} . . I OFT ORNL P 3083 9 w ww. som internet EEEFEEFE 3 MICROCOPY RESOLUTION TEST CHART NATIONAL BUREAU OF STANDARDS -1963 KO ..5 , CFSTI PRICES INCIPIENT AND STABLE BOILING SUPERHEAT FOR POTASSIUM BOILING ON A SURFACE CONTAINING RENTRANT CAVITIES ORNL 308.3. Conf. 670909... J. A. EDWARDS H. W. HOFFMAN JUN 2 2 1969 HC. so 1.1' ! r Department of Engineering Mechanics North Carolina State University Raleigh, North Carolina, U. S. A, i : ' Reactor Division : Oak Ridge National Laboratory Oak Ridge, Tennessee, U. S. A. MASTER . . -..*... - INTRODUCTION The use of alkali metals as the heat- transfer fluid in stationary power plants, such as those incorporating nuclear reactors, and in auxiliary power systems, such as those intended for space probes and underwater exploratory crafts, appears to have good potential. As a direct consequence of some of their properties, alkali merkils have, under certain conditions, some very distinct .. advantages over the more familiar fluids such as water. For example, from a thermal stand- point, they are well-suited for high temper- ature use in high heat flux heat-transfer systems; however, they do impose some prob- lems due to their chemical activity and heat- transfer characteristics. boil sodium and potassium. The results re- ported in Refs. 3 and 4 were obtained with a porous surface (sintered stainless steel), a surface in which circular cylindrical cav. ities had been formed, a polished surface, and arì as-received (commercially smooth) sur- face. Superheats for alkali metals have also been reported by Balzhiser,5 Marto and Rohsenow, and Petukhov, Kovalev, and Zhukov.' The superheats reported by Marto and Rohsenow were smaller than those given in Refs. 3, 5, and 7 for similar surface conditions and saturation temperatures. However, as pointed out by these experi- menters, the alkali metal used in their experiments may have been contaminated by particles of graphite or dirt which could cause nucleation at a lower superheat than found with pure fluids. ....: .......... .. . ::...:.... In this paper, we direct our attention to certain aspects of heat transfer with boiling potassium. Experimental results are given for the superheat required to boil potassium on a stainless steel surface which contains reentrant cavities (surface opening smaller than body of cavity within wall). The effectiveness of these cavities in re- ducing the stable boiling superheat – as compared to the superheat needed to boil potassium on a smooth surface – is demon- strated by the data. In addition, the depen- dence of superheat on heat flux for both the stable boiling and bubble nucleation (incip- ient boiling) modes is shown. Further, a qualitative description of cavity quenching is given which is consistent with data ob- tained during incipient boiling. Since surface conditions play such a vital role in the nucleation process for all fluids, it is essential that the precise in- Iluence surface cavities have on the nucle- ation process be understood. For alkali metals, the surface condition may be the predominate factor in fixing the amount of superheat needed to initiate and maintain stable boiling. SUPERHEAT EQUATION Superheats for boiling potassium and sodium have been reported previously by the authors. These results, which were obtained both in a reflux capsule and in a natural- circulation loop, showed that the surface conditions of the boiler had a substantial : influence on the amount of superheat needed for stable boiling. In both types of experi ments, it was found that the stable boiling superheat was correlated fairly well by the following equation when the boiler contained well-defined cylindrical cavities: ... :9 ; - The influence of surface. conditions on nucleate boiling of liquids (nonmetals as well as metals) has been reported by numer- ous investigators. Thus, Corty and Foust? showed the effect of surface roughness for R-113, diethyl ether, and n-pentane when boiling on copper, nickel, and Inconel sur- faces. Bonilla, Grady, and Averys found 4772 parallel scratches reduced the surface .. nr. : . RT I , 20 - i v sat la . - + BLANK PAGE 2 '. rr1 - - AL transfer characteristics. cruse nucleation at a lower superheat than found with pure fluids. t than Since surface conditions play such a vital role in the nucleation process for all fluids, it is essential that the precise in- fluence surface cavities have on the nucle- In this paper, we direct our attention to certain aspects of heat transfer with boiling potassium. Experimental results are given for the superheat required to boil potassium on a stainless steel surface which contains reentrant cavities (surface opening smaller than body of cavity within wall). The effectiveness of these cavities in re- ducing the stable boiling superheat - as compared to the superheat needed to boil potassium on a smooth surface - is demon- strated by the data. In addition, the depen- dence of superheat on heat flux for both the stable boiling and bubble nucleation (incip- ient boiling) modes is shown. Further, a qualitative description of cavity quenching is given which is consistent with data ob- tained during incipient boiling. metals, the surface condition may be the predominate factor in fixing the amount of superheat needed to initiate and maintain SUPERHEAT EQUATION Superheats for boiling potassium and sodium have been reported previously by the authors. These results, which were obtained both in a reflux capsule and in a natural- circulation loop, showed that the surface conditions of the boiler had a substantial influence on the amount of superheat needed for stable boiling. In both types of experi ments, it was found that the stable boiling superheat was correlated fairly well by the following equation when the boiler contained well-defined cylindrical cavities: The influence of surface. conditions on nucleate boiling of liquids (nonmetals as well as metals) has been reported by numer- ous investigators. Thus, Corty and cust- showed the effect of surface roughness for : R-113, diethyl ether, and n-pentane when boiling on copper, nickel, and Inconel sur- faces. Bonilla, Grady, and Avery found that parallel scratches reduced the surface temperature, and thus the superheat, for water and for mercury containing 0.1% sodium. Edwards and Hoffman and Hoffman and Krakoviak“ have demonstrated the effective- ness of various types of surface cavities in reducing the amount of superheat needed to Rrr RErsat (q Revisat 2 * FP sat 20 T, - Tsat = 1 h2g r P sat V A , where T., is the vapor temperature within the bubble, 'Tzat is the liquid saturation temper ature, Peat is the corresponding liquid sat- uration pressure, he is the latent heat of vaporization, o is the surface tension, r is the cavity radius, R is the gas constant, ang dimensionally consistent units are used. T. . : : :. - WI.. Research sponsored by the U. S. Atomic Energy Commission under contract with the Union Carbide Corporation. DISTRIBUTION OF THIS, DOCUMENÉ S UNLUYLTED :: . BLANK PAGE . . 1. * in Equatsiin (1) was shown to correlate the data over altemperature range of 1000 to 1600*F when the heat Clux was relatively small $(<50,000 Btu/tar •ft2) with cavities ranging from 10-3 to 10-8 in. diameter. The results reported herein extend the previous work in that data for both incip. ient and stable boiling were obtained with cavities having a short cylindrical section of nearly circular shape opening at the boiler surface above a flared region in the shape of a truncated cone within the wall. The geometry o:f these reentrant cavities is shown in the inset of Fig. 1. not exactly circular in cross section. The cylindrical portion of the cavities was approximately 0.01 in. long, and the total depth was of the order of 0.03 in. The cavities numbered 1 to 4 are approximately twice the diameter of those numbered 5 to 7. I This arrangement allowed us to investigate the influence of cavt sy size on the boiling • process. Open junction chromel-alumel thermocouples formed the base of eight of • the reentrant cavities; the other six cavi- ties were not instrumented with thermo- couples. EXPERIMENTAL APPARATUS The boiler was heated radiatively by a set of five surrounding, cylindrical, resis- tance heaters. These heaters, which were 4 in. long, were arranged such that each section of the boil.er could be used indepen- dently as the boiler heat-transfer surface. Iwo instrumented cavities were located in each heater region. The results presented were obtained with boiling in the sections containing the cavities designated 3A, 4A, 4B and 5A, 5B, 6B (see Fig. I inset). The results discussed were ootained with potassium boiling in a vertical pipe under natural-circulation conditions. The experimental system is shown schematically in Fig. 1. . 1000 Weds 347 stal ID) and in hisha OPERATING PROCEDURE The loop was fabricated from 1/2-in., schedule 40, type 347 stainless steel pipe (0,840-in. OD X 0.622-in. ID) and was approximately 34 in, wide and 78 in. high. Chromel-alumel thermocouples were attached to the outside pipe surface at 6-in. inter- vals, permitting continuous monitoring of surface temperatures. In adäition, thermo- couples were located at the base of some of the reentrant cavities. A movable thermo- couple probe, mounted in the vertical leg of the loop which contained the boiler section, was used to measure the liquid and vapor temperature in the boiler. An electro magnetic flowmeter was inctalled at the center of the bottom cross meriber of the loop, and a precision pressure gage was attached to the top horizontal section of the loop. An air-cooled condenser, located in the upper half of the vertical leg of the loop opposite the boiler, consisted of six annular sections to which cooling air could be supplied independently. This type of: condenser proved very satisfactory, since it provided accurate control of the cooling rate in the condenser section. Prior to boiling and securing superheat data, the entire loop was brought to a uni- form temperature equal to the saturation tem- perature at which the test was to be run. After thermal equilibrium had been achieved, the power to one of the heaters on the boiler was suddenly advanced to initiate boiling; the power setting for the heater was pre- determined to yield a specific heat flux. From a simultaneous recording of the temper- ature measured by the thermocouple located at the base of the cavities to which the heat flux had been applied and the tempera- ture measured by the thermocouple probe (which was located in the vapor space above the liquid metal surface), the amount of : superheat required to initiate boiling was obtained. During the early phases of this investigation, it was ascertained that the vapor temperature indicated by the thermo- couple probe usually agreed with that cal- culated from the measured vapor pressure to within 3°F. The results reported are all based on a saturation temperature measured by the thermocouple probe. A schematic of the boiler 13 shown in the inset of Fig. 1. The reentrant cavities, arranged in two opposing (though offset) rows of seven, were evenly spaced at 2-in. inter- vals. The cross-sectional dimensions of the cavities are given in Table 1. Note that two The saturation pressure, and thus the .. saturation temperature, was regulated by controlling the condensation rate in the air- cooled condenser. It was found that exceed- ingly close control of the saturation pres- sure could be maintained. Table 1. Dimensions of Reentrant cavities Cavity No. Diameter (in.) Cavity Diameter No. (in.) JA 24 0.007 X 0.008 0.007 X 0.008 1B 2B 0.007 X 0.0077 0.007 X 0.007 The state of flow within the system was observed from the output of the electro- magnetic flowmeter. This device proved to be extremely valuable as it gave an indica- tion when stable boiling ceased and incipient hoilini heman (flow ceases and the boiler . . ...... " .::. .. . .. ": :: :.. " "" " "" . A BLANK PAGE ET T center of the bottom cross member of the loop, and a precision pressure gage was attached to the top horizontal section of the loop. An air-coolec condenser, located in the upper half of the vertical leg of the loop opposite the boiler, consisted of six annular sections to which cooling air could be supplied independently. This type of condenser proved very satisfactory, since it provided accurate control of the cooling rate in the condenser section. at the base of the cavities to which the heat flux had been applied and the temper&- ture measured by the thermocouple probe (which was J.ocated in the vapor space above the liquid metal surface), the amount of : superheat required to initiate boiling was obtained. During the early phases of this investigation, it was ascertained that the vapor temperature indicated by the thermo- couple probe usually agreed with that cal- culated from the measured vapor pressure to within 3°F. The results reported are all based on a saturation temperature measured by the thermocouple probe. A schematic of the boiler is shown in 1 . arranged in two opposing (though offset) rows of seven, were evenly spaced at 2-in, inter- vals. The cross-sectional dimensions of the cavities are given in Table 1. Note that two The saturation pressure, and thus the saturation temperature, was regulated by controlling the condensation rate in the air- cooled condenser. It was found that exceed- ingly close control of the saturation pres- sure could be maintained. Table 1. Dimensions of Reentrant cavities - 1 Cavity Diameter No._ (in.). IA 0.007 X 0.008 24 0.007 X 0.008 34 0.0085 X 0.0085 4 0.007 X 0.007 5A 0.003 * 0.003 6A 0.003 x 0.003 7A 0.0045 X 0.005 Cavity Diameter No. (in.) 1B 0.007 X 0.007 2B 0.007 X 0.007 3B 0.008 X 0.0085 4B 0.0075 x 0.011 5B 0.0045 x 0.0045 GB 0.0035 X 0.004 7B 0.004 X 0.005 The state of flow within the system was observed from the output of the electro- magnetic flowmeter. This device proved to be extremely valuable as it gave an indica- tion when stable boiling ceased and incipient boiling began (flow ceased and the boiler wall temperature began to rise). 1 During the course of the experiments, it was observed, when boiling on the surface which contained the 0.0045-in.-diam cavities (i.e., cavities 5A, 5B and 6B), that cavity 5B controlled the incipient boiling process. It is noted that this cavity is larger than the other two cavities (5A and 6B) anů, con- sequently, should play the predominate role during incipient: boiling; this follows from Eq. (1). Thus, the superheat data reported for the 0.0045-in. cavities are based on temperature measurements made at the base of dimensions, representing the maximum and min- imum diameter of the cavity, are given; un- fortunately, the method used to form the cavities often resulted in a hole which was 22 BLANK PAGE ..... W W W U . n .. . L'A ... "I'Y LA 'T IR A cavity 5B and, for the larger diameter cavi- ties, on the temperature recorded at the base of cavity 3A. SI ti .. Figure 2 shows representative records of the temperatures measured at the base of a reentrant cavity and in the vapor above the boiling alkali metal during incipient and stable boiling. It is eminently evident that the wall temperature fluctuations during in- cipient and stable boiling are drastically different. It was observed that the wall temperature increases at a fairly uniform rate during the early stages of incipient voiling (Path A to B in Fig. 2) and that there was no detectable fluid motion in the system. When the temperature reached Point B, the thermocouple probe', which was located in the vapor above the liquid surface, sud- denly increased in temperature. Simulta- neously, the EM flowmeter indicated that a very unsteady type of motion had been ini- tiated in the system. These events showed that a volume or vapor had been released from the surface of the boiler. At the same time, the temperature at the base of the cavity suddenly decreased (B to C). After this sudden decrease in temperature, the cavity temperature again rose for a period of time (C to D) and then decreased to a temperature near the saturation temperature. Motion persisted in the system during these latter phases of the incipient boiling cycle. We then examine the above equations for the circumstance of small contact angle such as exists for alkali metals. IL O is smaller than 45 deg, Pv – P is larger for a bubble interior to a cavity than for one forming exterior to the cavity. Further, we imagine that a bubble formed exterior to the cavity has just been released from the surface. If the new interface between the vapor and 'liquid is within the cavity, then a higher temperature would be required to maintain equilibrium between the vapor-liquid inter- face than needed for equilibrium exterior to the cavity. If the temperature is not high enough, condensation will take place at the interface, the liquid will penetrate into the cavity, and boiling will cease. This sug- gested mechanism of cavity quenching is sup- ported by observations made in our experiments of the cavity temperature during incipient boiling. Thus, we found that immediately after the release of vapor the cavity temper- ature first decreased as indicated by the Path B-C in Fig. 2 and then increased again along C-D before finally falling to the orig- inal level, E = A. This performance was un- expected and caused us to question initially the validity of the data. However, the same behavior was observed subsequently with other cavities; and we were led to envision the quenching mechanism outlined above. At Point B, the vapor bubble is released from the surface and rapidly grows absorbing the superheat stored in the liquid within the boiler. The sudden drop in temperature at the cavity base (B-C) reflects the transition between the interior and exterior bubble and the sudden expansion of the bubble in the "pool", of superheated liquid. The vapor is then replaced by cooler liquid entering the boiler, and the boiler inside wall tempera- ture decreases. However, uuring this period, the temperature at the base of the cavity in- creases again, since the heat transfer in this region is to the vapor remaining in the cavity rather than to liquid. Finally, as discussed above, when the vapor in the cavity! is completely condensed, the temperature at the cavity base drops abruptly (D-E). . Sometimes, after the occurrence of the above described events, the boiling would continue in a stable mode; while at other times, the incipiency would' persist and the above cycle would be repeated. It was ob- served that incipient boiling would most likely or.cur on a repetitive basis when the saturation temperature and heat flux were small. However, incipierit boiling would occasionally be encountered after an extended period of stable boiling. This latter occur- rence is related to cavity quenching which is . not understood perfectly at this time. How- ever, based on our observations, we believe that the initiation of cavity quenching is related to the overall stability of the system which, in turn, may depend on heat Llux, saturation temperature, system geometry, and fluid properties. Based on the data re- ported in this paper, we offer the following tentative description of the mechanism of cavity quenching with the qualification that (it is subject to revision when we secure additional data, . ....- .. The superheats associated with stable and incipient boiling of potassium at various saturation temperatures and heat fluxes from both the 0.0045- and 0.0085-in.-diam reentrant cavities are shown in Figs. 4 through 6. For comparison, the results obtained in the pres- ent study are shown along with those reported in earlier investigations in Fig. 7. mimi'ww.all here. e ' r First, let us consider the condition of a bubble in equilibrium with its surroundings. We will assume that the bubble has one of the two shapes shown in Fig. 3. Then, for a spherical pubble forming exterior to a sur- face cavity of circular cross section, the equilibrium condition is given by: L... If we consider first the stable boiling results obtained with the 0.0045-in.-diam cavity, we note (Fig. 6) that, as the heat flux increased from 32,000 to 71,000 Btu/ hr.fts, the superheat also increased. Fol- lowing Hsu, if we postulate a thermal bound- ary layer adjacent to the boiler surface (albeit with potassium, the temperature gra. dient is relatively small) and require that . banhei m 20 sino 2 :,: * . - !! * - 2 -- 4 - T- * * 1.1 . il . BLANK PAGE is . ". * RE به دست abov داد دیتات دست درمان قرار دارند مانند دندان بدانند ای - continue in a stable mode; while at other times, the incipiency would persist and the above cycle would be repeated. It was ob- served that incipient boiling would most likely occur on a repetitive basis when the saturation temperature and heat flux were small. However, incipient boiling would occasionally be encountered after an extended period of stable boiling. This latter occur- rence is related to cavity quenching which is not understood perfectly at this time. How- lever, based on our observations, we believe that the initiation of cavity quenching is related to the overall stability of the system which, in turn, may depend on heat : flux, saturation temperature, system geometry, and fluid properties. Based on the data re- ported in this paper, we offer the following tentative description of the mechanism of cavity quenching with the qualification that it is subject to revision when we secure additional data. dan Chee CVTey base (B) reflects the transition between the interior and exterior bubble and the sudden expansion of the bubble in the "pool". of superheated liquid. The vapor is then replaced by cooler liquid entering the boiler, and the boiler inside wall tempera- ture decreases. However, during this period, the temperature at the base of the cavity in- creases again, since the heat transfer in this region is to the vapor remaining in the cavity rather than to liquid. Finally, as discussed above, when the vapor in the cavity is completely condensed, the temperature at the cavity base drops abruptly (D-E). The superheats associated with stable and incipient boiling of potassium at various saturation temperatures and heat fluxes from both the 0.0045- and 0.0085-in.-diam reentrant cavities are shown in Figs. 4 through 6. For comparison, the results obtained in the pres- ent study are shown along with those reported in earlier investigations in Fig. 7. First, let us consider the condition of a bubble in equilibrium with its surroundings. We will assume that the bubble has one of the two shapes shown in Fig. 3. Then, for a spherical bubble forming exterior to a sur- face cavity of circular cross section, the equilibrium condition is given by: 20 sin o = - - P , (2) If we consider first the stable boiling results obtained with the 0.0045-in.-diam cavity, we note (Fig. 6) that, as the heat flux increased from 32,000 to 71,000 Btu/ hr•ftę, the superheat also increased. Fol- lowing Hsu, if we postulate a thermal bound- ary layer adjacent to the boiler surface.... (albeit with potassium, the temperature gra- dient is relatively small) and require that the superheat within this layer be as large as that needed to maintain a bubble in thermal equilibrium, then the superheat (Arsat) must increase with increasing heat flux. Thus, Fig. 6 shows at Tsat ~ 1200°F, Atgat ~ 11°F for g/A = 31,700 Btu/hr.ft2 and ~30°F for g/A = 70,800 Btu/hr.fta. At a somewhat higher saturation temperature (1400°F – the normal boiling temperature of potassium), Alcat varies from ~3°F to ~15°F over the same range of heat flux. At the lowest heat flux level, the data (Fig. 4) are in good agreement with the superheat predicted by Eq. (1), suggesting boiling at a cavity of approximately 0.004-in. where Py, is the pressure within the bubble, Pl is the liquid pressure outside the bubble, and o is the contact angle as defined in Fig. 3. For a bubble interior to a cavity, the equilibrium condition requires: . 20 cos Py - P = (3) suri BLANK PAGE " SM- 2. 2 .- + . CONCLUSIONS From the results presented in this paper, the following conclusions are offered con- cerning the boiling and superheat character- istics of potassium. 1. Reentrant cavities are effective in reducing the stable boiling superheat needed 'to boil potassium. diameter; note that the parameter in Figs. 4 through 7 is the cavity radius rather than the cavity diameter. At the other extreme (Fig: 5 showing data obtained at g/A = 70,865 Btu/hr.ft? with the 0.0045-in.-diam cavity), the experimental data correspond to super- heats predicted for a cavity of 0.002-in. diameter. We know, however, from the exper- imental data that, even at this higher heat Ilux level, the 0.0045-in.-diam cavity was still the active site. An analysis currently under way, but still incomplete, explains this apparent anomaly by including the heat- flux dependence in the theory. Further ex- perimental studies at higher heat fluxes to verify this conclusion are planned. Figure 5 also shows clearly the influence - with stable boiling - of cavity size on AT cat; specifi- cally, we observe about an order of magnitude decrease in the superheat as the cavity size increased from 0.0045 in. to 0.0085 in. 2. For stable boiling, the effective- ness of reentrant cavities in reducing super- heat is no better than the effectiveness of circular cylindrical cavities of approxi- mately the same mouth area. 3. Incipient boiling is more apt to occur at small vapor pressures. 4. Incipient boiling superheat does not show any definite dependence on heat flux. For the incipient boiling mode, no appar- ent trend of the superheat with heat flux can be discerned; and it appears that the super- heat is now more closely related to satura- tion temperatures than to cavity size or heat-flux level. We also see (Figs. 4 – 6) that the superheat associated with stable boiling is only about one-tenth of that ob- served with incipient boiling. Further. studies in this interesting and important area are needed. 5. The magnitude of the stable boiling superheat is related to cavity size and heat · flux; in general, it increases with increasing heat flux and decreases with increasing cavity diameter. " ACKNOWLEDGMENTS The authors wish to acknowledge R. E. Dial and B. J. Sutton for their support in the construction and operation of the experi- mental loop described and Dolores Eden for her conscientious typing of the manuscrípt. pial come back insation for their support ser REFERENCES 1. C. Corty and A. S. Foust, Surface Vari- ables in Nucleate Boiling, AICHE Heat Transfer Conference, St. Louis, Missouri, pp. 1-12, Chemical Engineering Progress Symposium Series, 1955. 2. . Finally, the results obtained in this study for stable boiling from 0.0045-in.-diam reentránt cavities are compared in Fig. 7 with our previous datas, 4 for boiling in a small natural-circulation loop and in a re- Ilux capsule on an as-received surface and on a surface with 0.006-in.-diam cylindrical cavities. All of the results shown in Fig. 7 were obtained in the heat-flux range of..... 25,000 to 35,000 Btu/hr.ft. We conclude that the particular reentrant cavities studied were no more effective in reducing the super- heat than the circular cylindrical cavities. Further, very limited data on incipient Doiling from the cylindrical cavities (Ref. 3, Fig. 6, constituting four data points at a saturation temperature of 1100 to 1150°F showing superheats of 250 to 400°F) yield the same general conclusion. This result can be understood in terms of our previously dis- cussed observations and analysis. If, during stable boiling, the liquid-vapor interface never penetrates the cavity, then the super- heat magnitude is controlled by the surface opening. With incipient boiling, there is indication that the cylindrical neck of the reentrant cavity dominates the performance; again further experiments are needed. ._-. - C. F. Bonilla, J. J. Grady, and G. W. Avery, Pool Boiling Heat Transfer from Scored Surfaces, AICHE Sixth National Heat Transfer Conference, Boston, Mass., pp. 280-288, Chemical Engineering Progress Symposium Series, 1965. 1 - - .- - . . -. ..' J. A. Edwards and H. W. Hoffman, Superheat with Boiling Alkali Metals, pp. 515-534 in Proceedings of the Conference on Applica- tion of High Temperature Instrumentation to Liquid-Metal Experiments, AEC Report ANL-7100, Argonne National Laboratory, 1965. i' . Wh i z , . .- Also shown in Fig. 8 are the results obtained with an as-received commercial stainless steel pipe. The temperature nat- tern in this cases was similar to that ob- served for incipient boiling in the large, H. W. Hoffman and A. I. Krakoviak, Con- vection Boiling with Liquid Potassium, pp. 19-33 in Proceedings of the 1964 Heat Transfer and Fluid Mechanics Institute, Stanford University Press, Stanford, California. BLANK PAGE i LG . . *Pri 7 PAS * 1.biat ILU, **** ! he ..17 :. .. ANTA ... L 4 ie ... * hati LAL MA " *12: 11 13. : 17 - 11 IL T 1 .t P . . wea are needed. The authors wish to acknowledge R. E. Dial and B. J. Sutton for their support in the construction and operation of the experid mental loop described and Dolores Eden for her conscientious typing of the manuscript. REFERENCES 1. C. Corty and A. S. Foust, Surface Vari- ables in Nucleate Boiling, AICHE Heat Transfer Conference, St. Louis, Missouri, pp. 1-12, Chemical Engineering Progress Symposium Series, 1955. Finally, the results obtained in this study for stable boiling from 0.0045-in.-diam reentrant cavities are compared in Fig. 7 with our previous datas, 4 for boiling in a small natural-circulation loop and in a re- flux capsule on an as-received surface and on la surface with 0.006-in.-diam cylindrical cavities. All of the results shown in Fig. 2 were obtained in the heat-flux range of 25,000 to 35,000 Btu/hr.mts. We conclude that the particular reentrant cavities studied were no more effective in reducing the super- heat than the circular cylindrical cavities. Further, very limited data on incipient boiling from the cylindrical cavities (Ref. 3, Fig. 6, constituting four data points at a saturation temperature of 1100 to 1150°F showing superheats of 250 to 400°F) yiela the same general conclusion. This result can be understood in terms of our previously dis- cussed observations and analysis. If, during stable boiling, the liquid-vapor interface never penetrates the cavity, then the super- heat magnitude is controlled by the surface opening. With incipient boiling, there is indication that the cylindrical neck of the reentrant cavity dominates the performance; again further experiments are needed. C. F. Bonilla, J. J. Grady, and G. W. Avery, Pool Boiling Heat Transfer from Scored Surfaces, AICHE Sixth National Heat Transfer Conference, Boston, M888., pp. 280-288, Chemical Engineering Progres Symposium Series, 1965. 3. J. A. Edwards and H. W. Hoffman, Superhea with Boiling Alkali Metals, pp. 515-534 1 Proceedings of the Conference on Applica- tion of High Temperature Instrumentation to Liquid-Metal Experiments, AEC Keport ANL-7200, Argonne National Laboratory, 1965. H. W. Hoffman and A. I. Krakoviak, Con- vection Boiling with Liquid Potassium, pp. 19-33 in Proceedings of the 1964 Heat Transfer and Fluid Mechanics Institute, Stanford University Press, Stanford, California. Also shown in Fig. 8 are the results obtained with an as-received commercial stainless steel pipe. The temperature pat- tern in this case was similar to that ob- served for incipient boiling in the large, artificial cavities with the exceptions that the large amplitude wall temperature oscilla- tions were continuous and occurred at high saturation temperatures. Thus, we can con- strue this behavior as a form of stable boiling involving very high superheats as predicted by Eq. (1) with rapid quenching o the natural shallow, very small diameter cavities. The data also demonstrate that either type of artificial cavity signifi- cantly reduces the superheat. 5. R. E. Baizhiser, Investigation of Boiling Liquid Metal Heat Transfer, Report RTD-TDR-63-4130, University of Michigan, 1963. P. J. Marto and W. M. Rohsenow, Effects of Surface Conditions on Nucleate Pool Boiling of Sodium, ASME Paper No. 65-HT-51, ASME-AICHE Heat Transfer Con- ference, Los Angles, California, 1965. de video tits we vinden ... HT BLANK PAGE . . ..ru ' :D T- . 22 . .' TW -Ar. UFFLE 7. B. S. Petukhov, S. A. Kovalev, and V. M. Zhukov, Study of Sodium Boiling Heat Transfer, Proceedings of Third Inter- national Heat Transfer Conference, vol. V, ! pp. 20–91, American Institute of Chemical Engineers, New York, 1966. Y. Y. Hsu, On the Size Range of Active Nucleation Cavities on a Heating Surface, Trans. ASME, Journal of Heat Transfer, Series C, 84: 207-216 (1962). LEGAL NOTICE The report was prepared as an account of Government sponsored work. Nolther the United status, nor the Commisslon, nor any person soting on behalf of the Commission: A. Makes my warranty or representation, exprussed or implied, wit, respect to the accu- racy, completeness, or wefulness of the information contained in this report, or that the use of any information, apparatus, method, or procou discloned in the report may not latringe privately owned righto; or B. Asmmus May Ilabiliuos with respeot to the um of, or for damages resulting from the un of any information, apparatus, method, or procou disclound in wis report. A. vand in the abovo, "person acting on behalf of the Commisslon" includes way on- ployee ur contractor of the Commission, or employ of such contractor, to the extent that auch employs or contractor of the Commission, or employme of such contractor preparı, disseminates, or provider aucou to, any information pursuant to do employmeat or contract with the Commission, or die employment with such contractor. ADJUSTABLE THERMOCOUPLE PROBET PURGE LINE PRESSURE "LINE -1/2" SCH 40 PIPE CONDENSER I • BACK-FILL PLUG 1AZ 2 in. 2B 2422 in. 11 Z3B 3AZU 2 in. 4B 4A2 in. 15B SAU 2 in. | -- 6B GAT2 in. - UITB . . . ............ ..... 7A7Wiin. BOILER THERMOCOUPLE JP DETAILED VIEW OF REENTRANT CAVITIES -FILL LINE : HEATER D -SUMP UT LE.M. FLOW METER " (1) INCIPIENT STABLE CAVITY TEMPERATURE . . E 4% fixe Pin Tsot -T sat FLOWMETER OUTPUT TIME TIME . 175-0 100ML. 1.3.4.45-14 Pure of 3 / 5 67-4304 . ..in 2 n ܥܥܥܥܥܥܥܥܥܥܢ EXTERIOR · BUBBLE INTERIOR BUBBLĘ c. 7-1300 . resuce to 3% 68 % found to lcão). . MEN OP. V I . .- --- A IV _ O INCIPIENT © STABLE . v=0.00001 in. Lego 80000000 0.00003 lo . 0.0001 SUPERHEAT, Tw-Tsat (°F) 0.0003 .. 0.001 - 0.003 0.01 1000 9200 1400 SATURATION TEMPERATUR 1600 (9) UL A RNIEREM ER . . . - 100 O INCIPIENT 0.0045 in. DIAMETER A INCIPIENT 0.0085 in. DIAMETER O STABLE 0.0045 in. DIAMETER - A STABLE 0.0085 in. DIAMETER p = 0.00003 in.! <0.0001 SUPERHEAT, Tw-Tsat (°F) 0.0003 0000 AA - AAAAA - <0.001 0.003 0.01 1000 1600 1200 1400 SATURATION. TEMPERATURE (°F). (bi 8 8 CQ/A (B/hr-f12) STABLE INCIPIENT , 31,700 48,000 . 58,700 70,800 DOO 8 *r = 0.00003 in. - 8 0.0001 8 0- SUPERHEAT, Tv-Tsai (°F) !1°12 III YLLLL 0.0003 8 nes AV-AA=0 APAA A+ - A-A-A-A4- A A <0.001 0 - 0.003 9 0.01 8- 1000 1200 1400. .: 1600 SATURATION TEMPERATURE (°F) ... - - ... .1917. - ...WT, VT " - O AS-RECEIVED O CYLINDRICAL CAVITIES, 0.006 in. DIAM A REENTRANT CAVITIES, 0.0045 in. DIAM = 0.00001 in.--- 0.00003 K0.0001 . SUPERHEAT, TW-Tsai (°F 0.0003 48 8 Stone RAZILIZO toe_0001- 0.003 1000 1200 . 1400 . . 1600 CATUDATION TEMDCONInner .....: . . . U r ...? i . , . . .. 5 . : . ' - 1 . :. : .'. ins " P NU . . CA 7 > EUR LL SU in M 8 / 9 /67 DATE FILMED END m2 71 R- . . . ?