https://hdl.handle.net/1721.1/61396 2026-04-09T03:57:45Z https://hdl.handle.net/1721.1/62142 Film boiling on the inside of vertical tubes with upward flow of the fluid at low qualities Dougall, R. S.; Rohsenow, Warren M. Flow regimes, local heat transfer coefficients, and temperature distributions along the wall have been studied for film boiling inside a vertical tube with upward flow of a saturated liquid. The area of interest has been limited to cases of constant heat flux from the tube wall, small inlet liquid velocities, and film boiling which completely covers the entire heated portion of the tube. The last restriction means that there is no large region of nucleate boiling prior to the film boiling section. A visual test section made of electrically conducting glass tubing has been used for flow visualization studies at low qualities. Visual observations with this test section have indicated that the flow regime is annular with liquid in the center and vapor along the walls of the tube. Based on interpretations of temperature distribution data, it has been concluded that the annular flow regime changes at higher qualities to one of dispersed flow--where the liquid is dispersed in the form of drops through a predominately vapor flow. Two different diameter test sections made of stainless steel and heated electrically have been used to obtain experimental data of temperature distributions along the tube wall and local Nusselt numbers for different heat fluxes and flow rates. The fluid used in all the experimental tests has been freon 113. For the larger tube, 0.408" I.D., heat fluxes have been varied from 14,400 to 25,600 Btu/hr-ft2 , and mass velocities have been varied from 482,000 to 818,000 lbm/hr-ft2 . For these conditions, values of heat transfer coefficients from 24.0 to 41.4 Btu/hr-ft 2 -*F and values of Tw-Ts from 407 to 697*F have been obtained. These conditions have resulted in exit qualities up to 10 per cent. For the smaller tube, 0.180" I.D., heat fluxes have been varied from 22,500 to 41,800 Btu/hr-ft2 , and mass velocities have been varied from 332,000 to 398,000 lbm/hr-ft". For these conditions, values of heat; (cont.) transfer coefficients from 27.0 to 87.5 Btu/hr-ft2 -*F and values of Tw-T from 261 to 883*F have been obtained. These conditions have resulted in exit qualities up to 50 per cent. A theoretical derivation based on an annular flow model with turbulent flow in the vapor film has given good agreement with the experimental data at low qualities. A dispersed flow theory using a modified form of the Dittus and Boelter-McAdams equation seems to be an asymptote which the experimental data approaches with increasing qualities. 1963-01-01T00:00:00Z https://hdl.handle.net/1721.1/62141 The mechanism of void formation in initially subcooled systems Griffith, P.; Snyder, George A. When an initially subcooled, water filled system undergoes a transient in heat flux or pressure such that bubbles form, the most important variable which determines the volume of the resulting void is the number of bubbles that is formed. In this report the number of bubbles that are formed is shown to be a function of the surface micro-configuration, the contact angle and the history. A method of specifying the history is developed, experiments are run and the general correctness of the history specification is shown to be correct. Order of magnitude values of the limiting wall superheats as a function of the surface history and configuration are presented, but the reproducibility of the experiments is not found to be high. 1963-01-01T00:00:00Z https://hdl.handle.net/1721.1/62023 The mechanism of void formation in initially subcooled systems Griffith, P.; Snyder, George A. When an initially subcooled, water filled system undergoes a transient in heat flux or pressure such that bubbles form, the most important variable which determines the volume of the resulting void is the number of bubbles that is formed. In this report the number of bubbles that are formed is shown to be a function of the surface micro-configuration, the contact angle and the history. A method of specifying the history is developed, experiments are run and the general correctness of the history specification is shown to be correct. Order of magnitude values of the limiting wall superheats as a function of the surface history and configuration are presented, but the reproducibility of the experiments is not found to be high. 1963-01-01T00:00:00Z https://hdl.handle.net/1721.1/61503 Model of critical heat flux in subcooled flow boiling Fiori, Mario P.; Bergles A. E. The physical phenomenon occurring before and at the critical heat flux (CHF) for subcooled flow boiling has been investigated. The first phase of this study established the basic nature of the flow structure at CHF. A photographic study of the flow in a glass annular test section was accomplished by using microflash lighting and a Polaroid camera. The results showed that the flow structure at CHF for high heat flux (1 x 106 - 5 x 106 Btu/hr-ft2), high subcooling (50-110 *F), at low pressures (less than 100 psia) was slug or froth flow depending on the mass velocity. Nucleation was shown to exist in the superheated liquid film. Pin-holes in the burned-out test sections suggested that the CHF condition was extremely localized. Flow regime studies in tubular and annular geometries, using an electrical resistance probe, provided further evidence of the slug or froth nature of the flow, and also showed that dryout of the superheated liquid film was not responsible for CHF. Since this evidence was contradictory to previously formulated models of CHF,a new model was proposed: Near the CHF condition, nucleation is present in the superheated liquid film near the surface. As a large vapor clot passes over the surface, these nucleating bubbles break the film and cause a stable dry spot which results in an increased local temperature. As the vapor finally passes the site, the dry spot is quenched by the liquid slug, and the temperature drops. At CHF, the volumetric heat generation, slug frequency, and void fraction are such that the temperature rise resulting from the dry spot is greater than the temperature drop during quenching. An unstable situation results where the temperature of this point continues to rise when each vapor clot passes the site until the Leidenfrost temperature is reached, at which point quenching is prevented and destruction is inevitable.; (cont.) A new method of measuring surface wall temperatures, in conjunction with high speed (Fastax) 16 mm movies, confirmed the microscopic features of the proposed model. At CHF, the wall temperature cyclically increased with the same frequency as the slug-vapor bubble passage. Destruction finally resulted as the temperature increased beyond the Leidenfrost point. An analytical investigation based on an idealized model demonstrated that the cyclical nature of the temperature increase at CHF could be predicted with appropriate flow pattern inputs. A parametric study using the program indicated that heater thickness and heater material should affect the CHF. It was shown that the proposed model appears to be consistent with parametric trends, i.e. mass velocity, pressure, subcooling, diameter, length, and surface tension. The model indicated that the CHF for thicker walled tubes, keeping all other conditions the same, would increase. CHF tests were conducted which confirmed that thicker walled tubes (0.078 vs. 0.012 in. ) had CHF up to 58 percent higher than thin walled tubes. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61502 The influence of return bends on the downstream pressure drop and condensation heat transfer in tubes Traviss, Donald P.; Rohsenow, Warren M. The influence of return bends on the downstream pressure drop and heat transfer coefficient of condensing refrigerant R-12 was studied experimentally. Flow patterns in glass return bends of 1/2 to 1 in. radius and 0.315 in. I. D. were examined visually and photographically using a high frequency xenon light source. Local pressure drop and heat transfer measurements were made along a horizontal 14 1/2 ft. test section immediately following the return bend. The refrigerant mass flux ranged from 1.32 X 105 to 4.58 X 105 lbm/hr-ft 2, saturation temperature from 90 to 107*F, and return bend quality from 0.24 to 1.0. The pressure drop and heat transfer data were compared to previous data for condensation without return bends. Effects on the downstream pressure drop and heat transfer were found to be small, if not negligible. 1971-01-01T00:00:00Z https://hdl.handle.net/1721.1/61501 The mechanism of heat transfer in nucleate pool boiling Han, Chi-Yeh; Griffith, P. A criterion is developed for bubble initiation from a gas filled cavity on a surface in contact with a superheated layer of liquid. It is found that the temperature of bubble initiation on a given surface is a function of the temperature conditions in the liquid surrounding the cavity as well as the surface properties themselves. It is also found that the delay time between bubbles is a function of the bulk liquid temperature and the wall superheat and is not constant for a given surface. By consideration of the transient conduction into a layer of liquid on the surface, a thermal layer thickness is obtained. With this thickness and a critical wall superheat relation for the cavity, a bubble growth rate is obtained. Bubble departure is considered and it is found that the Jakob and Fritz relation works as long as the true (non-equilibrium) bubble contact angle is used. The effect on the departure size of the virtual mass in the surrounding liquid is found to be negligible at one gravity. That is, at one gravity the primary effect of bubble growth velocity on bubble departure size is found to be due to contact angle changes. The initiation, growth and departure criterions are each experimentally, individually, checked. They are then used to compute the heat transfer near the knee of the boiling curve using only an experimental determination of the number of bubbles as a function of wall superheat and other known quantities. Finally the q vs. Tw - Tsat relation is computed and measured and compared. The comparison is satisfactory. 1962-01-01T00:00:00Z https://hdl.handle.net/1721.1/61500 The effects of surface instabilities on laminar film condensation Gerstmann, Joseph; Griffith, P. Heat transfer rates for laminar film condensation of Freon-1l3 were measured on the underside of horizontal surfaces, inclined surfaces, and vertical surfaces. Several distinct regimes of flow were observed. On the underside of horizontal surfaces, the interface is best described as a fully established Taylor Instability. At slight angles of inclination there are three regimes of flow. Near the leading edge, the interface is smooth and waveless. Next there is a region of developing waves which are best described as longitudinal ridges. As the ridges grow in amplitude, drops are formed at the crests and subsequently fall from the surface. Beyond the point at which drops first fall, a third regime exists which can be considered to be a fully established state, independent of distance from the leading edge of the surface. At moderate angles of inclination and up to the vertical, "roll waves" appear a short distance from the leading edge. An analysis is presented which considers the surface waves to be fully established flows, resulting from bounded instabilities. It is shown that the shape of the interface can be determined without investigating the stability of the unperturbed film. The analysis results in an equation for the shape of the interface which is used to determine the average reciprocal film thickness, hence the heat transfer coefficient. The results of the analysis are valid for condensation on the underside of horizontal surfaces and slightly inclined surfaces. The wavelengths predicted by the analysis are in fair agreement with the experimentally observed wavelengths. The observed heat transfer rates agree quite well with the theory. 1965-01-01T00:00:00Z https://hdl.handle.net/1721.1/61499 Two phase pressure drop in inclined and vertical pipes Griffith, P.; Lau, Chun Woon.; Hon, Pou Cheong.; Pearson, John Franklin. A method of calculating the pressure drop in inclined and vertical oil-gas wells is proposed. The data used to establish the method is from a variety of sources but is largely from air and water flowing in systems close to one atmosphere in pressure and in pipes from 1 to 2 inches in diameter. All inclinations from vertical to almost horizontal are included. The method proposed is used to calculate the pressure distribution in ten oil and gas wells. The predictions for the overall pressure drop are good to +/- 10% for these wells. 1973-01-01T00:00:00Z https://hdl.handle.net/1721.1/61498 Dispersed flow film boiling Yoder, Graydon L.; Rohsenow, Warren M. Dispersed flow consists of small liquid droplets entrained in a flowing vapor. This flow regime can occur in cryogenic equipment, in steam generators, and during nuclear reactor loss of coolant accidents. A theoretical analysis of dispersed flow film boiling has been performed using mass, momentum and energy conservation equations for both phases. A numerical solution scheme, including wall-to-drop, vapor to drop, and wall-to-vapor heat transfer mechanisms was used to predict wall temperatures for constant heat flux, vertical upflow conditions. Wall temperature predictions were compared to liquid nitrogen, Freon-12 and water data of four separate investigators with reasonable results. A local conditions solution was developed by simplifying the governing equations, using conclusions from the numerical model. A non-dimensional group was found which solely determined the non-equilibrium with the flow, and allowed hand calculation of wall temperatures. The local conditions solution was compared to data taken by five investigators with good results. 1980-01-01T00:00:00Z https://hdl.handle.net/1721.1/61497 Investigation of heat transfer augmentation through use of internally finned and roughened tubes : final summary report Bergles A. E.; Brown, G. S.; Lee, R. A.; Simonds, Richard R.; Snider, Warren Dean. This report summarizes a three-year program concerned with obtaining basic design information for tubes having a random roughness on the inside wall (RID) and tubing having continuous internal fins (Forge Fin). Test apparatus and procedures were developed to obtain accurate heat-transfer and friction data for a wide variety of tube geometries using water as the test fluid. For the random roughness the heat-transfer coefficient was above the smooth tube value, for comparable flow conditions, by over 60 percent at a Reynolds number of 30,000. Larger percentage improvements can be expected for higher Reynolds numbers and for fluids having higher Prandtl numbers. Improvements in performance, based on equal pumping power for augmented and smooth tubes, of about 50 percent were observed. The heat-transfer characteristics for tape-generated swirl flow through rough tubes were investigated in order to determine the interaction of swirl flow and roughness effects. For the particular range of parameters covered, for equal flow rates, the maximum improvement in heat transfer with swirl flow in smooth tubes was 70 percent, whereas with swirl flow in rough tubes, the improvement was as much as 100 percent. The heat-transfer coefficient for rough tube swirl flow was accurately correlated by a modification of an additive expression previously suggested for prediction of smooth tube swirl flow data.; (cont.) The test program for internally finned tubes established that short spiralled fins produce the greatest improvement in heat transfer. On the basis of equal flow conditions, the heat transfer was improved by over 200 percent; while at equal pumping power, the performance was as high as 170 percent. These improvements, which are attributed to increased area and turbulence promotion, appear to equal the improvements displayed by any of the schemes used to augment heat transfer inside tubes. In order to bring the augmentation problem into perspective, a discussion of data for other types of roughness and finning is included. 1970-01-01T00:00:00Z https://hdl.handle.net/1721.1/61496 Unsteady momentum fluxes in two-phase flow and the vibration of nuclear reactor components Yih, Tien Sieh; Griffith, P. The steady and unsteady components of the momentum flux in a twophase flow have been measured at the exit of a vertical pipe. Measured momentum flux data has been machine processed by standard random vibration techniques to obtain the power spectral density curves. From these curves, the predominant frequency and the rms value of the unsteady momentum flux have been obtained. The effects of the average flow velocity, volumetric quality, system pressure, flow channel size and geometry on the unsteady momentum fluxes have been observed. It has been found that the fluctuation of momentum fluxes is important only in the low frequency range. The maximum values of unsteady momentum fluxes appeared in either the high void slug flow or the low void annular flow regime. The experimental results have been correlated and suggestions have been made for constructing the power spectral density curve of momentum fluxes under untested conditions. In the sample problems, using the experimental results, the effect of the unsteady momentum fluxes on a steam generator U-tube and a reactor fuel rod has been studied. The amplitudes of the structural vibrations resulting from the two-phase excitation have been found. In addition, it has also been found that there is a possibility of unstable vibrations owing to a nonlinear restoring force on the mechanical system. This nonlinearity is due to the unsteady component of the momentum flux in the flow past the system. In both examples, the major vibrations occurred in a narrow frequency band around the natural frequency of the mechanical system. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61495 Heat transfer and pressure drop in tape generated swirl flow Lopina, Robert F.; Bergles A. E. The heat transfer and pressure drop characteristics of water in tape generated swirl flow were investigated. The test sections were electrically heated small diameter nickel tubes with tight fitting full length Inconel tapes of twist ratios from 2. 48 to 9. 2 inside diameters/180 of tape twist. Heat transfer coefficients and friction factors were determined for non-boiling forced convection heating and cooling while overall pressure drop information and curves of heat flux versus wall superheat were determined for surface boiling conditions. Improvements in heat transfer for equal flow rates of up to 851c were observed for the non-boiling swirl flows with heating, but the improvement with cooling was substantially less. Compared on the basis of equal pumping power, improvements in heat transfer of up to 351c were observed for the tighter tape twists. A method for predicting the heat transfer coefficient for non-boiling swirl flows was developed. It was based upon the theory that the improvement was due primarily to: 1) the increased flow path created by the tape, 2) the increased circulation created with heating due to the buoyancy effect set up by the large centrifugal force present, and 3) the fin effect of the tape. The experimental results of this and previous swirl flow investigations were in good agreement with the analytical prediction. The surface boiling characteristics of swirl flow were found to be similar to those observed in straight flow. The boiling curves for various velocities were asymptotic to a fully developed line at high wall superheats, and the visually observed point of incipient boiling and the transition to the fully developed boiling asymptote were predictable by conventional straight flow methods. It was concluded, therefore, that the dominant surface boiling heat transfer mechanism was similar for both swirl and straight flow.; (cont.) For non-boiling swirl flows, the decrease in the pressure drop with heating was slightly less than is usual with straight flows, while the increase in the pressure drop with surface boiling was substantially less. A method for predicting the difference in each case is presented. 1967-01-01T00:00:00Z https://hdl.handle.net/1721.1/61494 Liquid and gas distributions in a two-phase boiling analogy Wallis, Graham B.; Griffith, P. This report contains a description of the design and operation of an experimental appats for the analysis of two-phase flows similar to thom occuwring in boiler tubes at low pressures. Velocity and denuity profiles of air-water mixtures are determined across a passage in which boiling Ocditiosis are simulated by puping air through porous walls into a water strea. Photographs of the flow patterns are also presented as a qualitative check on the quantitative data. The results obtained are: 1) Vslocity and concentration profiles of the two phases for various values of the flow rates of each. 2) A classification of the flow into several patterns or regimes with quantitative data describing each regims. The data is suitable for use In acapering the physical phenen with athematical models and for developing a more accurate theoretical treatment of the flow of boiling fluids in heated channels. 1958-01-01T00:00:00Z https://hdl.handle.net/1721.1/61493 Mechanism of dropwise condensation Umur, Aydin; Griffith, P. From a study of surface phenomena, information is obtained about conditions under which net condensation can occur. An experimental examination of the surface, using an optical method capable of detecting thin films of molecular dimensions, shows that no film greater than a monolayer in thickness exists on the area between the drops. Wetted pits and grooves in the surface are considered to be the most probable drop nucleation sites. A model for drop growth gives results that are compatible with experimentally observed values, and show the growth rate to be a function of the vapor pressure. 1963-01-01T00:00:00Z https://hdl.handle.net/1721.1/61492 A study of system-induced instabilities in forced-convection flows with subcooled boiling Maulbetsch, John S.; Griffith, P. A combined analytical and experimental program was carried out to investigate the problem of hydrodynamic stability of forcedconvection flows with boiling. The study was restricted to the flow of water in small channels (<".250 I.D.), high length-to-diameter ratios (25-200), moderate temperature and pressure (Ti ~ 70[degree] F; pex< 60 psia), and primarily directed toward subcooled, local boiling (hex< hsat.liq.). Two types of instability were recognized: The first, a nonrecurring excursive instability, and the second, an oscillatory instability in which the operating point varies in a sustained, repetitive way around the original condition. The excursive behavior was predicted on the basis of a comparison of the slopes of the pressure drop vs. flow rate curve of the heated section and the external system. The criterion for stability was simply that the slope of the external system characteristic be more strongly negative than that of the heated section. This was verified experimentally. This excursion was found to be the limiting condition on the heat flux in a parallel channel system. The oscillatory instability was investigated analytically through the use of a linearized, lumped parameter model in which steady-state measurements of the pressure drop were assumed to be valid in characterizing the transient characteristics of the heated section. For cases in which the energy storage mechanism was a compressible volume either upstream of, or within, the heated section, a critical slope of the pressure drop vs. flow rate curve in the heated section could be computed, as well as an associated frequency. The analytical results were verified on a single-tube apparatus with controlled compressibility, and the onset of the instability agreed well with the theoretical predictions. The criterion was applied to the data of a number of other investigations and gave excellent correlation in all cases where the restriction of subcoole 1965-01-01T00:00:00Z https://hdl.handle.net/1721.1/61491 Burnout in forced convection nucleate boiling of water Reynolds John Mitchell Data are presented for burnout in forced coivection nucleate boiling of water at pressures above 500 psia. A dimensionless correlation is devised for. the M.I.T. data which is found to be valid for certain recent data reported by Argonne National Laboratory. 1957-01-01T00:00:00Z https://hdl.handle.net/1721.1/61490 Forced-convection, dispersed-flow film boiling Hynek, Scott Josef; Rohsenow, Warren M.; Bergles A. E. This report presents the latest results of an investigation of the characteristics of dispersed flow film boiling. Heat transfer data are presented for vertical upflow of nitrogen in an electrically heated tube, 0.4 in. I.D. and 8 ft long. Heat fluxes up to 18,000 Btu/ft 2-hr and mass fluxes up to 200,000 lbm/ft 2-hr were investigated. By variation of the startup procedure, it was possible to operate in two distinct regimes of film boiling. By preheating the tube before introducing the flow, film boiling was observed throughout the test tube. If the flow was established before applying power, film boiling was initiated downstream of the inlet. For similar conditions, the local heat transfer coefficients were different in the two cases due to the different degrees of thermal nonequilibrium. The data for both regimes were satisfactorily predicted by a modified version of the nonequilibrium model presented in earlier reports. The model was also applied to available data for methane, propane, and water. By modification of the empirical constant governing the direct wall-to-droplet heat transfer, these data were generally predicted to within 10 percent. The tests were repeated with tight-fitting, full-length twisted tapes installed in the test tube. Considerable augmentation of the heat transfer was achieved, with the heat transfer coefficient being increased by as much as a factor of 3 with the tightest tape twist. For the higher mass fluxes, it was observed that the tape promoted droplet deposition to such an extent that a continuous liquid film could be reestablished on the wall near the test section exit. The tape-generated swirl flow did not improve the "burnout" condition, primarily due to the fact that a liquid streamer forms on the twisted tape. The semiempirical model gave a reasonable prediction of the heat transfer coefficient when the effects of the swirl flow were included. 1969-01-01T00:00:00Z https://hdl.handle.net/1721.1/61489 The deterioration in heat transfer to fluids at supercritical pressure and high heat fluxes Shiralkar, B. S.; Griffith, P. At slightly supercritical pressure and in the neighborhood of the pseudo-critical temperature (defined as the temperature corresponding to the peak in specific heat at the operating pressure), the heat transfer coefficient between fluid and tube wall is strongly dependent on the heat flux. For large heat fluxes, a marked deterioration takes place in the heat transfer coefficient in the region where the bulk fluid temperature is below and the wall temperature above the pseudo-critical temperature. An analysis has been developed, based on the integration of the transport equations, to predict the deterioration in heat transfer at high heat fluxes, and the results have been compared with the previously available experimental results for steam. Experiments have been performed with carbon dioxide for additional comparison. Limits of safe operation in terms of the allowable heat flux for a particular flow rate have been determined both theoretically and experimentally. Experiments with twisted tape inserted in the test section to generate swirl have shown that the heat transfer rates can be improved by this method. Qualitative visual observations have been made of the flow under varying conditions of heat flux and flow rate. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61488 The bubbly-slug transition in a high velocity two phase flow Griffith, P.; Snyder, George A. A possible mechanism for the transition between bubbly and slug flow is proposed and tested in a simulated slug flow system. No sudden collapse of slug flow with increasing velocity is found and it is concluded that: a. Slug flow is generally stable at voids greater than 35%. b. Bubbly flow at voids higher than this is a result of entrance conditions. c. Visual observations of bubbly flow in unheated systems at higher voids are most likely faulty. No simple asymptote limit or criterion that would predict the location of the bubbly-slug transition was found. 1964-01-01T00:00:00Z https://hdl.handle.net/1721.1/61487 Description of boiling project burnout detector Raymond, M. W.; Reynolds John Mitchell Introduction: In order to effectively utilize the high heat flux available through the mechanism of nucleate boiling in forced convection heat transfer, it is of primary importance that the maximum flux or "burnout" conditions be known for the liquid under consideration. It is a characteristic of the boiling heat transfer process that, as attempts are made to exceed the burnout heat flux, the conditions at the heat transfer surface become such that the heat transfer coefficient decreases with increasing temperature difference between the wall and fluid. If the apparatus in which this process occurs is not of the type in which a constant temperature is imposed, another equilibrium point will be reached at a significantly higher wall temperature. In the case of water at pressures of atmospheric and higher, the wall temperature assumed in the new equilibrium state is high enough to cause failure in all but the most conservatively designed apparatus. Because of the unstable nature of the boiling process beyond the maximum vs. temperature difference on the q/A vs. L T curve, once the burnout temperature difference is exceeded, small power reductions will not save the heat exchanger from the major portion of the incipient temperature jump. Power must be reduced to a relatively low level to insure that excessive temperature will not be developed in the equipment. The time in which this power reduction must be accomplished depends on the particular flux and the heat capacity of the system being used; however, in most practical cases, this time can be expected to be extremely short. 1957-01-01T00:00:00Z https://hdl.handle.net/1721.1/61486 The Measurement, interpretation and use of unsteady momentum fluxes in two-phase flow. Yih, Tien Sieh; Griffith, P. The steady and unsteady components of the momentum flux in a two-phase flow have been measured at the exit of a vertical pipe by means of an impulse technique using a turning tee and beam. Different electrical filters have been tried in the recording circuit for eliminating the signals around the natural frequency of the beam system. A special filter set has been designed to approximate the inverse of the transfer function of the beam system. Thus the signals recorded after passing through the beam-filter combination can be considered as the excitation times a constant. Two different analog methods have been used to analyze the random signals for obtaining some statistical quantities such as the predominant frequency and the rms value of the unsteady momentum flux. These quanities are useful in some applications involving two-phase flow. In the preliminary measurements of the unsteady momentum flux for the adiabatic up flow of an air-water mixture in a 5/8 inch pipe, the greatest unsteadiness of momentum flux appeared in the quality range of one per cent to six per cent. Above ten per cent quality no appreciable fluctuation has been detected. In an example problem, using the preliminary results, the effect of the unsteady momentum flux on a fuel rod has been studied. The amplitude of the vibrations resulting from the two-phase excitation has been found. In addition it has also been found that there is a possibility of unstable vibrations due to a nonlinear restoring force on the fuel element. This nonlinearity is due to the unsteady component of the momentum flux in the flow past the rod. 1967-01-01T00:00:00Z https://hdl.handle.net/1721.1/61485 Film boiling of saturated liquid flowing upward through a heated tube : high vapor quality range Laverty, W. F.; Rohsenow, Warren M. Film boiling of saturated liquid flowing upward through a uniformly heated tube has been studied for the case in which pure saturated liquid enters the tube and nearly saturated vapor is discharged. Since a previous study at the M.I.T. Heat Transfer Laboratory covered the case in which only a small percentage of the total mass flow is vaporized, this investigation has been concentrated on film boiling in the region where the vapor quality is greater than 10 percent. Visual studies of film boiling of liquid nitrogen flowing through an electrically conducting pyrex tube have been made to determine the characteristics of the two-phase flow regimes which occur as a result of the film boiling process. It was found that the annular flow regime with liquid in the core and vapor between the liquid and the tube wall, which exists at very low qualities, is broken up at higher qualities to form a dispersed flow of droplets and filaments of liquid carried along in a vapor matrix. A stainless steel test section having a .319 inch ID and heated electrically, has been used to obtain experimental data of wall temperature distributions along the tube and local heat transfer coefficients for different heat fluxes and flow rates with liquid nitrogen as the teit fluid. Heat flux has been varied from 3500 2 to 30000 BTU/hr-ft and mass velocity from 70000 to 210000 lbm/hr-ft2. From these tests, values of wall superheat, (Tw -Ts ), from 200 to w S 975[degree]F and heat transfer coefficients from 11.1 to 65.5 BTU/hr-ft2-[degree] F have been obtained. A theoretical derivation using the Dittus-Boelter equation as an i,7mptote for the heat transfer to pure vapor has demonstrated that a significant amount of vapor superheat is present throughout the film boiling process. The mechanism of the heat transfer process in the dispersed flow region has been described by a two step theory in which 1) all ofthe heat from the wall is transferred to the vapor and 2) heat i 1964-01-01T00:00:00Z https://hdl.handle.net/1721.1/61484 Thermal contact conductance in a vacuum Yovanovich, M. Michael The object of this work is to develop analytically equations by which one could predict the thermal contact conductance between contiguous surfaces operating in a vacuum environment. In this work the solution to the problem is obtained by considering that any surface can be modelled as being either: 1) nominally-flat but rough, 2) a smooth surface having cylindrical waviness, 3) a smooth surface having spherical waviness, or 4) a surface having either cylindrical or spherical waviness plus roughness. Since the radiative heat transfer and the conduction through the interstitial fluid are negligible, the conduction of heat across the metal contact spots is the dominant mechanism. It is considered that the prediction of thermal contact conductance must be approached by: 1) examining the surface geometry, 2) proposing mathematical models for the solution of the heat transfer problem, 3) determining the surface parameters from deformation analysis, and 4) obtaining experimental data to substantiate the proposed models. The surface analysis is actually a critical examination of profiles of real surfaces as obtained by profilometers. From such profiles it is proposed that real surfaces can be idealized by assuming that any surface is a combination of a wavy and rough component. The thermal analysis is based upon the models proposed and the solutions for the steady-state condition are obtained for the various models and the appropriate boundary conditions. Certain surface parameters appear in the thermal contact conductance equations, which require that an analysis of the deformation of the surface under load be undertaken. The deformation analysis is separated into two regimes: 1) purely elastic and 2) purely plastic. The surface parameters are then determined as functions of the applied load for the proposed models under the restrictions of pure elastic or pure plastic deformation. 1965-01-01T00:00:00Z https://hdl.handle.net/1721.1/61483 Influence of surface roughness and waviness upon thermal contact resistance Yovanovich, M. Michael; Rohsenow, Warren M. This work deals with the phenomenon of thermal resistance between contacting solids. Attention is directed towards contiguous solids possessing both surface roughness and waviness. When two such surfaces are brought together under load, they actually touch at isolated microcontacts, and the resulting real area is the sum of these microcontacts. Because of the waviness the microcontacts are confined to a region called the contour area which may occupy some fraction of the total available area. The non-uniform pressure distribution over the contour area results in microcontacts which vary in size and density. In the absence of an interstitial fluid and negligible radiation heat transfer, all the heat crossing the interface must flow through the microcontacts. A thermal analysis, based on size and spatial distribution, results in a thermal resistance equation which differs from previously developed theories. The equation is verified by liquid analog tests which show that the size and spatial distribution are very significant.; (cont.) A surface deformation analysis considers the influence of surface roughness upon the elastic deformation of a rough hemisphere. An equation is developed which shows the extent of the contour area as a function of the surface geometry, the material properties, and the applied load. The equation is compared with existing theories and qualitatively checked against experimental results. Experimental heat transfer data were obtained to verify the thermal and deformation theories. The agreement between theory and test is quite good over a large range of surface geometry and applied loads. 1967-01-01T00:00:00Z https://hdl.handle.net/1721.1/61482 The effects of surface conditions on boiling characteristics Lorenz, J. J.; Mikic, B. B.; Rohsenow, Warren M. A unified model relating surface variables to the nucleate pool boiling characteristics was developed. A simple vapor trapping mechanism was postulated and a geometrical model constructed for idealized conical cavities relating the effective radius for nucleation to cavity radius, cone angle and contact angle. This model for individual cavities was extended to the entire surface providing an expression for the cumulative site density in terms of geometrical parameters. A gas diffusion technique was developed to measure the effective radius for natural cavities and was used successfully to verify the nucleation criteria AT = 2T sv /h p. A transient heat conduction model was experimentally verified for watf an organics at atmospheric pressure and was incorporated into a unified expression showing explicitly the role of surface geometry and contact angle. 1972-01-01T00:00:00Z https://hdl.handle.net/1721.1/61481 Pressure drop with surface boiling in small-diameter tubes Dr̲mer, Thomas; Bergles A. E. Pressure drop for water flowing in small-diameter tubes under isothermal, nonboiling, and surface-boiling conditions was investigated. Experimental results for local pressure gradient and heattransfer coefficients are presented. Heat-transfer results for nonboiling and surface boiling are in agreement with previous investigations. Isothermal friction factors compare favorably with conventional smooth-tube data. Nonboiling friction factors were well correlated with a wall-to-bulk fluid viscosity ratio. It is concluded that boiling pressure gradients cannot be correlated on the basis of local conditions alone. The axial build up of nonequilibrium vapor in the tube produces an increase in pressure gradient even when all other local parameters are constant. The heat-transfer - pressure-gradient analogy was investigated in the boiling region. For the chosen boiling-to-nonboiling ratios, the analogy was found to be valid only under limited conditions. Over-all pressure-drop data are presented for numerous geometries and a range of flow conditions. Diameters of 0.062 to 0.180 in. and L/D's of 25 to 200 were considered. Exit pressures ranged from 30 to 80 psia and velocities ranged from 5 to 50 ft/sec. The majority of the data was taken for an inlet temperature of 80 OF. Heat fluxes were increased from zero to near the burnout condition unless the saturation condition was reached first. These results were correlated by a relation which is independent of all parameters except geometry. This correlation is presented graphically for all the geometries used. Either this plot or the original data plots can be readily used for design purposes 1964-01-01T00:00:00Z https://hdl.handle.net/1721.1/61480 Correlation of maximum heat flux data for boiling of saturated liquids Rohsenow, Warren M.; Griffith, P. 1955-01-01T00:00:00Z https://hdl.handle.net/1721.1/61479 Survey and evaluation of techniques to augment convective heat transfer Bergles A. E.; Morton, Harmon Lindsay. This report presents a survey and evaluation of the numerous techniques which have been shown to augment convective heat transfer. These techniques are: surface promoters, including roughness and treatment; displaced promoters, such as flow disturbers located away from the heattransfer surface; vortex flows, including twisted-tape swirl generators; vibration of the heated surface or the fluid near the surface; electrostatic fields; and various types of fluid additives. Natural and forced convection situations for nonboiling, boiling, and condensation heat transfer are included. The conditions under which heat transfer is improved are summarized, and the efficiency of each technique is presented in terms of a performance criterion where possible. 1965-01-01T00:00:00Z https://hdl.handle.net/1721.1/61478 Transition boiling heat transfer from a horizontal surface Berenson Paul Jerome An experiment, utilising a condensing fluid as the heat source, was performed to determine the heat flux vs. temperature difference curve for transition pool boiling from a horisontal surface. The boiling cure was determined as a function of surface roughness, material, and cleanliness for n-pentane at atmospheric pressure. The results of the experiment show that the liquid contacts the solid heating surface in transition boiling. The burnout heat flux and the film boiling curve are independent of surface properties. For commercial heating surfaces, and probably provided that the combination of surface energies which exist do not result in spreading of the liquid on the solid heating surface, the location of the minimum point is independent of surface properties. It is concluded that transition boiling is a combination of unstable nucleate and unstable film boiling alternating at a given location on the heating surface. The heat transfer data in the transition region was found to be correlated by a straight line on log-log graph paper which connects the burnout point and the minimum point. The bubble spacing and growth rates in film pool boiling from a horizontal surface are shown to be determined by Taylor Hydrodynamic Instability for temperature differences near the minimum. An analytical expression for the heat transfer coefficient in film pool boiling from a horizontal surface is derived. Comb--aing this equation with the equation for the minimum heat flux yields an analytical expression for the temperature difference at the minimum, which defines the location of the minimum point. The above equations agree with the experimental measurements made on n-pentane and carbon tetrachloride within +/- 10%. 1960-01-01T00:00:00Z https://hdl.handle.net/1721.1/61477 The Prediction of multiple heated channel flow patterns from single channel pressure drop data Eselgroth, Peter Ward; Griffith, P. Pressure drop and burnout data taken on a single tube apparatus using Freon at one atmosphere has been used to predict the flow patterns, burnout and points at which flow reverses in a five tube array. All the behavior which might have been expected from the single tube experiments was found in the five tube apparatus but quantitative predictions of the details were not possible because of uncertain bubble nucleation and substantial departures from thermal equilibrium. Application of the techniques suggested in this report is outlined for two reactor problems. One application concerns determining when the flow reverses in a channel of the core of a reactor in which a loss of pumping power accident has occurred. The other application concerns determining when a natural circulation loop will be set up during the quenching in a reactor which has already lost its coolant. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61476 Forced-convection condensation inside tubes Traviss, Donald P.; Baron, Anton G.; Rohsenow, Warren M. High vapor velocity condensation inside a tube was studied analytically. The von Karman universal velocity distribution was applied to the condensate flow, pressure drops were calculated using the Lockhart- Martinelli method, and heat transfer coefficients were calculated from the momentum and heat transfer analogy. Subsequently, the analysis was reduced to an accurate, but simplified form, to facilitate calculations. Experimental data for refrigerants R-12 and R-22 condensing in a 0.315 in. I. D. tube were obtained for mass fluxes from 1.2 x 105 to 11.3 x 105 lbm/hr-ft 2, qualities from 0.02 to 0.96, and saturation temperatures from 75 to 140*F. On the basis of the data and analysis, a simplified non-dimensional presentation of the results evolved. The agreement between the majority of the data and the analysis was within + 15 percent. 1971-01-01T00:00:00Z https://hdl.handle.net/1721.1/61475 The prediction of low quality boiling voids Griffith, P. Slug flow theory is used to predict the density in heated channels of various shapes. In order to make this calculation possible, measurements are made of the bubble rise velocity in annuli, tube bundles, and channels. It is found that the large dimension is most important in channels and the shroud dimension most important in annuli and tube bundles. It is also found that no rotationally symmetrical bubble shapes are obtained in annuli and tube bundles. Finally, a comparison is made between the theory, which contains no free constants, and the experiments. The comparison is good. The results, as presented, apply only to vertical heated channels of various shapes with up flow in the low quality region. 1963-01-01T00:00:00Z https://hdl.handle.net/1721.1/61474 Refrigerant forced-convection condensation inside horizontal tubes Bae, Soonhoon; Maulbetsch, John S.; Rohsenow, Warren M. Condensing heat transfer rates inside a horizontal tube were investigated -for large quality changes across the tube. The proposed correlation is a modification of the work of Rohsenow, Webber and Ling [29]. The result of the investigation is modified through new variables which include the effect of the true axial pressure gradient in a tube. Experimental data are presented for a range of flow conditions. A 0.493 in. ID, 19.75 ft. long nickel tube was used for condensing Refrigerant-12. The saturation temperature was varied from 84.6*F to 118*F and flow rates of vapor-liquid mixture ranged from 151,000 lbm/ft 2hr to 555,000 lbm/ft 2hr. The inlet quality was essentially 100% at saturation and exit qualities ranged from 50% to zero and subcooled liquid. The test results for average heat transfer coefficient ware correlated by the analysis within 15%. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61473 The effect of surface roughness and waviness upon the overall thermal contact resistance Mikic, B. B.; Yovanovich, M. Michael; Rohsenow, Warren M. A thermal contact conductance equation was developed which considers both the effect of surface roughness and waviness. It was shown that the overall thermal contact conductance is determined by the roughness at large contact pressures or rough surfaces. It is also shown that surface roughness increases the contour radius over that predicted by the theory of Hertz. The surface roughness influences the magnitude of the waviness resistances by spreading the load at the contact over a larger region. The theory was seen to be in very good agreement with experimental data. 1966-01-01T00:00:00Z https://hdl.handle.net/1721.1/61472 Refrigerant forced-convection condensation inside horizontal tubes Bae, Soonhoon; Maulbetsch, John S.; Rohsenow, Warren M. High vapor velocity condensation inside a tube was studied theoretically. The heat transfer coefficients were calculated by the momentum and heat transfer analogy. The Von Karman universal velocity distribution was applied to the condensate flow. Pressure drop was calculated by the Lockhart-Martinelli method and the Zivi void fraction equation. Experimental data was obtained for the mass velocities from 150,000 to 555,000 lbm/ft2 hr for R-12 and R-22 condensing in a 0.493" I.D. 18 ft. long test section. The measured heat transfer coefficients agreed with the prediction within 10% except a few points in the very low quality region. 1970-01-01T00:00:00Z https://hdl.handle.net/1721.1/61471 The correlation of nucleate boiling burn-out data Griffith, P. A dimensionless correlation is developed for nucleate boiling buzrnout data including the following ranges of variables. Fluids - Water Bensene n - Heptane n - Pentane Ethanol Pressure - 0.0045 to 0.96 of critical pressure Velocity - 0 to 110 ft/see. Subcooling - 0 to 280 OF Quality - 0 to 70% The data is drawn from a variety of sources and has been collected on widely varying types of systems. Over 300 points are correlated with 94% of the points included between the ł33% envelope drawn aroumd the best line through the points, The correlation includes only fluid properties and quantities which can be calculated on the assumption of eqnilibrium conditions at the burnout point 1957-01-01T00:00:00Z https://hdl.handle.net/1721.1/61470 Heat transfer and pressure drop data for high heat flux densities to water at high subcritical pressures Rohsenow, Warren M.; Clark, John A. Local surface ooeffioients of heat t-ansfer, overall pressure drop data and mean friction factor are presented for heat flamms up to 3.52106 BtuAr ft2 for water flowing in a nickel tabe isder the following conditions: mass rates of flow up to 5.6x 106 lb.m/hr ft2 (or inlet velocities up to 30 ft/sec), absolute pressures up to 2000 pula, and liquid suboooling between 50 F and 250 T. The test section dmnsins were 0.160 inoh I.D. and 9.4 inches long. 1951-01-01T00:00:00Z https://hdl.handle.net/1721.1/61469 Heat transfer during film condensation of a liquid metal vapor Sukhatme, S. P.; Rohsenow, Warren M. The object of this investigation is to resolve the discrepancy between theory and experiment for the case of heat transfer durirnfilm condensation of liquid metal vapors. Experiments by previous investigators have yielded data which are extremely scattered and markedly below the predictions of both the classical IUusselt theory and more recent modifications to it. All theoretical treatments so far have taken account only of the thermal resistance presented by the condensed film. However, calculations from kinetic theory show that with liquid metals a significant thermal resistance can exist at the liquid-vapor interface. This resistance increases with decreasing vapor pressure and is dependent on the value of the "condensation coefficient." Experimental work to back up this hypothesis of a liquid-vapor interfacial resistance is presented. The working fluid for the experiments is mercury condensing at low pressures in the absence of non-condensable gases on a vertical nickel surface. Data of previous investigators are analyzed, and possible reasons for being unable to interpret these results meaningfully are cited. 1964-01-01T00:00:00Z https://hdl.handle.net/1721.1/61468 The effect of nonuniform axial heat flux distribution on the critical heat flux Todreas, Neil E.; Rohsenow, Warren M. A systematic experimental and analytic investigation of the effect of nonuniform axial heat flux distribution on critical heat rilux was performed with water in the quality condition. Utilizing a model which ascribes the critical condition, to either a nucleation-induced disruption of the annular liquid film or annular film dryout, the experimental results taken at low pressures (50-200 psia) were confirmed. Application of this model to higher pressure conditions (500-2000 psia) indicated qualitative agreement with available data of other investigators. Experimental data was obtained for flux distributions representing cosine, linear increasing and decreasing, inlet and exit peaked, spike and uniform shapes. These flux distributions were achieved by electrical resistance heating of test sections whose outside diameter had been machined to the required dimensions.. In each case the critical location as well as the total critical power was obtained by testing the tubes in vertical upflow to failure. The analytic prediction of the results for all flux shapes has been achieved by development of a model which considers the effect of nucleation within the annular film. It is thown that the occurrence of the critical condition is related to the local degree of nucleation (the ratio of the local flux to the flux required to cause nucleation at the local conditions) and the local film flow rate. Both the experimental total critical power and the critical location are confirmed by this model. The results indicate that the total critical power for the outlet peaked flux distributions tested (ratios of maximum to minimum flux of 2, 4, and 5.75 to 1) can be 15 to 30% lower than for uniform flux distributions at comparable hydrodynamic operating conditions.; (cont.) In addition, from this model for given operating conditions, a locus of critical conditions can be constructed from uniform flux distribution data which will enable prediction of the performance of nonuniform flux distributions at similar conditions of mass velocity, pressure and diameter. 1965-01-01T00:00:00Z https://hdl.handle.net/1721.1/61467 Dryout droplet distribution and dispersed flow film boiling Hill, Wayne S.; Rohsenow, Warren M. Dispersed flow film boiling is characterized by liquid-phase droplets entrained in a continuous vapor-phase flow. In a previous work at MIT, a model of dispersed flow heat transfer was developed, called the Local Conditions Solution, which is amenable to hand calculation of wall temperatures. This solution identifies a single nonequilibrium parameter which depends solely on conditions at dryout, particularly a characteristic droplet diameter. Previous to the current study, no simple model including mechanisms occurring upstream of dryout had succeeded in predicting the droplet distribution at dryout. The Local Conditions Solution is rederived to identify which droplet diameter characterizes the distribution of droplets at dryout for purposes of dispersed flow heat transfer analysis. Based on mechanisms of droplet entrainment and deposition, the dryout droplet distribution is derived. This distribution is integrated to obtain the characteristic droplet diameter. A simple method of calculating the characteristic droplet diameter is presented. With the droplet distribution model, the Local Conditions Solution is compared with three correlations and seven data sets. In general, the Local Conditions Solution predicts wall temperature data to within about 5% better than the three correlations. It is found that some data display a type of behavior not predicted by the Local Conditions Solution. This may be caused by the enhancement of droplet heat transfer by free stream turbulent fluctuations. 1982-01-01T00:00:00Z https://hdl.handle.net/1721.1/61466 Nucleate boiling bubble growth and departure Staniszewski, Bogumil E. The vapor bubble formation on the heating surface during pool boiling has been studied experimentally. Experiments were made at the atmospheric pressure 28 psi and 40 psi, using degassed distilled water and ethanol. The heat fluxes and heating surface temperatures have been measured simultaneously by taking high, speed motion pictures of growing bubbles. The diameter time curves of the bubbles and their diameter at the departure moment were obtained in these investigations. Bubble growth rates and bubble departure sizes have been compared to existing theories. It has been found that existing growth theories do not agree very well and that the departure size of the bubble is a function of the growth velocity. 1959-01-01T00:00:00Z https://hdl.handle.net/1721.1/61465 Mechanism of nucleate pool boiling heat transfer to sodium and the criterion for stable boiling Shai, Isaac A comparison between liquid metals and other common fluids, like water, is made as regards to the various stages of nucleate pool boiling. It is suggested that for liquid metals the stage of building the thermal layer plays the most significant part in transfer heat from the solid. On this basis the transient conduction heat transfer is solved for a periodic process, and the period time is found to be a function of the degree of superheat, the heat flux, and the liquid thermal properties. A simplified model for stability of nucleate pool boiling of liquid metals has been postulated from which the minimum heat flux for stable boiling can be found as a function of liquid-solid properties, liquid pressure, the degree of superheat, and the cavity radius and depth. Experimental tests with sodium boiling from horizontal surfaces containing artificial cavities at heat fluxes of 20, 000 to 300, 000 BTU/ft hr and pressures between 40 to 106 mm Hg were obtained. At relatively low heat fluxes, convection currents have significant effects on the period time of bubble formation. An empirical correlation is proposed, which takes into account the convection effects, to match the experimental results. Some recorded temperature variations in the solid close to the surface during stable nucleate boiling are presented. 1967-01-01T00:00:00Z https://hdl.handle.net/1721.1/61464 Bubble growth rates in boiling Griffith, P. The conditions determining the growth rate of a bubble on a surface in boiling are considered and a mathematical model framed in the light of these conditions. The growth rate is then calculated for bubbles growing under a range of conditions of pressure, wall superheat and bulk fluid temperature. The average growth rate of a bubble is found to decrease with increasing maximum size' and to decrease with increasing pressure. At high pressure the maximum size of the bubble is found to be independent of pressure and primarily a function of the thickness of the superheated layer near the surface. The calculated bubble growth velocities are then used to correlate some burnout data for a variety of fluids under a range of pressures in pool boiling. Bubble growth pictures are presented for water at atmospheric pressure under a variety of conditions. 1956-01-01T00:00:00Z https://hdl.handle.net/1721.1/61463 Two phase flow in capillary tubes Suo, Mikio; Griffith, P. The flow of two phases, gas and liquid, has been studied in horizontal tubes of capillary diameter. The flow has been primarily studied in the regime where the gas flows as long bubbles separated from the wall of the tube by a liquid film and from each other by slugs of liquid. In this regime the pressure drop, density and, to a certain extent, the thickness of the liquid film around a bubble have been correlated. The conditions under which the long bubble flow can exist and under which the correlations are valid have been determined. Of special interest is that the correlations should be valid in a zero gravity field. 1963-01-01T00:00:00Z https://hdl.handle.net/1721.1/61462 Cooling of high-power-density computer components Bergles A. E.; Bakhru, N.; Shires, J. W. This report summarizes work carried out during the first two years of a research program sponsored by IBM Corporation. This study has elucidated a number of the heat-transfer characteristics of several fluorochemicals which have potential application as coolants for high power density computer components. Single-phase heat-transfer coefficients have been determined for Freon-113 and FC-78 flowing in a rectangular channel with one of the wide sides heated. Heat-transfer coefficients for the short channels are significantly higher than those predicted by conventional correlations due to entrance effects and superimposed free convection. Uncertainties as to the actual values of certain thermophysical properties introduce corresponding uncertainty in correlation. Subcooled flow boiling data were taken with the rectangular channel for a wide variety of fluid-surface combinations, and over a wide range of flow conditions. These data should facilitate a close estimate of surface temperatures in actual systems. Data taken with an annular test section delineated hysteresis effects which were tentatively identified in the rectangular channel tests. Pool boiling tests indicated large temperature overshoots in the boiling curve with Freon-113. It was found that conventional correlations do not adequately describe critical heat flux data for horizontal circular heaters. The critical heat flux increases as diameter decreases, and increases as the wall thickness increases. The observed hysteresis effects in pool and flow boiling were examined in detail. It is suggested that the phenomenon of temperature overshoot hysteresis is due to two causes: a) the existence of metastable bubbles which are triggered only at sufficiently high disturbance level, and b) the deactivation of larger nucleation sites by displacing the vapor by liquid during subcooling.; (cont.) Under conditions of low velocity or pool boiling, a) is probably most important, but at high velocity b) should be controlling. Increased heat-transfer coefficients observed prior to nucleation appear to be due to thermocapillary circulation induced by the metastable bubbles. It is demonstrated that a random array of Teflon pits on the surface is effective in augmenting heat transfer for pool boiling or low velocity flow boiling of water; however, there is no improvement in heat transfer with boiling of Freon-113 or FC-78. The mechanism of augmentation is examined from the standpoint of incipient boiling, and a plausible explanation for this behavior is suggested. An investigation of the factors controlling the void fraction for subcooled boiling was carried out with a horizontal visual section using Freon-113 and Freon TA. The observed point of net vapor generation was much farther upstream than that indicated by a recent prediction, due primarily to the dominance of the buoyant force. An examination of the void collapse in the unheated downstream portion of the tube revealed that bubble collapse theory cannot be used to predict the void fraction. The large void fractions produced by boiling fluorochemicals can be minimized by using twisted-tape inserts. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61461 Thermal contact resistance with non-uniform interface pressures McMillan, Robert; Mikic, B. B. This work considers the effect of roughness and waviness on interfacial pressure distributions and interfacial contact resistance. It is shown that for moderate roughness the contour area could be substantially different from the contour area calculated using the Hertzian theory. The model for pressure calculation assumes plastic deformation of surface irregularities and elastic deformation of a spherically wavy base. The calculations of pressure distributions cover the range of parameters of practical interest. Experimental contact resistance values have been determined and are compared with theoretical predictions. It was calculated that contact conductance for wavy surfaces can be increased for certain ranges of parameters by making surfaces rough. 1970-01-01T00:00:00Z https://hdl.handle.net/1721.1/61460 The transition from two phase bubble flow to slug flow Radovcich, Nick A.; Moissis, Raphael. The process of transition from bubble to slug flow in a vertical pipe has been studied analytically and experimentally. An equation is presented which gives the agglomeration time as a function of void fraction, channel diameter, initial bubble diameter and liquid purity. A dependent function which also appears in the equation has been evaluated using experimental data. A reasonably good correlation of the data has been achieved. 1962-01-01T00:00:00Z https://hdl.handle.net/1721.1/61459 The thermodynamics of bubbles Clark, John A. This paper outlines those concitions annanded by the laws of thermodynamics for equilibriza betwoen the vapor in a bubble and the surrounding liquid and then employs these concepts with a nucleation theory in an atteapt to arrive at an expression relating the nucleating superheat in a pure liquid to other fluid properties. The first part of the paper deals with the phase equilibrium across a curved interface and same of its consequences. In the second part recourse is made to the kinetic theory of wHeterophaseR thermal fluctuations given by Frankel for nucleus formation which is extended and employed to derive a relationShip between a superheat at nucleation and other fluid properties. 1956-01-01T00:00:00Z https://hdl.handle.net/1721.1/61458 Forced-convection surface-boiling heat transfer and burnout in tubes of small diameters Bergles A. E.; Rohsenow, Warren M. A basic heat-transfer apparatus was designed and constructed for the study of forced-convection boiling in small channels. The various regions of forced-convection surface boiling were studied experimentally and analytically. In the region of low wall superheat, the heat flux can be predicted by available correlations for forced convection. Data indicate, however, that these correlations do not properly account for the radial variation of properties for water at high temperature difference. The conventional Dittus and Boelter-McAdams relationship is recommended for design purposes on the basis of its simplicity and conservative predictions. An analysis for the prediction of the inception of first significant boiling was developed. Experimental results are in good agreement with analytical predictions. The analysis provides information necessary for the prediction of the complete forced-convection surface-boiling curve. Data for a small-diameter tube indicate that the bubbles formed at incipient boiling can trip the laminar or transition boundary layer to a fully-developed turbulent boundary layer. The region of vigorous boiling coincides approximately with the extrapolation of the pool-boiling curve in one set of experiments. In other experiments, pool-boiling data were strongly influenced by fluid and surface conditions, as well as by bubble-induced convection in the pool. Due to the complexities in these pool-boiling data, it is impossible to make a conclusive comparison with forced-convection-boiling data. The heat flux obtained by a superposition of pool boiling and forced convection is close to the apparent asymptote for fully-developed boiling. For design purposes, it is concluded that fully-developed -forced-convection boiling can be related to pool boiling by either direct extrapolation or superposition of forced convection.; (cont.) The burnout heat flux under conditions of forced convection and surface boiling is shown to be a complicated function of subcooling at low values of subcooling. This appears to be due to the velocity increase caused by the relatively large volume fraction of vapor. Burnout flux is shown to increase with decreasing tube diameter. This effect can be attributed to an increase in void fraction with decreasing tube diameter. Entrance effects are significant in forced-convection surface boiling as shown by the decrease of burnout flux with increasing length. Flow oscillations caused by system compressibility can greatly reduce the burnout heat flux in the subcooled region. This instability is particularly difficult to avoid with tubes of very small diameter. 1962-01-01T00:00:00Z https://hdl.handle.net/1721.1/61457 The deterioration in heat transfer to fluids at super-critical pressure and high heat fluxes Shiralkar, B. S.; Griffith, P. Introduction: Several supercritical steam generators in the American Electric Power system have shown evidence of tube overheat in the lower furnance at the point where the water bulk temperature is about 670 0 F. The evidence is of two kinds. First thermal fatigue has occurred and caused tube failures long before a failure of any kindwas to be expected. Second, pairs of cordal thermocouples have shown very high wall temperatures and, extrapolating back to the inside of the tube, evidence reduced inside heat transfer coefficients. It was suspected that a possible cause of the high tube temperature was a supercritical "burnout". The primary purpose of this investigation is to determine the cause and conditions leading to a supercritical "burnout" such as might occur in a supercritical steam generator. Before focusing on this aspect of the problem it is worthwhile to mention several other possible causes for the high tube wall temperatures which have been observed. In this context high means higher than the design temperature. Let us just list these possibilities. 1. Scale inside the boiler tubes. 2. Hot spot factors in the design procedure which are too low. 3. Higher heat transfer from the combustion gases than expected. Better design procedures or better control of the water purity might be sufficient to cause the problem to disappear without changing the water flow conditions inside the tube. Because the three factors which are listed above are really rather vague, it appeared that the most promising approach is to eliminate the excessive temperatures inside the tube at supercritical pressure is to eliminate the "burnout". Therefore, only the burnout aspect of the problem has been studied here. The undesirable behavior of the Nusselt number, which is of interest, is indicated in Fig. 1. In particular we want to find out when the supercritical Nusselt number is less than one would expect from the affects of simple porperty variations alone. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61456 Refrigerant forced-convection condensation inside horizontal tubes Bae, Soonhoon; Maulbetsch, John S.; Rohsenow, Warren M. High vapor velocity condensation inside a tube was studied theoretically. The heat transfer coefficients were calculated by the momentum and heat transfer analogy. The Von Karman universal velocity distribution was applied to the condensate flow. Pressure drop was calculated by the Lockhart-Martinelli method and the Zivi void fraction equation. Experimental data was obtained for the flow rate from 240,000 to 485,000 lbm/ft2 hr. R-22 was condensed in a 0.493" I.D. 18 ft. long test section. The measured heat transfer coefficients agreed with the prediction within 10% except a few points in the very low quality region. 1969-01-01T00:00:00Z https://hdl.handle.net/1721.1/61455 An experimental and theoretical study of radiative and conductive heat transfer in nongray semitransparent media Eryou, N. Dennis; Glicksman, Leon R. One dimensional temperature profiles and heat fluxes within a slab of molten glass were measured experimentally. The glass slab was contained in a platinum foil lined ceramic tray inside a high temperature furnace. An optical method of temperature measurement was developed in which a helium-neon laser beam was directed along an isothermal path through the glass. The attenuation of the beam was a strong function of temperature and was used to evaluate the local temperatures within the glass slab. In order to perform a theoretical analysis the spectral absorption coefficient of the glass was measured between .6328 microns and 2. 75 microns from 2000 0F to 23000F. Two analyses were performed; one for a diffuse platinum-glass boundary and the other for a specular boundary. The experimentally measured temperature profiles and heat fluxes agreed with the predicted profiles within 50F and the heat fluxes to within ten percent. 1969-01-01T00:00:00Z https://hdl.handle.net/1721.1/61454 Tables for solution of the heat-conduction equation with a time-dependent heating rate Bergles A. E. Tables are presented for the solution of the transient onedimensional heat flow in a solid body of constant material properties with the heating rate at one boundary dependent on time. These tables allow convenient and rapid estimation of the temperature distribution in the many practical cases where the mathematical model applies. Examples illustrating use of the tables are given. 1962-01-01T00:00:00Z https://hdl.handle.net/1721.1/61453 The role of surface conditions in nucleate boiling Griffith, P.; Wallis, John D. Nucleation from a single cavity has been stuied indicating that cavity gemtry is aportant in two ways. The mouth diameter determines the superheat nmeded to initiate boiling and its shape determines its stability one boiling has begun. Contact angle is shown to be important in bubble nucleation primarity thrugh its effect on cavity stability. Contact angle measurements made on "clean" and paraffin coated stainIess steel murftces with water shcw that the contact angle varies between 20 and 110* for tenperatures from 20[degree] to 170[degree] C. On the basis of single cavity nucleation theory, it is proposed to characterie the gross nucleation properties of a given surface for all fnds under all conditions with a single group having the dmnions of length. Finally, it is shown experimentally that this characterization is adequate by boiling water, methmnol and ethanol different copper surfaces finished with 3/0 emry, and showing that the number of active centers per unit area is a function of this variable alone. 1958-01-01T00:00:00Z https://hdl.handle.net/1721.1/61452 Areas of contact and pressure distribution in bolted joints Gould, Herbert Hirsch; Mikic, B. B. When two plates are bolted (or riveted) together these will be in contact in the immediate vicinity of the bolt heads and separated beyond it. The pressure distribution and size of the contact zone is of considerable interest in the study of heat transfer across bolted joints. The pressure distributions in the contact zones and the radii at which flat and smooth axisymmetric, linear elastic plates will separate were computed for several thicknesses as a function of the configuration of the bolt load by the finite element method. The radii of separation were also measured by two experimental methods. One method employed autoradiographic techniques. The other method measured the polished area around the bolt hole of the plate's caused by sliding under load in the contact zone. The sliding was produced by rotating one plate of a mated pair relative to the other plate with the bolt force acting. The computational and experimental results are in agreement and these-yield smaller zones of contact than indicated by the literature. It is shown that the discrepancy is due to an assumption made in the previous analyses. In addition to the above results this report contains the finite element and heat transfer computer programs used in this study. Instructions for the use of these programs are also included. 1970-01-01T00:00:00Z https://hdl.handle.net/1721.1/61451 Natural convection flows in parallel connected vertical channels with boiling Eselgroth, Peter Ward; Griffith, P. The steady-state flow configuration in an array of parallel heated channels is examined with the objective of predicting the behavior of a reactor during a loss of flow accident. A method of combining the results of single tube experiments with a calculation scheme for the multiple tube array is developed. Only single ~single tube experiments were run. These are three possible flow solutions for a channel. They are as follows: gas and liquid flowing up, gas flowing up and liquid down, and both gas and liquid flowing down. This investigation eliminates the latter as a possible solution for a channel operating in an array of channels. For the two remaining flow cases the constraints of a common pressure drop across the channels and the satisfaction of continuity in the closed end plenum, determine a solution for each heat flux distribution over the array. The data in this investigation was obtained using Freon 113 as the coolant for an electrically heated glass tube. The data presented in not intended for use other than verification of the method of solution proposed. It is recommended that the coolant intended for use in the application be used with a prototype, rather than a model, channel in obtaining the single channel data. 1967-01-01T00:00:00Z https://hdl.handle.net/1721.1/61450 Heat transfer during film condensation of potassium vapor Kroger, Detlev Gustav The object of this work is to investigate theoretically and experimentally the following two phases of heat transfer during condensation of potassium vapore, a. Heat transfer during film condensation of pure saturated potassium vapor on a vertical surface. b. Heat transfer during film condensation of potassium vapor in the presence of a small quantity of non-condensable gas. Investigators have until recently been unable to explain the large discrepancy between theory and experiment for the case of heat transfer during film condensation of pure liquid metal vapors. Calculations from kinetic theory and irreversible thermodynamics suggest that this difference may be due to a significant thermal resistance at the liquid-vapor interface in addition to that of the condensate film. This interfacial resistance becomes especially noticeable at low vapor pressures and is dependent on a "condensation" coefficient to be determined experimentally. The presence of a trace of non-condensable gas during condensation may greatly reduce the heat transfer rate. An analysis predicting the heat transfer rate during condensation of potassium vapor in the presence of a non-condensable gas is presented,and experiments are performed to test the validity thereof. 1966-01-01T00:00:00Z https://hdl.handle.net/1721.1/61449 Thermal contact resistance Mikic, B. B.; Rohsenow, Warren M. This work deals with phenomena of thermal resistance for metallic surfaces in contact. The main concern of the work is to develop reliable and practical methods for prediction of the thermal contact resistance for various types of surface characteristics under different conditions. In particular, consideration is restricted to the following cases: (i) rough nominally flat surfaces in a vacuum environment; (ii) rough nominally flat surfaces in a fluid environment; (iii) smooth wavy surfaces in a vacuum environment (with either of the following three types of waviness involved; spherical waviness, cylindrical waviness in one direction and cylindrical waviness in two perpendicular directions) and (iv) rough wavy surfaces in a vacuum environment. The problem is divided into three parts: thermal analysis, surface analysis and deformation analysis. The thermal analysis, based upon the proposed models, investigates the analytical solutions for the thermal contact conductance under steady state conditions. It was found convenient, due to the extensive analytical work connected with various models and different methods used here, to present all details of the thermal analysis separately in the appendices. The surface analysis, treating the surfaces as random processes with Gaussian distribution of height, relates the interface geometry to the actual contact area. The method suggested in this analysis has been checked against some autoradiographical experimental data. The deformation analysis, in its two parts, gives dependence between the load supported by the interface and (i) the actual contact area and (ii) the contact spots distribution for rough spherically wavy surfaces, respectively. The result of the first part of; (cont.) the analysis is based on the plastic deformation of the surface asperities. The second part considers, through the model of the equivalent contour area, the combined effect of spherical waviness and roughness on the problem of contact spots spreading at the interface. Limitations and possible deviations of the proposed models are discussed. Prediction of the thermal contact conductance is compared with experimental data obtained in this work (in a vacuum environment) together with some data obtained by other investigators (for which necessary surface parameters were available). Agreement between the measured and predicted values was good in the whole tested range of system variables. 1966-01-01T00:00:00Z https://hdl.handle.net/1721.1/61448 A study of boiling water flow regimes at low pressures Fiori, Mario P.; Bergles A. E. "A comprehensive experimental program to examine flow regimes at pressures below 100 psia for boiling of water in tubes was carried out. An electrical probe, which measures the resistance of the fluid between the centerline of the flow and the tube wall, was used to identify the various flow regimes. This probe proved to be an ideal detection device, because of its simplicity, reproducibility, and accurate representation of the flow pattern within the heated test section. The major flow regimes observed were bubbly, slug and annular flow. Under certain conditions at high flow rates, a wispy-annular flow patern was observed. The effects of mass velocity (0.2 x 10 - 2.4 x 100 lbm/hr-ft2), inlet temperature (100, 150, 2000F), exit pressure (30, 100 psia), quality (x = -10 - +7 percent), purity (9, 40 PPM NaCl; 1-3 megohm-cm), length (L/D-30, 6Q, 90), diameter 0.094, 0.242 in.), and orientation (vertical and horizontal on the flow regimes were studied. Flow regime maps on coordinates of mass velocity and quality are presented for these conditions. Bubbly and slug flow occurred primarily in the subcooled region, while fully developed annular flow was reached at equilibrium qualities between 2 and 4 percent. The transitions between the different flows were shifted to regions of increased subcooling when velocity, pressure, and heat flux increased, and when inlet temperature decreased. Purity and geometry had little affect on the flow regime boundaries.; (cont.) The shifting of the transitions is related to the agglomeration point, which is that point at which the bubbles so coalesce that slug flow is first observed. The agglomeration point depends on the point of incipient boiling, the number of bubbles in the flow, and the number of collisions per bubble. These latter quantities in turn depend on velocity, temperature, pressure, and heat flux. The flow regime information obtained in this study s~hould be of value in correlating and interpreting low pressure heattransfer data. The flow regime data were found to be useful in explaining the effect of inlet temperature on burnout heat flux. 1966-01-01T00:00:00Z https://hdl.handle.net/1721.1/61447 Mechanism and behavior of nucleate boiling heat transfer to the alkalai liquid metals Deane, Charles William; Rohsenow, Warren M. A model of boiling heat transfer to the alkali liquid metals is postulated from an examination of the events and phases of the nucleate boiling cycle. The model includes the important effect of microlayer evaporation which causes a wave of temperature depression to penetrate into the heating solids; calculated results predict the periodic boiling behavior in the heating solid as a function of the heat flux, the system pressure, the cavity size, and the thermophysical properties of the liquid and the solid. An experimental program was designed to examine the microscale boiling behavior of sodium and to verify the calculated predictions of the boiling model. Artificial cylindrical cavities are used in most of the test sections; a thermocouple is placed close to the boiling surface and adjacent to the cavity wall, and microscale temperature measurements were obtained for stable boiling of sodium from artificial cavities and also from a natural cavity. Horizontal heating surfaces were made from nickel "A", stainless steel 316, and molybdenum-1/2%-.titanium; the range of saturation pressure is from 20 to 780 mm Hg. Favorable comparisons with the predictions of the boiling model are obtained with data for the bubble period and for the amplitude of temperature oscillation at a thermocouple close to the boiling surface; these results indicate the importance of the microlayer in a model for boiling sodium.; (cont.) The effect of pressure-temperature history on incipient superheats for boiling was examined; the presence of inert gas in a cavity can lower the incipient superheat. Also, an analysis of unstable boiling indicates that outgassing from the heating solid at high temperatures can cause erratic temperature fluctuations by sporadically triggering nucleation at a previously inactive site. 1969-01-01T00:00:00Z https://hdl.handle.net/1721.1/61446 Boiling and condensation in a liquid-filled enclosure Bar-Cohen Avram; Bergles A. E. A combined experimental and analytical investigation of boiling and condensation in a liquid-filled enclosure, with water and Freon- 113 as the working fluids, is described. The operating characteristics of a boiling system, utilizing a condenser submerged in the fluid, are presented and related to specific operational modes and thermal transport mechanisms. A lower bound of operation, corresponding to natural convection heat transfer at both the heated and condenser surfaces, is identified. Similarly, for the commonly encountered range of system operation, a condensive upper bound is identified and shown to correspond to vapor space condensation. A nondimensional vapor bubble collapse length, L c/W, is found to govern the rate and mechanism of heat transfer at the submerged condenser surface. LValues of wC << are associated with natural convection heat transfer at the L c submerged condenser. For -~ I the presence of a substantial vapor frac- w tion in the bulk liquid leads to augmented convection, while for values of L C >> 1 condensation is found to dominate thermal transport at the condenser surface. 4 possible technique for augmenting condensation heat transfer on horizontal surfaces is examined in an attempt to raise the condensive upper bound of submerged condenser operation. A doubly-rippled surface with small, constant radius of curvature undulations is shown to yield a factor of two increases in the rate of vapor space condensation based on the projected area of the condenser surface.; (cont.) A systematic design procedure for submerged condenser systems utilizing the proposed models and correlations is described and related to typical design considerations. 1971-01-01T00:00:00Z https://hdl.handle.net/1721.1/61445 The Momentum flux in two-phase flow Andeen, Gerry B.; Griffith, P. The average momentum flux at a section of a pipe with twophase upflow has been measured by the impulse technique. Steamwater and air-water mixtures were tested in one-inch and onehalf inch nominal pipes. Homogeneous velocities ranging from 150 to 1200 ft/sec. and qualities from 5% to 85% were tested. The results are compared to the results of models currently in practice for predicting pressure drop and critical flow. The influence of the void fraction, the velocity profile, phase distribution and fluctuations upon the momentum flux are discussed. 1965-01-01T00:00:00Z https://hdl.handle.net/1721.1/61444 Heat transfer during film condensation of potassium vapor on a horizontal plate Meyrial, Paul M.; Morin, Michel L. The object of the investigation is to analyze the following two features of heat transfer during condensation of potassium vapor: a. Heat transfer during film condensation of a pure saturated potassium vapor on a horizontal surface. b. Heat transfer during film condensation of potassium vapor in the presence of a small quantity of non-condensable gas. Until now, the discrepancy between theory and experiment concerning the condesnation of pure liquid-metal vapors has been explained by a thermal resistance at the liquid-vapor interface. This interfacial resistance was analyzed by means of the kinetic theory, and the results depended on the use of a condensation (or accommodation) coefficient. This coefficient was found to decrease in value at higher pressures. This work presents a more refined analysis of the interfacial temperature distribution, including the effect of subcooling in the vapor in the region of the liquid-vapor interface. Furthermore, a theory predicting the temperature drop in the condensate film on a horizontal plate is presented. Experiments were performed to verify the theory. In addition, data from previous investigators were analyzed. Experiments with non-condensable gases in the saturated vapor were made using the horizontal plate facing upward. These results supported Kroger's (19) theory. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61443 Slug flow Griffith, P.; Wallis, Graham B. Introduction: When two phases flow concurrently in a pipe, they can distribute themselves in a number of different configurations. The gas could be uniformly dispersed throughout the liquid in the form of small bubbles. There could be large gas bubbles almost filling the tube. There could be an annulus of liquid and core of vapor with or without drops of liquid in it. The interface could be smooth or wavy. When one describes how the phases are distributed, one is specifying the flow regime. Such a description is necessary before any mathematical model can be constructed which will predict a quantity such as pressure drop It is naive to expect that a single mathematical model would adequately encompass all possible two-phase flow regimes, even for a single geometric configuration. Therefore, we shall begin by saying that for this work the results that have been obtained and the conclusions that have been drawn apply only to fully developed slug flow in a round vertical pipe. Slug flow is characterized by large bubblesalmost filling the tubewhich are separated by slugs of liquid. The nose of the bubble is rounded and the tail generally flat. One may or may not find small bubbles in the slug following the large bubble. A number of typical slug flow bubbles are pictured in Figures 4-10. Bubbles very similar to these have been studied by Dumistrescu (1), and Davis and Taylor (2). Both these references consider the same problem. How rapidly will a closed tube full of liquid empty when the bottom is suddenly opened to the atmosphere. The approach used by both authors is to assume that the asymptotic rise velocity (for large times) can be calculated from potential flow theory. The boundary condition at the pipe wall is that the velocity is axial. At the bubble boundary it is assumed that the pressure is constant, The problem is then to find the shape of the bobble that would satisfy the constant pressure boundary condition.; (cont.) This was done approximately and in both cases the comparison with experiment was satisfactory though the deviations became large for small tubes. The work of Davis and Taylor, and Dumitrescu served as the starting point for this investigation. The boundary condition at the bubble wall for large bubbles, constant pressure, was still valid to an excellent approximation and the finiteness of the slug flow bubbles did not appear to make much difference in their rise velocity. In the next section, the fluctuation period, the mean density, and the pressure drop will be expressed in terms of the pipe area, the Taylor bubble rise velocity and the flow rates of the two phases. In subsequent sections the observations rade of bubble shape, length and velocity will be described and then a comparison of computed and measured pressure drops given. 1959-01-01T00:00:00Z https://hdl.handle.net/1721.1/61442 Thermal contact resistance in a non-ideal joint Roca, Richard T.; Mikic, B. B. The contact conductance at an interface can be determined by knowing the material and surface properties and the interfacial pressure distribution. This pressure distribution can be influenced strongly by the roughness of the mating surfaces but until now this effect has been ignored in studies of joint conductance. This thesis considers this effect and shows the circumstances when it is an important factor. Furthermore, it is shown that one can either raise or lower the total resistance of a joint by changing the surface properties in the proper manner for the particular system being considered. Specifically, this thesis deals with three systems: the contact of two rough, wavy surfaces; the contact of two rough but nominally flat plates pressed together over a concentrated area; and the contact of two rough but nominally flat plates bolted together. In each case the pressure distribution is calculated as a function of the surface properties. In the case of wavy surfaces it is found that all necessary information for any combination of parameters can be reduced to one master graph. In the other two cases one such graph is needed for each geometry used. The resulting pressure distributions are used in a specific heat transfer example and the total joint resistance versus roughness is presented. It is shown how one can actually decrease the resistance by increasing the roughness - a seemingly contradictory phenomenon. Heat transfer experiments performed by Joseph Pigott qualitatively confirmed the theoretical findings. 1971-01-01T00:00:00Z https://hdl.handle.net/1721.1/61441 The effect of surface conditions on nuceate pool boiling heat transfer to sodium Marto, P. J.; Rohsenow, Warren M. A simplified theoretical model for bubble nucleation stability has been proposed, and an approximate stability criterion has been developed. This criterion contains both fluid and surfqce properties, and it predicts that nucleation for sodium should be unstable. Commercial grade sodium was boiled from a horizontal disc at pressures of 65 mm, 200 mm and 400 mm Hg absolute, with sodium temperatures ranging from 1200 0F to 1500 0F. Heat fluxes as high as 236,000 BTU/hr.ft.2 were attained. Boiler surface finishes ranged from highly polished mirror finishes to coarse, porous coatings. By following a prescribed cleaning and filling procedure, nucleate boiling results were generally reproducible for a given type surface. The effect of roughness as well as any aging and hysteresis effects were experimentally determined. Incipient nucleate boiling results are discussed as well as the effect of pressure and pool depth on the nucleate boiling curve. The effect of surface material, chemical treatment, heat flux and cavity geometry on nucleation stability was measured, and the experimental results agreed with the predictions of the proposed stability model. 1965-01-01T00:00:00Z https://hdl.handle.net/1721.1/61440 Thermal boundary layer development in dispersed flow film boiling Hull, Lawrence M.; Rohsenow, Warren M. Dispersed flow film boiling consists of a dispersion of droplets which are carried over a very hot surface by their vapor. This process occurs in cryogenic equipment and wet steam turbines. It is also of interest in the analysis of a nuclear reactor loss of coolant accident and many other applications. The integral approach to boundary layer theory is first used to analyze heat transfer to the turbulent flow of single phase vapor in the entrance region of a circular tube. The single phase predictions compare well with published correlations and data. The integral analysis is then extended to analyze the heat transfer to turbulent dispersed flow. A numerical solution and a simplified explicit solution for the dispersed flow case are presented. These analyses are verified by comparisons to experimental data. An experiment was designed which allowed control over the key parameters appearing in the analysis: vapor Reynolds number, quality, and drop diameter. The heat transfer tests were carried out at vapor Reynolds numbers of 2x10 4 and 4x10. The quality varied between 10% and 50%. A photographic study of the dispersion provided the estimates of the drop sizes. For each pair of Reynolds number and quality, a test with relatively large drops and one with relatively small drops was carried out. This provided an experimental parametric study in addition to the data required for the model verification. 1982-01-01T00:00:00Z https://hdl.handle.net/1721.1/61439 Subcooled flow boiling of fluorocarbons Murphy, Richard Walter; Bergles A. E. A study was conducted of heat transfer and hydrodynamic behavior for subcooled flow boiling of Freon-113, one of a group of fluorocarbons suitable for use in cooling of high-power-density electronic components. Problems arising from the excellent wetting characteristics and large solubility constants of fluorocarbons were also examined. The primary configuration was vertical upflow through a 0.500-in. ID stainless steel tube with direct resistance heating of the tube wall. Operating parameter ranges included up to 4.28 ft/sec velocity, 22.3 psia pressure, 61*F subcooling, 0.40 void fraction, 1.08 X 10-3 moles/mole dissolved gas, and 105 Btu/hr ft2 heat flux. Single-phase heat transfer was adequately correlated by standard methods. Boiling curves had a unique form dominated by large, discontinuous jumps in wall temperature at the incipient point on increasing heat flux traverses. Effects of velocity and subcooling on two-phase heat transfer followed conventional trends. Techniques were devised for accurate determination of the temperature dependence of the air-Freon-113 solubility constant and for measurement and control of dissolved gas content in the main loop. Dissolved gas effects were found to increase heat transfer significantly in the partial boiling mode. Data in the fully-developed boiling mode were successfully described by modifications of existing correlations. A conventional correlation provided, at best, an upper bound for the critical heat flux data. Models and analyses were formulated for predicting delayed nucleation and dissolved gas effects on incipience. Delayed nucleation and hysteresis were successfully eliminated by means of a special surface coating. Transition in gassy boiling heat transfer from gas-dominated to vapor-dominated modes was postulated with reference to adjusted saturation temperatures.; (cont.) Single-phase pressure drop was adequately correlated by standard methods. Parametric effects on two-phase total pressure drop were investigated and described. Three novel techniques--photographic, trap, and capacitance-- were employed to obtain accurate void fraction measurements. It was found that dissolved gas drastically retarded bubble collapse rates. Parametric effects on void fraction were examined and approximately correlated on quality coordinates. Modification of an existing analysis for predicting the point of net vapor generation gave reasonable agreement with void data. Void information was used to estimate the gravity component of pressure drop. The remaining friction-acceleration component data were plotted on coordinates suggested in an existing correlation. Alteration of the coordinates to account for issolved gas resulted in fair agreement of data with the correlation curv-. A qualitative description of the gas-dominated to vapor-dominated transition in pressure drop performance, analogous to that for heat transfer, was developed. 1971-01-01T00:00:00Z https://hdl.handle.net/1721.1/61438 Heat transfer to impacting drops and post critical heat flux dispersed flow Kendall, Gail E.; Rohsenow, Warren M. Heat transfer to drops impacting on a hot surface is examined in context of dispersions of flowing, boiling fluids. The liquid contribution to heat transfer from a hot tube to a two-phase dispersion is formulated in terms of heat transfer contributions due to surface impacts of individual drops. High heat transfer rates are associated with liquid wetting of the surface at surface temperatures between saturation and the minimum stable film boiling (transition) temperature. Low heat transfer rates are associated with non-wetting, or dry, impacts at surface temperatures above the transition temperature. In the wetting region, experimental measurements of heat transfer rates to sparse streams of drops impacting on a hot surface showed complete evaporation of the drops. In the non-wetting region, an analysis of impact dynamics and heat transfer for deformable drops was performed using simple, idealized shapes to model the deformation. Lagrangian methods were used to derive equations of motion and deformation for impacting drops. Resulsts compare well with available information on drop dynamics and heat transfer. The analysis of heat transfer to impacting drops was formulated for incorporation into dispersed flow analysis, given the statistical distribution of drops in the dispersion. Applications include the prediction of local heat transfer and vapor generation rates. The liquid contribution to dispersed flow heat transfer must be included for the prediction of rewet in nonuniformly heated tubes. 1978-01-01T00:00:00Z https://hdl.handle.net/1721.1/61437 The Effect of vapor subcooling on film condensation of metals Fedorovich, Eugene D.; Rohsenow, Warren M. This work presents an analysis of the interfacial "vapor-condensate" temperature distribution, which includes the effect of subcooling (supersaturation) in the vapor. Experimental data from previous investigators for different metals were analyzed. It is shown that taking into account this subcooling effect permits the thermal interphase resistance to be described with assumption of [sigma] = Constant = 1.0 where[sigma] is the condensation (mass accommodation) coefficient. 1968-01-01T00:00:00Z https://hdl.handle.net/1721.1/61436 Post critical heat transfer to flowing liquid in a vertical tube Plummer, David N. 1974-01-01T00:00:00Z https://hdl.handle.net/1721.1/61435 Transition from film boiling to nucleate boiling in forced convection vertical flow Iloeje, Onwuamaeze C.; Plummer, David N.; Rohsenow, Warren M. The mechanism of collapse of forced cnnvection annular vertical flow film boiling, with liquid core, is investigated using liquid nitrogen at low pressures. The report includes the effect of heat flux from the buss bar. Tests include runs with mass fluxes varying from 44,000 lbm/hr-ft2 2 2 to 186,000 lbm/hr-ft , and buss bar heat fluxes from 0 to 107,000 BTU/hr-ft The channel was a 0.4 inch I. D. by 0.5 inch O.D. by 8 feet long Inconel 600 tube. Two modes of collapse were isolated, in the absence of rewet by dispersed cooling within the mist flow region. These were axial conduction controlled collapse originating at the entrance to the test section for zero or negative buss bar heat flux; and impulse cooling collapse originating downstream of the entrance for heat into the test section from the buss bar. Collapse heat flux was found to be a function of mass flux, but the collapse wall temperature difference (T - T ) was independent w S of mass flux and could be successfully predicted within 6% by pool boiling minimum transition correlation (e.g. Berenson [5]). 1972-01-01T00:00:00Z https://hdl.handle.net/1721.1/61434 Entrance effects in a developing slug flow Moissis, Raphael; Griffith, P. The kinetics of a Taylor bubble, as it rises behind a series of other bubbles in a gas-liquid slug flow, have been determined. The rise velocity of a bubble is expressed as a function of separation distance from the bubble ahead of it. Using this information, the pattern of development of bubbles which initially enter a tube at regular intervals is determined by means of finite difference calculations. The density and, to a first approximation the pressure drop, of the developing flow are then calculated from continuity considerations The density distritution in the entrance region is found to be a function of flow rates of the two phases, of distance from the inlet, and of initial bubble size0 Density calculated by the present theory is compared with experimental measurements by the present and other investigators. Theory and experiments are generally in good agreement. 1960-01-01T00:00:00Z https://hdl.handle.net/1721.1/61433 Void volumes in subcooled boiling systems Griffith, P.; Clark, John A.; Rohsenow, Warren M. Introduction: Knowledge of the pressure drop in a channel and the resulting flow redistribution is essential in predicting the performance of a nuclear reactor. The pressure drop in a channel which is experiencing boiling is, in part, dependent on the void volumes present. In this report the results of an experimental investigation of the void volumes in systems at 500, 1000, and 1500 psia are presented. A method of predicting the void volumes is then presented. Finally, a comparison of measured end predicted void volumes is made with data collected at 2000 psia under completely different conditions. Agreement with the data is very good. 1958-01-01T00:00:00Z https://hdl.handle.net/1721.1/61432 The transition from the annular to the slug flow regime in two-phase flow Haberstroh, Robert D.; Griffith, P. Experiments were conducted to determine the transition from annular to semiannular flow regimes for two-phase, gas-liquid upflow in vertical tubes. The influencesof liquid flow rate, tube diameter, liquid viscosity, surface tension, and density ratio were tested. The transition location, giving the least gas flow which will support annular flow, is correlated for low liquid flows. The effects of surface tension and liquid viscosity are small for tubes of 0.5" and greater diameter. For large liquid flows the transition occurs at constant quality. Over the range of variables tested, the void fraction at transition lies between 80% and 90%. The criterion of transition is the appearance of liquid plugs which intermittently bridge across the tube. The pressure drop and volume void fraction at transition are reported but not correlated. Recommendations for extension of the experimental results to untested conditions are presented. 1964-01-01T00:00:00Z https://hdl.handle.net/1721.1/61431 A method of correlating heat transfer data for surface boiling of liquids Rohsenow, Warren M. A method based an a logical uxplanation of the meani of beat transfer associated with the boiling process is presented for correlating heat transfer data for nucleate boiling of liquids for the case of pool boiling. Tbe suggested relation is [mathematical equation hand written in] iere the various fluid properties awe evaluated at the saturation temperature correspondinag to the local pressure and Csf is a function of the particular heating surface-fluid combination. 1951-01-01T00:00:00Z https://hdl.handle.net/1721.1/61430 Film condensation of liquid metals -- precision of measurement Wilcox, Stanley James; Rohsenow, Warren M. Major differences exist in results published by investigators of film condensation of liquid metal vapors. In particular, the reported dependence of the condensation coefficient on pressure has raised questions about both the precision of the reported data and the validity of the basic interphase mass transfer analysis. An error analysis presented in this investigation indicates that the reported pressure dependence of the condensation coefficient at higher pressures is due to an inherent limitation in the precision of the condensing wall temperature measurement. The magnitude of this limitation in precision is different for the various test systems used. The analysis shows, however, that the primary variable affecting the precision of the wall temperature measurement is the thermal conductivity of the condensing block. To verify the analysis, potassium was condensed on a vertical surface of a copper condensing block. The copper block was protected from the potassium with nickel plating. Condensation coefficients near unity were obtained out to higher pressures than those previously reported for potassium condensed with stainless steel or nickel condensing blocks. These experimental results agree with the prediction of the error analysis. In addition, a discussion of the precautions used to eliminate the undesirable effects of both non-condensable gas and improper thermocouple technique is included. It is concluded from the experimental data and the error analysis that the condensation coefficient is equal to unity and that the pressure dependence reported by others is due to experimental error. 1969-01-01T00:00:00Z https://hdl.handle.net/1721.1/61429 Some hydrodynamic characteristics of bubbly mixtures flowing vertically upward in tubes Rose, Sewell C.; Griffith, P. An investigation of bubbly flow has been conducted in vertical plexiglass tubes using air and water at atmospheric pressure. The bubbly flow pattern is an entrance condition or a non-fully developed flow. A spontaneous changeover to slug or annular flow usually occurs if the channel is long enough. The experiments were performed in turbulent flow with superficial liquid velocities ranging from 5 to 30 ft/sec. The friction, hydrostatic, and momentum pressure drop have been separated and analyzed individually with the aid of two new experimental measurements. These measurements were of the wall shear force and the momentum flux. The validity of these measurements was verified with numerous single-phase tests. Several different air-water mixing methods, with the air always being introduced at the wall, had no affect on the results. Recommendations are presented for the use of these results when applied to steam-water mixtures. 1964-01-01T00:00:00Z https://hdl.handle.net/1721.1/61428 An experimental and theoretical study of density wave oscillations in two-phase flow Yadigaroglu, George; Bergles A. E. 1969-01-01T00:00:00Z https://hdl.handle.net/1721.1/61427 Thermal non-equilibrium in dispersed flow film boiling in a vertical tube Forslund, Robert Paul; Rohsenow, Warren M. The departure from thermal equilibrium between a dispersed liquid phase and its vapor at high quality during film boiling is investigated, The departure from equilibruim is manifested by the high resistance to heat transfer between the dispersed and continuous phases, which result in much higher vapor temperatures and a defect in the amount of vapor generated. The effect on the overall heat transfer is to raise the tube wall temperature, and incomplete evaporation occurs within the tubes. Film boiling tests with liquid nitrogen (70,0 90 SG 190,000 lbm/hr/ft2 and 5000sq/A!25,000 Btu/hr/ft ) were made with 0.228, 0,323, and 0.462 inch ID tubes, 4 and 8 foot long. Visual observations showed that complete evaporation occurs at heat inputs much greater than the required heat of evaporation based on thermal equilibruim (A Hinput >Hfg); in terms of quality, the heat inp t was as large as 300% quality for G = 70,000 lbm/hr/ft . The departure from equilibruim is principally a function of the total mass velocity, being less at higher mass velocities. The non-equilibruim quality was measured experimentally by a helium tracer ggs technique; reliable quality data at G = 70,000 lbm/hr/ft was found to be in agreement with the departure from equilibruim calculated by applying a modified single phase heat transfer coefficient to the film boiling data. A kinematic-heat transfer analysis of the core flow, which takes into account the acceleration, evaporation and breakup of a droplet, confirmed the trends in the departure from equilibrium. A Weber number criterion (Wec = 7.5) was found to adequately describe the breakup of droplets over a partial range of test conditions. Film boiling pressure drop is also reported. 1966-01-01T00:00:00Z