incentive, the selection of a design which, while complying with the
specifications in all ways, can be built at the shops for the least money ex¬
pended in time and material, and which can be erected at the particular
location with the least cost and risk.
The bridge engineer acting for the railroad company, while giving
all due weight to the aims of the contractor, is seeking to get a structure
which will give the best results after its completion, for operation and
maintenance, without undue cost.
As the engineer of the contractor may at some other time be the
engineer acting for the railroad and the reverse, these different views
are constantly reacting on each other, to the great benefit of our bridge
structures.
The competition to-day between the different bridge companies has
been largely reduced to the question of shop management, or the
relative cost of turning out so many tons of bridge work in a certain
limited time.
Practical bridge engineers have given little weight to the theo¬
retical examination of the economical truss. They have recognized as
fallacious the deductions usually drawn by theorists from the neglect
of most important factors in such investigations. True economy does
not necessarily mean minimum weight of material. The relative weight
of the various classes of iron in a bridge, the cost of manufacturing each
form, rolled, riveted or forged, facility of transportation and erection
and other factors must be given their due weight; and even these factors
must change from time to time, and do vary even for the same time at
different manufactories. While the tendency towards economy of
material due to changes in forms and proportions has been fully
recognized, the comparison of alternate or competitive plans has
led to the proper proportions of true economy with far greater
certainty.
The engineer who is constantly employed in preparing competitive
designs, can choose almost instinctively the proper proportions to give
his design, selecting the panel lengths, depth of truss and skeleton system
best suited to give the proper relations between the floor, web and
chord systems of the bridge, with due regard to a suitable arrange¬
ment for the lateral and sway bracing.
No fixed proportions can be permanently established suitable for all
cases; for in addition to the varying elements before mentioned, the
question of head room, waterway and skew often necessitate special
changes. There must also be a certain undefined harmony in the
relation of the several parts, in order that the disproportion or
inefficiency of the connection of parts may not render the design a
defective one.
26
A successful bridge engineer, from the American point of view, must
be something more than a mere calculator of strains. That is the most
elementary part of the duty, and does not come within the province of
designing. After the selection of the skeleton form and relative pro¬
portions of panels, depths and widths of span, a very moderate knowl¬
edge of mechanical mathematics would enable any one to determine the
strains in an American bridge. He must, in addition to his knowledge
as to the effects of varying forms and proportions, have a full knowl¬
edge of the capacity of his forms and their connections, and also of the
practical processes of manufacture and erection. He must know how his
design can be made and put together, and whether it is so harmonized
in all its parts and connections, that each part may do its full duty under
all possible conditions of service.
In addition to knowing all the elements that make up a perfect
design, he must have the instinct of designing or the power of adapting
his knowledge to any individual case, in order to obtain the best or the
desired result.
Then experience, observation and a sharp competition with men of
like knowledge and instinct will give him his position as a bridge
engineer.
The equivalent uniform loads given in the following tables (1 and 2),
represent in a general way the live loads used in designing bridges dur¬
ing the past, and the maximum live load now being adopted by certain
of our railroads.
/
I o CDCDC
") (1
!w
y
>
\L
6 ,6' 6
n (T)0
1 6' *' : f
o o 1 t**1 fat-
»£ PI I 1 « !
1 1 1 ! ! M I
»'< *> r.'i f.'r
Mill
Fig. 1.
1> »■'< f-! - *'/
i ! 1 f !
4/. T.'t *'/ i-T . 4./
ML
CDCYYt
) m (
<
CD CM
xb o <1
| it*0 &4.JU€ Cm.jvct.
* I 1 I I ! | I 1
v § § \
$ W £ c
1 1 ! 1 1
Fig. 2.
27
TABLE No. 1.
Fbom Moments—Loads fob Longitudinal Girders.
Spans.
Feet.
Uniform Load
of One Ton
per foot.
Cincinnati
Southern
Train Load,
1873.
Fig. 1.
Erie Train
Load.
1878.
Fig. 2.
101 Ton Consoli'
dation, followed
by 3 000 pounds.
1884.
Fig. 3.
Heavy Grade
Lehigh, follow
ed by 4 000
pounds.
1889.
Fig. 4.
5
2 240
8 000
8 800
12 000
10 000
10
4 790
5 290
6 750
9 610
15
ii
4 470
5 280 ,
6 670
9 600
20
n
4 320
4 640
6 190
8 560
25
ii
3 970
4 400
5 850
7 950
30
II
3 770
4 200
6 470
7 290
35
<<
3 540
4 080
5 110
7 040
40
u
3 400
3 850
4 760
6 690
45
3 230
3 700
4 550
6 000
fO
<<
3 140
3 590
4 350
5 870
55
II
3 050
3 450
4 220
5 680
60
«<
2 960
3 400
4 110
5 480
70
II
2 920
3 220
3 940
5 110
80
n
2 920
3 260
3 950
4 800
90
«i
2 860
3 250
3 970
4 560
100
II
2 860
3 230
3 910
4 520
125
K
2 780
3 150
3 830
4 740
150
ii
2 670
3 070
3 740
4 760
175
ii
2 590
2 990
3 650
4 670
200
II
2 530
2 910
3 590
4 640
225
II
2 440
2 850
3 520
4 550
250
II
2 390
2 800
3 480
4 470
275
«i
2 360
2 750
3 430
4 380
300
II
2 340
2 710
3 390
4 290
350
II
2 240
2 660
3 230
4 250
400
II
2 200
2 610
3 180
4 200
450
II
2160
2 570
3 160
4 160
500
i<
2 120
2 540
3 150
4 120
CD(T)(t)CD
ft) (I) <h th /A
.T r I' r i r r'1
CXXbCD f> f> o n
& & S
I i
Fig. 3.
s 1 - «
I § I
*">
f,< -It.
k i k. 'j Vt <.'/
S.7~* k$\ S.'i i.'f 5 's k>. «>
v,
i (be
r
IT
) J
i r
) <
) c
/
!> A
> (
be1
xlx
") r
) c
> (
> c
,
i^ <^«2 yOtf/
§ § § S ? § £ § 2 £
S S § § 1 % t % $ I
Fig. 4.
TABLE No. 2.
Equivalent Loads for Erie Train Load.
Span—Feet.
From Moments.
From Double Shear.
From Single Shear.
5
8 800
5 280
9 680
10
5 290
4 620
7 200
15
5 280
4 210
6 240
20
4 640
3 830
5 400
25
4 400
3 588
5 040
30
4 200
3 400
4 740
35
4 080
3 270
4 540
to
3 850
3 175
4 300
45
3 700
3 090
4 220
50
3 590
3 030
4 240
55
3 450
3 000
3 930
60
3 400
2 970
3 830
70
3 220
2 960
3 680
80
3 260
2 870
3 620
90
3 250
2 820
3 570
100
3 230
2 750
3 450
125
3 150
2 570
3 280
150
3 070
2 450
3 180
175
2 990
2 400
3 090
200
2 910
2 370
3 000
Table No. 1 gives the approximate equivalent uniform loads for the
moments due to several forms of typical live loads upon spans of dif¬
ferent lengths. The approximate dates on which these loads were in¬
troduced are also given.
These equivalent loads represent in a general way the growth in the
required strength of our best bridges, from the earlier days of iron bridge
building to the present time.
This table will make clearer than any other method of description the
difficulties American bridge builders have had to contend with. In face
of the constant progress and bettering of our bridges we have had to
meet rapidly increasing demands for greater loads; with too often an
increasing reluctance on the part of the authorities to build for anything
but the immediate present requirements.
It will be readily seen from this table why our shorter spans are being
overtaxed, and why in all our older bridges the floor and counter sys¬
tems are too weak. The above table represents simply the equivalent
loads, which will render the same maximum moments as the correspond¬
ing train loads. To provide for the maximum action of any miscel¬
laneous loading, we must also consider the equivalent loads, which rep-
29
resent the maximum shears. For this purpose we need two additional
equivalent loads—one to represent the maximum end-shear upon a
single span of girders or trusses, which we will call the "single shear;"
and another to represent the maximum shear which can occur at the
supports of two adjacent spans or panels of girders or trusses, which we
will call the " double shear."
The single shear equivalent loads give us the loads on the end mem¬
bers of the trusses and girders. The double shear equivalent loads give
us the loads on the supports, whether they be columns or cross-girders.
Table No. 2 gives for one engine the equivalent loads corresponding
to the maximum moments, and maximum double and single shears.
They differ from one another for each different span of truss or panel.
"We must therefore use at least three series of uniform loads to represent
even approximately one engine load.
The maximum moment of a miscellaneous loading being generally at
some other point than the center of the girders, the equivalent load
giving the maximum moment will be different for a girder with uniform
flanges and for one with flanges of uniform strength.
To represent the action of a train load properly by this method, we
should need four different series of equivalent loads, all varying for
each span.
The apparent simplicity, therefore, of using equivalent uniform loads
for proportioning our structures, is a fallacious one; when applied to
partial loads it becomes far more confusing and untrustworthy.
The old method of computing the strains by use of panel loads and
equivalent uniform loads (and too often using only the one derived from
the moments) has gradually given way to a system much simpler, more
accurate and also as easy of application.
The principles and application of the new method were worked out
independently but simultaneously by Mr. Robert Escobar, C. E., of the
Union Bridge Company, and the author, some years ago (1880).
\
\
\
Y§ l s
l\ s i
S \ 1 8.
- I\I v 1
«\s § 8
1 \l I
5 Y &
i i ?
5 5 §
b
\c
5 § V t
1 I C\ i
•c $ \ t
)000\
! i i\t
I S t; \S
r> r> n \
1 ! IM
§ a s \
OOOOs.
| | I
ion
i * \a
Fig. 5.
30
Let Fig. 5 represent a skeleton truss of a single system of triangula-
tion and n panels.
The maximum shear at any point a will occur, when one wheel of any
miscellaneous series is on the panel point, and when the total weight of
train on the bridge, IF and n times the loads in front of a (not including
the one at a), are the nearest to an equality; or, in other words, calling
P the loads in front of a, when TF — n P is a minimum.
Similarly the maximum moment at any point a will occur when the
train is so place ! that TF : P : : L : y; L being length of span.
Application of the method: Lay down the engine and train load
on a card b to the same scale as is used for the diagram of the truss. We
will assume that the train is headed to the left hand. Commencing at
the left hand, write over each wheel the summation of the weights up to
that point ; on another line above write the moments of all the wheels
to the left of the particular wheel about that wheel.
The first line will give the weights P and TF for any position of the
train, so the placing the diagram to obtain TF — n P = a minimum, or
TF y — P L is readily done.
The second line above will in like manner give the moments (m P)
or (m TF) or the moments of the loads about a, or the end of the span.
Where the last moment does not correspond to the end of the span, it
must be increased by TF X d, the distance from the last wheel to the
end of the span.
m, • 1 i I \ -11 il 1 (m TF) ,l (m P)
The maximum shear at (a) will then be: ^
The maximum moment at (a) (different position of train) will be—
(m TF) , _
1—j—'v — (m P)
An absolutely mathematically correct result can thus be obtained
in a very few minutes, after a little practice. The preparation of
a diagram for any engine and train load is only the work of a few
minutes' time.
The method has therefore the merit of being no longer than that by
uniform loads, and is far more exact, for it gives perfect accuracy for
all positions of the loads, without any of those allowances for excess, etc.,
which are needed in the old methods.
This method has never been published before, but is in general
use among the American bridge companies, by gradual transfers. The
method cannot be applied to multiple system trusses, but it can be to
girders with a continuous web.
31
The advocates of the method by uniform loads make the claim that
from the rapid increase of the loads upon our cars we are fast approaching
the time when the loads on the wheels of our cars will equal the loads
on our engine wheels, totally forgetting that the distance between the
wheels is as important an element as the weights. As long as the engine
drags the train by the adhesion of its wheels, its weights and their con¬
centration on a short wheel base must produce effects that will always
exceed those produced by the car loads.
The problem before us, therefore, is not the selection of some
maximum uniform load which will give us imperfectly proportioned
structures; but how far can the designers of engines go in their loads
and spacings of the driving wheels? Is there any maximum, or can they
go on indefinitely? The want of harmony or rather the independence
of the different departments of our railroad organization, has left this
important question to be solved by the tentative system, rather than by
any scientific study or investigation. That all of the roads in our vast
system should adopt one maximum train load for designing their struc-
tures^ even could such a maximum be decided upon, cannot be expected.
The peculiarities of each road as to grade, alignment, traffic and
financial strength must lead to variations.
That every road should build the best and strongest bridges that a
sound and far sighted financial management will admit, is true. But
we cannot expect more than this. Bridges, therefore, will continue to
be built of various capacities. Our duty as engineers, is to strive to ad¬
vance their capacity as far ahead of present needs as the funds available
will permit.
"We may accomplish this purpose best by bearing in mind certain
principles:
First.—By selecting for our live loads typical train loads, where
the engine loads will bear a reasonable relation to the following train
load. The growth of the traffic and development of the train loads will
then be more likely to increase the strains on all parts of our structures
proportionately, and they, therefore, can be used till they have reached
their maximum justifiable duty; instead of removing them, as is now so
frequently the case, because the floors, counters or detail parts are over¬
worked.
A recently constructed bridge is known to the author, which was
designed for a very heavy load, so badly arranged that parts of the
32
structure are only of a capacity equal to the present service of the road,
■while other parts have double this capacity. The material in this bridge,
properly arranged, would have given a structure at least fifty per cent,
stronger for future traffic, than it can now do.
Second.—By not increasing our unit strains in proportion to our
faith in new forms and new material, where necessity does not compel us.
To-day we are able to get some steel forms at a less price than iron
ones, and some others at the same price. Before long we will get all
steel cheaper than iron. Why not take advantage, when we can, of this
advance towards a stronger material to increase the capacity of our
bridges? Why endeavor to push the strains up to the limits of our faith
in its capacity for the various forms?
Camber.—The desired camber is obtained by making the length of
the top chord panels to exceed those of the lower chord by an amount
slightly in excess of the sum of the compression and elongation of the
top and bottom chord panels respectively, under the working strains.
For strains in the top chord of 8 000 lbs. per square inch and
10 000 lbs. for the tension of the lower chord and a modulus of elasticity
of 24 000 000 lbs., the sum of the changes in the two chords would be
18 000 _ 1
24 0001)00 1 333.
This amount of increase would, however, only provide for the deflec¬
tion due to the chords. To provide for the additional deflection due to
the web system of the trusses, and to allow some excess, this amount is
usually increased one-third, which makes j-qqq> which corresponds very
closely to "i inch for every 10 feet of panel."
This becomes a very convenient rule for the shops where English
measures are employed. It is entirely independent of the depth of the
truss and only requires modification when the unit strains differ from
the above averages.
Lateral and Transverse Bracings.—The general practice in America
is to brace our trusses for a definite amount of lateral force per lineal
foot of the structure, instead of assigning so many pounds of wind
pressure per square foot of exposed surface.
This has arisen from two causes: 1st. The indefiniteness of the
methods of estimating the exposed area, thus giving unfair advantages
to the less scrupulous competitor. 2d. The recognition by practical
33
engineers that there are other forces requiring lateral resistance, even in
districts that may be windless.
The rule formulated for the Erie specifications in 1878 and now
generally adopted, was as follows:
"To provide for wind strains and vibrations, the top lateral
bracing in deck bridges and the bottom lateral bracing in through
bridges shall be proportioned to resist a lateral force of 450 pounds for
each foot of the span, 300 pounds of this to be treated as a moving
load.
"The bottom lateral bracing in deck bridges and the top lateral
bracing in through bridges shall be proportional to resist a lateral force
of 150 pounds for each foot of the span."
"In no case shall any lateral or diagonal rod have a less area than I
of a square inch."
While the above rules correspond approximately to 30 pounds pres¬
sure on the projected surface of a train of cars and the trusses, it was
only selected after comparing the results with existing bridges up to
200 feet spans, which gave satisfactory lateral action under rapidly
moving trains.
5. Strength of Material and Parts of Skeleton Structures.
In the early days of iron bridge building, the knowledge of the
strength of materials was very limited. The early experiments of
Fairbairn and Hodgkinson comprised about the extent of our knowl¬
edge of the strength of iron.
Crude tests made upon grooved specimens of a small size, lead to
much misconception of the capabilities of American bar iron. In a
pamphlet by Mr. Wendell Bowman on his Harper's Ferry Bridge, built
in 1851-52, he states
"the tensile strength of the best American bar iron tables as 80 000
pounds per square inch. Its practical value is generally rated at about
one-fourth the nominal value. In the diagram (his strain sheet) the
highest value of iron is 16 000 pounds, being reduced below any prob¬
able rate of fibral separation in any previous data."
Even as late as 1870, Captain Eads was assured by the manufacturers
of bar iron that there was no difficulty in furnishing bars of any size,
capable of standing 60 000 pounds per square inch, and he made the
contract for the St. Louis Bridge with this expectation, only to be dis¬
appointed.
In 1878, when the requirements for the strength of bar iron as con¬
tained in the Erie specifications, making a reduction in the ultimate
strength as the size of the bars increased, was submitted to the most
u
experienced iron masters for their criticism, they were condemned by
all with but one exception, Mr. Andrew Ivloman, of Pittsburgh: 1st, as
being entirely too low in tensile requirement for good bar iron; and 2d,
in making any allowance for increased section of the bars; " there
being no reason why the fibers in large bars should not be as strong as
in the smaller."
It may be unnecessary to state, that when these specifications were
enforced, and the acceptance of the material became dependent upon
the test made on either the bars themselves or specimens cut from
the same, it was found that only the best and highest priced bar iron
could meet the requirements.
The general tendency, since then, has been to relax them somewhat
for a broader competition.
The Phoenix Iron Company and the Keystone Bridge Company had
crude testing machines at their shops previous to 1870.
They were used to test the strength of eye-bars and wrought-iron
columns. In some of the more important structures, all the eye-bars
were tested up to a strain of 20 000 pounds per square inch.
"While these machines were crude, they did serve to develop the de¬
tail proportions of the eye-bar and column used by these companies.
The use of steel and new forms of members in the St. Louis Bridge,
in connection with the importance of the structure and the yet untried
method of erecting as a cantilever, impelled the engineer and the con¬
tractors to unusual efforts to determine the capacity of the material and
of the forms to be used.
Of the many thousand tests made during the "construction of this
work, but a limited number have ever been published; the general de¬
ductions, however, soon passed into accepted doctrine, and developed a
keener desire for a more refined knowledge of the influence of form and
proportion upon all our bridge members.
The eye-bar had been well developed under the crude tests previ¬
ously made by the early bridge companies.
Pull sized tests had also been made on a limited number of wrought-
iron columns. But no marked progress was made in the forms of com¬
pression members, till after the tests of Mr. Bouscaren on the forms
then in use. His tests not only showed the inapplicability of Gordon's
formula, but also the possibility of much better results by improving
the detail of the open and box columns; the only forms of columns
35
without a proprietary claim and the ones best adapted to the American
style of bridges; as they gave ready facilities for connections without
the use of cast-iron.
From this time forward rapid progress has been made in the detail
and proportion of all our full sized bridge members.
To-day every first-class bridge manufactory has its complements of
testing machines, to test with all the refinements either samples of the
material or full sized members in compression, tension or transverse
strain. Our knowledge of the strength and capabilities of our material
and of the usual forms employed in the American style of bridge is
such, that no first-class bridge company in America hesitates to accept
the clause now general in all specifications, that " full sized members
may be tested to destruction," with the sole proviso that the expense of
testing and cost of the piece shall be paid for by the purchaser if it
satisfies the requirements of the usual specifications.
This positive knowledge of the capacity of our full sized members
marks one of the great advantages of our system of bridges over all
others. Our "factor of ignorance" has been reduced to this extent;
a no mean portion. ,
We, therefore, have a right to claim, that as our working strains are
as low and in many cases lower than those used in Europe, with our
more perfect knowledge of the strength of our members, we have in our
first-class structures a greater factor of security than prevails in
European bridges.
In the appendix will be found tables of the abstracts of the more
recent tests on eye-bars and compression members.
6. Manufacture of Bridges.
The great demand for iron bridges in a country as vast as the
United States caused the formation of companies for the special pur¬
pose of manufacturing and erecting bridge structures. They estab¬
lished special plant and created a corps of engineers, who have made
this branch of engineering a special one.
There are to-day in America more than forty bridge building com¬
panies manufacturing railroad and highway bridges.
Of these, at least a dozen are capable of constructing bridges of every
size in a first-class manner.
The shops, which can be called especially shops capable of construct-
36
ing the largest class of railroad or highway bridges, are capable of turn¬
ing out by the year 125 000 tons of bridge work.
It would be safe to estimate that all the shops could turn out
200 000 tons, or approximately 80 miles of 100 feet spans of single track
railroad bridges, per year, if their plant were devoted exclusively to this
work. Other iron work, as roofs, iron buildings, piers, elevated rail¬
roads, etc., however, make a considerable figure in their yearly output.
Some few of the bridge shops have been constructed to do riveted work
almost exclusively.
The typical American bridge shops are, however, fitted to do any class
of bridge, girder or roof work, whether it be exclusively riveted, or com¬
bined riveted and pin-connected work. Each company has, therefore,
the following arrangements for receiving the iron, and putting it through
all the processes to the completion ready for shipment:
First.—Receiving yard, where the iron for each bridge is properly
classified and stored.
Second.—Department for straightening, where pieces can be straight¬
ened with more accuracy than can be obtained directly from the rolling
mills.
Third.—Template and pattern shop, for preparing the templates for
the rivet and pin holes and crude shapes and dimensions of all pieces,
with the proper allowance for final tool finish.
Fourth.—Laying out shops, where each individual piece of iron is
carefully marked in accordance with the templates.
Fifth.—Punch and shear shop, where all the iron is punched and
sheared.
Sixth.— Fitting up shop, where all the iron for riveted members is
assembled and bolted together ready for the riveting machines.
Seventh.—Riveting shop, with its proper complement of air, steam
and hydraulic riveting machines.
Eighth.—Machine shop for planing, boring, turning, etc., to com¬
plete the finished bearing surfaces of all the members.
Ninth.—The upsetting and forge shops for making eye-bars and all
forged parts required.
Tenth.—Painting and shipping sheds and yard.
The aim in the construction is to pass the material from the time of
its receipt from the rolling mill to its final shipment through the neces¬
sary steps with as little waste labor in handling as possible and to per¬
form all work by machinery in preference to hand labor.
37
While the laying out and riveting are done with all care and accuracy,
the lengths of the members and the fitting of the same together do not
depend upon the accuracy or neglect of these processes.
The length of all abutting members and distances between centers of
pin holes are determined finally at the machines for planing the abutting
ends or for boring the pin holes.
Machines for operating on each end of such members are provided
with iron beds and extremely accurate methods of setting the same to
any required distances.
Such machines once set and operated with proper precautions in
regard to uniform temperature guarantee great accuracy of duplication
of all similar parts.
It is sometimes claimed that there is an error in lengths of parts due
to the usual allowances for play of the pins in the pin holes, varying
from tr4 to 3~2 of an inch, according to the size of the pins. This is an
error; this allowance is provided for by either measuring from out to out
of the pin holes for tension members and from the inside of the pin holes
for compression members, or by taking it into consideration where the
lengths are given from center to center.
The surety of the fitting of the members of even the largest structures
after they have passed through a properly organized bridge shop is such
that no assembling of the finished members of a structure is ever made
at the shops, except in extremely crooked and complicated structures,
and even then not as a whole, but only sufficient to test the fitting of
specially intricate connections.
The rapidity of the erection of our structures and the satisfactory
manner in which they come together in the field without any tool
work prove the certainty of the American methods.
A pin connected truss is composed of the following class of members:
1. Top chord sections.
2. End posts.
3. Intermediate posts.
4. Pedestals.
5. Lower chord bars.
6. Diagonal web bars.
7. Pins.
The top chord sections and end posts are usually similar in form. They
are usually formed of two or more vertical channels, either of rolled
forms, or built of angles and plates, connected on top with a cover plate
and on the bottom with lattice bars.
3S
In the best practice most of the material for these members is con¬
centrated in the vertical channels; the top plate besides acting as a
cover and stiffener during transportation, gives an unbalanced section
sufficient to counteract the tendency to deflect due to the weight of the
piece. The pin holes are reinforced by additional plates to reduce the
bearing pressure to the allowed limits. The amount of rivets in these
pin plates and at the ends of the sections is made large enough to dis¬
tribute the localized pin pressure over the section of the whole chord.
The chords are spliced at one side of the pin hole to insure a full pin
hole for more convenient boring and to enable the rivets in the splice
to be driven after the parts are assembled.
After these chords are riveted together in the shop the abutting ends
are carefully tooled to square surfaces and exact lengths and the pin
holes bored. The accuracy of the finished lengths and diameter of pin
holes and the character of the tool work having been passed upon by
the inspector, the piece is marked, painted and stored for shipment.
All posts are made with pin bearings only. The intermediate posts
are usually made of the open form, consisting of two channels, either
rolled or built up of plates and angles, latticed on both sides. The sizes
of the end or batten plates and the lattice bars have been evolved from
experience in transportation and from the study of the results of full
sized tests.
The ends of the posts are usually forked (see Fig. R, Plate XXVII),
in order to pack between them the tension bars to produce a more
compact joint.
These ends are stiffened by thickening with plates of sufficient length
to make the compressive strength of these ends fully up to the capacity
of other parts of the same post.
The pedestals are riveted members formed of plates and angles. They
are proportioned for the proper bearing pressure and with rivets suffi¬
cient to meet all the strains. Their base is determined by the allowed
pressure upon the masonry.
Under the pedestals of one end of the span, nests of friction rollers
proportioned according to an accepted rule are placed, to provide for the
expansion and contraction of the trusses.
Lover chord bars.—These are flat bars with forged eyes at each
end. These bars have, in the past, been made by various methods.
But economy in their manufacture has developed special upsetting
39
and die forging machines of great power, using either steam or
hydraulic pressure, by which they can be made with great certainty and
at a very moderate cost. Only by such perfected machinery and great
experience in the working could such accurate results as are demanded
by our ordinary specifications be attained. The center of the heads
must be in the center line of the bar and the heads symmetrical to this
line. The heads must be clean, smooth and of proper thickness. The
necks must be of proper shape and upset so as to be full in all dimen¬
sions. The lengths between centers of the heads must be very close to
correct length or when bored there will not be the proper excess for the
eye. To produce such bars economically all reheating and reworking
to correct errors must be avoided.
The process, of course, requires the best material, whether of iron
or steel.
All steel bars are annealed after they have been forged.
Our summary of some recent tests made in the ordinary course of
the execution of contracts shows how perfectly these bars answer to the
full capacity of the material.
Web-bars.—These are similar to the lower chord bars, except the
counter-rods, which are made in two pieces and connected by sleeve nuts
or turnbuckles. It is the best practice to secure these turnbuckles or
sleeve nuts firmly in their position after these rods have been properly
adjusted.
Pins.—These are made of round iron or steel, turned perfectly
smooth, and to accurate diameters, with an allowance of ia inch for pins
under inches diameter, and 3V inch for larger pins, less diameter than
the bore of the pin holes. These pins have wrought-iron nuts at each
end, which are tightly screwed up and secured after the bridge is con¬
nected together.
The Iron Floor System of our bridges consists of cross floor beams at
each panel point, with a pair of stringers under each track. These floor
beams and stringers are usually plate girders. This system is connected
to the trusses usually in one of two ways ; either by suspending from
the pins or by riveting to the vertical posts for through bridges ; or
for Deck Bridges, either resting on the top chords at the panel points,
or by riveting to the posts. Each particular case must be worked out to
accomplish the purpose best, considering the local requirements of
clearance for floods or navigation, head room, etc.
40
Lateral and Sway Bracing. — The horizontal and vertical bracing
between the trusses have generally in the past been similar to the other
characteristics of our type of bridge, rectangular systems of tension
rods and compression struts connecting to the pins of the main trusses.
Of late years the preference for more rigid forms for the parts of these
comparatively light trusses has led to the use of angle iron bracing
instead of tension rods.
7. Erection op Bridges.
The great risk involved in erecting important structures over rivers
subject to sudden floods, ice jams and similar dangers, emphasizes the
importance of having a style of structure which can be rapidly and
surely assembled without detriment to the perfection of the workman¬
ship or accuracy of the connections.
The American pin-connected bridge is especially suited for this pur¬
pose, the connection of the comparatively few compact members being
readily completed after they have been hoisted into position by the
driving into place the connecting pin and securing the same by setting
up the nuts.
The only rivets to be driven in the main trusses are usually those in
the plates connecting the top chord and end post sections. But these
do not in any manner affect the perfect action of these members, which
depends entirely upon the machine surfaces of the abutting ends and
the turned pins. They serve solely to insure the bearing surfaces from
side displacements through any accidental blows. They can be con¬
nected by bolts, or the rivets may be driven at any time, even after the
bridge is in use for trains.
The rivets in the floor system can likewise be driven after the bridge
is free from the false work.
The erection, therefore, of a pin-connected bridge is very readily
performed, and with men experienced in this class of work with mar¬
velous rapidity.
The importance of the erection with rapidity compels a careful
design for the detail of all connections, and also often gives weight in
favor of a special style over all others.
The additional security of the cantilever method of erection favors
that form of bridge for long spans over deep gorges or rivers where the
risks are unusually great.
41
The selection of a cantilever style of bridge for spans less than 500
feet is entirely unjustified, except where the special merit of this
method of erection outweighs the objectionable features of the system—
excessive deflection and reversal of strains.
The rapidity with which our pin-connected trusses can be assembled
and swung clear of the supporting false works has been so often
demonstrated, it would be useless to enumerate examples.
A 250-foot span railroad bridge has been erected in sixteen working
hours, taking the material from the storage yard 1 000 feet from the
bridge, without any emergency demanding unusual exertions.
It would be no exaggeration to assert that any span up to 250 feet
could be erected within the limits of one day-light, so as to be self-sus¬
taining and independent of all risks from loss of the false works.
The erection of the two channel spans of the Cairo bridge is an
example of rapid erection, which illustrates the possibilities of the
American system of construction.
These spans were each 518 feet 6 inches center to center of end pins;
61 feet deep center to center; 25 feet wide center to center of trusses;
panel lengths 30 feet 5} inches. Total weight of one span, 2 055 200
pounds.
The first span was erected in six days. After this span was erected,
the false works were taken down, the supporting piles drawn and
redriven for the second span, the false works again put up and the
second span erected.
The whole time covering the erection of the two spans and moving
the false works was one month and three days, including five days lost
time, waiting for the completion of the masonry.
The false works and traveler were in position ready to commence the
erection of the second span by 2.30 p. m., October 30th, 1888. At 2.50
p.m., November 3d, the trusses of this span and the top bracing were
all connected. No work was done at night. The material was run on
trucks about 1 025 feet from the storage yard to the nearest end of the
span being erected. About twenty-four men were employed in deliver¬
ing the material and fifty in erecting and connecting together.
The false works stand about 104 above low water. The bents being
72 feet high above the capping of the piles. The depth of water at its
low stage is about 20 feet. The piles were from 50 to 75 feet long.
The erection of these spans was done by William Baird & Co., sub-
42
contractors, under the Union Bridge Company, for this part of the
work. Plates XX to XXIV are reproductions of photographs taken
every twenty-four hours to show the daily progress of the erection
of this second span.
8. Typical American Railroad Bridges and their Relative Merits.
While the pin-connected bridge is recognized as the one which
can be especially called the American type, American bridge engi¬
neers have not failed to "search all things and hold fast that which
is good," not only within our own experience but within that of
European practice. We have had one great advantage over our
European brothers, that their works and practice are so much
prompter and fuller given in the technical literature of the day.
We have had another advantage from our allying the several branches
of designing, manufacturing and erecting more intimately than is usual
in European countries. We have been less liable, therefore, to give
undue weight to the claims of theorists or the riders of hobbies. Nearly
every imaginable form of structure has been tried in the past by some
enthusiast.
After an experience unequaled for its variety and extent, we are ar¬
riving at a very general uniformity in the styles of bridges to be adopted
for different spans.
Ten years ago, of a thousand bridges of which no two would be
alike, a practical bridge engineer could almost invariably determine the
designer and manufacturer, either by the general style of the bridge or
by the use of some peculiar forms of member or detail.
To-day the use of a peculiar form of truss, member or detail is due
more to the special necessities of the case than to the idiosyncrasies of
any particular individuals.
While no two bridges to-day are exactly alike, there is a general
approximation to certain broad types in styles and detail.
Plate Girders.—These are very generally used for spans up to 65
feet, or lengths which do not require more than two ordinary flat cars 33
feet long to transport from the shops to the bridge site.
Some railroad companies have extended their maximum length for
plate girders to three car lengths, or about 100 feet.
These girders are riveted completely at the shops. Plate girders
completed at the shops in this manner give excellent satisfaction. For
43
equivalent strengths, they are cheaper up to at least 65 feet spans than
lattice girders. Their maintenance is also less. Their relative security
is far greater ; one per cent, of faulty rivets will make a much greater re¬
duction of strength in a lattice girder than in a plate girder to do the
same duty. Plate girders are clepner than lattice girders, being free
from the numerous recesses and corners peculiar to this latter type,
and hence less exposed to the accumulation of dirt, moisture and their
accompanying oxidation.
Lattice Girders.—Riveted lattice girders have been used quite gen¬
erally on all our railroads for short spans and on certain lines of railroad
for all spans. For the shorter spans, as before stated, preference is
now strongly in favor of plate girders. For the larger spans the
necessity of performing so much riveting of important connections at the
bridge site, where that care and accuracy attainable at the shops cannot
be depended upon, does not render them acceptable to many engineers.
And the necessarily increased risks and cost of the erection lead most
manufacturers to prefer some other style of bridge, except for special
cases.
While lattice bridges have given good satisfaction where they have
been well proportioned, they have no merits which cannot be also had
from our pin-connected types. They have not, therefore, made any
progress for the longer spans.
On most of our roads they are, now generally limited to the spans
between the longest plate girders and the minimum pin-connected
spans; a limit not definitely fixed, the lower limit being determined
by the relative cost of plate and lattice girders and the ability of the
roads to transport long plate girders ; and the upper limit by the per¬
sonal preference of the engineer. These limits may be broadly stated
as somewhere between 60 and 120 feet. The riveted form of connection
has an advantage over the pin form for the lower spans, to be deter¬
mined by the circumstances of each case, from the additional stiffness
of the connections and the form of the members. This is partially
attained by some designers by using stiff members for both tension and
compression in their short pin-connected trusses. It would be possible
to make an equally stiff pin-connected bridge for these lower spans, and
at the same time attain that perfection of fitting and lengths of parts
peculiar to this form of truss, if those connections, at which reversal of
strains occur, were also rigidly bolted or riveted together. This
44
*
might be objectionable to the theorists who object to secondary strains.
The author does not consider these, if restrained within proper bounds,
as so objectionable ; pin-connected trusses cannot be considered as free
from them, as long as there exists such a thing as friction.
Pin-connected Spans.—With the exception of not over a few hundred
spans, the longest of which are 2G0 feet (lattice bridges), the iron rail¬
road bridges of the United States for spans over 100 feet, amounting to
about 7 000 spans, aggregating 210 miles in length, are of the pin-con¬
necting type.
The American people are of too practical a character to have built
this quantity of bridges of a peculiar type, with the knowledge of other
forms in use, both here and in European countries, and after thirty
years' experience in their use, to still adhere to them, if their merits
were not strong and positive ones.
Through all the changes of styles of trusses and forms of parts, the
connection of the members by means of inns, and the attaining accu¬
racy of lengths and fittings by machining the joints, has been persistent.
Forms of trusses with more than a single system of triangulation
have gradually been rejected, and have now disappeared from first-class
bridge designing.
Such subdivisions of the trusses as will give the greatest concentra¬
tion of parts, with long panels, and without any theoretical or practical
ambiguity of strain under a miscellaneous series of wheel loads, are
now only adopted.
Forms of members requiring cast-iron joint boxes or other uses of
this material have become obsolete.
The old forms, like the Bollman, Fink, Lowtliorp and Post trusses
(see Plates XII, XIII, XIY, XV, XXV, XXVI), have disappeared from
American practice. The double intersection Whipple or Linville is
rapidly following tliem. Generally bridges are made with parallel
chords and equal panels; this gives the minimum number of different
lengths of parts, which leads to greater economy and accuracy of
manufacturing. For the longer spans, the depth may be reduced with
advantage at the end panels.
The longest pin-connected truss span in existence is the recently
completed channel sp>an of the Ohio River Bridge at Cincinnati,
545 feet from center to center of end pins, 84 feet deep at the
center and 60 feet at the end posts; panels, 27 feet l£ inches; trusses, 30
45
feet apart, center to center. It carries a double-track railroad between
the trusses, and has on each side a wagon-way and foot-walk, 16 feet wide.
The persistence of the pin-conneeted type of structure, now that the
bridge engineer acting for the railroad companies has become an im¬
portant factor in the problem of bridge construction, shows that it is
not solely due to the preferences of the manufacturers; but that the
operation and maintenance of such structures are also in favor of this
type.
The typical American railroad bridge is a skeleton structure, pin-con¬
nected at all the principal articulations. Its essential characteristics
in addition to being connected by pins are :
First.—So formed as to reduce all ambiguity of strains to a mini¬
mum ;
Second.—Concentration of parts ;
Third.—Facility of manufacture ;
Fourth.—Perfection of lengths and fitting of all the members, so as
to reduce to a minimum all riveting or mechanical work in the field ;
Fifth.—Readiness with which the individual members can be
assembled during erection.
9. Amount and Kind op Bkidges on the Railroads of the
United States.
The author attempted to collate from the official reports of the State
Railroad Commissioners the data from which to form an estimate of
the amount and kinds of bridges on our railroads.
He found that many States had no bureaus which collected this
information. Also, that many reports were indefinite and evidently
incorrect.
In order, therefore, to obtain more definite and reliable information,
and also to get data which could be considered as truly representative
of all the geographical divisions of the country, he sought it directly
from the principal large systems of railroads. Many of these systems
promptly placed at his disposal very full and reliable information as to
their bridges.
Taking from the Railroad Commissioners' reports such data as ap¬
peared correct and complete, and adding thereto this additional informa¬
tion, avoiding as far as possible any duplication of the same roads,
data covering nearly 60 000 miles of railroad, fairly distributed over the
whole of the country, were obtained.
4C
In order to make these data as nearly comparable as possible, the
length of all double-track bridges was reduced to the corresponding
length in single-track bridges.
In like maimer, the length of all main tracks, omitting sidings and
turnouts, was put into terms of single track.
Elevated railroads, which are composed mostly of bridges or trestles,
have been omitted entirely.
Table No. 3 gives the general data as to quantity of bridges and
trestles and the average rate per mile of track.
It shows that the relative amount of bridges and trestles varies in
different districts from 58 feet per mile to 231 feet per mile. This last,
however, is excessive, from including the crossing of Lake Pontchartrain,
near New Orleans, on a trestle 22 miles long. Omitting this, we would
get only 162 feet per mile as the maximum.
These variations are not entirely due to geographical location, as
might appear at first thought. They are also affected by principles
governing the original location of each road or division of a system.
The alignment and grade may have been sacrificed to the avoidance of
bridges or trestles, or the contrary.
From the large mileage covered by our table, we can rely with con¬
siderable confidence upon our average.
TABLE No. 3.
System of Railroad, or State.
Miles of
Road.
Total length
of Bridges
and Tresties
in feet.
Lineal feet
of Bridges
and Trestles
per Mile
of Road.
New Yoik Central and West Shore Railroads
2 894
364 722
126
New York, Lake Erie and Western Railroad
1 514
95 509
63
Other Roads in New York
3 586
445 900
130
Roads in Pennsylvania
4 352
336 957
77
Roads in New England
2 199
176 700
80
Wabash System
1 G36
160 025
98
Missouri Pacific System
4 707
566 953
120
Chicago, Milwaukee and St. Paul Railroad
5 727
614 736
107
St. Louis and San Francisco Railway
1 441
130 075
90
Denver and Rio Grande Railroad
1 458
102 195
70
Union Pacific Railroad.
4 754
276 032
58
Louisville and Nashville Railroad
2 495
322 679
123
Queen and Crescent System
1 139
299 222*
231*
Roads in Illinois
8 639
707 535
83
Roads in Michigan
4 151
249 345
60
Roads in Iowa
7 778
1 049 386
135
Central Railroad and Banking Company of Georgia....
1 487
173 975
117
Totals
59 857
6 071 946
101
* Included the crossing of Lake Pontchartrain, a trestle 22 miles long.
47
Taking, therefore, 100 feet per mile as our basis of estimate, we have
for the 1G0 000 miles of railroad in the United States, 16 000 000 feet or
3 030 miles of bridges and trestles.
TABLE No. 4.
Miles of Bail-
road.
Trestles and
Spans under
20 feet.
Spans from 20-
50 feet.
Spans from 50-
100 feet.
Spans from 100-
150 feet.
A
o
g,
a 2
£ °
« CS
ft
ca
Spans from 200-
300 feet.
Spans from 300-
400 feet.
Spans from 400-
500 feet.
Spans over
500 feet.
Total.
Average per
Mile of Koad.
26 228
2 299 758
85 181
94165
149 121
80 551
29 542
5 677
1 211
1040
2 746 246
104.7 feet.
Table No. 4 gives the distribution of the bridges upon 26 000 miles
of railroad into spans of different lengths.
Using this as a basis of estimate, the 3 030 miles of bridges and
trestles in the United States should be distributed as follows:
Trestles and spans under 20 feet 2 424 miles = 727 200 spans.
Spans from 20-50 feet 121
50-100 " 130
100-150 " 190
150-200 " 109
" over 200 feet 56
3 030
= 18 150
= 9 100
= 8 000
= 3 300
= -■ 1150
= 766 900
The above includes all bridges of either wood or iron.
Using the detailed information in the author's possession as to the
bridges on the above 26 000 miles of railroad, which are of wood, com¬
bination and iron, we obtain the following estimate for the amount of
iron bridges on our railroad system:
Iron spans under 20 feet 17 miles = 5 100 spans.
20-50 feet 86 " = 12 900
50-100 " 66 " = 4 600
100-150 " 93 " = 3 900
150-200 " 69 " = 2 100
" over 200 feet 49 " = 950
Total 380 " = 29 550
48
The author's own office records contain the following iron bridges over
200-foot spans, and they are very incomplete for spans less than 300 feet:
Spans over 500 feet 2.5 miles.
" 400-500 " 3.9 "
" 300-400 " 5.0 "
" 200-300 " 15.0 "
Total 26A "
From the previous figures it would appear that there are now in ex¬
istence on our railroads the following wooden and combination trestles
and bridges:
Trestles and spans under 20 feet 2 407 miles.
" 20-50 " 35 "
" 50-100 " 64 "
" 100-150 " 97 "
" 150-200 " 40 "
" over 200 " 7 "
Total 2 650 "
Of the 2 400 miles of wooden trestle, we can consider one-quarter
as only temporary, to be filled in as embankment. Of the remaining
1 800 miles at least 800 miles will be maintained in wood. This leaves
1 000 miles to be replaced gradually with iron bridges and trestles of
such spans as may be most suitable for each location, probably from 50
to 200 feet.
This would ultimately make 1 600 miles of bridges for the 160 000
miles of railroad now in existence, of which only 380 miles are in iron
at the present time.
The substitution of iron for the existing wooden bridges will, of
course, be a gradual one. The amount of work involved in this, with
that required in building iron bridges for new roads, additional crossings
of our large rivers to connect present and future systems of railroad,
and the demands for city and highway bridges, will furnish plenty of
employment for our bridge building firms.
The author has no reliable basis for estimating the amount of highway
bridges demanded by the present or the future. A comparison of the
relative crossings of streams by railroads and highways in the more
developed parts of our country, leads him, however, to the belief, that
the lineal feet of highway bridges will be four to five times that
required for railroads only.
49
10. Failure op Bridges.
In consideration of the facts already presented, it need not be a
matter of surprise, that bridges have failed under train service in
America.
The data in regard to the number of such failures have been magni¬
fied and distorted.
That such distorted evidence should have been submitted to a
foreign society, without any opportunity for those acquainted with the
truth to counteract the false impressions and conclusions, is to be re¬
gretted.
In (he paper referred to, 251 bridges are given as the number of
failures in the last ten years, classified as follows, by the author of the
paper: 57 knocked down by derailed trains; 30 "square falls"; 96 un¬
certain; 5 occurred during replacement or repairs, and 63 unclassified.
That the great majority of these bridges were old and defective
wooden bridges is not mentioned.
In addition to this generalization, there are given a number of
photographs with incomplete statements of the failures they represent.
Of the nine railroad bridges, in regard to which their identity or any
reliable statements are given, we are able to summarize as follows, with
the aid of other facts in our possession:
Three knocked down through derailment of the train, one from a
broken axle, one by a cow on the track, and one from neglect to spike
the newly laid rails.
One split open by two trains colliding within the trusses; this was an
old-fashioned structure with cast iron top chords.
One knocked down by a steam derrick too large to pass through the
bridge.
One wood and iron combination bridge, seven years old, which failed
in its wooden members under a load far in excess of its designed load.
One, the Bussey Bridge, an abortion in design and construction, and
a bridge in which no engineer had any part.
One which is stated to have been culpably overloaded, and to have
withstood this excessive overloading for a long period without showing
any apparent signs of distress.
One showing a lattice bridge distorted by a train collision and which
still carried the train.
The above is a summary of the evidence presented to show that
50
American pin-connected bridges are not so safe and reliable as those
of a lattice form. It is fair to presume that this is the strongest evidence
that the facts would admit.
The same authority, quoted by the author of the paper referred to,
gives more recently a fuller account of the bridge failures for the ten
years ending January 1st, 1889; total number, 265, of which only 38 are
known to be of iron.
In his total he includes 27 wooden bridges burned, 39 carried away
by floods, 8 failing during repairs, 60 knocked down, 31 square falls,
97 unknown.
Of the 31 square falls, 10 only are of iron.
The one iron bridge recorded during 1888 as a square fall was a
through plate girder, failing from the breaking of a plate carrying the
cross flooi; beams. More detailed information of the other nine is not
given, so we are unable to determine the kinds of iron bridges and the
causes of the failures.
It may be necessary to state that these reports of accidents are not
based entirely on official reports; but are collated from the newspapers,
and are undoubtedly full of sensation and inaccuracies.
Does any one advocate the designing and building of bridges to
withstand the impact of a railroad train, or the bursting effect of piling
two trains on one another inside of the trusses? Are such accidents to
be classed as bridge failures or as failures of management?
That bridges of the lattice type have given good results in many cases
of derailment, etc., proves nothing. Scores of similar cases, showing
that pin-connected bridges do stand much abuse, and are as cajiable of
resisting occasionally these extraordinary strains as any other kind of
bridge, could be given by every practical bridge engineer in America.
They prove nothing in regard to the merit of any particular form of
truss. They simply point to the need of better means to prevent and
remedy cases of derailment.
It is a very gratuitous claim to assume that the bridge accidents on
railroads in America are in any manner due to the accepted forms of
our bridges, or that other forms of bridges would have given different
results under the same circumstances.
The number of bridge accidents on American railroads, due to causes
or forces which a bridge should be designed to meet or resist, is an ex¬
ceedingly small one, when compared, as it should be, with the total
51
number of bridges on our railroads, and when the circumstances
under which many of them have been built be given proper consid¬
eration.
The total number of spans of railroad bridges in the United States
over 50 feet span, is 21 550, of which number 11 550 are of iron.
Due weight must be given to the rapid development of our railroad
system, and the want of appreciation in the past of the demands for
heavier rolling stock and correspondingly heavier bridge structures.
The American idea of building cheap railroads far in advance of
the immediate demands of the regions through which they run, to settle
those districts and build up a future paying traffic, has compelled the
use of cheap and light bridge structures.
Even on the roads in the older portions of our country, the earlier
bridges were built for loads far below present necessities.
Economy has also, in many cases, favored the lowest bidder, re¬
gardless of the incapacity or unscrupulousness of the bidder. There
have been and still are many bridges of an inadequate strength in use
on American railroads.
They are disappearing very rapidly, especially upon our better
managed and first-class railroads.
Under these circumstances, it is a great credit to our style of bridge
construction, that so few, comparatively, of our bridges have failed.
The writer does not know of a single case of the failure of a modern
all-wrought-iron American railroad bridge, which has failed under legiti¬
mate train service. He has had knowledge of many bridges which have
for years done far beyond their intended duty with satisfaction.
That the bridges built in the past have been lighter than the de¬
mands of the future must not be entirely credited to the engineer. As
a rule he has done fully as well as the means given him would admit.
The American engineer has reaped one great advantage from the
defective and inadequate structures of the past, not only from those that
have actually failed, but also those which have been worked beyond
their designed capacity—a knowledge of the relative merits of the
various forms and proportions of the members and their details.
These structures may be said to be full sized tests, from which much
that is valuable has been learned.
TVe can feel that our knowledge of bridges has passed far beyond the
study of strains and tests of specimens. We have not remained content
52
with this limited field of knowledge, covering our ignorance under a so-
called "factor of safety." We have determined tlie capacity and relia¬
bility of the full sized parts of our structures. We have also been
enabled to study the working under excessive loadings, even up to rup¬
ture, of structures and their details which had outlived their intended
purpose or which were inadequately designed from the beginning.
American bridge engineers and manufacturers have therefore legiti¬
mate grounds upon which to base their claims as to the great merits of
our system of construction and proportioning.
The author desires to emphasize the statement that failures of bridges
in the United States have no more bearing upon the relative merit of
different systems of their construction, than the failures in other kinds
of machines. The old-timed boiler explosions on our Western rivers did
not convince any sensible engineer that cylindrical boilers were not suit¬
able for high pressures. That machines inadequately proportioned for
the demands which will be put upon them break down ultimately, espe¬
cially when they fail to receive intelligent care and attention, is to be ex¬
pected.
The intelligent investigator does not decide upon the merits of any
developed system by the failures which are necessary steps in its develop¬
ment. Without variations and failures there would be no evolution or
the survival of the fittest.
These variations and failures are interesting and instructive, but
should not blind us to the merits of the final result.
The following remarks by Professor Unwin are worth careful con¬
sideration in this connection :
"If an engineer builds a structure which breaks, that is a mischief,
but one of a limited and isolated kind, and the accident itself forces him
to avoid a repetition of the blunder. But an engineer who from de¬
ficiency of scientific knowledge builds structures which don't break down,
but which stand, and in which the material is clumsily wasted, com¬
mits blunders of a most insidious kind."
The author desires to draw attention to a series of articles on Ameri¬
can and English bridges, by Mr. T. Claxton Fidler, in " Engineering,"
November 9th to December 28tli, 1888. They are a valuable contribution,
and are exceedingly fair in their treatment of the subject, considering
that their author on some points has not the real facts of our best prac¬
tice before him, and that he has mistakenly accepted a particular bridge
company's specification as representing the general practice in our
country.
53
He summarizes the following results as derived from the American
system of competitive bridge construction :
"First.—It has developed much ingenuity of design, which has been
directed, however, to the improvement of trusses in detail rather than
to such broader features of design as are to be seen in the relative out¬
lines of the Britannia Bridge, the Saltash Bridge and the Forth Bridge.
"Second.—It has attained to a great economy in details and the advan¬
tage obtained by the use of pin-connections, together with the attendant
facility of erection has sometimes enabled the American maker to obtain
the lead in the market.
" Third.—The system has necessitated and has produced an earnest
inquiry into the theoretical principles of bridge construction, whose
results have been embodied in certain specifications comprising a series
of rules and regulations which are intended to insure the safety of the
bridge.
"Fourth.— The defects and possible abuses of the competitive system
are, however, acknowledged; and have been attended with numerous
bridge failures.
"Fifth.—In quite recent times there has been a strenuous endeavor
to repair the defective features of the system, and to obtain more ample
guarantees for the safety of the bridge in the shape of more stringent
rules; but so long as the rules are drawn with reference only to the real
stresses, they must always be insufficient for the purpose in view, unless
they are supplemented by the independent judgment of the engineer
proceeding from merely arbitrary methods,
" Sixth.—In order that the rules may be effective for the purpose in
view, they must be based upon a complete theory of safe construction,
and not merely upon a theory of stresses; and must provide for all those
requirements of stability and solidity which are instinctively recognized
by the practical engineer, and which cannot be complied with by merely
using a large factor of safety.
"Seventh.—If these requirements are to be formulated in mathemati¬
cal shape, it would be necessary to enter upon that wider inquiry which
has been referred to in these articles, and to specify a certain series
of empirical and unreal distorting forces, which are to be assumed as
acting at different points in the structure, and to be resisted by members
possessing the strength which may be requisite for that purpose in ad¬
dition to the strength which may be demanded by the real stresses. Such
54:
a theory of unreal stresses may perhaps be dispensed with in English
practice as it has been hitherto; but there is no reason why it should
not be framed upon common sense principles, and adapted to the re¬
quirements of the American competitive system, for whose purposes it
seems to be urgently required."
It is true that our system of bridge building has not advanced in the
direction of such designs as the Britannia and Saltash Bridges during
the past. Whether it will be towards designs like the Forth Bridge in
the future is very doubtful. The author believes that problems of this
magnitude will be solved by the American bridge builder, when the op¬
portunity occurs upon the broad principles of our past experience.
We are always ready to admire " big things," as we have each of these
structures in its day, and to reap benefit from the valuable contributions
which have been given us by their able engineers in the line of new in¬
vestigations and experience.
WTe are ready to acknowledge the evils of the old competitive system,
where " the independent judgment of the engineer," or " those require¬
ments of stability and solidity which are instinctively recognized by the
practical engineer," do not enter sufficiently to obtain a safe system of
construction.
The most perfect system of rules to insure success must be inter¬
preted upon the broad grounds of professional intelligence and common
sense.
While our competitive system, guided by a general specification, is
more apt to produce good bridges than dependence upon designs pre¬
pared by any one with narrow and limited practical knowledge, its best
results follow when supplemented by the independent judgment of the
practical bridge engineer.
APPENDIX.
ABSTRACT OF RECENT TESTS ON FULL SIZE
BRIDGE MEMBERS.
TABLE No. 5.
Abstract of Tests on Wrougiit-iron Columns, Watertown Arsenal.—Symmetrical Sections—Pin Connected.
Set.
*
Test
Number.
Ratios.
Area of
Section.
Pin
Bearing.
Diameter
of Pin.
Ultimate
Strength per
Square Inch.
Mode of Failure.
1
~d
I
r
Square
Inches.
Square
Inches.
Inches.
Pounds.
X
17
10
27
12.
9.1
31
tt
35 070
Channels buckled near end of column.
X
1109
10
27
12.2
9.1
33 420
Channels buckled near end of column.
A'
15
15
40
12.1
9.1
tt
33 830
Channels buckled near end of column.
A'
16
15
40
12.5
9.1
it
35 490
Channels buckled near end of column.
L
13
15
41
9.7
5.6
35 550
Filled by direct compression.
L
14
15
41
9.6
5.6
35 350
Channels buckled near end of column.
C
372
20
49
10.
8.8
31 698
Plates buckled between rivets.
C
371 .
20
49
9.1
8.5
31 302
Plates buckled between rivets.
L
1640
20
62
12.
13.
it
33 740
Deflected perpendicular to pins.
L
1G41
20
52
12.3
13.
it
34 670
Deflected diagonally.
M
1638
20
52
12.4
13.
t i
31 130
Deflected perpendicular to pins.
M
1639
20
52
12.7
13.
it
31 990
Deflected perpendicular to pins.
L
469
20
52
13.
8.1
3
28 970
I By deflection in line of pins; column had an initial bend
} of | inch in that direction.
I
740
20
53
7.6
7.
31
•«
34 340
Deflected perpendicular to pins.
I
741
20
53
7.6
7.
33 530
Deflected perpendicular to pins.
II
746
20
63
7.6
7.
i*
33 910
Deflected perpendicular to pins.
II
1111
20
53
7.5
8.2
a
34 810
Channels buckled.
II
1112
20
53
7.5
8.2
<•
35 240
Deflected perj)endicular to pins.
L
11
20
55
9.6
5.6
a
33 840
Channels buckled.
L
12
20
55
9.6
5.6
34 000
Channels buckled.
X
9
20
56
12.
5.6
tt
34 360
Channels buckled.
X
10
20
56
12.3
5.6
it
36 610
Channels buckled.
X
1107
24
61
4.6
5.2
t (
34 740
Deflected perpendicular to pins.
X
1108
24
61
4.6
5.2
it
34 160
Deflected perpendicular to pins.
X
1
25
64
4.6
3.5
tt
35 880
Deflected perpendicular to pins.
X
2
25
64
4.7
3.5
it
32 380
Deflected perpendicular to pins.
A
752
20
64
9.8
8.7
a
30 220
Deflected perpendicular to pins.
D
360
20
65
15.4 •
12.7
30 965
Deflected in plane of pins.
D
361
20
65
15.4
12.7
A
767
20
G5
10.2
8.75
B
367
27.5
65
11.4
5.7
B
368
27.5
65
11.4
5.7
X
470
25
65
13.
8.1
L
467
24
66
9.2
10.8
L
1113
25
67
7.5
8.2
L
1114
25
67
7.5
8.2
X
7
25
67
12.1
9.1
X
8
25
67
12.0
9.1
E
358
28
67
17.8
9.3
E
359
28
67
17.2
9.3
L
3
25
69
9.7
9.1
L
4
25
69
9.8
9.1
M
1636
30
70
12.
13.
M
1637
30
70
12.
13.
G
370
30
74
9.2
8.6
C
369
30
74
9.5
8.6
X
1231
30
77
4.5
5.2
X
1232
30
77
4.6
5.2
L
1634
30
78
12.1
13.
L
1635
30
78
12.2
13.
H
747
30
78
8.
7.
H
748
30
79
7.6
7.
H
749
30
79
7.6
7.
J
739
30
79
7.5
7.
I
750
30
79
7.5
7.
H
1115
30
80
7.6
8.2
H
111G
30
80
7.5
8.2
X
28
30
80
12.2
5.6
X
29
30
80
12.5
5.6
H
465
30
80
7.75
7.8
L
5
30
83
10.
5.6
L
6
30
83
10.
5.6
L
468
30
83
9.2
10.8
B
357
37
87
11.3
5.7
B
356
37
87
11.4
5.7
X
1229
35
89
4.7
5.2
X
1230
35
89
4.7
5.2
E
1653
37
90
17.9
9.4
E
1654
37
90
19.5
9.8
H
1123
35
93
8.0
3.8
H
1124
35
93
7.7
3.8
* See
Plate XXV11.
31
494
31
380
32
329
33
205
30
000
29
980
33
970
33
610
32
940
34
240
32
074
32
253
33
880
33
660
32
830
32
740
31
650
3U
720
33
820
32
380
33
630
32
440
36
580
34
120
33
410
33
390
34
390
32
160
32
140
31
610
31
340
27
750
34
130
31
930
32
000
29
186
29
947
32
750
31
120
26
480
25
290
32
230
31
370
Deflected perpendicular to pine.
Sheared rivets in bearing plates.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
(Deflected perpendicular to pins; column had an initial
I bend in that direction.
Deflected perpendicular to pins.
Deflected in line of pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Channels buckled near end.
Channels buckled near end.
Deflected perpendicular to pins.
Deflected diagonally.
Deflected perpendicular to pins.
Deflected diagonally.
Deflected diagonally.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected diagonally.
Deflected in plane of pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Channels buckled near end.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
(Deflected perpendicular to pins; column had initial de-
I flection of 1 inch in that direction.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Vi
TABLE No. 5.—(Continued.)
Test
Number.
1645
1650
755
756
26
27
751
1642
1651
1652
354
355
466
463
1117
1118
24
25
1643
1644
1648
1649
1119
1120
22
23
1121
1122
464
753
Ratios.
40
40
30
30
35
35
30
30
40
40
40
40
37.5
40
40
40
40
40
40
40
40
40
45
45
45
45
50
50
50
40
96
96
96
96
96
96
98
98
98
98
98
98
1()0
102
102
102
107
107
107
107
107
107
115
115
120
120
128
128
128
128
Area of
Section.
Square
Inches.
16.2
16.2
10.
10.
9.3
9.6
16.
16.3
21.
20.6
9.2
9.4
7.75
4.7
4.7
4.6
7.8
7.8
7.7
7.7
7.7
7.7
4.6
4.7
7.8
7.8
4.7
4.6
4.7
9.7
Pin
Bearing.
Square
Inches.
10.
10.
8.6
8.6
9.2
9.2
13.
13.
12.
12.
8.75
8.75
7.8
4.2
5.2
5.2
9.4
9.4
7.
7.
7.
7.
5.2
5.2
9.4
9.4
3.5
3.5
4.2
8.6
Diameter
of Pin.
Inches.
3*
3*
3
Ultimate
Strength per
Square Inch.
Pounds.
28 020
27 910
25 160
21 105
32 180
29 380
26 430
22 540
25 770
25 950
28 950
29 879
26 000
25 000
29 180
30 990
31 350
27 850
30 840
30 770
31 610
29 870
30 590
31 050
24 850
26 920
23 350
25 360
15 260
19 380
Mode of Failure.
Deflected in plane of pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
(Deflected perpendicular to pins; strength reduced by
\ bending of pin.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected diagonally.
Sheared rivets in bearing plates.
Triple flexure.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins; weakened by separators.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected in plane of pins.
Deflected diagonally.
Deflected in plane of pins.
Deflected diagonally.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins.
Deflected perpendicular to pins; initial bend of 0.3 inch.
Deflected perpendicular to pins.
CJT
oc
a
754
40
128
9.7
8.6
3*
16 220
Deflected perpendicular to pins.
d
1046
40
131
16.2
13.
19 700
Deflected perpendicular to pins.
d
1647
40
131
16.1
13.
17 670
Deflected perpendicular to pins.
x
20
55
141
4.7
5.2
21 850
Deflected perpendicular to pins.
x
21
55
141
4.7
5.2
20 810
Deflected perpendicular to pins.
x
18
GO
154
4.7
7.
**
14 740
Deflected perpendicular to pine.
x
19
60
154
4.7
7.
15 900
Deflected perpendicular to pins.
un symmetrical sections.—pln connected.
n
1630
29
75
17.6
9.3
3£
26 190
(Deflected perpendicular to pins. Pin in center of gravity
\ of section.
n
1631
29
76
17.2
9.3
n
28 150
(Deflected perpendicular to pins. Pin in center of gravity
( of section.
k
352
29
77
12.6
9.1
t*
27 668
( Deflected perpendicular to pins. Pin in center of gravity
| of section.
k
353
29
77
12.8
9.1
a
30 596
t Deflected perpendicular to pins. Pin in center of gravity
) of section.
n
1632
29
78
17.6
9.3
17 420
(Deflected perpendicular to pins. Pin in conter of
( channels.
n
» 1633
29
78
17.7
9.3
*i
17 240
(Deflected perpendicular to pins. Pin in center of
1 channels.
k-
350
29
79
12.5
9.
a
16 242
f Deflected perpendicular to pins. Pin in center of
( channels.
k
351
29
79
10.8
9.
a
19 207
(Deflected perpendicular to pins. Pin in center of
\ channels.
Square Ended Columns.
e
e
b
b
f
f
g
379
380
377
378
1059
1000
346
347
318
20
20
20
20
20
20
21
21
21
47
47
47
47
51
51
51
51
51
17.
17.8
12.
11.1
4.8
4.7
15.7
15.8
21.
34 950
Plates buckled between rivets.
35 595
Triple flexure.
31 722
Plates buckled between rivets.
33 564
Plates buckled between rivets.
36 720
Channels buckled.
35 330
Channels buckled.
32 846
Plates buckled.
35 050
Plates buckled.
33 682
Plates buckled.
* See Plate XXVII.
TABLE No. 5.—(Continued.)
Set,
*
G
F
F
G
G
N
N
K
K
Test
Number.
349
342
344
341
343
337
338
339
340
Ratios.
I
d
21
30
30
30
30
30
30
29
29
I
r
r»i
75
75
75
70
70
77
79
79
A reft of
Section.
Square
Iuches.
21.5
15.7
15.0
21.2
21.5
17.
17.4
12.0
12.7
Tin
Bearing.
Square
Iuehes.
Diameter
of Pin.
Inches.
Ultimate
Strength per
Square Inch.
Pounds.
33 001
33 000
34 505
33 019
33 943
34 279
33 333
32 006
33 802
Mode of Failure.
Triple flexure.
Deflected in plane A B.
Pla'es buckled.
Deflected in plane A B.
Deflected in plane A B.
Deflected diagonally.
D< fleeted diagonally.
Deflected upwards.
Deflected upwards.
O
Forked Ends of Open Posts.—Pin Connected at One End.
r
391
p
392
R
393
R
394
11.5
11.5
11.8
11.8
0.4
6.4
4.9
4.9
3}
I •
4 <
it
Total
Ultimate
Strength.
393 000
415 000
302 000
302 200
Pin ends bent inwards
Pin ends bent inwards
Ends buckled.
Ends buckled.
* 8oe Plate XXVII.
Vol. XXI, p. 60.
TABLE No. 6.
TESTS OF EYE-BARS.
Date.
May 10th, 1887
May 16th, 1887
May 17th, 1887
June 1st, 1887
July 16th,1887
August 3d, 1887
November loth. 1887
November 21st, 1887..
December 14th, 1887 ,
March 1st, 1888
June 29 th. 1888
July 24th, 1888
October 25tli, 1888...,
December 7th, 1888...
December 10th, 1888.,
December 13th, 1888.,
December 13th, 1888.,
December 20th, 1888..
January 30th, 1889
February 21st, 1889..
February 2Gth, 1889...
February 26th, 1889 .,
February 26tli, 1889.,
January 10th, 1888
January 10th, 1888..,
Januarv 10th, 1888...
April 26th, 1888
April 26th, 1888
April 26th, 1888
March 5tli, 1888
March 29th, 1888
April 3d, 1888
April 3d, 1888
April 7th, 1888
April 7th, 1888
March 12th, 1888
March 12th, 1888
March 13 th, 1888
March 13th, 1888
March 14th, 1888
March 13tli, 1888
March FfJtli, 1888
July 10th, 1888
July 10th, 1888
July 10 th, 1888
July 10th, 1888
July 10th, 1888
July 10th, 1888
August 30th, 1888
August 30th, 1888...
August 31st, 1888
August 31st, 1888
August 31st, 1888.....
August 31st, 1888
August 31st, 1888....,
August 31st, 1888....,
August 31st, 1888.....
August 31st, 1888
September 1st, 1888...
September 15th, 1888
September 15th, 1888
September 17th, 1888
September 17th, 1888
September 17th, 1888
September 17th, 1888.
September 18 th, 1888
September 18th, 1888
September 18th, 1888.
September 18th, 1888.
September 22d, 1888..
September 22d, 1888..
September 24th, 1888,
October 25th, 1888....
Makers.
Kind of Steel.
Union Bridge Company
Edgemoor Iron Company
Bessemer
Open Hearth
Bessemer
Open Hearth
Bessemer ...
Open Hearth
Bessemer
Size of
Bars.
Inches.
8 x 11
8 x 2t\j
8 x 13
8 x 2j
6xl}§
8 x 1|
6x2
5 x 1J
5x1]
6xl&
8x2
6x1]
6x1
6 x l^g
4] x 1
7x1]
6x1
6x1]
5 x 1 jl
8 x 1§
7 x I]
6x1]
6x1]
5 x l/s
5 x 1«
5x1^
4x1
3x1]
5x1
3x1
4x1
4x3§
4x1]
6x1]
5 x l/e
5x1
5x1
5 x l/8
5xl&
5xl&
6x1?
5 x 13
5x1
5x1
4 x 13
4 x if
21 xf
2 X 4
21 x-I
21 x
7 x 13
7 x If
I x w
7x2
7 x 2j*g
xii?
7xl]g
4x1]
4 x U
4 x H
4 x |-
4x2
4 x 1/*
4 x £
4 x g
4 x ]|
4 x ]|
"Ml
6 x 11
6x1]
6xlA
Length of
Bar Tested.
Excess of
Material in
Eyes.
Elastic
Limit in
Pounds per
Ultimate
Strength in
Pounds per
Percentage of
Elongation.
Percentage
of Reduc¬
tion of
Area at
Fracture.
Size of
Pin Holes.
Elongation
of Pin
Holes.
Square Inch
Square Inch
Feet and
Inches.
Per Cent.
on Original
Section.
on Original
Section.
At
Fracture.
Over
Whole Bar.
Inches.
Inches.
32
4.8
41.7
36 900
65 800
38.
16.
49.3
7
1.2
14
3
49.
37 200
63 700
39.
16.8
43.9
7
1.125
30
10.8
48.
38 600
64 680
39.
14.2
45.
?
1.0
9
10
40.6
31 160
62 790
28.3
13.5
63.5
1.48
16
3
34.
36 460
61 340
37.5
17.3
44.6
6
1.17
23
3
32.
37 100
60 770
30.8
15.8
40.5
8
1.2
20
5
45.
40 560
65 930
35.4
17rl
47.3
6
1.04
28
8
43.3
38 849
62 578
35.8
It.7
47.
4]
1.0
36
11
47.6
38 560
62 144
30 8
10.0
43.6
4
0.51
29
8
52.3
40 063
63 597
35.8
15.2
55.9
5}
1.08
31
0
41.9
38 048
63 410
28.3
12.4
41.7
7
1.35
31
0
62.2
37 170
62 880
34.2
15.9
48.4
7
0.62
27
2
43.7
31 803
58 770
35.8
13.6
47.7
6
0.99
19
9
74.6
32 440
58 100
35.
16.7
48.8
5}
0.60
22
0
60.
35 300
59 600
30.
161.1
45.2
4';
0.49
17
5
49.3
34 600
55 370
32.5
18.8
48.2
61
1.38
31
6
52.
35 370
59 480
30.
IS.6
42.8
6*
0.96
25
0
51.1
29 530
56 000
36.6
16.6
56.6
4f
0.80
10
5
55.5
34 170
59 250
35.8
21.2
48.4
5
0.67
22
9
43.8
30 880
55 650
44.2
14.0
50.6
7
1.26
22
3
41.9
31 000
57 350
41.6
14.1
49.1
6r»j
1.02
31
1
46.6
31 500
55 600
40.
is;, l
49.6
41
1.04
31
1
49.5
31 280
55 340
41.6
17.7
50.1
1.17
20
2
41.2
44 330
68 230
38.7
13.3
54.4
1.15
20
2
42.2
41 470
64 5U0
45.4
14.9
53.8
1.08
29
2
41.2
41 240
68 230
40.3
14.7
56.0
"
1.24
20
0
28.7
41 000
62 800
39.7
13.8
57.
"
0.54
20
0
36.4
42 520
65 400
36.6
16.8
54.4
**
0.48
20
0
52.2
43 860
63 000
37.1
14.8
45.2
4ts
0.44
15
0
39.3
41 370
65 630
41.
19.8
41.3
m
0.46
15
0
28.9
42 890
66 090
40.3
11.6
60.9
0.54
18
4
31.6
44 810
67 210
36.3
12.3
51.1
3]
0.36
18
4
32.9
43 750
64 390
42.0
15.1
55.
0.81
29
2
43.7
41 810
65 370
42.5
12.0
43.6
6]
0.55
29
2
41.5
39 180
65 490
41.8
11.0
48.8
1.25
24
0
36.6
32 670
63 310
43.8
15.
48.3
0.77
24
0
41.7
34 650
66 600
40.
16.4
47.4
0.98
24
0
40.0
32 960 •
64 460
45.
17.2
47.2
"
0.74
24
0
40.0
37 650
68 670
45.6
17.7
46.6
"
0.84
20
0
35.4
33 180
66 360
38.5
12.
44.3
"
0.93
22
6
54.6
35 190
61 000
47.5
11.8
50.
5
0.72
22
6
54.8
36 840
65 320
48.1
17.
51.8
5
0.48
17
2
35.
31 840
59 400
43.7
14.9
50.5
ffi
0.76
17
2
37.
31 000
58 800
42.5
16.1
53.2
1.05
17
2
37.2
32 000
58 200
45.0
15.1
50.0
"
1.07
17
2
40.
33 000
59 300
42.0
14.6
50.6
««
1.06
31
2
55.1
32 100
59 600
44.0
15.4
53.3
"
0.64
31
2
48.
32 470
61 500
39.5
15.5
50.0
"
0.67
17
0
41.3
39 000
63 000
34.5
16.3
60.6
n
0.38
17
0
44.2
39 430
63 090
28.5
14.5
52.6
0.56
17
0
42.4
39 430
61 120
33.0
12.8
55.3
"
0.54
17
0
41.8
39 710
64 370
32.0
13.6
59.1
0.42
22
2
44.
38 860
60 860
35.
18.
43.4
CIS
1.21
27
10
46.
38 080
60 500
35.
13.9
39.5
1.39
27
10
44.4
39 540
61 800
39.2
13 5
47.2
"
1.39
27
10
48.6
37 360
59 160
36.6
14,7
50.3
"
1.30
27
10
46.7
36 800
60 400
38.3
14.8
50.8
"
1.29
29
0
45.4
38 710
60 000
40.
14.5
61.0
"
1.50
29
0
43.1
38 900
62 300
26.
15
46.8
1.04
18
22
0
31.6
40 680
36 630
68 940
Stf
0
34.0
63 820
39.2
17.4
56.4
0.67
22
0
27.7
41 620
67 630
37.5
13.9 »
51.9
"
0.43
22
0
34.6
41 370
66 840
36.4
14.
42.6
«»
0.88
22
0
27.3
42 270
65 240
37.2
14.
47.2
"
0.58
28
5
28.
39 350
66 650
40.3
13.2
57.5
"
0.79
28
5
30.8
40 230
64 610
33.
11.6
48.8
"
0.23
28
5
34.4
40 000
63 760
36.7
14.
50.0
"
0.45
32
32
2
30.3
37 070
41 440
59 510
65 610
2
33.8
32.7
10 4
45.
«
0.69
24
8
47.2
36 000
64 300
38.7
13 8
44.2
0.98
24
8
47.5
37 400
62 680
46.5
14 0
50.9
0.90
24
8
47.4
32 890
60 760
41.2
13 4
48.7
0.81
22
6
44.0
38 600
65 400
43.7
13 7
47.5
1.01
Where
Fractured.
Body of bar.
In eye.
Body of bar.
In eye.
Body of bar.
Character of Fracture.
Silky cup.
100 per cent, silky.
Silky cup.
50 per cent, silky and 50 per cent, granular.
Silky.
Bagged, partly cupped, fine crystal on edge.
90 per cent, silky, 10 per cent, crystalline.
Ragged, silky, few crystals.
55 per cent, silky, 45 per cent, crystalline.
Silky cup.
85 per cent, silky, 15 per cent, fine granular.
Silky cup.
Silky.
Silky cup.
90 per cent, silky, 5 per cent, granular.
Silky cup.
Silky cup.
80 per cent, silky, 20 per cent, granular on edges.
Silky cup.
Silky.
Silky.
Silky.
Silky cup.
Silky cup.
Silky cup.
Silky cup.
Silky cup.
Silky cup, slightly granular.
Silky cup, slightly granular.
Silky cup, slightly granular.
Silky cup.
Silky cup.
Silky, ragged.
Silky, cupped.
Silky.
Silky, with granular edges.
Silky, with granular edges.
Silky, cupped.
45 per cent, silky.
Silky, with granular edges.
45 per cent, silky.
Silky.
Silky cup.
45 per cent, silky.
Silky.
Silky.
Silky cup.
45 per cent, silky.
Silky, cupped.
45 per cent, silky.
45 per cent, silky.
Silky, cupped.
Silky, 35 per cent, granular.
Silky, 45 per cent, granular.
Silky.
Silky, cupped.
Silky, cupped.
8ilky, 15 per cent, granular.
Silky, granular edges.
Granular, not properly annealed.
Silky, cupped.
Silky, cupped.
45 per cent, silky.
45 per cent, silky.
Silky, cupped. x
45 per cent, silky.
45 per cent, silky.
Granular.
45 per cent, silky.
Silky, with granular edges.
Silky, cupped.
Silky, cupped.
45 per cent, silky.
Vol. XXI, p. 60.
TABLE No. 7.
Tests of Eye-Bars, with Corresponding
'J ESTS
Maker of Material.
Carnegie, Pliipps & Company
Maker of Eye-Bars.
Keystone Bridge Company.
Kind of Material.
Bessemer Steel
Iron
Size of Bars. Length Tested,
Inches.
'{ round.
1.245 x 0.5
5xlJ
j* round.
1.224 x 0.7
6xl&
* round.
1.317 x 0.637
6 x 1£
=* round.
0.75 x0.4418
4x1
$ round.
1.050 x 0.5
5x1^
0.987 x 0.55
5x1
1 x 0.62
6x1
2 round.
4 x 1{
i round.
4xl§§
| round.
5 X lyg
1 round.
4X1A
2 round.
5 x lf|
jf round.
5 x 1&
2 round.
| round.
6xl&
Feet and
Inches.
0
0
21
0
0
13
0
0
12
0
0
20
0
0
17
0
16
0
11
0 8
19 0
8
0
0
16
0
20
0
31
0
19
0
19
0
22
0
17
Elastic Limit.
Pounds per
Square Inch
on Original
Section.
42 560
42 720
41 086
40 380
40 150
38 288
39 800
35 760
38 390
41 950
38 280
48 550
41 650
44 380
42 671
39 240
42 105
37 990
38 133
29 730
29 900
28 970
29 344
29 300
29 329
29 780
30 282
29 990
27 109
28 150
31 540
29 500
27 160
29 570
23 603
Pounds per
. Square Inch
on Original
Section.
on Samples.
Ultimate
Strength.
Elongation
in 8 inches.
64 640
67 330
59 000
63 780
64 880
61 080
61 150
60 760
60 210
66 320
67 780
65 000
65 890
67 910
65 744
58 960
61 580
60 780
58 716
52 450
49 400
52 180
47 309
49 650
49 939
50 960
50 470
50 653
46 820
49 490
49 280
50 850
46 681
51 380
46 570
Per Cent.
26.2
25.0
28.7
21.9
28.7
28.7
27.2
26.2
24.2
28.
24.
27.5
19.5
25.
25.5
21.5
18.7
24.3
Elongation
over Whole
Bar.
Per Cent.
10.2
20.
16.4
20.7
17.7
15.4
16.1
"n.c"
Reduction of
Section.
Per Cent.
52.2
47.6
57.6
36.9
54.1
53.5
47.3
43.7
53.4
54.6
41.5
56.2
50.3
43.7
51.3
48.
50.6
50.7
33.2
21.6
43.
36.
24.7
24.3
24.2
33.
32.8
29.6
25.9
27.3
26.8
20.6
31.4
26.5
Character of
Fracture.
Silky.
Silky, perfect cup.
Silky, perfect cup.
Silky, angular.
Silky cup.
Full cup.
Silky.
Silky cup.
" angular.
Silky, angular.
•' cup.
Fibrous.
Where Fractured.
In body of bar.
A
PERMANENT BRIDGE.
3uilt ly Timothy Palmer.
tn western
I verity
,»ra cv
PLATES IVANDV.
TRANS.AM.SOC.CIV.ENG'RS.
VOL. XXI N9 ^18.
COOPER ON
AMERICAN RAILROAD BRIDGES
Seclwn- of bridge al centre.
■
L,
rTrr
f l'iA „
Wt '^'|i|i; ■
HIGHWAY. BHh
OVER THE
HUDSON R
between
WATEHFOBD and LAN
BUILT BY THE OB QBE E,
North wt . t-jrrt
University
Libra, cy
PLATE VI
TRANS-AM.SOC.CIV. ENG'RS.
VOL. XXI N9 -A IB.
COOPER ON
AMERICAN RAILROAD BRIDGES
17*'- o' Ctr.tr etr. of (end fish.
S1NGBURGH.
JHR.
3A-0'-
jSca^ofJat^
NT-O"-
PLATE VI!
TRANS. AM.SOC.CIV.ENG'RS.
VOL. XXI N9 ^1 8. m
COOPER ON m
ICAN RAILROAD BRIDGES 1
.Sctdt of feet.
_ao 3o {>Q soft.
lOLh
Northwestern
University
Library
PLATE VIII
TRANS. AM.SQC.CIV. ENG'RS.
VOL. XXI N° A-1 8.
COOPER ON
AMERICAN RAILROAD BRIDGES
diawz/ia Onn4itc4
M( HAWK BRIDGE SCHENECTADY NY
Built W Theodore Burr.
J
IBID GE greeted OVVC tb DELM£
CUV
hope -
Built Tn^ Lw« Wernwaq.
Northwestern
University
Library
PATENT MIDGE "COLOSSUS
Across tlifc TUwr S(^i1Ml at T}\x\Meb(dticu.
Smqlf' Arch, 340 fee-h 3 ^%iruhes~
ILuilt "bx< L ems Wernm
PLATE IX
TRANS. AM. SOC.CIV.ENG'RS,
VOL. XXI N9 A I 8.
COOPER ON
AMERICAN RAILROAD BRIDGES
I
Li
[
PLATE X.
TRANS. AM. SOC.CIV.ENG'RS.
VOL. XXI N9 ^ I 8.
COOPER ON
AMERICAN RAILROAD BRIDGES
clear jpuast.-
TDWN. LATTICE TRUSS.
1 1 1
i n
I
1 , 1 , 1
i i
-Lr-
I
TFE BOLLIMAN PATENT
IBOI SUSPENSION TliUBS BUDGE
BUILT BY
WEMBEB3L BOLLMAN
PLATE Xii
TRANS.AM.SOc r>n« ,,
VOL XXI N? XIAGRS-
COCjper „J,8-
AMERICAN *A,«0ad*BR|dges
Jrta ils of .Buttress.
PLATE 'SWns'ty
TRANS. AM.V.ENG'RS.
VOL. XXI N9 AI8.
COOPER ON
AMERICAN RAILROAD BRIDGES
JRQi BRIDGE over
Constructed under i3.H.Latro|f Chief Engineer ,inl8ul-2.
Designed, & patented by Albert Fink Assis 'ant Engineer.
Fig. I
Section ab
Section ef
Section rs
Fig. 9.
Section xv
Fii 7
EHE3BHE3H
to ng. s 7.8.9mn.
Sn.u to np ta.'ks./s.
X
Northwestern
University
Library
FINK COMBINATION BRIDGE
jig I
ur-ft.
m
-3"
S.
i5
-=i
2E_
S
e
t
<=t
BJ
£•
£
— A
r ^
T.
A.
s?
<
Si
xL
,T£S
-sST
3&»
£5
:
-=
3
sis-
(
3:1s:
-)
.4
^=1
PLATE XIV.
TRANS.AM.SOC.CIV.ENGRS.
VOL XXI N° AI 8.
COOPER ON
AMERICAN RAILROAD BRIDGES
Fig 14.
IT
Ju
I
Sl Fig- <:-
r-T^'
^ , . Kir
.j . .
^ Il
|: lUi
■ . m - , II,
i< n ... - ..
PLATE XV.
TRANS AM.SOC.CIV.ENG'RS.
VOL. XXI N° AH 8.
COOPER ON
EEiE: AMERICAN RAILROAD BRIDGES
PLATE XVI
TRANS.AM.SOC.CIV,ENG'RS,
VOL. XXI N? -A I 8.
COOPER ON
AMERICAN RAILROAD BRIDGES
I
»<*•««♦" t&c'-SJ.n' tafiffjft+ttjtvf'
"tf c. . ..i—, ..*■ . tt-*v rr , rr"ii r*Si *
plate: xvm .
trans.am.soc.civ. eng'rs.
vol. xx! N9 ^18.
cooper on
american railroad bridges
iSfiamnd' e)-ria,n
' ■ 'zr^-
<?)lxi(l'eni'itYe- <s3 ziJae*
S$63~ 6%.
On: J. gjp. dlsri'l'iffe (?■ &.
tie i>{' .• cr &_&±'f-i-tv:-
FIG c5. END VI EW,
AT LONG post AT BRACE C D
»
FIG.4. END" VI EW
PLATE.XVIII.
TRANS. AM. SOC. CIV. ENG'RS.
VOL. XXI N9 AI3.
COOPER ON
AMERICAN RAILROAD BRIDGES
■
FIG. 2. PARTIAL PLAN OF BOTTOM FRAME.
FIG ,«5. END VI EW,
at long post at brace c d.
built BY THE louisville BR i DOE &r i RON CO.
GHNERAL PLAN
blue; river bridge.
J. M. 6. I. R. R.
LoutivvUc BrxdLj* 8. lr<i» Co.
MarcJ*. |7 1871
PLATE -XIX
TRANS.AM.SOC.CIV. ENG'RS.
VOL XXI N9 a-i a.
COOPER ON
AMERICAN RAILROAD BRIDGES
00\OT M fm « \WJSS«> SRNXS.
Sc«L«.» fntt %ixa.
7;'th^«ste
'-'n/v
rn
"sity
Llbrnry
PLATE XXV
TRANS.AM.SOC.CIV.ENGRS.
VOL. XXI N9 418.
COOPER ON
AMERICAN RAILROAD BRIDGES
®&>Urri&) <z<nJ^ ftezte<rz,-teJ i'v
<£Sow^Sv/uz.
<2£go&,.
<&£t&rztorz ,
=/*.
Northwestern
University
Library
PLATE XXVI
TRANS.AM.SOC.CIV. ENGRS.
VOL.XXH N<? ^18.
COOPER ON
AMERICAN RAILROAD BRIDGES
Scale ol Elevation fe ol an Inch to one Foot
■ t o ie>(b
UppeT Chord and Top Lateral Bracing.
Square End Post
Round End Post and Compensation for Expansion
Connections oS Lower Chord at A.
Connections of Top Chord.
tssnttsuieus plate t2'-%'.
JY/i 3Vdi'a.
JT=0 isi. P? 3fi?\>3f/.
x/,e - ~.3}l*JSi~
Vi revets.
punched tides.
Plate IP h* 9 '.
'JSI and bi'ttmi plates 9 "Si
ebu'e/t 'lis
Plates inside and outside
-Plate/r-to'-tr.
Pole: fdus/ins,J 'hare s/aelteil sides.
% si gets, the/lundied titles.
e? but or SWABBER DIMENSIONS.
C Similar in form
Plate/3.1-V/. PL
%'rivets.
■H "punctied tides.
li "rivets.
Pf'htse/s.
-pTu /t/a/fS 7% ">%/7%
Pitifilabs 7%'*%t,
Put Jtkjft/i.
J*m 3% "dut
Pin-3'/iWur
Jo/ei {p7tfritnfpJl 7/tr/'f jwe77ff7 s?/7ff.
PLATE XXVII.
TRANS. AM. SOC. CIV. ENC'RS.
VOL. XXI. No. 4 18.
COOPER ON
AMERICAN RAILROAD BRIDCES.