THERMAL BOUNDARY LAYER DEVELOPMENT IN DISPERSED FLOW FILM BOILING Lawrence M. Hull Warren M. Rohsenow . Report No. 85694-104 Contract No. CME-76-82564 A02 I.- Heat Transfer Laboratory Department of Mechanical Engineering Massachusetts Institute of Technology Cambridge, Massachusetts June, 1982 ENGINEERING PROJECTS LABORATORY ,NGINEERING PROJECTS LABORATOR IGINEERING PROJECTS LABORATO ~INEERING PROJECTS LABORAT' NEERING PROJECTS LABORA 'EERING PROJECTS LABOR ERING PROJECTS LABO' RING PROJECTS LAB TNG PROJECTS LAB IG PROJECTS L PROJECTS PROJECTS ROJEC')JEr 02139 MITLibraries Document Services Room 14-0551 77 Massachusetts Avenue Cambridge, MA 02139 Ph: 617.253.5668 Fax: 617.253.1690 Email: docs@mit.edu http://Jibraries.mit.edu/docs DISCLAIMER OF QUALITY Due to the condition of the original material, there are unavoidable flaws in this reproduction. We have made every effort possible to provide you with the best copy available. If you are dissatisfied with this product and find it unusable, please contact Document Services as soon as possible. Thank you. Pages are missing from the original document. Pages 90-92, and 106 are missing. THERMAL BOUNDARY LAYER DEVELOPMENT IN DISPERSED FLOW FILM BOILING Lawrence M. Hull Warren M. Rohsenow Report No. 85694-104 Contract No. CME-76-82564 A02 Heat Transfer Laboratory Department of Mechanical Engineering Massachusetts Institute of Technology Cambridge, Massachusetts 02139 June, 1982 Heat Transfer Laboratory THERMAL BOUNDARY LAYER DEVELOPMENT IN DISPERSED FLOW FILM BOILING by LAWRENCE M. HULL Submitted to the Department of Mechanical Engineering on May 11, 1982 in partial fulfillment of the requirements for the Degree of Doctor of Philosophy in Mechanical Engineering ABSTRACT 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. Thesis Supervisor: Title: Warren M. Rohsenow Professor of Mechanical Engineering -4- ACKNOWLEDGEMENTS I sincerely thank Professor Warren M. Rohsenow for all his helpful guidance and support throughout my stay at M.I.T. I wish to express thanks to Professors Peter Griffith, Borivoje Mikic and Ain Sonin for their careful suggestions and review of this work. I greatly appreciate the daily assistance and support of Wayne Hill and the other members of the M.I.T. Heat Transfer Laboratory. Technical assistance was provided by W. Finley, F. Johnson and J. Caloggero. Thanks to Dr. Harold Edgerton for his help with the photography. Special thanks to Sandy Tepper for all her cheerful help and guidance. This work was funded by a grant from the National Science Foundation. Thanks to Ms. Gisela Rinner for her aid with contract business. -5- TABLE OF CONTENTS PAGE ABSTRACT 2 ACKNOWLEDGEMENTS 4 LIST OF FIGURES 8 11 NOMENCLATURE CHAPTER 1 - 1.1 1.2 1.3 1.4 CHAPTER 2 - INTRODUCTION Dispersed Flow Film Boiling Thermal Boundary Layer Development Review of Related Work Objectives ANALYSIS OF SINGLE PHASE FLOW 2.1 Introduction 2.2 Integral Energy Equation 2.3 Dimensionless Variables 2.4 Single Phase Velocity and Temperature Distributions 2.5 Single Phase Boundary Layer Development 2.6 Average Temperature 2.7 Wall Friction and the Effect of Developing Velocity Profiles 2.8 Comparisons to Other Investigations CHAPTER 3 - ANALYSIS OF DISPERSED FLOW FILM BOILING 3.1 Introduction 3.2 Vapor Energy Balance 3.3 Radiation 3.4 Drop-Wall Interactions 3.5 Temperature Profile for Dispersed Flow 3.6 Evaluation of the Loading Parameter B 16 16 17 19 22 23 23 25 26 27 31 33 35 39 47 47 48 51 52 53 56 -6- PAGE 3.7 General Solution Procedure 3.8 Simplified Solution CHAPTER 4 - EXPERIMENTAL APPARATUS 4.1 Introduction 4.2 4.3 Steam Delivery System Liquid Delivery System 4.4 4.5 Inlet Section Liquid Film Removal System 84 88 94 4.6 Set Points 97 4.7 Free Spray Test Section 97 4.8 Heated Test Section 99 CHAPTER 5 - EXPERIMENTAL RESULTS AND MODEL PREDICTIONS 5.1 Introduction 5.2 Drop Size Measurement 5.3 Single Phase Tests and Heat Loss Calibration 5.4 5.5 Parametric Behavior Dispersed Flow Data and Model Predictions CHAPTER 6 6.1 6.2 CONCLUSIONS AND RECOMMENDATIONS Conclusions Recommendations 80 82 105 105 105 124 127 133 150 150 151 153 REFERENCES APPENDIX A - AXIAL CONDUCTION NEAR DRYOUT 156 APPENDIX B - INTEGRAL APPROXIMATION 160 APPENDIX C - INSTRUMENT SPECIFICATION 164 APPENDIX D - VENTURI CALIBRATION 165 APPENDIX E - ROTAMETER CALIBRATION 173 -7- PAGE APPENDIX F - SET POINTS 176 APPENDIX G - WALL TEMPERATURE ERROR ESTIMATE 192 SINGLE PHASE DATA 195 TWO PHASE DATA 204 APPENDIX H - APPENDIX I - -8- LIST OF FIGURES NUMBER TITLE PAGE 1.1 Dryout in Annular Flow 18 2.1 Single Phase Boundary Layer 24 2.2 Single Phase Temperature Profile 32 2.3 Single Phase Friction Factor 38 2.4 Comparison to McAdams Equation 41 2.5 Comparison to Heinmann Equation 42 2.6 Comparison to Deissler's and Sparrows' Analyses 44 2.7 Comparison to the Data of Mills 45 2.8 Effect of Developing Velocity Profile 46 3.1 Dispersed Flow Boundary Layer 49 3.2 Differential Element for Temperature Profile 54 3.3 Temperature Profile for Dispersed Flow 57 3.4 Force Balance on a Drop 59 3.5 Variation of the Loading Parameter 63 3.6 a,b Comparison of the Simple Model to the Local Conditions Solution 3.7 Distribution Function $l 74 3.8 Distribution Function $2 75 3.9 Parametric Behavior of the Simple Model 79 4.1 Conceptual Experimental Design 81 4.2 Flow Schematic 83 4.3 Steam Supply System 85 4.4 Steam Measurement System 86 71 - 72 -9- TITLE NUMBER PAGE 4.5 Liquid Delivery System 87 4.6 Schematic of Inlet Section 89 4.7 Photographs of Nozzle and Inlet Section 90 4.8 Film Removal System 95 4.9 Pressure-Flowrate Characteristic for the Film Removal System 96 4.10 Spray Photography Set-Up 98 4.11 Heated Test Section 100 4.12 Thermocouple Attachment 101 4.13 Inlet of Heated Test Section 104 5.1 Photograph of Spray 106 5.2 Histogram for Drop Sizes 108 Drop Size Distribution 111 5.15 Correlation for Volume Mean Diameter 123 5.16 Loss Coefficient 126 5.17 Single Phase Test 128 5.18 Single Phase Test 129 5.19 Parametric Behavior of Dispersed Flow 130 5.20 Parametric Behavior of Dispersed Flow 131 5.21 Parametric Behavior of Dispersed Flow 132 5.22 - 5.33 Dispersed Flow Test 134 5.34 Dispersed Flow: Effect of Properties 147 A.1 Axial Conduction at Dryout 157 B.1 Effect of Laminar Sub-layer on Integral 161 B.2 Effect of Laminar Sub-layer on Wall Temperature 163 5.3 - 5.14 -10- NUMBER TITLE PAGE D.1 Venturi Calibration 166 D.2 Venturi Calibration 168 D.3 Velocity Coefficient 169 D.4 Venturi Calibration 171 D.5 Velocity Coefficient 172 E.1 Film Flow Rotameter Calibration 174 E.2 Liquid Supply Rotameter Calibration 175 G.1 Resistance Network for Temperature Error Estimates 193 - M111.MInMM0INUMIMMnINI Ifi .u1W -11- NOMENCLATURE English A Drop size distribution parameter a Acceleration B Droplet loading parameter c Drop size distribution parameter CD Drag coefficient on a drop cp Vapor specific heat d Drop diameter DT Tube diameter f Friction factor F Force g Acceleration of gravity G Mass Flux h Heat transfer coefficient hfg Latent heat of vaporization i Local specific enthalpy of vapor k Thermal conductivity of vapor kd Deposition velocity K Non-equilibrium parameter m Mass N Total number of drops n Droplet number density or number frequency P Pressure -12- English (Cont'd.) q Local vapor heat flux q" Wall heat flux r Radial position from tube centerline S Slip ratio t Time or thickness T Temperature Uloss Loss coefficient u Local vapor velocity U* Friction velocity V Average vapor velocity w Mass flow rate We Weber number x Quality y Distance from wall z Axial position from dryout Greek Void fraction or thermal diffusivity U Emissivity Thermal boundary layer thickness 6m Momentum boundary layer thickness Drop-wall effectiveness n Surface tension -13- Greek (Cont'd.) Non-equilibrium parameter i Vapor viscosity x Kinematic vapor viscosity Distribution parameter p Density (of vapor unless subscripted) Stephan-Boltzman constant Shear stress $1,2 Distribution functions Subscripts a Atomizing av Average b Dryout or burnout bl Average over boundary layer 6m Edge of momentum boundary layer d Drop dw Drop-wall e Equilibrium or entrained f Film properties g Saturated vapor li Liquid inlet 1 Liquid p Patch point -14- Subscripts (Cont'd.) rad Radiation s Saturation or sink T Tube or turbulent tot Total vd Vapor to drop v Bulk vapor or "to vapor" w Wall wd Wall to drop Fully developed 0 Wall or dryout Count mean 2 Area mean 3 Volume mean Dimensionless Variables Nu Nusselt number Pr Prandtl number Re Reynolds number (of vapor unless subscripted) DT+ Tube diameter, DTu*/v I+ Vapor superheat parameter, Eq. (2.25c) r+ Radius, ru*/v T+ Local vapor temperature, (T-Ts)/(q"DT/k) u, Friction velocity, u Local vapor velocity, u/u* /T 0 /P ljlim W11111,1,14 111d 1.1 mu. miII ig~ -15- Dimensionless Variables (Cont'd.) y+ Distance from wall, yu*/v y Distance from wall, y/DT z Axial position from dryout, z/DT Boundary layer thickness, 6u*/v 6 Boundary layer thickness, 6/DT -16- CHAPTER 1 INTRODUCTION 1.1 Dispersed Flow Film Boiling Boiling heat transfer encompasses all heat transfer processes in which the liquid to vapor change of phase is involved. Throughout this work it is assumed that only one component is present; pure water, pure nitrogen, etc. When the heated surface temperature is high enough, no liquid may wet the wall and film boiling conditions are said to exist. (The term "wetting" is used here to indicate intimate contact between the liquid and the heated surface, rather than the phenomenon associated with surface tension.) Heat transfer rates are low as compared with wetted wall boiling processes and wall temperatures may be high enough to cause material damage or failure. Film boiling can exist with any imaginable two-phase flow regime; i.e., the wall is dry but the core may be in slug flow, churn flow, bubbly flow, etc. The flow regime considered in this work is the dis- persed flow regime. In this regime the liquid is present in the form of droplets which are entrained and carried over the heated surface by their own vapor. The heat transfer is mostly by vapor convection, but this is augmented by the presence of the liquid via three mechanisms: 1) Vapor to drop. Since the vapor may become superheated and the liquid remains at the saturation temperature, heat may be transferred from the vapor to the drops. -17- 2) Wall to drops. As a drop comes close to the wall, there is vapor generation from the facing surface of the drop which provides a cushion of vapor that prevents liquid-wall contact. This is termed the drop wall interaction and is modeled as a heat transfer mechanism. 3) Radiation. If the number density of droplets in the dispersion is such that there is a significant view factor and the wall temperatures are high enough, radiation from the wall to the drops may be important. It is of interest to predict wall temperatures in dispersed flow film boiling in such applications as the analysis of the reflood phase of a nuclear reactor loss of coolant accident. Other areas of applica- tion include wet steam turbines and once through steam generators. 1.2 Thermal Boundary Layer Development A common situation which produces dispersed flow film boiling is shown in Figure 1.1. Liquid is fed to the bottom of a vertical heated tube. When the heat flux is moderate, nucleate boiling conditions will prevail in the lower portion of the tube. At some point, called the dryout point, enough liquid has been evaporated so that the rate of deposition of liquid entrained in the core is not sufficient to prevent the liquid film on the wall from drying out. Downstream of this point, the wall temperature increases rapidly and film boiling conditions exist. Since the wall is wet upstream of the dryout point, the vapor at the dryout point is essentially at the saturation temperature. Since the liquid does not wet the wall downstream of the dryout point, the -18- o a 6 I DRYOUT POINT 6 -/O FIGURE 1.1 Dryout in Annular Flow TEMPERATURE -19- vapor superheats and the mechanisms described in the previous section come into play. The region immediately downstream of the dryout point may therefore be modeled as the development of a thermal boundary layer. The investigation of this region, both experimentally and theoretically, is the subject of this work. Flow regimes in which the liquid is not in the form of a dispersion at and downstream of the dryout point are not considered. 1.3 Review of Related Work Dispersed flow film boiling has been the subject of many investi- gations. A very thorough review of the available literature has been performed by Groeneveld [1]. A few of the investigations which are of interest in this study are discussed below. The available experimental data are considered first, and then the prediction methods. There is a fairly large data base for dispersed flow film boiling which includes many heating methods and geometries: flat plate, cir- cular tube, heated rods, etc. The circular tube is considered in this investigation and the following selections from the data base are for this geometry and a constant wall heat flux: Bennett [2] et al., and Era [3] et al., have published data for water, Groeneveld [4] for freon 12, Forslund [5] and Hynek [6] for nitrogen and Koizumi [7] et al. for freon 113. Relatively few experimental investigations of dispersed flow film boiling have included measurements of drop sizes. Recently, ex- perimental data for two-component dispersed flow film boiling have been published in which drop diameters were measured. However, to the knowledge of the author, only the work of Koizumi [7] et al., reports measured drop diameters in single component dispersed flow film boiling in a circular tube. Unfortunately, the dryout point occurred well downstream of the point at which the drop sizes were measured, and therefore allowed the entrainment and deposition processes to further affect the drop size distribution before dryout. Another important parameter which is difficult to measure is the vapor superheat. Forslund [5] used a helium tracer technique to measure actual quality and more recently Chen [8] has developed a technique of measuring the vapor temperature directly. No experiments known to the author have paid particular attention to the region in which the thermal boundary layer would be developing. Very often there is only one wall temperature measurement in this region. There are a large number of correlations for predicting dispersed flow film boiling, and the majority are based on the data mentioned. The most accurate include the effect of thermal non-equilibrium; they account for the vapor superheat in some fashion. Of these, the Chen [9] or the Groeneveld and Delorme [10] correlation currently appear to be the choice selections. Since these are based on the data mentioned above, and use fully developed single phase correlations to estimate the vapor convection, they cannot be expected to do well in the "entrance region" just beyond the dryout point. This is well represented in the model compari- sons presented by Hill [11]. It has been suggested that a single phase -21- entrance length correction be used, but it is not clear that this is the correct procedure. Many workers have developed phenomenological models of dispersed flow film boiling. These analyses model the individual mechanisms mentioned in Section 1.1, combine them with differential mass, energy and momentum balances, and integrate stepwise downstream from the dryout point. Recently, Yoder [12] has presented a simple graphical solution of a one-dimensional phenomenological model. This model, which does not require a computer, is termed the Local Conditions Solution and applies to points well downstream of dryout. This work uses the integral approach of boundary layer theory to analyze the region just beyond dryout. It is directly linked to the Local.Conditions Solution of Yoder [12]; there is a simple transition from the entrance length calculation to the fully developed flow calculation. Both the Local Conditions Solution and the Droplet Laden Boundary Layer model require that the conditions, including the characteristic drop diameter, be known at the dryout point. An analysis of the annular flow regime by Hill [11] has resulted in a method for calculating the proper drop diameter for dispersed flow film boiling. To the knowledge of the author, the only other analysis of thermal boundary layer development in dispersed flow film boiling has been performed by Yao [13]. He used a finite difference technique to solve the two- dimensional energy equation. While the analysis method used by Yao [13] is probably the most accurate, the simplified model presented in Chapter 3 is easy to use and yields reasonable results. -22- 1.4 Objectives As mentioned in the previous section, the experimental and theoretical investigations of dispersed flow film boiling have not concentrated on the "entrance region" just downstream of the dryout point. This investigation adds to the information, both experimental and theoretical, available about this region. The following table lists the objectives of this work. 1) Perform a detailed analysis of thermal boundary layer development in dispersed flow film boiling. 2) Present a simple method of estimating the heat transfer in this region, and indicate the link to the Local Conditions Solution. 3) Collect data specific to the entrance region. In particular, obtain experimental evidence of the parametric behavior by controlling independantly and systematically varying the variables deemed to be of importance by the analysis. -23- CHAPTER 2 ANALYSIS OF SINGLE PHASE FLOW 2.1 Introduction The heat transfer to a single phase vapor in the entrance region of a circular tube is analyzed using the integral approach of boundary layer theory. The flow is assumed to be turbulent and is divided into two zones: a wall zone and a turbulent zone. Velocity and temperature distributions are postulated for each zone and then patched at a point near the center of the buffer layer. The case of a fully developed initial velocity profile is worked out in detail and then a method for including the effect of a developing velocity profile is presented. Figure 2.1 indicates diagramatically the transport quantities to be considered in the single phase analysis. The following assumptions are used repeatedly: 1) The wall heat flux is constant. This assumption neglects conduction through the tube wall material for the case of a constant internal heat generation. However, it is shown in Appendix A that this is a good assumption for many such cases. 2) The properties of the vapor are not a function of radial position but may vary with axial position. It is shown that properties evaluated at the average temperature of the boundary layer yields adequate results even when there are severe radial temperature gradients. 3) Axial conduction of heat through the fluid is negligible. WeJ I FIGURE 2.1 Single Phase Boundary Layer M06IMNAft IWI -25- In the subsequent sections the single phase analysis is presented according to the following outline. The integral energy equation is derived from first principles. Then the dimensionless variables normally used for the analysis of turbulent flow are introduced. Next, velocity profiles applicable to turbulent flow are used to derive temperature distributions. These distributions are then substituted into the integral energy equation and the result solved to find the variation of the Nusselt number with axial position. The method for estimating the effect of developing velocity profiles is then given and all of the results compared to data and analyses published by other authors. 2.2 Integral Energy Equation From Figure 2.1 the following balances are written Energy: dI Iz + I + q" DTdz I = + (2.1) dz Mass: dw wz +w ze or w = wz + z dw = The flow entrained, we (2.2) z dz dz dz , is at temperature Ts so that dw w is = dzi (2.3) -26- Making use of the formulas r w rz 0 p u 2Trrdr (2.4) (> u iz-ardr (2.5) r00 z r0-6 Equations (2.1) to (2.3) are combined to give the integral energy equation r r0 2.3 pu cp(T-Ts) 21lrdr = q"TDT (2.6) Dimensionless Variables The dimensionless variables used in this analysis are Velocity: u = u/u* (a) where u* = /1 O/p (b) Temperature: T+= (T-T)/ k (c) (2.7) -27- Radial Position: r = ru*(d) or y+ yV= etc. (e) Axial Position: z = z/D T (f Using the above definitions the integral energy equation becomes dz 0 d u+T+(r DT -y+)dy+ = 1 2.4 Single Phase Velocity and Temperature Distributions The following assumptions, suggested by Prandtl for turbulent flow over a flat plate, are used to derive temperature distributions from velocity distributions known to fit experimental data. 1) The turbulent Prandtl number is unity. This should be a reasonable assumption, especially for the region near the wall. See Schlichting [14], 7th edition, pp. 706-712. 2) In the laminar sublayer the eddy diffusivities are negligible; only the molecular transport properties are important. (2.8) -28- 3) In the turbulent zone the molecular transport properties can be neglected. 4) The ratio, q/T is constant over the boundary layer and may be evaluated at the wall. For the turbulent zone the velocity distribution 1/7 (2.9) = 8.74 y is based on the Blasius friction factor, which is used since it yields simple and familiar results. For a derivation of Eq. (2.9) see Schlichting [14], 7th Edition, p. 600. Equation (2.9) is used for de- veloping velocity profiles where it must be remembered that +I -I u = u6m for y 4 +4 The shear stress and heat flux are expressed, for the turbulent zone, by du = ( .10) (2 q = -pc( T (2.11) Dividing Eq. (2.11) by Eq. (2.10) and invoking assumptions 1 and 4 results in = - c d/dy (2.12) -29- Substituting Eq. (2.9) into (2.12), integrating with the boundary condition T = Ts when y = 6 and writing the result in terms of the dimensionless variables gives T 6 L 8.74 - 4Pr T T 1/7 1/7 (2.13) T For the laminar zone the linear velocity distribution is used u+ + (2.14) Equations (2.14) and (2.9) cross at y' = 12.54 , determining the patch point for both the temperature and the velocity distributions. The heat flux and shear stress for this case are du T = (2.15) pdyv q = -p ca d (2.16) Invoking assumption 4 again yields "= T0 - cp Pr dT/dy du/dy Substituting Eq. (2.14) into (2.17) and integrating with the boundary condition T = Tw at y = 0 results in (2.17) -30- w-T q" DT = k y (2.18) DT Equations (2.13) and (2.18) are now patched at the point y+ = 12.54 If 6+ is less than 12.54 then only the laminar equation need be considered. At the patch point, Eq. (2.18) gives 12.54 D+ T Tw -T p =q kDT (2.19) and Eq. (2.13) gives qk DT Tp - Ts Equating T 874 D+6/7 Pr P DT _ T 1/7 12.54 1/7 (2.20) [ in these equations results in the expression for the wall temperature 8=74 DT Pr 6 1/7 T + (2.21) 12.54 (Pr - 1) DT+ Pr Using Eq. (2.18) in (2.21) yields the temperature distribution for the laminar sublayer + 74 DT Pr T 6 1/7 DT 12.54 6 1/7 12.54 - y DT T (2.22) -31- Figure 2.2 is a plot of the distribution provided by equations (2.2) and (2.13) for typical values of the parameters 2.5 D+ , Pr , and 6/D Single Phase Boundary Layer Development Substitution of the velocity and temperature distributions into the integral energy equation and integration in the radial and axial directions results in an expression for the thermal boundary layer Use of Eq. (2.21) gives the thickness as a function of axial position. corresponding wall temperature. It is shown in Appendix B that in the evaluation of the integral energy equation the turbulent zone temperature distribution, Eq. (2.13), may be extrapolated to the wall with negligible effect on the predicted This approximation, which results in a considerable Nusselt numbers. algebraic simplification, is used in the following presentation and all of the single phase predictions. Appendix B also shows that it is using the extrapolated possible to make a significant error in T+ w profile. Therefore, Eq. (2.21) is used to estimate Tw When Eq. (2.13) is substituted into Eq. (2.8) and the axial and radial integrations carried out, the relation between boundary layer thickness and axial position is obtained. Thus z DT = 7.43 D+2/ 7 ( T (2.23) - T Nusselt numbers for single phase flow are usually based on bulk temperature. This definition and the dimensionless variables give 1.0 Dr .8 = 1500 Pr = S/Dr - .6- a + 6 8 10 1+ T+ 2 FIGURE 2.2 Single Phase Temperature Profile Sp3 -33- Nu Tw = b (a) (2.24) Nu 2.6 = (b) + + T - Tb Average Temperature Two average temperatures are presented in this section: temperature and the bulk boundary layer temperature. the bulk The bulk tempera- ture is obtained by averaging over the whole tube and is used in the definition of the single phase Nusselt number. The bulk boundary layer temperature is obtained by averaging over only the boundary layer and is the temperature at which the properties are evaluated. For the case of developing flows the bulk temperature can be represented as an increase of average temperature over the inlet temperature: pcuT-T )2irrdr Tb - T s = 0 r0 p (2.24) s Jpcpu 2ffrdr 0 Since the fluid properties are assumed constant, the denominator of Eq. (2.24) is just the flowrate through the tube. Additionally, the numerator of Eq. (2.24) is zero for y > 6 variables, Eq. (2.24) can be rewritten as . Using the dimensionless -34- 6+2Pr + + + kD T )dy+ y Tf u+T+(r0 c wT pv 0 DT kI T (a) (b) cw (2.25) where + = u+T+(r 0 - y+)dy+ (c) T Integrating Eq. (2.8) in the axial direction and using Eq. (2.25c) results in I+ = z/DT . Substituting this result into Eq. (2.25b) gives kD (a) c w z/D T pv + 4 z/DT b Re Pr (2.26) (b) For single phase flow Eq. (2.26) can be obtained from a one-dimensional energy balance (w c (Tb-Ts) = 7DTzq"). However, Eqs. (2.25) apply equally well to single phase or dispersed flow. The bulk boundary layer temperature is defined as -35- Tbl - O pcpu(T-Ts)2rdr (2.27) ~~= Ts pc pu27nrdr S In dimensionless form and using Eq. (2.25c) T Tbl = { I (2.28) u+(r+- y+)dy+ ~ DT Assuming that Eq. (2.9) may be extrapolated to the wall in the evaluation of the integral in Eq. (2.28) gives 1+ bl 2.7 6 8/7 T --T )(6) 8/7 7.65 PrDT+ 16(2.29) 15 DT Wall Friction and the Effect of Developing Velocity Profiles The wall shear stress is intrinsic in the preceeding equations through the dimensionless variables. For a fully developed velocity profile, the wall shear stress is estimated from the Blasius friction factor correlation. The velocity and temperature distributions used for the fully developed flow field case are assumed to apply to developing flow field cases if the variation of wall shear stress with distance is accounted for. -36- The definition of the friction factor is f TO Using the definition of u0= (2.30) PV2 u* gives (2.31) V/f77 /7~ 10'* < Re < 105 where the factor is used to correct the fully developed friction /Tf factor for the entrance region. The Blasius correlation is used for f: .0791/Re 1 4 Combining Eqs. (2.31) and (2.32) and the definition of = (2.32) gives .199 Re7/ 8 /fT7T- (2.33a) For fully developed flow f =fCf and Eq. (2.33a) reduces to f D =. 199 Re7/ 8 (2.33b) MONNIM111111,1411111111W I -37- is estimated from the For a developing velocity profile, f/fC results of Deissler [15], Figure 20 of reference [15]. f/fm = 491 1 1 -1.53(z/DT) (2.34) Equation (2.34) essentially replaces the solution of the momentum equation. Figure 2.3 shows the points used to correlate Deissler's analysis. The correlation is for a Prandtl number near 1. The velocity distribution may be obtained by combining Eqs. (2.9), (2.31), and (2.32): u 1.38 ( ) - 1 /7 T (2.35) where u = u6m for y > 6m ' The momentum boundary layer growth is found by requiring the velocity profile to satisfy the mass balance: pfrV Since u = u6m nr 2 = 0 for r0 O (2.36) .pu2rdr y > 6m r0 r0 6 u 2 rdr + r0-6 2ardr 2Tr0 r (2.37) -38- S oo 1.5 A - ANALYSIS OF DEISSLER. -EQUATION 2.34 L> .0 - 2. 4 6 8 FIGURE 2.3 Single Phase Friction Factor 10 Imwil WIN ,,, -39- Substituting Eq. (2.35) into (2.37), integrating and solving for u /V results in u6m - (1 -- 1 f 26 2 [1D- .2 2 (7) -) 7 4/7 {g( 8/7 -) -T (2.38) T 14m14 6m)15/7 T Equation (2.35) can be applied at y = 6m u6m4/7 = .38( ) - (6 M1/7 ( (.9 (2.39) T Equations (2.34) and (2.38) can be solved for the variation of 6m/DT with axial position Z/DT . For the case of single phase vapor, the Prandtl number is near 1 and so 6m 6 Therefore, the only added complication in the calculation of the heat transfer for developing velocity profiles arises in the evaluation of D+ ; Eqs. (2.33a) and (2.34) should be used rather than just Eq. (2.33b). 2.8 Comparisons to Other Investigations The general calculation procedure is to solve Eq. (2.21) to find wall temperature and Eq. (2.23) to determine the corresponding axial position. ted at Tbl The calculation is iterative since the properties are evalua, Eq. (2.29). The bulk temperature and Nusselt number are '. , ,- iM00---- - -40- then calculated from Eqs. (2.26) and (2.24). In the following compari- sons, the properties were assumed constant and the velocity profile fully developed, resulting in an explicit calculation for Nusselt number versus axial position. At the point where the thermal boundary layer reaches the center of the tube, Ts becomes the centerline temperature, and the analysis provides a prediction for fully developed flow. Figure 2.4 compares this analysis to the widely accepted McAdams [16] correlation. Nub .023 Re 8 Prj 4 (2.40) The Heinmann [17] correlation '04 D Nuf = .0157 Re 84 3( ) (2.41) 6 < z/DT < 60 is for steam and applies over a portion of the developing region. A comparison for a specific case is given in Figure 2.5. Deissler [15] has carried out an analysis of the entrance region heat transfer, also using the integral approach. This analysis differs from his in that he calculated velocity and temperature profiles from eddy viscosity distributions and allowed for temperature dependent fluid properties. Sparrow [18] analyzed the entrance region heat transfer, for a fully developed flow field, using a differential formulation along with -41- 2.5 Pr =.73 z .73 20 - dj.j -- 4.5 FIGURE 2.4 Comparison to McAdams Equation - -- -Mc AAms -42- 1500 - LL 0 HULL - 1000 HEINMANN Soo 2, FIGURE 2.5 4 8 Comparison to Heinmann Equation c Z , nc.h -43- Deissler's eddy viscosity distributions. Figure 2.6 compares this work with that of Deissler and Sparrow. Mills [19] reported heat transfer coefficients for air, drawn in from his laboratory, for a variety of initial conditions. Figure 2.7 compares this analysis to the data of Mills where a value of .015 Btu/hr/ ft/F has been assumed for thermal conductivity in the evaluation of his data. All of the above comparisons have been for the case of a fully developed velocity profile. The predictions of this model for a develop- ing velocity profile are compared to that for a fully developed velocity profile in Figure 2.8. It was assumed that case of developing velocity profiles. 6m = 6 in calculating the Nu Nu.. 1.4 Re = 10 5 1.3 PEISSLEA 1.0 HULL l.0 1.0 2.0 3.0 4.0 I I 5.0 6.0 7.0 I I I 8,0 9.0 10.0 P1FIGURE 2.6 Comparison to Deissler's and Sparrows' Analyses 300 )ATA OF MILLS o 10.Z6 x 104 8.150 x 104 C) 4.860 A Re o 3.012 x x 104 104 1.667 x 10+ v zoo 0 0 0 200 O 1.0 2.0 FIGURE 2.7 3.0 o 4.0 5.0 6.0 Comparison to the Data of Mills 7.0 8.0 9.0 t0.0 1.6 N1 Re = )0 5 1.5 Pr = .73 14 1DEVELOPING VELOCITY PROFILE 1.3 FULLY DEVELOPEP VELOCITY PROFILE 1.1 1.0 Z/PT FIGURE 2.8 Effect of Developing Velocity Profile -47- CHAPTER 3 ANALYSIS OF DISPERSED FLOW FILM BOILING 3.1 Introduction In this chapter the concept of a volumetric heat sink is used to analyze boundary layer development in dispersed flow film boiling. The heat sink is due to the presence of the liquid, which remans at the saturation temperature. The model presented in this chapter reduces to that presented in Chapter 2 when the liquid loading goes to zero. It is assumed that it is valid to superpose heat transfer mechanisms with,a wall heat balance: q1o The terms q ad and qna (3.1) + q1w + q" q w are estimated as if they were totally indepen- dant of each other and the vapor velocity and temperature fields. The following assumptions, in addition to those stated in Section 2.1 are used in this chapter. 1) The liquid is in the form of a mono-dispersion of spherical droplets. 2) The void fraction is near 1. 3) The vapor velocity profile is unaffected by the presence of the liquid. In the subsequent sections the analysis for dispersed flow film boiling is presented according to the following outline. The integral -48- energy equation is modified to account for the heat transfer from the vapor to the liquid and the result written in terms of the dimensionless variables of Chapter 2. The methods used for evaluating drop-wall interactions and radiation from the wall to the drops are presented. Next, the velocity profiles used in Chapter 2 along with the heat sink formulation of this chapter are used to obtain temperature profiles for dispersed flow. Then I method for evaluating the sink strength, using accepted correlations, is presented. The general solution procedure, a simplified solution of the equations, and the connection to the fully developed flow model are given. The comparisons of the model to experimental data are saved for Chapter 5. 3.2 Vapor Energy Balance The integral energy equation must be modified to account for the heat transferred from the vapor to the drops. With reference to Figure 3.1, the vapor energy balance is dz j r0-6 uc (T-Ts) 2frdr + r0 nrd2hvd (T-Ts) 2ffrdr = q"DT r0-6 (3.2) where assumptions 1 and 2 of Section 3.1 have been employed. Some assumption must be made about the radial distribution of liquid in the flow. Since film boiling conditions exist, there is no liquid on the wall. One distribution function which satisfies this con- dition is - av 1/7 n~ = 1. 2 5 ( y ) 0 (3.3) 0 o 0CG o 00 62 0 0 0 | 40 a IIf' FIGURE 3.1 Dispersed Flow Boundary Layer G -50- Equation (3.3) is a convenient choice since it has the same form as the velocity distribution. Using Eqs. (3.3), (2.9) and the dimensionless variables of Section 2.3 in Eq. (3.2) yields d [q" u+T+ (r - y+ y+] dz 0 D+0 T 158 q + (3.4a) B 0 2 Pru+T+(r+ - y+)dy+ +8/7 v T T Pr Introducing the definition of I+ , Eq. (2.25c), d (q" I ) + ~ dz v .158 B ( I+) 8/7 Pr(qI B 2 navd av Nu dd T (3.4b) q" where (3.5) The droplet loading parameter, B, can be thought of as a sink strength. When B=O the analysis reduces to that of Chapter 2. The integral appearing in Eq. (3.4) was defined as I+ by Eq. (2.25c) and is termed the vapor superheat parameter. This parameter is related to the actual and equilibrium qualities. A one-dimensional energy balance, written for the total flow, leads to ill@|lllilllt||ululillalu lll I il a -51- x -x x i hfg (3.6) iv - i , is related to The bulk vapor superheat, I+ and actual quality through their definitions: 4q" =GI+ i -i (3.7) Substituting this into Eq. (3.6) gives the relation between I+ and the equilibrium and actual qualities: 4q" x 3.3 = x e -- Gh fg (3.8) I+ Radiation The formulation used here is the same as used by Yoder [12]. The radiant flux from the wall to the drops is q1a rad= aF wd (T'- T) S where a is the Stefan-Boltzman constant. Fwd = 1 1 Yd Yw (3.9) From the analysis of Sun [20] (3.10) where Yd and yw are the emissivities of the drops and wall, respectively. These are evaluated from -52- Yd and yw = .76 DT -1.1(-a = 1-ed(3.11) for inconel. 3.4 Drop-Wall Interactions The total heat flux due to drop-wall interactions is formulated in terms of the droplet heat transfer effectiveness for one interaction. For each impact of a drop Qd d P h e(3.12) The product of deposition velocity, density, kd , and average droplet number nav , determines the number of impacts per unit time and area. Thus, the total drop-wall heat flux is w dw d3 ~~- h hfg Ana Kdlv The results of Liu and Ilori [21] show that (3.13) kd should be linear with the friction velocity for the conditions of interest in dispersed flow. Iloeje [22] suggests Kd = .15 u* = .15 --DT D+ T (3.14) Both the data of Watchers [23] and the analysis of Kendall [24] indicate that the effectiveness is nearly constant and of the order 10-3 for wall superheats above about 200*F. Even for the developing flows considered -53- in this work wall superheats attain values well above 200*F a very short Therefore, a value of c = 0.0025 distance downstream of dryout. was chosen from the data of Watchers. 3.5 Temperature Profile for Dispersed Flow The assumptions of Section 2.4, except number 4, are used in the following presentation. For the case of dispersed flow, it is assumed that the liquid does not affect the velocity distribution, but it is expected that the presence of the heat sink will affect the shape of the temperature profile. The approach used here is to use the velocity distribution to derive an eddy diffusivity for use in the determination of the temperature profile. Deissler found that the radial distribution of shear stress and heat flux had little effect on velocity and temperature profiles in turbulent flow. The eddy viscosity for the turbulent zone is found from Eqs. (2.9) and (2.10) and the assumption of a constant shear flow: 6/7 D+6/7 (Y - ) 8.74 T DT 7 vT v Assuming aT vT (3.15) results in - = T 7 8 .74 +6/7 Pr DT 6/7 - ) T Considering Figure 3.2, the following energy balance is written: (3.16) -54- tL III ) S lyf-dy I FIGURE 3.2 .< .dy Differential Element for Temperature Profile -55- a q"" + + 3 (pui (3.17) 0 q"'. , is formulated in the same way as in The volumetric heat sink, Eq. (3.2), and the z-gradients are neglected in comparison to the y-gradients. Thus, Bk (T - T d dy D2 T n (3.18) s nav Notice that if B=O and Eqs. (3.16) and (2.11) are used, Eq. (3.18) reproduces the single phase turbulent zone temperature distribution, Eq. (2.13). Combining Eqs. (2.11), (3.3), (3.16) and (3.18) gives the equation for the turbulent zone: d dy 7 8.74 D6/7 Pr DT .38 B y1/7 6/7 dT + 9 (3.19) dy where D DT For the laminar zone, Eqs. (2.16), (3.3) and (3.18) give d2T+ 2 dy = 1.38 B y 117 + T (3.20) -56- For specified values of B, D Pr , and 3 , the temperature distribu- tion is obtained by numerical integration of Eqs. (3.20) and (3.21), subject to the boundary conditions T 9 at = 0 = 6 and (3.21) y = 0 at -1 dy The two zones are patched at the point y+ = 12.54; the heat flux and p temperature are continuous at y+ p A plot of the distribution for typical values of the parameters is shown in Figure 3.3. increasing B . As expected, the temperature is suppressed for The shooting technique was used to solve the equations of this section (see Hildebrand [25], pp. 290-297). 3.6 Evaluation of the Loading Parameter B In this section a method for estimating the loading parameter, B = nidNu d T ,2from accepted correlations is presented. The droplet Nusselt number is calculated from the hard sphere correlation suggested by McAdams [16] for gases Nudf = .7 Re for 17 < Redf < 70,000 (3.22a) \.0 DT =1000 S/DT - Pr .6 = .5 =1 B =2000 13 = 0 L, 8 2+ FIGURE 3.3 10 1z 14 Temperature Profile for Dispersed Flow 16 zo 18 T+ x 103 and the result for pure conduction Nudf = 2 Redf < (3.22b) for 17 The droplet Reynolds number is based on the relative velocity, Redf Pvf(VPp - d vf V, - V (3.23) The relative velocity is estimated with the help of a force balance on a drop. Considering Figure 3.4, FD - mdg = (3.24) mdad The drag force FD D~ 1 P 'v W dV d2(V - V )2C V~)D (3.25) is estimated using a hard sphere correlation: CD = Red (1 + .142 Re 698) for Red < 2000 (3.26a) Mlliftfl Will 10, 1'., -59- I VA v t FIGURE 3.4 j1tt)Ilt Force Balance on a Drop v -60- with a lower limit of = .45 for (3.26b) Red 2000 > The maximum acceleration that the liquid could experience occurs for an equilibrium flow (x = xe) and the minimum for a complete nonequilibrium flow (x = xb). The final results are not overly sensitive to liquid acceleration and the choice of the maximum acceleration has been shown to yield reasonable results for the case of fully developed flow by Yoder [12]. Therefore, the maximum acceleration case (x = xe) is used here. The one-dimensional liquid velocity is written in terms of the slip ratio v S Gx V e (3.27) (3.28) l The liquid acceleration is therefore dVV d = d Gx V d ( PvS V2Qdx x dz (3.29) where it was assumed that 1 dx Saj dz d 1/Sa N x' dz (3.30) -61- and that G/p is constant to obtain the last result. The assumption, Eq. (3.29), is valid since S and a are both asymptotic to 1 and nearly equal to 1 for all cases encountered here and was also shown to be valid for other data sets by Yoder [12]. Thus, the maximum acceleration is dV V2 dx dxdz (3.31) Substituting Eqs. (3.25) and (3.31) into (3.24) results in an expression for the slip ratio 2 S 1+ 1/2 ()] (3.32) The average number density is related to the void fraction through nav = (3.33) 1 -ca The void fraction is related to quality and the slip ratio a 1 + LX (3.34) p29 S and the equilibrium quality is given by Xe = Xb + 4q"I tot z/D fg (3.35) -62- The preceeding equations are solved simultaneously according to the following procedure. Given: d, G, X, DT , and qtot 1) Assume an S 2) Use Eqs. (3.34), (2.38), (3.23), (3.26) and (3.35), in that order, to calculate values of a , V and , Red, CD dX /dZ . 3) Use Eq. (3.32) to calculate S . If this value of S does not agree with the assumed value, then return to Step 1. 4) When the equations have balanced out, calculate from Eq. (3.22), nav from Nud Eq. (3.33) and B from Eq. (3.5) B = nay 'dNuD 2 (3.5) Figure 3.5 shows the result of this calculation for typical values of X , Re, d, DT , and fluid properties. In the general solution procedure, the calculation of the loading parameter is updated at each axial position. Since evaporation is taking place, the drops will decrease in size as they move downstream. The drop diameter can be related to actual quality and the drop diameter at dryout. The liquid flux is oinil~mloililllilliilll1 illi -63X=.I los X=.3 X =.5 10+ Re =2 x104 Re= 4- x 10 io2 - 10 FIGURE 3.5 100 Variation of the Loading Parameter 1000 -64- G(l-x) = navV p (3.36) If no coalescence or breakup occurs, the number of drops is conserved. This may be expressed by = constant navV (3.37) Equation (3.36) may be applied at the dryout point Ad3 G(1-xb) = navV kpk (3.38) 6 Dividing Eq. (3.38) into (3.36) and using (3.37) results in d = db(1 - x b 1/3 ) Equations (2.33) and (2.34) are used to calculate (3.39) D+ The Reynolds number is based on the vapor velocity Gx DT Re (3.40) = v 3.7 General Solution Procedure In this section the method by which the equations of the previous sections may be used to predict wall temperature as a function of axial position is presented. -65- The solution procedure, which involves numerical integration of Eqs. (3.4), (3.19), (3.20) and (2.25c), begins at the dryout point. The boundary layer thickness is zero at this point and therefore the heat transfer coefficient is infinite. Since the vapor is at the saturation temperature and the heat transfer coefficient is infinite, the wall will also start at Ts . Since there is no temperature difference between the wall and the drops, the drop-wall flux and the radiant flux are zero and The calculations are iterative because of the temperature q" = qo. dependance of the fluid properties and the coupling of the wall temperature to the radiant heat flux. The following steps outline the procedure. Given: G, Xb, db, DT, and qtot 1) The loading parameter, B , is calculated for the dryout conditions according to the method described in Section 3.6. 2) The Reynolds number is calculated from Eq. (3.40) and then D from Eq. (2.33). 3) Equation (3.4b) is integrated by one increment in the axial direction. This results in a value of q" I+ corresponding to an axial position Z/DT 4) Guess a wall temperature, Tw * 5) The radiant flux is calculated as described in Section 3.3 using the current estimate of wall temperature, void fraction, and drop diameter. -66- 6) The drop-wall flux is calculated as described in Section 3.4, again using the current values of the required parameters. 7) Equation (3.1) is used to get q" and then V I+ can be calculated from the current value of q" I+ 8) In this step the temperature distribution is found and then integrated to find I+ . I+ calculated here matches that is adjusted until the value of found in step 7). The boundary layer thickness This is described in the following sub- procedure. Given: I+, B, DT , and Pr a) Guess a value of ~ b) Find the temperature distribution by numercial integration of Eqs. (3.19) and (3.20) with the boundary conditions (3.21) and the patch at y+ = 12.54. p c) Use this distribution, Eq. (2.9), and the assumed value of (2.25c) to find 6 , and numerically integrate Eq. I+ d) Check the value of against the required value. return to step a). I+ calculated in step c) If they do not agree, If they doagree the iteration stops. The above sub-procedure yields, in particular, the dimensionless wall temperature, Tw , and the boundary layer thickness, . 111MININ -67- 9) Tbl is calculated from Eq. (2.29) for the value of S found in step 8) and the current values of the other parameters. 10) The properties at Tbl are found. The properties found in this step are used throughout the procedure. 11) Calculate the wall temperature from the results of step 8), the current properties and q" . If it agrees with the value assumed in step 4), qo to step 17). 12) If they do not then qo to step 12). Calculate the quality from Eqs. (3.8) and (3.35) and then the drop diameter from Eq. (3.39) 13) Recalculate the loading parameter, using the current d, G, X, and fluid properties. 14) Recalculate the Reynolds number from Eq. (3.40) and DT from Eqs. (2.33) and (2.34). 15) Reintegrate Eq. (3.4b) to obtain a new estimate of q" I+ but at the same axial position Z/DT 16) Return to step 4). 17) The correct value of wall temperature corresponding to an axial position has been found. The calculation returns to step 3) and repeats until the boundary layer reaches the center of the tube. The results of this calculation are compared to data in Chapter 5. 3.8 Simplified Solution It is desirable to have a solution which does not require numerical integration. The following assumptions result in an explicit solution -68- and have been found to be reasonable by examination of the results of the general solution procedure. a) The heat flux to the vapor is uniform with distance. b) The temperature profile is unaffected by the drops. c) The group B/RePr is constant. d) The velocity profile is fully developed. Using assumption a) and Eqs. (2.25c) and (2.33), Eq. (3.4) becomes, for a fully developed velocity profile, -I + I+ (3.41) = 1 dz With the help of assumption c), this is integratable I+ - B Z/D T) eRePr RePr (3.42) Use of the single phase temperature profile in Eq. (2.25c) results in I+ 9/7 = 7.4373.D+2/7 T The functional dependance of (3.43) T T I on 6/DT in Eq. (3.43) may be approximated by 9/7 ( ) [ - ] 10DT ~ .645( +) DT (3.44) IMMINININIII'l -69- Substituting this into Eq. (3.43) and the resulting expression for I+ into (3.42) gives B Z/D = .366 Re-. 277 - .908 (3.45) B/RePr Neglecting the second term in Eq. (2.21) and using (3.45) to eliminate 6/D T results in .13 B -Z/D T+ 30.23 Re. 78 Pr w (3.46) 1 - e B/RePr The Nusselt number is related to T+ w through their definitions Nu = T (T w s (3.47) 1 TW Therefore, from Eq. (3.46) 1 .13 B Nu = .0331 Re. 78 Pr RePr (3.48) RePr /DT Equation (3.48) relates the wall superheat, Tw - Ts , to the vapor heat flux. Equation (3.1) is used to relate this to the total flux in cases where the radiative and drop wall interactions are not negligible. The group B/RePr is calculated at the dryout point and remains constant. However, the Reynolds number dependance in the leading factor -70- of Eq. (3.48) results from the Reynolds number dependance of the temperature and velocity profiles and is not constant with axial position. The Reynolds number may increase because of the increase in actual quality or decrease because of increasing viscosity. A simple expression for actual quality as a function of equilibrium quality is obtained by substitution of Eq. (3.42) into (3.8) and using Eq. (3.35) to write the result in terms of quality - (xe -x _"/q_ X = - Xe tot[ 1 - e b (3.49a) where A = B Gh f RePr 4q 34b (3.49b) tot The parameter A is closely related to the non-equilibrium parameter, K, derived by Yoder [12] for use in the Local Conditions Solution for fully developed flow. A comparison of their definitions leads to (1-xb)7/12 A =)3/ 2 K x3/4 b (3.50) The Local Conditions Solution for fully developed flow is therefore directly linked to the developing flow solution through Eq. (3.50). A comparison of Eq. (3.49) to the Local Conditions Solution is provided in Figure 3.6. From the figure it is seen that Eq. (3.49) may be used for a change in equilibrium quality of about 10%. This provides a simple .7=. .52 T .3 2.- .1 ~ .2- FIGURE 3.6a .3 .q . 7 3 Xe Solution Comparison of the Simple Model to the Local Conditions -9 1.0 .0 Hull Yoder Hull .5. .E .7 .9 Xe FIGURE 3.6b Comparison of the Simple Model to the Local Conditions Solution -73- criteria for switching from the developing flow model to the Local Conditions Solution: The developing flow model is used until the boundary layer reaches the center of the tube or a change in equilibrium quality of 10%, whichever occurs first. The two analyses have an overlap region. The fluid properties are evaluated at the average temperature of the boundary layer, Tbl . may be ob- A simplified expression for Tbl tained by using Eqs. (3.42), (3.43), (3.44), (3.45), (2.29) and an additional approximation 852 8/7 ( DT [1 -1 ) (3.51) .471 ( ' ) DT 15 DT Thus, .23 ReB T+ bl Re14.13 .07Pr Z/DT RePr 1 - B/RePr 2 (3.52) T When annular flow precedes dryout, the analysis of Hill [11] may be used to calculate conditions at dryout. The necessary results are taken directly from [11] and presented here for completeness. The non- equilibrium parameter, K , is found from $1$2 Go ) 5/6 v x + 1)5/2 Pr2/3 Re4b 3 ( K = .0013 ~p 5/4_b5/12 7/4 1/12 ) (3.53) -74- 557 4.0 !Th {$1i bt I+4A I 6<t 10 10 4 T " 5<i 44 4 I - ; 4x10 4 -~. - 44 u -- 14 7+4 l tt 3 * 6x1 I -ti - 3.04 W2x10 ' Vt -- - +1 -- t 3 2.0 2x1 0T 3 106XO 1j II :jf3t Fu tl;i - Qi + 250 iT-h} 4, tit4 3 .44+ f ~ ~ r , I-- t 7ftI' i -- r t~ + -- -17- - -- - -- ,-+ ---..-.-- 1.0 0.3 0.5 . Figure 3.7 1.0 2.0 3.0 5.0 Distribution Function$ 10 20 30 -75- 2.0 - 1.0 1.9 1.8 1.7 1.6 .90 1 1.5 1.4 .80 1.3 .70 1.2 .60 .50 .40 .30 1.1 1.0 10 2 Figure 3.8 14 We Distribution Function $2 105 106 xb -76- where 2 are read from the graphs in Figures 3.7 and 3.8 and I and We = G2 x2D bT (3.54) py i .0338 xb Rel1/8 b (3.55) to Ghfg The slip ratio is taken as unity and the void fraction calculated from Eq. (3.34). K =.554( In cases where the drop diameter is known d( 5/4 T q1 3/4 tot ) Pr2/3 Re12 Rb Gh ( 5/12 12 (1-X b) T (3.56) Xb1 The following is a summary of the calculation procedure. It is assumed that annular flow precedes dryout and that G, Xb, DT , and q1ot 3.8. are known. Calculate K using Eqs. (3.53), (3.54), (3.55) and Figures 3.7 and Calculate A from Eq. (3.50) and then B/RePr from Eq. (3.49b). Calculate the drop diameter from Eq. (3.56), and the void fraction from (3.34) with S = 1 . The radiation shape factor, Fwd , is now found from Eqs. (3.10) and (3.11). The drop-wall interaction may be written in terms of the void fraction and Reynolds number by combining Eqs. (3.13), (2.33b), and (3.33): INUMUNIMINIIIINNI, -77hfg 1Iw pv DT - .03 = Pv E(1-at)Re 7/ (3.57) Equation (3.57) and the following equations are solved simultaneousTw, Tbl, X, X , Re, q", ly to obtain , and qa as a function of Z/DT ' = Re -v (3.40) T Re. 78 .0331 = a ~ RePr Z/DT + qII '1I ot Xe + = 1 qad 4q"I + Ghtot fg Xb q"/q"_ X =X v tot [1- -A(X -xbI e e ~ RePr s RePr (3.48) (3.35) Z/DT B bl K (3.1) q1I + JT 1.13(Tw-T) B/RePr B Pr Pr i B/RePr (3.49) b] .23 Z/D T q" D v T KV (3.52) -78- qrad = Fwd(T4 - (3.9) T ) The properties are evaluated at Tbl . Figure 3.9 shows the predicted Nusselt number variation for a few specific cases. -79- 160 140 B/RePr= 1.0 I2o B/Re Pr = .1 100 80 G0 40 Pr Ldw = c11~c 4 Figure 3.9 8 I -1 0 ~ 0 1a 16 Parametric Behavior of the Simple Model -80- CHAPTER 4 EXPERIMENTAL APPARATUS 4.1 Introduction This chapter contains descriptions of the apparatus and the pro- cedures used to acquire the necessary data. The objective of the experi- mental work is to obtain data in the entrance region for dispersed flow film boiling. Additionally, it is of interest to test the sensitivity of the heat transfer in this region to the parameters appearing in the analysis: vapor Reynolds number, drop diameter and quality. Figure 4.1 in- dicates the conceptual design of the experiment. To obtain some control over the drop diameter present at dryout, it was decided to use an atomizing nozzle to produce the dispersion. The ratio of water to steam supplied to the nozzle determines the drop size produced. To control the overall vapor flowrate, and thus the vapor Reynolds number, the nozzle sprays the dispersion into a pipe carrying an independently controllable flow of vapor. To obtain a data set which fulfills the objectives in mind, five sets of experiments were required: 1) Calibration of the flow measurement equipment. 2) Determination of set points giving flows at thermal equilibrium at the dryout point. 3) Measurement of drop sizes produced by the nozzle. 4) Calibration for heat losses from the test section. 5) Single and two-phase heat transfer tests. 11=01101hiiii ,..U -81 - 4ea4+e4 Nozzle Atornizing S1eamP" Water Primxxr P Figure 4.1 9em Conceptual Experimental Design -82- As can be seen from Fig, 4.1, the experiments require two steam flows and a water supply. Figure 4.2 is a schematic of the system used to deliver the required flows. The assumptions used in the analysis of dispersed flow film boiling, both that presented here and that of other workers, become more questionable as the liquid loading increases. As a practicality, if the quality is much less than 10% in a "naturally occurring" dryout, the liquid will probably be in the form of slugs or ligaments rather than a dispersion of droplets. It was therefore de- cided to acquire data spanning a quality range of 10% to 50%. The analy- sis of Chapter 3 predicts that, at constant quality and drop diameter, the augmentation of the heat transfer due to the liquid decreases with increasing Reynolds number. It was therefore decided to test Reynolds numbers of 2x10 4 and 4x10 4 . These are on the low side for turbulent flow, yet the factor of 2 is enough to adequately test the Reynolds number sensitivity of the analysis. Pue to the practical reasons mentioned in Section 4.4, a 7/8 inch tube diameter was chosen. The above require- ments allowed the maximum steam and water flowrates to be estimated to aid in the sizing of the supply systems. The following sections describe the steam and water supply systems, the test sections, and procedures used in the experiments. The specifications of the instruments mentioned in the text are listed in Appendix C. 4.2 Steam Delivery System The steam available in the laboratory is at 60 psi with a maximum demand of 3200 pph. The flowrates needed by this experiment are compara- VAC-(ckU?4 T Figure 4.2 FLOW SCHEMATIC, -84- tively small and 60 psi is enough to operate a commercial atomizing nozzle. Figure 4.3 shows the steam supply system. The wet steam is fed to the separation tank on the left. The velocity in this tank is very low, allowing the condensate to be separated. Also, the majority of the crud is knocked down along with the condensate. All pipes and fittings carrying the steam used in the experiment are made of brass to reduce rust problems. Vapor is drawn from the top of the separation tank and split into three lines. One line is used for the outer side of the double pipe heat exchangers, the other two for the two flows required by the experiment. The heat exchangers and electric heaters allow the steam to be superheated slightly before being fed to the venturis for measurement. It is difficult to prevent the steam in the venturi tap lines from condensing and the differential pressure gauges must not be overheated, so the system shown in Fig. 4.4 was used. Each tap line must be level to prevent errors due to the condensate slugs which form. The liquid head on each side of the differential pressure gauge is maintained constant by the overflows. The upstream pressure is measured at the corresponding condensate pot, and the temperature and pressure are measured downstream of the venturis. Guard heaters prevent heat losses between the venturis and the nozzle. The calibration and venturi system checks are presented in Appendix D. 4.3 Liquid Delivery System The liquid delivery system, Fig. 4.5, supplies a measured flow of GUARP HEATERS --4 SEPARATION TANK DOUBLE PIPE HEAT EXCRANGERS STEAM PRAIN PRAIN Figure 4.3 Steam Supply System TO VENTURIS PRESSURE GAUGE -CONDENSING POTS-S OVEAFLOW OVERFLOW RE.SERVOl K RESERVOIA PIFF ER.E NTIAL P- P RE5URE ~RM SUPERHEATERS VENTU SECTION l Figure 4.4 Steam Measurement System GAUGE TO NOZZLE N ITROGEN THERMOCOUPLE ROTAMETER VLNT Q )4 FiLTER FILL I00 ELECTRIC HEATER ELECT RIC HEATER Figure 4.5 Liquid Delivery System PRAIN -88- water at a specified temperature and pressure. All of the piping and fittings are made of brass. The electric heaters submerged in the tank are used to set the temperature of the water in the tank just below the desired delivery temperature. The heater between the filter and the rotameter is used to make any necessary fine adjustments. Regulated nitrogen is used to pressurize the tank. The tank is made from a 2 foot length of 30-inch diameter steel pipe welded to 1/2inch steel plates to close the ends. It was designed to safely operate at pressures up to 120 psi, but 60 psi was found to be a convenient pressure to run at. A rotameter was used to measure the flowrate and its calibration is in Appendix E. The water temperature is measured at the delivery point with a chromel-alumel thermocouple and the Omega trendicator. 4.4 Inlet Section The atomizing nozzle chosen is a Spraying Systems Company 1/4 J-SU42. This is intended to be fed with 1/4 NPT pipes and mounted in a thick wall with a 3/4" NPT bushing. The inlet section, Fig. 4.6, consists of four main components: a mount for the nozzle, a mixing chamber, a bellmouth contraction, and a porous tube. Photographs of these components are shown in Fig. 4.7. All of the components are made of brass except the porous tube, which is made of bronze. The mixing chamber is made from a short length of 2-inch brass pipe. -89- 1~~ I SHAPE 2" brass pipe MIXI NG CHAMBER I.]LI NOZZLE MOUNT Figure 4.6 PRIMAKY STEAM HEAPER Schematic of Inlet Section MITLibraries Document Services Room 14-0551 77 Massachusetts Avenue Cambridge, MA 02139 Ph: 617.253.5668 Fax: 617.253.1690 Email: docs@mit.edu http://ibraries.mit.edu/docs DISCLAIMER OF QUALITY Due to the condition of the original material, there are unavoidable flaws in this reproduction. We have made every effort possible to provide you with the best copy available. If you are dissatisfied with this product and find it unusable, please contact Document Services as soon as possible. Thank you. Pages are missing from the original document. Page Pages 90-92 missing. 90 are is missing. -93- The bottom is threaded to fit into a 2-inch brass cross. The primary steam is fed to the bottom of the cross and the atomizing steam and water are fed to the nozzle with 1/4-inch brass pipes that enter the cross through the cross-flow legs. The purpose of the mixing chamber is to allow the spray produced by the nozzle to mix, and attain thermal equilibrium with, the primary steam flow. A disk serves as both a header plate for the primary steam and a mount for the atomizing nozzle. There are eight 1/8-inch holes drilled around the periphery through which the primary steam flows. The center is drilled and tapped for the nozzle mount, 3/4-NPT. The disk was press fitted into the bottom of the mixing chamber and then silver soldered in place. The bellmouth contraction was constructed according to the ASME flow nozzle shape. The piece that the porous tube is mounted in was machined to fit with the bellmouth contraction. These were silver soldered together and then pressed into the mixing chamber and soldered in place. It was decided to use a 7/8-inch test section diameter to avoid large amounts of liquid deposition in the contraction. The purpose of the contraction is to provide a uniform velocity profile at the entrance of the heated length. The uniform velocity profile was chosen as the entrance condition because a long velocity development section would allow large amounts of deposition. This is undesirable for two reasons: The entrainment/deposition process affects the drop size distribution and the presence of a liquid film on the wall will either -94- force the dryout quality to be very high or the required dryout flux to be very high. Thus, control over two of the major variables of interest would be lost. The purpose of the porous tube is to remove the liquid film just before the start of the heated length. The reasons for doing this are much the same as stated above for the contraction. The film would affect the drop size and the dryout flux, and probably cause a quench of the test section. 4.5 Liquid Film Removal System As mentioned in Section 4.4, a porous tube is used to remove the liquid film from the wall just before the flow enters the heated section. A pressure drop must be applied across the porous tube to cause the liquid to flow through. Further, no vapor should be withdrawn and the liquid flow must be metered. The system shown in Fig. 4.8 fulfills these requirements. The flow from the porous tube is first passed through a sight glass for visual observation. flashing. Next, a heat exchanger cools the water to prevent The flow is measured with a rotameter (see Appendix E for the calibration). The water is then collected in the cold trap. A suction is applied to the cold trap with a vacuum pump. The pressure in the cold trap is maintained at about 8 inches Hg vacuum by allowing the pump to draw some air through the vent valve. The method used to determine the point at which all of the liquid film is being removed, and none of the vapor, can be understood with the help of the pressure drop-flowrate characteristic of the film removal system, Fig. 4.9. As the pressure drop is increased, bubbles are first ob- VENT SIGHT GLASS HEAT EY4CHlANGER ROTAMETER -* SUCTION COLP WATER FROM POROUS TUBE. GOLD TRAP -- Figure 4.8 Film Removal System -96- STEAM BUBBLES A 0 o n0 An AO Qf Cc/s STEAM BUBBLE S 6- 0 OA 20 40 Figure 4.9 60 80 to 20 Pressure Drop-Flowrate Characteristic for the Film Removal System 1+0 160 A P Cm HzO -97- served in the sight glass at the knee and this is associated with the desired operating point. Therefore, all that is necessary is to adjust the pressure drop to the point at which the bubbles are first observed. 4.6 Set Points To determine the best way to operate the apparatus, the following facts were taken under consideration. The test section pressure was always very nearly atmospheric. steam in the nozzle is at a relatively high pressure. The Both the primary and atomizing steam are slightly superheated to prevent condensation in the venturis. Flashing of the liquid will have a strong effect on the drop size distribution of the spray. However, the performance of the nozzle is unstable when cold water is used. Therefore, the water subcooling is determined such that neither condensation or flashing should occur. The performance of the nozzle is steady in a fairly large region around this point. Use of this set point should also ensure that the flow is as close as possible to thermal equilibrium at the entrance of the test section. The set points, which cover the range of conditions mentioned in Section 4.1, are tabulated in Appendix F. All of the experiments were carried out with the apparatus operating at one of these set points. 4.7 Free Spray Test Section The set point determination tests and the photographic study were carried out using a free spray. for the photographic tests. Figure 4.10 shows the arrangement used The spray issues out of the top of the inlet ACTUATING STEAM 4. STEAM EJECTOR I4 I,,, I - SPRAY CAMERA MICKOFLASH PIFFUSE.R Figure 4.10 Spray Photography Set-up UNIT -99- section, travels across a gap of about 1-inch as a free spray, and is then collected by the suction of the steam ejector. This allows optically undistorted visual observation. The steam ejector was constructed from standard pipe and pipe fittings. The photographs were taken with diffuse back lighting. was a Canon Fl and the lens a 135 mm macro. flash duration of about .3 microseconds. The camera The microflash unit has a Kodak plus-X film was used since it has a very fine grain and a fairly slow speed. The photographs were taken by setting the camera on bulb, opening the shutter and then firing the strobe by hand. This ensured that the shutter was fully open when the light pulse reached it. Clouding of the film was prevented by dimming the lights in the laboratory during the photographic sessions. The film was slow enough that total darkness was not required. 4.8 Heated Test Section The heated test section is a 14. inch length of 7/8 OD, .015 to .018 wall thickness inconel 600 tubing, Fig. 4.11. Inconel was chosen as the test section material since its electrical resistivity varies little with temperature. This allows the heat input to be assumed constant, even when there are severe axial temperature gradients present. The heat is applied by imposing a voltage across the electrical connectors. The electrical connectors are heavy pieces of copper flat stock silver soldered to the outside of the tube. The power source was the 15 kw DC motor-generator set available in the laboratory. The electrical power input was found by measuring the voltage between the electrical connectors on the test section and the voltage drop across a shunt resistor -100- F LI ELE..TACAL CONNECTOR. DISTA NCE T HERMOCOLUPLE FROM BOTTOM NUMBiER LLZ 1:3.0 25 I 2.5 20 11.5 19 10.5 18 17 9.5 9.25 24 8.5 16 8.0 15 7.5 14.5 " 7.0 13 6.0 if .5.5 5.25 5.0 i0 23 9 4.5 8 4.0 7 3.5 6 3.0 5 2.5 4 2.0 3 I.5 2 22 0- ).0 I '' \ Figure 4-11 ELECTRIc-AL CONNECTOR Heated Test Section 1 1 wlllalihw -101- SANTOCE L THERMOC.OUPLE - -1/4 INCH ASBESTOS 1116 INCH INCONEL - ASDESTOS MICA STRINCt GLASS TAPE Figure 4.12 Thermocouple Attachment ROPE -102- in the line. These voltages were recorded on the Perkin-Elmer minicomputer available in the laboratory. The arrangement used to attach the chromel-alumel thermocouples to the tube wall is shown in Fig. 4.12. The thermocouple is electrically insulated from the tube by a thin slice of mica, about .001 inch thickness, and then tied onto the tube with asbestos string. A 1/4-inch dia- meter asbestos rope was wound over the string to give the assembly more strength. This prevented the thermocouples from coming loose while the test section was being installed into the apparatus. The thermocouple output was recorded on the minicomputer. Thermocouples 22, 23, 24, and 25 are connected to the Omega Trendicator so that the wall temperature may be observed during the experiments. A wooden box with asbestos end pieces was used to hold the powdered (Santocel) insulation. This provided a minimum thickness of 1-inch of santocel. The expected temperature measurement error is a maximum of about 50*F. The method used to arrive at this estimate is discussed in Appendix G. To prevent warping of the test section during the experiments, the tube was put into tension. This was done by first weighing the complete assembly: tube, connectors, box and insulation, and then applying a force equal to the weight plus about 5 pounds symmetrically to the top electrical connector with a spring scale. The top pipe fitting was designed to provide electrical isolation from the drain pipe and allow travel for the thermal expansion of the test section. -103- An insert, Fig. 4.13, made from 3/4-inch OD type 304 stainless steel, was used to prevent conduction controlled quenching of the test section. This allowed stable film boiling to exist at heat fluxes considerably less than the critical heat flux. This was necessary since the top of the test section would melt at the higher fluxes. The test procedure for the film boiling tests is as follows. With no power to the test section, the flows are set at the desired set point. The power is then set at the desired point. The liquid flow is turned off momentarily. When the wall temperature near the inlet increases above about 350*F, the liquid flow is turned on. The final set point adjustments can then be made and when the wall temperatures reach steady values, about 15 minutes, the computer is started and the data recorded. -104- 7/g" OD 0.0I8 -0.O5ISWALL INCONEL GOO TU BING 1/,/ v x VI VI: O-RINCS CoPPER Jc t E ECTRI CA L CON N ECTOR "00 30/ O.00 STAINLES5 IN.S ERT Figure 4.13 Inlet of Heated Test Section WALL STEEL -105- CHAPTER 5 EXPERIMENTAL RESULTS AND MODEL PREDICTIONS 5.1 Introduction This chapter contains the results of the drop size measurements and the single and two-phase heat transfer experiments. The methods used to extract the drop size information from the photographs is first discussed and the results presented in graphical form. The single phase heat trans- fer data and the corresponding predictions of Chapter 2 are presented along with the heat loss correction technique. Section 5.4 discusses the sensitivity of dispersed flow heat transfer to the major parameters: Vapor Reynolds number, quality and the drop diameter. The comparisons of the analyses of Chapter 3 and the data are presented and the differences between the simplified model and the numerical solution discussed. 5.2 Drop Size Measurement The accurate measurement of drop size distributions is a difficult task. Average drop sizes obtained from measured size distributions, for the same flow conditions, scatter considerably. For example, the data of Nukiyama et al. [26] show a factor of at least 2 scatter in the Sauter mean diameter. The purpose of the experiments performed for this study was to show that the atomizing nozzle controls the drop size and to provide an order of magnitude estimate of average drop sizes present. Photographs of the dispersion at each set point were taken according to the procedure outlined in Section 4.7; Figure 5.1 is an example. The MITLibraries Document Services Room 14-0551 77 Massachusetts Avenue Cambridge, MA 02139 Ph: 617.253.5668 Fax: 617.253.1690 Email: docs@mit.edu http://libraries.mit.edu/docs DISCLAIMER OF QUALITY Due to the condition of the original material, there are unavoidable flaws in this reproduction. We have made every effort possible to provide you with the best copy available. If you are dissatisfied with this product and find it unusable, please contact Document Services as soon as possible. Thank you. Pages are missing from the original document. Page 106 is missing. -107- black line across the corner is a .032-inch diameter wire which provided the scale factor (about 14x) and an easy object to focus on. The focal point (the wire) was set at a point about 1/4-inch from the front of the spray. The depth of field of the lens was about 1/2-inch so this arrangement prevented any out of focus drops from obstructing the view of those that are in focus. The number frequency distribution was obtained by counting the number of drops between selected size pairs and fitting an assumed distribution function to these data, Fig. 5.2. During the counting process, only those drops which were objectively in focus were counted. It was generally possible to discern a difference of 1/32-inch in the drop diameters seen in the pictures. Therefore, at least four data points could usually be obtained. The exceptions are the cases in which very small drops were produced by the nozzle. When the drops are too small, it is easy to confuse them with the photographic grain. The distribution function chosen is a specialization of the function used by Nukiyama et al. [26] to fit their air atomized sprays. This function was also used by Cumo [27] to fit drop size distributions found in mist flows. Thus, n dN = Ad e-cd (5.1) A least squares curve fit of the data to this equation yields the constants A and c. -108- Figure 5.2 Histogram for Drop Sizes IMNIIIN MINNININ liilmlmlliilllmllii 611MINIIINININIMIMI Wili -109- The average drop sizes are obtained by integration of Eq. (5.1). Since the minimum diameter was not measured and since there is no assurance that the largest drop was seen in any given picture, the limits of integration were assumed to be 0 and co . Thus, the total number of drops is ndd N = = A/c2 (5.2) 0 The number frequency distribution is, from Eqs. (5.1) and (5.2), c2 d e-cd n (5.3) The count mean diameter, or most probable drop diameter, is the maximum of Eq. (5.3), d = 1/c (5.4) The area mean diameter is found from N7Td = f0nlTd2dd (5.5) Thu0 Thus d2 = (5.6) (56c) -110- The volume mean diameter is found from 3 N Id 6 3 n J6 = (5.7) 7Td3dd 0 Thus d = r 4/c (5.8) Figures 5.3 to 5.14 show the frequency distributions and average sizes obtained for flow conditions spanning the range considered in this work. Nukiyama et al. [26] showed that the ratio of flowrates fed to an atomizing nozzle has a major effect on the average drop size produced. The same was found to be true in this study. The volume mean diameter is plotted as a function of the ratio of atomizing steam mass flow to water input mass flow in Fig. 5.15. The data shown cover all of the conditions tested and are seen to be independent of the vapor Reynolds number. Thus, the data are reasonably correlated as a function of d3 = 468( wa x 100)- 7 .24 W1 iI wa/Wli , only. (5.9) Figure 5.15 and Eq. (5.9) indicate that the nozzle controls the average drop diameter present. The experiments where "the drop diameter was held constant" were accomplished by holding the flow ratio wa/wli constant. The magnitudes predicted by Eq. (5.9) should not be taken too seriously. As can be seen in Fig. 5.15, there is probably a factor of 3 scatter in d3 for the same flow ratio. -111- 0 x = Qg; = dla da =62.8"" = 74.0,7 =-.30 X 40 80 Figure 5.3 x10 4 A Re 12.0 )20 Ibm/hr 160 Drop Size Distribution ZOO -112- Re = 2x104 \Ali = I2O Ibm/hr 'C) dl 2 = ds3 = 59,,A .34- A4 14 50o 12? 10 8 6 4 Z5 50 Figure 5.4 75 100 Drop Size Distribution 125 150 -113- 8 x Re 4 WIL 120 d13 138K X .46 x 104 lbm/hr 117 o 6 5 4 3 50 100 Figure 5.5 150 200 Drop Size Distribution 250 -114- 0 x Re = 4-x-10+ 120 Ibm/hr. o W I-= d 2 = 39,g 20 d3 =46-5,- X =.45 10 20 40 Figure 5.6 60 80 Drop Size Distribution 100 Re. = Zxi10 60- w' q=29 so- /7'A 10o Z9'fp lbw/Ar K.30- 2.0- /0- A 0 oo 200 300 A100 Soo d94 Figure 5.7 Drop Size Distribution 600 700 Re =x 2. xI 9o0- 2'0 b /r dz =92.5 so 6/0 0 zo 0 - 00 160 zoo Z'10 do) Figure 5.8 Drop Si zeDistribution 280 .320 A-17 Re ='xlo" bs/ = 2'O l0 rZi 'x = .47 diZ = zo,I d,3 = 27 0 x 100 200 300 'q00 500 d(t) Figure 5.9 Drop Size Distribution 600 700 e ='xI 1 wa=290 lbmw/A~r .27 60 C da=/0.5 d3 = s10 IZo 160 oo d (P) Figure 5.10 Drop Size Distribution 23 ZV0O 2PO 320 -119- 80 Re =2-x104 vr,=z360Ib/Mr d: 23qp d =276r x 5:lz 20 - 10. I0 00 200 Figure 5.11 300 '00 Soo Drop Size Distribution 600 -120- 0 Re = 2 x 104 W; = 360 lbv/hr X( = .Iz d. 130A d 3 =15 3 70 60 50 40 30 20 10 12D Figure 5.12 160 z0 Drop Size Distribution 2.40 Z80 320 -121- 90 Re = 4fx1o = 360 lbnV/r X= .20 d.= 1 15 r 10 d = ZO6 r 60- 50 >Cs 40- 30 20 i0 zoo 0 Fgr () Figure 5.13 Drop Size Distribution :OO /00-1 -122- Re t 360 X lbv-n/hr .2.0 70 j 9 . da 60I >04 L10 304\ 7r -5 I0 0IGO Figure 5.14 Zoo Drop Size Distribution 0 o 300 y 100) A 0 0 O '7 V (00 Symbol v- , (pp) 3r0 240 120 0 Creai x 100 Figure 5.15 Correlation for Volume Mean Diameter As can be seen in Figs. 5.3 to 5.14, there are no data points to verify the low end of the distribution. To estimate the magnitude of error possible due to the lack of low end data, the data were fitted to a second distribution function: n =ae-cd (5.10) The estimates of d3 using Eq. (5.10) average 14% lower than those previously given. The individual decreases range from 1 to 25%, the largest decreases occuring at the smallest average drop diameters. This indicates that the average drop size estimates for the cases with small drops may be too large but the estimates for the cases with large drops are fairly good. 5.3 Single Phase Tests and Heat Loss Calibration The single phase tests were used for comparison to the analysis of chapter 2 and to provide a calibration of the heat losses from the test section. This calibration was then used to correct for the heat losses in the two-phase tests. The electrical heat input was known from the voltage and current measurements recorded by the computer. The actual heat input to the steam was found by measuring the steam inlet and outlet temperatures and the flowrate. Thus, nDTL q Tctalp = w c (Tout - T.n) in l oss input actual (5.11) (5.12) 11011MIN Ij -125- Both of the temperature measurements were made well away from the test section. The piping between the measurement points and the test section was insulated and guard heated. The temperature distribution in the steam should therefore be flat and the measurement a good estimate of the bulk temperature of the steam. Due to the insulation arrangement, pipe fittings, and electrical connectors on the test section, the heat loss would be expected to be uneven. Regardless, the losses should be roughly proportional to the differ- ence between the average temperature of the test section and the room temperature: Uloss Tavewall oss - Troom) (5.13) The loss coefficient is also expected to be a (slight) function of the temperature difference. The loss coefficient found from the single phase tests are plotted in Fig. 5.16. A least square curve fit to these points gives U where Tavewall = .0085(Tave,wall - 80) - 2.65 is in F and Uloss (5.14) is in Btu/hr/ft 2/OF. The loss coefficient is correlated with a scatter of about 12%. The heat losses were a maximum of 25% of the total heat input in the two-phase tests. Therefore, the actual heat input for the two-phase runs is known to within about ±3%. -126- "p4) 6 ~ .4 0 5 V 700 800 800 900 P000 900 1000 TC Figure 5.16 1100 -80 Loss Coefficient F) 1200 j300 -127- Figures 5.17 and 5.18 show comparisons of the prediction of Chapter 2 to the single phase data. The heat input to the flow was calculated from the temperature rise of the steam (not the loss coefficient). 5.4 Parametric Behavior The dispersed flow heat transfer tests were run according to the procedure outlined in Section 4.8. The tests resulted in wall temperature as a function of axial position for two vapor Reynolds numbers, qualities between 50 and 10% and a case with large drops and one with small drops for each. It was not possible to predetermine the heat flux. The sensitivity of the heat transfer to changes in quality, drop diameter and vapor Reynolds number can be seen by plotting Nusselt numbers as a function of axial position. The Nusselt number chosen for the compari- sons is defined as Nutot where qt tot T (5.15) is the total heat flux to the flow (corrected for the heat losses). Figure 5.1.9 shows that the sensitivity to changes in quality at constant drop diameter and vapor Reynolds number is small. Figure 5.20 shows that the sensitivity to changes in drop diameter is somewhat larger. Figure 5.21 shows the sensitivity to changes in vapor Reynolds number is by far the largest. Therefore, to obtain accurate estimates of the heat transfer to dispersed flow film boiling, it is most important to know the vapor Reynolds 1000 Tw *F 900 800 7,00 -DT G = 1.0 x 104 )bm/ti+2k 0=9.3 x 103 Btu/-Ft-Inhr = 7/8 inch Ti= ZBO . 4 Figure 5.17 6 Single Phase Test 8 10 12 14 ollilgAiiiii1wilil1w WAIIIiI I, -129- 14 00r- 1300 Tw*F M0o0 1100 1000 |- 104 pttz 1.5 900 - , Dr Ti =2. 1 = 10 4 7/8 inch 281 *F 800 700 Figure 5.18 x Single Phase Test -130- 200 Re = XO q Ai L --)(=.5 q.L Q I-o* 100 D EFFECT OF QUALITY AT CONSTANT DROP/ DIAMETER AND VAPOR REYNOLDS NUMBER 2 96 la 810 Z) INCH Figure 5.19 Parametric Behavior of Dispersed Flow -131- 200 Re a X /o )(- ./0 S- J, =660 F -d,=130,A EFFECT OF PROP PIAMETER AT CONSTANT QUALITY AND VAPOR REYNOLDS NUMBER 100 0 El L 6 00 0 8 10 12- 2, INCH Figure 5.20 Parametric Behavior of Dispersed Flow -132- 0 RUN TPI6 Re = 4 x10 4 = 40, =.45S 200 6 RUN TP20 Re = - -104 di= 40/( > .31 EFFECT OF VAPOR REYNOLDS NUMBF-R AT CONSTANT DROP PAMETER AND QUALITY oo A D OE[ 1O f-1 El r-1L AAAA Z A A 2, inc.h Figure 5.21 Parametric Behavior of Dispersed Flow -133- accurately. If only the total mass flux is known then this implies that the dryout quality must be known accurately. The second most important parameter is the drop diameter. In any application, the drop diameter will not be known accurately and will probably be a function of the upstream flow history. In higher pressure systems, the sensitivity to quality may be greater. This is because of an expected sensitivity to void fraction. The radiation view factor and the coverage area for the drop wall interaction increases with decreasing void fraction. These tests were all at atmospheric pressure where the density ratio is about 1600. Because of this large density ratio, the void fraction changes little over the tested quality changes. 5.5 Dispersed Flow Data and Model Predictions In this section the dispersed flow film boiling data are compared to the models of Chapters 2 and 3. The wall superheat, as calculated by the three models, (single phase, complete two-phase, and the simplified two-phase) are compared to the data in Figs. 5.22 to 5.33. A short vertical line at the end of the curve marks the point at which the thermal boundary layer is fully developed, 6/DT - 1 In most cases the single phase and simple two-phase models did not predict fully developed flow in the distance considered. The single phase model was applied by assuming that the Reynolds number was constant and equal to the vapor Reynolds number at dryout. Further, the radiation and drop wall interactions were neglected. The complete model was applied according to the procedure outlined in Section 3.7. The velocity profile was assumed to be fully developed and -1342000 PHASE LL 0 NUMERICAL AND SIMPLE MODELS PAOPERTIES ATTu, 1500 RUN TPii 1000 c, -q.q x i0' 3. 0 X10'' Re Ibm/f/ hr Btu/h)/hr 2.xl0Y .08q 43 375<,, z.31 500 Z,INCH Figure 5.22 Dispersed Flow Test -135- 2.000 U- SINGLE PHASE NUMERICAL AND SIMPLE MODELS PRO PERTIES AT T |5.00 SIMPLE, cs (000 SI MPLL c, RUN TPlO G-=g.'9 X 10'' )bm/ fi/f '=2.3 X o 6lu/fe'/hr E00 Re= 2-X 10' d 1 5 8,a 3 =Ibst, Figure 5.23 K, =.013 K3 -. 10 Dispersed Flow Test -1362000 PHASE NUMERICAL AND SIMPLE MODELS PROPERTIES AT Ti LL I500 SIMPLE, RUN TP Iq 1000 G = 5g,(1 oN&i I he /f t'hr I 10 Bfv/iP/hr 7t,= 2.. VX Xb .16 Re = 2-XIDO /03m a3 = 3057 K,=./0 K3 =.y2 500 2. 6 "Z,I NLIH Figure 5.24 Dispersed Flow Test -137- 2000 SINGLE PHASE NUMERICAL AND SIMPLE MODELS PROPERTIES AT Tb 1500 SIMPLE, 100D K AALP ~L~ L RUN TPH1 G = 5.3 X10" I6m/ft"/hr q;+m-Z.7 X 10 Etu/ft/hr Xb Re 19 -.. 1.11 10 dS7= C3, =(7e'ob *6 ) K, .027 K3 =.Z22 500 z 4 7,INC H Figure 5.25 Dispersed Flow Test -1382000 SINGLE PHASE 6- 1500 .fSMPLE , da 'n,'n,'n RUN TP20 G =2.7 x 0'O 1000 y .-25 XJO" X6 =.31 Re =2X O d,I Y2. , K, d3 = IZ.0M , .032. k3 =2. PROPERT1ES AT TI 0 0o 2 4 6 10 g Z mch Figure 5.26 Dispersed Flow Test -139- Z000 SINGLE PHASE L0 NUMERICAL AND SIMPLE MODELS PROPERTIES AT T 1500 - d3 SIMPLE) d 1000 - RUN TP21 G = -. 5-S 10 4 lbmn// = 2.4 x 104 BtLkt 500 - /hr 2 /hr x = 33 Re = 2 x 104 d I = a22, 7 K , = .0090 = .077 d1 = 63 , K 3 2. + ZE, inch Figure 5.27 Dispersed Flow Test -140NUMEICAL AND SIMPLE MODE.LS PROPERTIES AT Tb 1500 SINGLE PHASE,,, IL 0 1000 Annnz RUN 500 C G, TP)3 =2.9 x 104 = 4 x 104 = 70^ , K.056 = dl 3 2. Figure 5.28 i6-//e/hr -19 Xb Re c 8.7 x 104 Btu/tl/hr 68 Dispersed Flow Test K= .2.8 2O, to 2 inch -1412000 SINGLE PHASE LL NUMERICAL AND SIMPLE MODELS PROPERTIES AT Ti, 1500 SIMPLE 1000 RUN TP3 G= 5.5 X 10 I bm/iL'/kr '=2..5 X1o0 8t A/hr/ff' SO00 Re= 2X10 d, 3, = K, =.007g 3 5q K3 = .07/ 2-. Z,INCH Figure 5.29 Dispersed Flow Test -14?- 5 I NGLE PH4ASE I500 LL M PLE , d3 SI MPLE ,,A 1000 -, NUMERICAL AND SIMPLE MODELS PRO PERT IES AT T RUlN TPI8 G=6.q X ty10 |bmfL1/hr -:3.4 X10' Bru/W/hr 4 2.Y Re =X iO 00- d, = "7 c)3 = 13i 2 9 ) , 6 Z, INCH Figure 5.30 Dispersed Flow Test = .33 - .0q0 111mm NOW 114111111111 Wild -143- Z000 NUMERICAL ANP SIMPLE MODELS PROPE-RTIES AT Tbi TP6 RUN F- CG = -, 1500 X= 3.6 x 104 1bm/ftz/hr ). 9 x 104 B/hr/ft2 .45 4 x 104 Re di = 22^ K, = .012da = 64A ) K 3 = .J0 1000 SINGLE. PHASE. 500 ZA nL Z 2 Figure 5.31 4 6 8 Dispersed Flow Test 10 -144- 2000 NUMERICAL ANP SIMPLE MODELS PROPERTIES AT Tb1 tL IF- 1500 SINGE PHASE 1000 n 6 I RUN TP12, 500 Q 8.2. x04 2.7 x )04 x 10 Ibm/ftZ/hr Btak/ftz/'hr Xb = .21 Re li d3 a 4.3 - 48 , =- 138, K,= .027 K 3 = .. 4 12 2, Figure 5.32 Dispersed Flow Test INCH kW , -145- 2ooo NUMERICAL ANP SIMPLE MODELS PROPERATIES AT Tbl U- SINGLE PHASE. 1500 SIMPLE) d3 1000 n ins 4n, n n RUN TP9 500 6.2.x 10" 1bst/W/hr G 2. 8 x 104 9Jt X6 2. 4 6 Re .27 4 x 104 d3 40, , d(3 ) 1 6 /Ak 8 8tu/ t Dispersed Flow Test hr K=.02.4 . K3 = . ?-3 10 ~ , inc.h Figure 5.33 / -146- the calculation was carried out for both the count mean and the volume mean drop diameters. These drop sizes were chosen since they represent a span of the size distribution. The pairs of runs in which the drop diameter was held constant were generally fit equally well by choosing the same drop size for both. The cases with large drops were fit best with the volume mean diameter. However, use of the count mean diameter most often gave the best fit to the data. It is not clear that this has any physical implications, and is discussed further in the conclusions, Chapter 6. The simplified model was then used to predict the data. To provide a consistent comparison, the same drop diameter which was used in the complete model was used in the simplified version. The complete model always predicts lower wall superheats than the simple model, but the differences are not very great. The major difference between the simple solution and the complete solution is that the complete model predicts shorter development lengths. This is because of the suppression of the vapor temperature profile, Fig. 3.3. The heat sink term in the energy equation tends to suppress the growth of the boundary layer. However, the suppression of the temperature distribu- tion tends to increase boundary layer growth and this effect dominates. Figure 5.34 shows the sensitivity of the predictions to the temperature at which the fluid properties are evaluated. In each case, all of the vapor properties were evaluated at the indicated temperature. Four temperatures were chosen for comparison: saturation temperature, wall temperature, film temperature (Tw + Ts )/2), and the average temperature of the boundary layer. WN1011011111IN11W Wmill WiNifti. -147- S500 SINGLE PHASE d3 , PROPERTIES I- IDOD &OL RUN TPib G = 3.7 X10 b6,/UL/hr Z-= 3.0X10' Btu/ML/hr Re = .I Xl10 d., = 106 , K3 S=.50 goo I. NUMERICAL MODEL Z INCH Figure 5.34a Dispersed Flow: Effect of Properties -148- PROPERTIES EVALUATED AT SIMPLE MODEL 1500 TN# Twa 66L 1000F RUN TP11 B4/RJhr s .0(3 1' / Re= 2. X /0 d =c =375 ,<K=.'6I 500 Z, rNC H Figure 5.34b Dispersed Flow: Effect of Properties -149- The average boundary layer temperature was used in the previous figures since it worked well for the single phase predictions and has some theoretical basis. However, the two-phase models appear to be the most successful when the film temperature is used. that the film temperature be used. It is therefore recommended -150CHAPTER 6 CONCLUSIONS AND RECOMMENDATIONS 6.1 Conclusions 1) The simplified solution yields reasonable wall temperature estimates, but maybe in error in development length estimates. 2) The calculation of the loading parameter is the most questionable part of the analysis. Use of the count mean diameter usually resulted in the best prediction of the dispersed flow data. There are several possibilities of why it was necessary to choose such a small drop: a) It is possible that the droplet Nusselt number should be increased due to the turbulent fluctuations in the flow field. b) It may be important to include the complete distribution of drop sizes. c) Different radial liquid flux distributions may have been produced at different set points. At high liquid flowrates it is possible for the liquid to be concentrated in the center of the pipe. d) The drop size data may be slanted toward the larger sizes, especially for the small drop cases. e) A radial distribution of drop sizes may exist; i.e. the small drops may collect near the wall. NOUNHOW -151- 3) Due to the suppression of the temperature distribution in dispersed flow, it may be reasonable to evaluate the fluid properties at a temperature higher than the average temperature of the boundary layer. The chosen temperature would most likely be a function of the liquid loading. 6.2 Recommendations Additional experiments could be used to clear up the questions raised in Conclusion 2. 1) There are techniques for producing mono-dispersions. Therefore, it should be possible to compare the heat transfer for a mono-dispersion to that for a wide distribution of sizes. 2) Run experiments in which the radial liquid flux distribution is controllable. 3) Make measurements of radial drop size distributions to determine if the drops have a preferred location in the flow field in dispersed flow. 4) Develop a technique to measure the radial vapor temperature distribution. 5) Develop a technique for measuring the radial velocity profile in dispersed flow. If this analysis is to be used to estimate the heat transfer just downstream of a quench, it may be necessary to include the effect of conduction in the tube wall material. Additionally, it may be desirable to eliminate the infinite heat transfer coefficient at the dryout point. This could be done by extrapolating a transition boiling heat transfer coefficient until it intersects with that predicted by this analysis. mw mw NMIW111941 -153- REFERENCES [1] Groeneveld, D.C., Gardiner, S.R.M., "Post-CHF Heat Transfer Under Forced Convective Conditions", AECL-5883. [2] Bennett,A.W., Hewitt, G.F., Kearsey, H.A., Keeys, R.F.K., "Heat Transfer to Steam-Water Mixtures in Uniformly Heated Tubes in Which the Critical Heat Flux has been Exceeded", AERE-R 5373. [3] Era, A., Gaspari, G.P., Hassid, A., Milani, A., Zaveltarelli, R., "Heat Transfer Data in the Liquid Defficient Region for SteamWater Mixtures at 70 Kg/cm 2 Flowing in Tubular and Annular Conduits", CISE-R-184, 1966. [4] Groeneveld, D.C., "The Thermal Behavior of a Heated Surface at and Beyond Dryout", AECL-4309, 1972. [5] Forslund, R.P., "Thermal Non-Equilibrium in Dispersed Flow Film Boiling in a Vertical Tube", Ph.D. Thesis, Massachusetts Institute of Technology, December, 1966. [6] Hynek, S.J., "Forced Convection Dispersed Flow Film Boiling", Ph.D. Thesis, Massachusetts Institute of Technology, 1969. [7] Koizumi, Y., Ueda, T., Tanaka, H., "Post Dryout Heat Transfer to R-113 Upward Flow in a Vertical Tube", Int. J. of Heat and Mass Transfer, Vol. 22, pp. 669-678, 1979. [8] Nijhawan, S.M., Chen, J.C., Sundaram, R.K., "Parametric Effects on Vapor Non-Equilibrium in Post-Dryout Heat Transfer", ASME 80-WA/HT-50. [9] Chen, T.C., Ozkaynak, F.T., Sundaram, R.K., "Vapor Heat Transfer in Post-CHF Region Including the Effect of Thermodynamic Non-Equilibrium", Nuc. Eng. Des. 51 (1979), 143-155. [10] Groeneveld, D.C., Delorme, G.G.J., "Prediction of Thermal NonEquilibrium in the Post-Dryout Regime", Nuc. Eng. Des. 36 (1976), pp. 17-26. [11] Hill, W.S., "Dryout Droplet Distribution and Dispersed Flow Film Boiling", Ph.D. Thesis, Massachusetts Institute of Technology, May 1982. [12] Yoder, G.L., "Dispersed Flow Film Boiling", Ph.D. Thesis, Massachusetts Institute of Technology, March 1980. -154[13] Yao, S., Rane, A., "Numerical Study of Turbulent Droplet Flow Heat Transfer", ASME Paper WAM 1980. [14] Schlichting, H., "Boundary Layer Theory", 7th ed. McGraw-Hill 1979. [15] Deissler, R.G., "Turbulent Heat Transfer and Friction in the Entrance Regions of Smooth Passages", Trans. ASME, November 1955. [16] McAdams, W.H., "Heat Transmissions", 3rd ed., McGraw-Hill, New York, 1954. [17] Heinmann, J.B., "An Experimental Investigation of Heat Transfer to Superheated Steam in Round and Rectangular Tubes", ANL6213, 1960. [18] Sparrow, E.M., Hallman, T.M., Siegel, R., "Turbulent Heat Transfer in the Thermal Entrance Region of a Pipe with Uniform Heat Flux", Appl. Sci. Res. Section A, Vol. 7. [19] Mills, A.F., "Experimental Investigation of Turbulent Heat Transfer in the Entrance Region of a Circular Conduit", J. Mech. Eng. Sci., Vol. 4, No. 1, 1962. [20] Sun, K.H., Gonzalez, J.M., Tien, C.L., "Calculations of Combined Radiation and Convection Heat Transfer in Rod Bundles Under Emergency Cooling Conditions", ASME Paper, 75-HT-64, 1975. [21] Liu, Y.H., Ilori, J.A., "Aerosol Deposition in Turbulent Pipe Flow", Environmental Science and Technology, Vol. 8, No. 4, April 1974, pp. 351-356. [22] Iloeje, 0.C., "A Study of Wall Rewet and Heat Transfer in Dispersed Vertical Flow", Ph.D. Thesis, Department of Mechanical Engineering, Massachusetts Institute of Technology, February 1975. [23] Watchers, L.H.J., Westerling, N.A.J., "The Heat Transfer from a Hot Wall to Impinging Water Drops in the Spheroidal State", Chemical Engineering Science 21, pp. 1047-1056, (1966). [24] Kendall, G.E., Rohsenow, W.M., "Heat Transfer to Impacting Drops and Post Critical Heat Flux Dispersed Flow", M.I.T. Heat Transfer Laboratory Report No. 85694-100, March 1978. [25] Hildebrand, F.B., "Introduction to Numerical Analysis", 2nd ed., pp. 290-297, McGraw-Hill, 1974. UMM1 111&1111 -155- [26] Nukiyama, Tanasawa, "Experiments on the Atomization of Liquids in an Air Stream", Trans. Soc. Mech. Eng. (Japan), vol. 5, 18, 1939. [27] Cumo, M., Farello, G.E., Ferrari, G., Palazzi, G., "On Two-Phase Highly Dispersed Flows", Trans. ASME, Nov. 1974, pp. 496-500. -156- APPENDIX A AXIAL CONDUCTION NEAR DRYOUT The most severe wall temperature gradients occuring in dispersed flow film boiling result from the rapid change in heat transfer coefficient near the dryout point. The simple model depicted in Figure A.1 is used to obtain an estimate of the distance beyond the dryout point over which axial conduction is an important heat transfer mechanism. The dryout point is the point z = 0 . For heat transfer coefficient is assumed infinite. z less than zero the For z greater than zero the heat transfer coefficient is relatively small. The vapor temperature is assumed constant and there is constant volumetric heat generation within The applicable energy equation is the tube wall material. h d2 dz where, w (A.1) = T-T with boundary conditions e 6 = Ts - T = The solution of Eq. (A.1) = e z = 0, 0 finite at z =o is -/h/ktiz 0 at . + wt -/h/kt z - - (1 - e ) (A.2) *-- TV h = oo TS h Tt wi +- Figure A.1 Axial Conduction at Dryout iNSULATION -158- An estimate of the length over which conduction is important is therefore Lcond (A.3) = The order of magnitude of h can be estimated from the McAdams equation Nu = .023 Re. 8 Pr'4 (A.4) For steam flowing in a one inch tube with a Reynolds number of 104 at 500*F and atmospheric pressure h ~ 10 Btu/hr/ft 2 /OF The tube used in the experiments has a wall thickness of .015 - .018 inches and the conductivity of Inconel 600 is 9 Btu/hr/ft/*F. Therefore Lcond ~ .4 inch Since the development length is usually much greater than .4 inches, it is reasonable to neglect axial conduction. In addition, many applica- tions involve flows where the Reynolds number is much larger than 104 and will therefore exhibit negligible axial conduction effects over the majority of the thermal boundary layer development length. However, the effect of -159- conduction should be checked if the wall material has a high thermal conductivity or a large. effective thickness. -160- APPENDIX B INTEGRAL APPROXIMATION In this appendix it is shown that the extrapolation of the turbulent zone temperature profile to the wall results in negligible error in the Pr / 1, the evaluation of the integral Equation (2.8) but that, for laminar zone profile should be used in the evaluation of the wall temperature. Substituting the turbulent and laminar temperature and velocity profiles, Eqs. (2.9), (2.13), (2.14) and (2.22), into the integral energy equation, Eq. (2.8), and carrying out the integrations results in: z/DT = 7.43 D+217 6 9/7 T 66 1-lOT 1203 77 1/7 T 13.4 DT T + 1535 DT + 986 DT 21650 + D+2 D+6/ DTT DT (Pr - 1)[I 687 - 16.7 DT -) [1 ( 61/7 T 657 Pr DT This result is compared to Eq. (2.22) in Fig. B.l. 6.7 DI (B.1) DT+ 18 DT Equations (2.21), (2.26b), and (2.24b) were used to calculate the Nusselt number. The approximation is seen to be within about 5 percent of the exact integral. 110 Nu. Re = Pr =.73 .667 x )04 90 70 APPROXIMATE JNTEGRAL zEXACT S34 INTEGRAL 5 6 7 8 9 Z/ Figure B.1 Effect of Laminar Sub-layer on Integral T -162- The error made by extrapolating the turbulent zone profile to the wall to obtain wall temperature may be more serious. The extrapolated turbulent zone profile and the patched distributions predict the same wall temperature when Pr = 1. The laminar zone and the extrapolated turbulent zone distributions are compared for the case of a fully developed boundary layer, 6/DT I , and Pr = .71 in Fig. B.2. In this case the extrapolated turbulent profile overestimates the wall temperature by about 18 percent. Therefore, the laminar zone distribution was used to calculate the wall temperature in all of the predictions shown at the end of Chapter 2, .014 .o2 .010 .008 .006 .004 .002 4 6 8 10 03x Figure B.3 Effect of Laminar Sub-layer on Wall Temperature -164- APPENDIX C INSTRUMENT SPECIFICATION Perkin-Elmer minicomputer 12-bit accuracy provides temperatures measurements to within ± 1F Omega Trendicator, Model 403A 1VF resolution -165- APPENDIX D VENTURI CALIBRATION The system shown in Fig. D.1 was used to calibrate the venturis. The steam was condensed completely and the flowrate found by measuring the volume of water collected and the corresponding time. For each flow- rate, the upstream pressure was measured at the corresponding condensate pot. The temperature and pressure of the steam were recorded at a position several diameters downstream of the venturi and the upstream temperature was assumed to be equal to the downstream temperature. The isentropic flow equations, along with a velocity coefficient, were used to correlate the pressure drop-flowrate data. These were cast in terms of the isentropic expansion factor and the density at the inlet of the venturi, station 1: w = cYA2 fl (D.1) 4 1 - 8 k-l P Y P2 2/k 2/k = -2/ k P 4]( k 3 is the diameter ratio of the venturi and heats for steam: 1/2 4 I2 (D.2) k is the ratio of specific k = 1.33. The atomizing steam flow was measured with a 1/2-192 Barco venturi. The expansion factor for this venturi is closely approximated by PRESSURE GAUGE -O CONPENSING PoTS-, OVEU\FLOW RF-SERVOI K vORFLOW RESERVOI A ' PIFFERENTIAL PRE5$URE- CA UGE on IJ VENTUKI ,,-CONPENSEPs Figure D.1 Venturi Calibration -167- Y = (D.3) .3898 + .6121 ( P2 over the range of operating conditions. A curve fit to the calibration data, Fig. D.2, gives: wa = 5.0648 Y /p-AP - 1.9942 The density is in lbm/ft 3 and AP (D.4) in inches of water to get mass flow Cv , can be calculated from v Eqs. (D.1), (D.3), and (D.4) if it is assumed that Cv is a function of rate in lbm/hr. The velocity coefficient, Re only and that Cv = y is constant: 975 1 - 41 + ] (D.5) This is plotted if Fig. D.3 and is seen to have the same general shape specified for ASME venturis. Since the magnitude is near 1, it is concluded that the venturi is reliably calibrated. The same procedure of calibration and check of velocity coefficient was done for the 3/4-390 Barco venturi used to measure the primary steam flowrate. The relevant equations are Y, = .3759 + .6257( ) P P1 wp 34.95 Y/p1AP - 4.489 (D.6) (D. 7) -168- 28- i/2 - 192 2.4- 20 d 8- 4 I I 1 2 | 3 YNd"A, Figure D.2 4 V 5 m9'mhH0 Venturi Calibration -169- 1.0 .q7 5 Cv W/v - M2 0.9 0.8 I 0.7 InRe Figure D.3 Velocity Coefficient - - -170- cV = .936[1 3] - 4.9x10 Re + 1 and the results are plotted in Figs. D.4 and D.5. (D.8) -171- 803/4 -390 70 50 - 40 - 30 - 20 - 10 - 3 2 Y-/ Figure D.4 Venturi Calibration , 3 inch HZO lbrn/f+ 16 -172- 1.0 3/19 - 390 0.9 0.2 S9 nRe Figure D.5 Velocity Coefficient -173- APPENDIX E ROTAMETER CALIBRATION Rotameters were used to measure the liquid input flowrate and the film removal flowrate. Both were calibrated by measuring the volume collected over a known period of time. Also, they were calibrated with water near the operating temperature. As expected, the rotameters show a very linear characteristic. The film flow rotameter calibration is shown in Fig. E.1 and is correlated by Q[cc/s] where = .1606 X + X is the reading. .5657 (E.1) The liquid input rotameter calibration data is shown in Fig. E.2 and is correlated by Q[cc/s] where = .5257 X - .07114 (E.2) X is the reading. The point on the float at which the reading was taken is shown on the corresponding graphs. FILM FLOW ROTAMETER TUBE: FP 1/2 - 17 - (A - 10/55 CUSVT 4OT60 FLOAT:-/Z EO K READI NG H E.RE gi U 10 1- 20 80 30 90 X (Reading) Figure E.1 Film Flow Rotameter Calibration 100 35 ( - ROTAMETER LIQUID SUPPLY TUBE: FLOAT: FP-1/2.-2.7-GA-I0/55 T6-603A943 30 - READING L I HF-RE. 20 15- 10 10 20 30 Figure E.2 40 so 60 70 Liquid Supply Rotameter Calibration 80 90 100 -176- APPENDIX F SET POINTS The set points used are tabulated in the following tables. -1774 x 104 120 ibm/hr 198 OF Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: Nominal Drop Size, d 3 116 yi Atomizing Steam Flow Primary Steam Flow P,, PSIG d P mnH2.0 P,, PSIG 260 1.0452 97 3.4 4.4 T *F P,, 16/ff r -60.5 Liquid Film Flow . READI NG 'T, *F 27 50 WoS l/r 38.9 Test Section Conditions Wy Ibm/hr 68.7 W XRe 81.1h .459 4.02xl04 W,. 16,/hf -178Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: : Nominal Drop Size, d3 4 x 10' 120 ibm/hr 197 OF 62 y Atomizing Steam Flow Primary Steam Flow Liquid Film Flow READING 26 IT *F 50 Wf Ibw/hr 37.6 Test Section Conditions -179Nominal Liquid Liquid Nominal Reynolds Number: 2 x 120 Input Flowrate: Input Temperature: 204 : 120 Drop Size, d3 101 lbm/hr OF yp Atomizing Steam.Flow Primary Steam Flow Liquid Film Flow READING T *F W; \bmVhr 29.5 50 41.8 Test Section Conditions -180Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: Nominal Drop Size, d 3 2 x 10' 120 ibm/hr 204 0F 120 i Atomizing Steam Flow Primary Steam Flow Liquid Film Flow Test Section Conditions -181- 2 Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: . Nominal Drop Size, d3 x 10' 120 ibm/hr 203 'F 62 p Atomizing Steam Flow Primary Steam Flow P,, FSIG 6l/fI AP, nH,0 1, P,.PSIG - .7 Liquid Film Flow READING T 37.5 Wf Ib/hr 50 51.9 Test Section Conditions Wy lb/hr 34.5 W Ibm/hr 68.1 X .336 I I W IL6/hrl -I 260 .0364 8.7 Y, T *OF Re 2.02x10' 15.0 -1824 x 10' Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: Nominal Drop Size, d 3 240 ibm/hr 205 OF 189 yi Atomizing Steam.Flow Primary Steam Flow P,, PSIG 4.3 P,F,. Plo 3.3 APm H,0 fee I M .0451 95 T*F -59.8 260 Liquid Film Flow READING 40 I4 *F Y Wf Ibwvhr 54.4 50 Test Section Conditions We ib./hr WNMEMIIIM 110IM1111, 1011111MEM11111111 i'11111 -183- Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: Nominal Drop Size, d 4 x 10' 240 lbm/hr 205 OF 189 yi 3 Atomizing Steam Flow Primary Steam Flow P, PSIG 4.3 P,. PSIG 3.3 A P,MH1 95 P, I6./* ' .0451 T, 'F Y, M 260 - Liquid Film Flow Test Section Conditions W. I&../hrI I 59.8 -184Nominal Liquid Liquid Nominal Reynolds Number: Input Flowrate: Input Temperature: : Drop Size, d 3 4 x 10 240 lbm/hr 204 "F 108 y Atomizing Steam Flow Primary Steam Flow P,, PsLG 3.2 P.. PSI G A inH'0 68 2.4 f,, I6./fi .0422 T F 262 Liquid Film Flow READING | T50F W4 \b"/hr 40 50 55.1 Test Section Conditions Y, -49.8 W, l,,/hr -185- Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: : Nominal Drop Size, d 3 2 x 10' 240 ibm/hr 209 "F 297 y Atomizing Steam Flow Primary Steam Flow P,, PSIG 1.6 P,. PIGI AP., mH,0 1.3 Pp, I6/ft' 26.5 Yi, .965 T *F 263 .0383 Liquid Film Flow READING Wf Ibw/he | 'TF 67.7 50 50 Test Section Conditions Wy lbnm/hr 34.0 W 171.9 / Re .165 I 1 . 99x104 W, b../hrI 29.5 -186Nominal Liquid Liquid Nominal 2x Reynolds Number: Input Flowrate: 240 Input Temperature: 208 - 98 Drop Size, d3 104 1bm/hr F V Atomizing Steam Flow Primary Steam Flow P,, PSIG .8 PF, PSIGow APmHzO fe I -7.6 1p'F .0367 258 Liquid Film Flow READING 50 IT 5 0 50 Wf \bm/hr 1 67.7 Test Section Conditions Y .988 VIe 160/hr 13.8 -187- 4 x 104 Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: Nominal Drop Size, d 360 ibm/hr 207 0F 203 y 3 Atomizing Steam Flow Primary Steam Flow Liquid Film Flow READING *F: Wf I /hr 71 50 94.3 Test Section Conditions -188Reynolds Number: 4 x 10' 360 ibm/hr Liquid Input Flowrate: Liquid Input Temperature 207 OF Nominal Nomival Drop Size, d 3 140 y Atomizing Steam Flow Primary Steam Flow Liquid Film Flow READIN, 62 T *F 50 Wf hm/hr 82.9 Test Section Conditions -189Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: Nomipal Drop Size, d 2 x 104 360 ibm/h r 210 *F 364 y 3 Atomizing Steam Flow Primary Steam Flow Liquid Film Flow READING T0*F 76 50 Wf 1b/hr 101 Test Section Conditions -190Nominal Reynolds Number: Liquid Input Flowrate: Liquid Input Temperature: Nominal Drop Size, d 2 x 10' 360 lbm/hr 210 *F 364 y 3 Atomizing Steam Flow Primary Steam Flow Liquid Film Flow Test Section Conditions IMM111 lit 11 -191- Nominal Reynolds Number: 2 X 104 360 ibm/hr Liquid Input Flowrate: Liquid Input Temperature: 210 OF : 164 Nominal Drop Size, d3 Atomizing Steam Flow Primary Steam Flow Liquid Film Flow |READING 7T,"F 69 50 Wf \bm/hr 92.2 Test Section Conditions -192- APPENDIX G WALL TEMPERATURE ERROR ESTIMATE The error resulting from the method of thermocouple attachment shown in Fig. 4.12 can be estimated from the handbook values of thermal conductivity for the Santocel, the asbestos and the wood. The thermal contact resistances should fall within the range 10<hc<1000 Btu/ft 2/hr/*F. The effective thermal conductivity of Santocel is .01 Btu/hr/ft/0 F when the intersticial gas is nitrogen and is at an average temperature of 350*R. This corresponds very closely to the conductivity of the nitrogen alone. Therefore, the effective conductivity used for the Santocel in these estimates was .03 Btu/hr/ft/*F. This corresponds to air at 800*F. From Fig. B.1, the heat flux through the insulation is Ul(Tmeasured - T ) ins (G.1) where U = t ()+ asbestos (z ) Santocel + ( (G.2) ) wood + 0 The temperature error is then estimated from ATerror qins/U2 where U2 t (G.3) )- mica + h contact MICA I/hc +/k ASBESTOS t/k SANTOCEL WOOP t/k t/k ThT I/ho To Figure G.1 Resistance Network for Temperature Error Estimates -194- Using these equations and the typical ranges of the values of the properties discussed results in estimates of wall temperature error between 0 < ATerror < 50'F . -195- APPENDIX H SINGLE PHASE DATA The single phase data are tabulated in the following tables. -196RUN SP 4 Mass Flux: 2.9xl03 Inlet Temperature: Outlet Temperature: Room Temperature: Power Input: I bm/ft 2 /hr 254 390 75 .42 *F 0F *F *F kw AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, *F 1.5 638 2 629 2.5 652 3 704 3.5 734 4 4.5 763 780 5 ~ 824 856 5.5 6 6.5 875 7 7.5 885 880 8 931 8.5 9.5 942 960 10.5 985 11.5 994 970 12.5 INA11111,1101 61 -197- RUN SP 5 Mass Flux: 7.8xlO Inlet Temperature: 275 Outlet Temperature: 45 80 Room Temperature: 1.2 Power Input: 1bm/ft 2 /hr 0F 0F kw AXIAL WALL TEMPERATURE DISTRIBUTION -198RUN SP 6 Mass Flux: 1.1x10 Inlet Temperature: 4 280 Outlet Temperature: 489 Room Temperature: 8Q Power Input: 2.0 1bm/ft 2/hr *F 0 F *F kw AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, *F 1 -- 1.5 903 2 966 2.5 1021 1098 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 1135 1174 1196 -- 1252 1294 1318 1328 1315 1385 1397 9.5 1417 10.5 1448 11.5 1464 12.5 1450 -199RUN SP 7 1.4xlO1 Mass Flux: 273 Inlet Temperature: Outlet Temperature: 421 Room Temperature: '80 .6 Power Input: l bm/ft'/hr kF kw AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, 1.5 633 2 653 2.5 3 683 736 3.5 762 4 789 4.5 802 5 5.5 6 6.5 7 7.5 8 -- 843 873 889 896 888 -- 8.5 948 9.5 10.5 964 11.5 996 12.5 979 989 "F -200-. RUN SP9 1.OX104 Mass Flux: Inlet Temperature: Outlet Temperature: Room Temperature: Power Input: lbm/ft 2 /hr 0F 280 oF 398 .80 kF kw .92 AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 Wall Temperature, *F -- 1.5 612 2 646 2.5 676 3 3.5 722 744 4 766 4.5 777 5 5.5 6 -- 810 6.5 837 849 7 7.5 855 846 8 -- 8.5 901 9.5 10.5 913 935 11.5 943 12.5 935 -201RUN SP 10 1.5x104 Mass Flux: Inlet Temperature: 281 Outlet Temperature: 474 Room Temperature: .8Q 2.35 Power Input: lbm/ft 2 /hr OF *F kw AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, *F 1.5 869 2 932 985 1061 2.5 3 4 1093 1131 4.5 1149 3.5 5 5.5 6 6.5 7 7.5 8 -- 1202 1242 1263 1272 1253 -- 8.5 1337 9.5 10.5 1359 1391 11.5 1406 1394 12.5 -202RUN SP 11 6.4x103' Mass Flux: 250 Inlet Temperature: Outlet Temperature: - 439 ".8.0 Room Temperature: 1.17 Power Input: lbm/ft'/hr *F kF kw AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1.5 2 2.5 3 3.5 4 4.5 Wall Temperature, *F 753 799 846 922 958 996 1015 5 5.5 1066 6 1107 6.5 7 7.5 8 8.5 9.5 10.5 11.5 12.5 1126 1134 1118 1186 1196 1219 1245 1257 1237 -203RUN SP 12 8.2x1O Mass Flux: Inlet Temperature: Outlet Temperature: Room Temperature: Power Input: 1bm/ft 2 /hr ' 252 418 ,..80 1.2 0F *F *F kw AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1.5 2 2.5 3 3.5 Wall Temperature, 715 759 802 872 905 4 4.5 940 952 5 1002 5.5 6 6.5 7 1039 1056 1063 7.5 8 1048 1111 8.5 1121 9.5 1138 10.5 1165 11.5 1174 1160 12.5 "F -204- APPENDIX I TWO-PHASE DATA The two-phase data are tabulated in the following tables. -205RUN TP2 FLOW RATES: ltm/hr 14.8 Primary Steam: 19.4 Atomizing Steam: 118 Liquid Input: 37.6 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.5 2.4x10' kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 1.5 2 2.5 3 Wall Temperature, F 1037 896 937 3.5 961 996 1012 4 1036 4.5 1038 927 1081 1108 5 5.5 6 6.5 1128 7 7.5 8 8.5 1137 1135 1186 1198 9.5 10.5 1217 1245 11.5 1262 12.5 1249 -206RUN TP3 FLOW RATES: ibm/hr 13.4 Primary Steam: 20.8 Steam: Atomizing 240 Liquid Input: 42.7 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.6 2.5x10 kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1 1081 1.5 929 2 2.5 1024 1071 3 3.5 1119 1129 4 4.5 1148 1144 5 5.5 1018 1165 1195 6 6.5 7 7.5 8 8.5 9.5 1203 1204 1192 1247 1253 1266 10.5 1293 11.5 1307 12.5 1292 Nkmlilfili 11111111111 hikuiflu -207RUN TP4 FLOW RATES: lbm/hr 60.0 Primary Steam: 8.2 Steam: Atomizing 120 Liquid Input: 33.8 Liquid Film: HEAT INPUT Electrical Power Input: 1.8 1.9x10 Corrected Heat Flux: kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 Wall Temperature, F 1.5 727 578 2 638 2.5 3 672 713 3.5 730 746 4 4.5 749 6 678 770 792 6.5 798 7 7.5 8 8.5 800 791 827 830 9.5 837 10.5 855 11.5 858 12.5 848 5 5.5 -208RUN TP6 FLOW RATES: lbii/hr 49.2 Primary Steam: Atomizing Steam: 19.7 120 Liquid Input: 37.0 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: kw 2 Btu/hr/ft 1.8 2.0x10' AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 Wall Temperature, F 762 1.5 ~584 2 609 2.5 622 3 642 3.5 651 4 4.5 666 668 5 5.5 609 690 6 6.5 707 7 716 719 7.5 715 8 8.5 742 739 9.5 713 10.5 251 11.5 229 12.5 224 -209RUN TP8 FLOW RATES: lbm/hr 51.6 Primary Steam: Atomizing Steam: 18.7 240 Liquid Input: 52.9 Liquid Film: HEAT INPUT Electrical Power Input: 1.8 4 2.0x10 Corrected Heat Flux: kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 1.5 2 Wall Temperature, F 763 588 6 6.5 7 7.5 8 8.5 658 700 741 758 777 778 707 798 816 822 823 815 847 849 9.5 857 10.5 868 11.5 871 12.5 855 2.5 3 3.5 4 4.5 5 5.5 -210RUN TP9 FLOW RATES: lbm/hr 51.6 Primary Steam: Atomizing Steam: 18.7 240 Liquid Input: 52.9 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.7 2.8x10 kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1 1017 1.5 2 764 854 2.5 907 3 3.5 963 983 4 4.5 1003 5 5.5 6 6.5 903 7 7.5 8 8.5 1003 1026 1048 1055 1055 1044 1083 1086 9.5 1091 10.5 1107 11.5 1110 12.5 1091 501101111441 -211RUN TP10 FLOW RATES: lbn/hr Primary Steam: Atomizing Steam: Liquid Input: Liquid Film: 19.3 15.3 360 44.0 HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.5 2.4xl0' kw Btu/hr/ft 2 AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1 1007 1.5 2 848 2.5 3 979 1033 3.5 1052 4 4.5 1077 5 5.5 954 1111 6 1141 6.5 1155 1158 7 7.5 8 8.5 933 1078 1149 1201 1209 9.5 1222 10.5 1245 11.5 1254 12.5 1236 -212RUN TP1l FLOW RATES: ibm/hr Primary Steam: 29.7 Atomizing Steam: 5.03 Liquid Input: 360 Liquid Film: 44.6 HEAT INPUT Electrical Power Input: Corrected Heat Flux: 3.5 3.1x104 kw Btu/hr/ft2 AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1236 1071 1.5 2 1151 1191 2.5 3 3.5 1242 1262 4 4.5 1294 1306 5 5.5 1158 1352 6 1401 1423 6.5 7 7.5 8 8.5 9.5 1435 1426 1499 1514 _1537 10.5 1570 11.5 1584 12.5 1562 110111101,10.11,11 1, -213RUN TP12 FLOW RATES: ibm/hr 52.7 Primary Steam: Atomizing Steam: 19.7 360 Liquid Input: 91 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.5 2.7x10" kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1 951 1.5 2 705 2.5 3 787 837 895 3.5 917 4 4.5 941 945 5 849 5.5 975 6 6.5 995 7 7.5 8 8.5 1009 1011 1002 1041 1049 9.5 1057 10.5 1060 11.5 1056 12.5 1027 -214RUN TP13 FLOW RATES: Ibm/hr 59.07 Primary Steam: Atomizing Steam: 11.56 360 Liquid Input: 65.6 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.7 3.0x10' kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1 997 1.5 2 762 826 2.5 859 3 3.5 906 927 4 953 4.5 5 5.5 962 862 997 6 1031 6.5 7 1045 1052 7.5 1041 8 1095 8.5 1105 9.5 1118 10.5 1135 11.5 1133 12.5 1109 -215RUN TP14 FLOW RATES:lbm/hr 30.0 Primary Steam: Atomizing Steam: 4.3 240 Liquid Input: 58.0 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 3.4 2.8x10" kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 Wall Temperature, F 1.5 1264 1117 2 1205 2.5 3 1255 3.5 1320 1345 4 4.5 1380 5 5.5 1226 1450 1500 6 6.5 1394 1527 7 7.5 8 8.5 1538 1522 1608 9.5 1651 10.5 1685 11.5 1712 12.5 1683 1625 -216RUN TPl5 FLOW RATES: ibm/hr 60.6 Primary Steam: 8.6 Atomizing Steam: 240 Liquid Input: 58.0 Liquid Film: HEAT INPUT Electrical Power Input: 3.5 3.5xl0 Corrected Heat Flux: kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 1.5 2 2.5 3 3.5 4 4.5 Wall Temperature, 1192 909 999 1049 1115 1138 1166 1177 5 5.5 1041 1216 6 1256 6.5 1273 7 7.5 8 8.5 1279 1262 1328 9.5 1356 10.5 1374 11.5 1378 12.5 1349 1339 F -217RUN TP16 FLOW RATES: ibm/hr 59.07 Primary Steam: Atomizing Steam: 9.31 120 Liquid Input: 35.1 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 3.03 3.1x10' kw Btu/hr/ft 2 AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9.5 10.5 Wall Temperature, F 1094 844 923 968 1020 1035 1054 1059 946 1087 1115 1127 1130 1114 1167 1174 1186 11.5 1197 1199 12.5 1176 -218RUN TPI7 FLOW RATES: lbm/hr 24.3 Primary Steam: Atomizing Steam: 10.6 120 Liquid Input: 37.6 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.7 kw 2 2.6xl04 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1 1.5 2 1123 988 1052 2.5 3 3.5 4 4.5 1081 1129 1136 1157 1157 5 5.5 6 1013 1196 1229 6.5 1249 7 7.5 8 8.5 9.5 10.5 ,1254 1238 1307 1319 1338 11.5 1356 1367 12.5 1340 -219RUN TP18 FLOW RATES: 55.2 Primary Steam: Atomizing Steam: 13.6 240 Liquid Input: 42.7 Liquid Film: ibm/hr HEAT INPUT Electrical Power Input: Corrected Heat Flux: 3.3 3.4x10'- kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1 1157 1.5 2 882 2.5 3 1024 1084 3.5 4 1101 1125 4.5 1130 5 1001 5.5 1158 6 6.5 1190 7 7.5 8 8.5 1204 1183 1245 1251 9.5 1262 10.5 1273 11.5 1265 12.5 1236 975 1203 -220RUN TP19 FLOW RATES: ibm/hr Primary Steam: Atomizing Steam: Liquid Input: Liquid Film: 23.3 10.5 240 51.0 HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.9 2.7xl 0' kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1 1146 1.5 2 996 1074 2.5 3 1113 1168 3.5 4 1182 1207 4.5 1218 5 5.5 1084 6 19 1300 6.5 1319 7 1326 7.5 1311 8 1381 8.5 1394 9.5 1416 10.5 11.5 1440 1453 12.5 1428 -221RUN TP20 FLOW RATES: Ibm/hr 26.6 Primary Steam: Atomizing Steam: 7.9 120 Liquid Input: 43.3 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.4 2.5x10' kw Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches 1 Wall Temperature, F - 1.5 914 2 994 2.5 3 3.5 1039 1101 1116 4 4.5 1140 1145 5 5.5 -1185 6 6.5 1221 7 7.5 8 8.5 1243 1221 -1302 9.5 1321 10.5 1343 11.5 1348 12.5 1314 1238 RUN TP21 FLOW RATES: ibm/hr 14.81 Primary Steam: Atomizing Steam: 19.5 120 Liquid Input: 50.3 Liquid Film: HEAT INPUT Electrical Power Input: Corrected Heat Flux: 2.5 2.7x10 kw 2 Btu/hr/ft AXIAL WALL TEMPERATURE DISTRIBUTION Axial Position, inches Wall Temperature, F 1.5 910 2 932 940 976 989 1012 2.5 3 3.5 4 4.5 1017 5 5.5 6 1064 1094 6.5 1118 7 1127 7.5 1118 8 -- 8.5 1193 9.5 1218 10.5 1236 11.5 1252 12.5 1229