AN ASSESSMENT OF TWO-PHASE PRESSURE DROP CORRELATIONS FOR STEAM-WATER SYSTEMS by William Idsinga Lieutenant, United States Navy B.S., United States Naval Academy (1967) Submitted in Partial Fulfillment of th Requirements for the Degrees of Naval Architect and Master of Science in Mechanical Engineering at the Massachusetts Institute of Technology May 1975 Signature of Author Department of Ocean Engineeriny, May 1975 abs ‘ Certified by wee anette Thes#s Supervisor 7) Océap Pagineering’ Department Reader ——~"“pipyianjohl Epgiheering Department Reader Accepted by Chairman, Departmental Coiifiittee on Graduate StudentsAN ASSESSMENT OF TWO-PHASE PRESSURE DROP CORRELATIONS FOR STEAM-WATER SYSTEMS by William Idsinga ABSTRACT Eighteen two-phase friction pressure drop models and correlations are compared to 2220 experimental steam-water pressure drop measurements under adiabatic conditions and 1230 in diabatic flow conditions. The data represents sev- eral geometries and has the following property ranges: Pressure 250 - 1500 psia Mass Velocity =2x10° - 3.2x10°1bm/hr-£t? Quality subcooled to 1.0 Equivale.t Diameters .09 - 1.3 in. The four models and correlations that coincided most nearly to the entire data collection were the Baroczy correlation, the Thom correlation and the homogeneous model two-phase friction multipliers, ve Vv 208 and * 0 The correlations are also evaluated with the data being subdivided into sets which are based on properties. Thesis Supervisor: Neil E. Todreas Position: Professor of Nuclear Engineering Thesis Reader: Peter Griffith Position: Professor of Mechanical Engineering Thesis Reader: Clark Graham Position: Associate Professor of Marine Systems 2ACKNOWLEDGEMENTS: The author wishes to thank Mr. Robert B. Bowring and Professor Neil E. Todreas for their suggestions, advice and counseling during the period that this thesis was written. This study was sponsored by the Electrical Power Research Institute. Appreciation is also expressed to Roland B. Knapp for his assistance and encouragement. The author is very grateful to his wife Sherry, to whom this work is dedicated, for her continuous support and understanding.NOMENCLATURE A flow area Ag flow area of liquid phase ay flow area of vapor phase a Parameter in equation (4.13) B parameter in equation (4.66) and given in table 4.7 c parameter in equations (4.25) and (4.66) and given in table 4.1 &p specific heat D diameter Dy equivalent diameter of the vapor phase flow De equivalent diameter De equivalent diameter of the liquid phase flow F(x) parameter defined by equation (4.62) £ friction factor friction factor based on actual liquid flow co friction assuming entire flow to be liquid fy friction factor based on actual vapor flow fep friction factor appropriate to two-phase flow condition G mass velocity Ge, mass velocity of the liquid phase 6 mass velocity of the vapor phase g gravitational acceleration Ge gravitetional constant 4P 8 Re Re. Rego Re 9 ‘in specific enthalpy specific enthalpy of saturated liquid latent heat of vaporization inlet specific enthalpy specific enthalpy loss in test section numerical coefficient numerical coefficient relevant to liquid phase numerical coefficient relevant to vapor phase length parameter in equation (4.13) numerical exponent. value of variable v, number of data points numerical exponent pressure volumetric flow rate Volumetric flow rate of the vapor phase Volumetric flow rate of the liquid phase example function for uncertainty analysis radius Reynolds number Reynolds number based on actual liqvid flow Reynolds number assuming entire flow to be liquid Reynolds number based on actual vapor flow slip ratio inlet temperature flow velocityin ‘out. velocity of the liquid phase velocity of the vapor phase maximum local velocity mean velocity specific volume specific volume of saturated liquid Yg - Ye specific volume of saturated vapor mean specific volume example variable for uncertainty analysis mass flow rate liquid phase mass flow rate vapor phase mass flow rate uncertainty interval for variable v. Lockhart-Marlinelli parameter mass quality inlet mass quality exit mass quality distance from duct boundary distance along flow path pressure gradient due to acceleration pressure gradient due to friction friction pressure gradient assuming actual liquid flowap ap. a ap. P OP esc AP friction pressure gradient assuming entire flow to be liquia friction pressure gradient assuming actual vapor flow friction pressure gradient in the liquid phase friction pressure gradient in the vapor phase pressure gradient due to static head void fraction void fraction at a point in a flow maximum local void fraction volumetric quality Chisholm property index ratio of liquid flow ares to area calculated using the liquid flow equivalent diameter pressure drop acceleration pressure drop friction pressure drop friction pressure drop in subcooled region static head pressure drop film thickness ratio of vapor flow area to area calculated using the vapor flow equivalent diameter uncertainty interval for variable __ uncertainty interval for quality due to uncertainty in variableuncertainty interval for two-phase friction multi- plier divided by multiplier, due to uncertainty in variable __ € discrepancy between data and correlation defined by equation (5.1) root-mean-square value of for N data points RMS ° angle of flow inclination a parameter used in figure 2.4 and given by figure 2.5 u viscosity saturated liquid viscosity saturated vapor viscosity mean viscosity ° density saturated liquid density saturated vapor density mean density o surface tension or standard deviation of variable i T shear stress Te shear stress based on actual liquid flow Ty, wall shear stress "hp shear stress under two-phase flow conditions o heat flux from boiler or preheater 2 two-phase friction multiplier based on actual £ liquid flow 2 two-phase friction multiplier assuming the entire fo flow to be liquid 2 two-phase friction multiplier based on the actual g vapor flow<1 9 parameter used in figure 2.4 and given in figure 2.5 correlation adjustment factor average two-phase multiplier for diabatic conditionsTABLE OF CONTENTS ABSTRACT 20.0... eee eee SEES ee cere Sees ACKNOWLEDGEMENTS peer NOMENCLATURE ......-...65 See eee eee Seer LIST OF TABLES 22... .eeeeseeeeeeeeee - LIST OF FIGURES 00... sees eeeeeeeees Bee Eee eee Chapter 1, INTRODUCTION .........4. aorr PRELIMINARY CONCEPTS .. 2.1 Void Fraction and Quality ........... 2.2 Flow Regimes ..... 2.3. Pressure Drop ........ See eee EEE Eee 2.4 Two-Phase Friction Multiplier . 2.5 Priction Factors ............. poanoso 3. THE BASIC TWO-PHASE FLOW ...... ou08 3.1 The Homogeneous Model ..... eee EEE Eee 3.2 The Separated Flow Model ...........+ 4. TWO-PHASE FLOW CORRELATIONS .......++-+e0000005 4.1 Introduction ...... See eee eter 4.2 The Armand Correlation ............48 4.3 The Lockhart-Martinelli Correlation 4.4 The Martinelli-Nelson Correlation . 4.5 The Armand-Treschev Correlation ....... 4.6 The Levy Momentum Exchange Model .... 10 16 1g 18 20 21 22 24 43 45 46ul Chapter Page 4.7 The Martinelli-Nelson-Jones Correlation ...... ee eee eeeeeeeeeee eens AB 4.8 The Bankoff Variable Density Method .... 49 4.9 The Sze-Foo Chien and Ibele Correlation .......... Be eee 4.10 The Thom Correlation ..............0.005 52 4.11 The Baroczy Correletion ............2005 53 4.12 The Becker Correlation ....... 55 4.13 The Borishansky Correlation ............ 55 4.14 The Chisholm Correlation .. 56 4.15 The CISE Correlation ............eee000e 57 4.16 A Summary of the Pressure Drop Correlations and Models .............. 58 5. THE METHOD OF EVALUATION OF PRESSURE DROP CALCULATIONS .......eseeeeeeeeeeeees sees 77 S.L General ...sscseesccccccascccceseecceses 77 5.2 Pressure Drop Data ...... See eee eee eee 78 5.3 The Reduction of Pressure Drop Data .... 79 5.4 The Evaluation of the Correlation . + BL 6. RESULTS OF THE EVALUATION ...............00- see 88 6.1 Adiabatic Data see BB 6.2 Results of the Comparison of Diabatic LECAlicppapoepodandeccdoousceaeenaancae | (it 6.3 Applicability of Results to Boiling Water Reactors ....eeeeeeeeeeeeeeeeees 92 7, CONCLUSIONS .........e000005 see 104 APPENDIXES 6... eee ceceececccccecs eer eenees 110 A. TWO-PHASE FRICTION MULTIPLIER UNCERTAINTY ..... 11012 Chapter Page A.l General ... Pee Eee ete reel L O) A.2 The Uncertainty in Recorded Data ....... 112 A.3 Uncertainty in the Adiabatic Two-Phase Friction Multiplier .................. 113 A.4 Uncertainty in a Diabatic Two-Phase Friction Multiplier ..... 11s DATA REDUCING PROGRAMS ..... Ree eee CEE eee . 126 B.1 The Programs ........seeeseeeeeee eee eee 126 B.2 A Sample Program .. Ree eer zo) C. THE CORRELATION EVALUATION PROGRAM ............ 136 C.1 The Program for Adiabatic Data ......... 136 C.2 A Sample Program ..............204 eee D. CORRELATION EVALUATION FOR ADIABATIC DATANGETS Bete tee EEE Eee Eee ttre enter Eyal s9) E. CORRELATION EVALUATIONS FOR ADIABATIC DA'TA BASED ON FLOW CONDITIONS ....... eer oS) F. CORRELATION EVALUATION FOR DIABATIC DATA BETS Ieee eee eee Eee eee EEE ene eee 2 5) G. VOID FRACTION REFERENCES ..... see 248Table LIST OF TABLES Page Lockhart-Martinelli Correlation Constants .... 60 Martinelli-Nelson Local Multipliers used in HAMBO ......cecececeeeeesecseeeeseeereees 61 Slip Ratio Values used by Thom ..........++++- 62 Thom Void Fraction Correlation . . . 63 values of ¢, for the Separated Flow Model as Given by THOM ... eee eee cece cece e eee ee cence 64 Baroczy Correlation Co-ordinates of Two-Phase Frictional Multiplier & and G=1x10°Lbm/hr-ft? 6... ee eee cece cece ee eens -. 65 Values of B for Equation (4.66) .........05 66 Constants for Equation (4.67) .......eeeeeeeee 67 A Summary of Two-Phase Correlations sees 68 Data used in this Study .......... bee eeeeeeees 83 Uncertainty Intervals for Measured Variables . . er The Ranges of Physical Properties used to Form Data Subsets for Evaluation by Properties .. 87 Two-Phase Friction Pressure Drop Correlation Identification .......eceeeeeeee - 94 Overall Results for Adiabatic Data Reduced using the Thom Void Fraction Correlation and the Single-Phase Friction Factor £=.046/Re"? 95 Overa!1 Results for Adiabatic Data Reduced using the Thom Void Fraction Correlation and the Single-Phase Friction Factor £2.079/REP> eee cece cece ee eeeeneenens 96 13Table 6.4 6.6 Overall Results for Adiabatic Data Reduced with the Thom Void Friction Correlation and the Smooth Tube Single~Phase Friction FactOr weeeeeeeeeeeeeeee Overall Results for Adiabatic Data Reduced with Martinelli-Nelson Void Fraction Correlation and Smooth Tube Single~Phase Friction Factor ....esee cece e eee cence eee Overall Results for Adiabatic Data Reduced with the Homogeneous Void Fraction Model and Smooth Tube Single~Phase Friction actor sere ern ee rere Two-Phase Pressure Drop Correlation and Models Having the Least Discrepancy with the Entire Data Collection ....seseeeeeeeeeeeees The Adiabatic Data Subsets Based on Physical Properties ........ Overall Results for Diabatic Data Reduced with the Thom Void Fraction Correlation and the Smooth Tube Single-Phase Friction Factor 2.1... eee cece eee e eee eee eee 14 Page 97 98 99 100 103LIST OF FIGURES Page Flow Patterns in Vertical Flow .......eeseeeee 26 Flow Patterns in Horizontal Flow .......ss+e++ 27 Vertical Flow Regime Map .....s+eeseeeeseeeees 28 Horizontal Flow Regime Map ......seeeeeeeeeees 29 Values of \ and ¥ for Steam-Water Systems for use with Figure 1.4 .....ccseeeeseeeeeee 29 Lockhart-Martinelli Correlation ............++ 70 Martinelli-Nelson, Two-Phase Friction Multiplier for Steam-Water as a Function of Quality and Pressure ......sseeeeeeeeeeee 71 The Martinelli-Nelson Friction Pressure Drop Correlation .........seceeeeeeeeeeeeees 72 Thom Void Fraction Correlation ........ . 73 Baroczy's Two-Phase Friction Pressure Drop Correlation ......eceeseeeeeeeeeeeeees 74 Mass Flux Correlation Versus Property odes estes eae aee sentient Ein S Borishansky Correlation ........seesseeeeeeees 76 Multiplier Uncertainty for Adiabatic Data .... 122 Major Components of Multiplier Uncertainty for a Typical Adiabatic Condition . 5 123 Typical Multiplier Uncertainty for Diabatic Data .......ceeeeeeeeeeeecereeeeeee 124 Components of a Multiplier Uncertainty for a Typical Diabatic Case ......eeeeeeeeeeee es 125 15Chapter 1 INTRODUCTION In the operation of fluid energy conversion systems, such as boilers and nuclear reactors, two-phase flow phenom- ena occur by design or can happen in an accident situation. Presently, nearly all such systems use water or water and steam as the working fluid. Consequently, the ability to accurately predict the pressure drop in a steam-water flow is important in the design of such systems. For nuclear systems, knowledge of the portion o1 the flow not occupied by the liquid is very critical to the proper design of the reactor core. A completely acceptable analytical model of the two-phase pressure drop has never been developed causing reliance to be placed on empirical methods as the means to predict the pressure drop. Several semi-analytical models and empirical correlations for two-phase pressure drop have been developed since World war II, most being stimulated by the growth of nuclear power systems. Eighteen prediction methods are reviewed in this study. These include the most common and reputed correlations and models. Some 3450 steam-water pressure drop data points were collected for comparison with the predictions. Even though this study does not cover void fraction 1617 models and correlations, their application in the reduction of pressure drop data and in some pressure drop correlations justifies covering them. Consequently, the review of two- phase correlations includes several void fraction predictions. This study also notes the effects of using different void fraction correlations to reduce the pressure drop data. The ultimate objective of this work is to provide recommendations regarding the suitability of the various methods of predicting the two phase pressure drop.Chapter 2 PRELIMINARY CONCEPTS 1 Void Fraction and Quality The local void fraction is the time averaged volu- metric fraction of the vapor phase at a point in a two-phase flow. The void fraction of the entire flow at a given cross section is the area average of the local void fractions for that section or xf z | 4, B J ,°LocaL aa. (2.1) Put in other words, it is the ratio of the time averaged area occupied by the vapor phase to the total area of the cross section, . (2.2) The requirements for mass continuity must hold for each phase of the flow. The mass flux for each phase is written = Pp Ag Up (2.3a) and W, ALU. 23b ig Pg 4g Mg (2-3) The volumetric flow rates for each phase are defined as Qe = Ug Ag (2.4a) 1819 and zu Ae 2.4b Qy ig Ag ( ) Dividing equation (2.3) by the total cross section area yields the mass velocities for each phase G =u ig Pg (1-3) (2.5a) and G = ig = Yg Pg % (2.5b) The flowing mass and volumetric qualities are defined as W ee) % 4 7 (2.6) and ers a, ‘ (2.7) a5 +o; respectively. The flowing mass quality is not necessarily equal to the thermal equilibirum mass quality as determined by an energy balance. They are equal in the case of thermal equilibrium between the two phases, and under non-equilibrium conditions they are very nearly so, except in cases of ex- treme thermal gradients such as occur in subcooled boiling and film boiling. Equation (2.3) can be rewritten as We = W (1-x) (2.8a) and Wy = Wx, (2.8b) which divided by the total flow area gives20 (2.9a) and Gy = Gx. (2.$b) The parameters defined by equations (2.2), (2.6) and (2.7) can be related to each other through appropriate substition of equations (2.3), (2.4), (2.5), (2.8) and (2.9) by wo 12 4 3 5 (2.10) (2.11) 1-a) and (2.12) Pa" The ratio of the gas phase velocity to that of the liquid phase defines the slip ratio, u, J, s: if. (2.13) A knowledge of the slip ratio and the flowing mass quality is required to determine the void fraction by equation (2.10). 2.2 Flow Regimes A two-phase flow appears in several different patterns depending on the relative amounts of liquid and vapor present, the velocities of the phases, pressure, flow orientation, and rate of heat addition. Figures 2.1 and 2.2 depict the appear- ance of several of the flow patterns. The reader is referred to basic two-phase flow texts [1, 2, 3] for details about the21 various flow regimes. Figures 2.3 and 2.4 are generally accepted flow regime maps for vertical and horizontal flows. Figure 2.5 give the values of the variables ¥ and \ that are to be used in figure 2.4 for steam-water systems. 2.3 Pressure Drop By manipulating the conservation of momentum or energy relations [1] for a steady state two-phase flow it can be shown that the pressure gradient for such a flow is the sum of the pressure gradients due to friction, acceleration resulting from a change in volume of the flow and gravity, a dz al +2 4. (2.14) ap fe Ht az lar az The friction component can be computed using the familiar Fanning equation ae d= - 2 fep SY SG) iz gD where f,,, is a friction factor which is relevant to the two- phase flow condition and Vv is the spatial mean specific volume. The acceleration term is e-Gw (2.16) and the pressure gradient caused by a change in elevation is given by (2.17) Further modeling of the flow is required to be able to evaluate the mean specific volume and friction factor.22 The two basic models are known as the homogeneous model and the separated flow model. They will be covered in greater detail in a subsequent chapter. Equation (2.15) is of the same form as the friction pressure gradient for a single phase flow. Consequently, the unknown terms f,, and V are some muitiple of the com- parable single phase terms and, thus, it has become conven- ient to express the friction gradient as that for a single phase flow multiplied by an appropriate value. This is true for both models of two-phase flow. The primary difference in the two models is the evaluation of the acceleration and gravity pressure gradients. 2.4 Two-Phase Friction Multiplier As noted in the previous section, it has become con- venient to express the two-phase friction pressure gradient as a single phase friction gradient multiplied by an appro- priate function of the flow parameters. This function has become known as the two-phase friction multiplier, three common forms of which are defined as (2.18) (2.19) and23 ey (2.20) where 2 a | 2.2 fto® ¥ (2.21) az "| £0 Ge P 2 & | =- 2s Se Ye (2.22) zools Ie and 2£ Gy = - ga "a sad, (2.23) the gradient in equation (2.2) presumes that the entire flow is liquid. Equation (2.22) is based on the actual liquid flow and equation (2.23) is based on the actual gas flow. If it is assumed that the friction factor is of the Blasius solution type (2.24) the two phase multipliers can be related to each other by appropriate substitutions, so that x : (2.25) 2 42 1.75 _ 42 [% Fo 7 oe (I-x) Oo (73) (xs24 Friction Factors Evaluation of equations (2.21), (2-22) and (2.23) requires the selection of an appropriate friction factor. As will be noted in subsequent chapters, experimenters have used different forms of the friction factor. The familiar Blausius solution friction factors (2.26) and (2.27) are good approximations for the smooth pipe friction factor which can be expressed as 2 = 4.0 log, [RevE] - (2.28) Equation (2.27) is the more valid approximation for Reynolds numbers up to 50,000 and equation (2.26) applies for greater Reynolds numbers. ‘Two-phase steam-water data reviewed in this study ranges in liquid-only Reynolds numbers from 20,000 to 600,000. At this upper limit the friction factor computed by equation (2.27) is 15 percent less than that obtained by equation (2.26). Some investigators have used friction factors that reflect the hydraulic roughness of their test apparatus. These friction factors are based on tests conducted on the equipment under single phase conditions. In order to make estimates of the pressure drop during design calculations for boiling systems, a selection25 of an acceptable friction factor must be made. Wallis [2] contends that a single phase friction factor of .005 is “adequate" to compute the friction pressure gradient by equations (2.21) or (2.22) for common two-phase systems. Experimenters, for instance Macbeth [22], have concluded that surface finish and deposits have negligible effects on the pressure drop of boiling systems. Over the range of data reviewed in this study, the liquid only smooth pipe friction factors range from .003 to .007. Collier uses smooth pipe approximations in the examples in his text [1]. There is certainly a degree of arbitrariness in the selection of a friction factor. In performing pressure drop calcula~ tions a friction factor equal to or greater than that of the smooth pipe case would normally be selected to make the computation. If the two-phase friction multiplier used is based on data which has been reduced using the smooth pipe condition, then the calculation should give results equiva~ lent to or more conservative than the data on which the correlation is based. Consequently, it is considered appropriate to use the smooth tube friction factor or the appropriate approximation in reducing two-phase pressure drop data and making predictions.Bubbly Slug Churn = Wispy — Annular annular Figure 2.1 Flow Patterns in Vertical Flow ClJ 26Annular Figure 2.2 Flow Patterns in Horizontal Flow LI7 27108 - TOT 7 777 105 4 ’ 104}— _ / 3 Wispy a wa! annular {Annular 4 ! 1 it ut it a ! 10 10% Figure 2.3 Vertical oe 10% 108108 a (SY Flow Regime Map CII 2829 TT ™ T -$+ | (ex 275 (4.25) * and for the void fraction is x * 2 1 5 > (4.26) p +204 \? x42 The values of C used in these equations for the four flow types are given in table 4.1 also. The data which supports this correlation is of low pressure adiabatic flows of air and various liquids. The pressure range is from 16 to 50 psia. A wide range of flow rates and dimensions were covered. The authors make no claim of the correlation not being valid in any particular flow conditions, however, they do recognize that data at higher pressures (up to the critical pressures) are needed to establish the validity of this correlation.43 4.4 The Martinell jelson Correla' The stated intention of this correlation is to facil- tate the prediction of the pressure drop and void fraction during the forced circulation boiling of water. Only one of the flow conditions, as defined by Lockhart and Martinelli [5], is considered, since practically all forced circulation boiling systems operate with a turbulent gas phase and turbulent liquid phase flow condition. This correlation is tailored to that flow condition. The Lockhart-Martinelli correlation was developed using data observed in flows at pressures near one atmos- phere. This pressure range is assumed to be a longer limit of engineering concern for steam-water systems. The Martinelli-Nelson correlation utilizes the Lockhart-Martinelli results for 14.7 psia. At the critical pressure (4.27) ap | (2 F (az fo and, thus, the two-phase multiplier as defined by equation (2.18) is unity. Having determined values for the limiting pressures, the curves for the intermediate pressures were interpolated with the aid of data taken by Davidson et al [34]. This data was taken in horizontal coils under diabatic conditions at pressures up to 3200 psia. The mass velocities tended 6 to be less than .5 X 10° 1bm/hr-ft” for most of this data.44 Martinelli-Nelson give their results as a plot of the two-friction multiplier, oer as a function of quality and pressure. It is given in figure 4.2. Bowring [10] has tabulated values of the multiplier which are given in table 4.2. The authors also presented their correlation as an average multiplier to be used to determine the pressure drop along a boiling length assuming a linear change in quality. These results will be ignored here since the adiabatic multiplier can be integrated for any desired heat flux distribution by point to point numerical solution on a computer. ‘These average multipliers can be located in reference [6] and basic two-phase flow texts [1, 2, 3]. Martinelli and Nelson also proposed a void fraction correlation. At 14.7 psia it is based on the Lockhart- Martinelli correlation. At the critical pressure the two phases have equal properties and therefore, the void fraction is equal to the quality. Intermediate values were rather arbitrarily interpolated between these extremes. ‘The void fraction correlation is given in figure 4.3. The authors indicated that their results are tenta- tive and based on only a meager amount of data and that further experimental verification is required before this work is assumed valid. This pressure drop correlation is used in the reactor code HAMBO [10] as an option to the homogeneous model or a polynomial input.45 4.5 reschev Correla’ n_[9] This correlation recognizes that the earlier work by Armand [4] is based on low pressure flows. From equations (2.2) and (2.4) it is seen that the ratio of the volumetric quality to the void fraction is the ratio of the mean vapor phase velocity to that of the total flow, Ya =; (4.28) aie This ratio can be shown to depend on the density of the two phases by equation (2.10) and (2.11). Since the densities are a function of pressure, a= £(P) > B. (4.29) By empirical methods the authors concluded that a = [.833 + .05 10g qq@ay 1 Be (4.30) The form of the Armand friction multiplier as given by equation (4.7) is retained for flows with lower void fractions. The empirically determined two-phase friction multipliers are 1.75 2, = ee for a < (4.31) 1.2 (1-a) and 1.75 2 _ .48(1-x) 2. = (4.32) fo (a-a)” - -3 |_P where n = 1.9 + 1.48 X 10 (x23) for a> .5 and 8 <.9. For high void fractions (more46 specifically B > .9) Tea] + +955 (a-x) 2°75 5 5 : (4.33) (1-8)"* +0025 2 2 Eq These relations are based on horizontal, adiabatic steam-water data at pressures ranging from 150 psia - 2700 psia and are considered by the authors to be valid in that range. This correlation supplements Armand's earlier work [4] which is given by equations (4.9), (4.10) and (4.11) and is considered valid at pressures up to 10 atmospheres. 4.6 ‘The Levy Momentum Exchange Model [12] Levy derived a theoretical model for the void fraction and friction multiplier. The Bernoulli equation for each of the two phases can be written betgdu, ‘ a ~ pg (sine) dz (4.34) lc and ap SP gt] az - A(Age gus 2) - eae ) g s g a Pg £°ete ae oR (4.35) pg (sind)az. It is assumed that the total pressure drops of the two phases over a given incremental length are equal. Subtracting equation (4.34) from equation (4.35) gives af aw? 3-3 9| ta) | (4.36) ap a se + (bg - pg)aine| as az ae a)47 after substituting equations (2.2), (2.6) and (2.7). If there is no heat addition or flashing the combined friction and elevation pressure drop for each phase must be equal, or ap . . ~ (a ool . (¢ fr] dz - pesinddz = (SP oF) - pysineaz (4.37) | and, consequently 2 = 0. (4.38) Equation (4.38) can be rewritten as a (1-20) +a/(1-2x) +0 [2—2(1-a) 240 (1-20) ] B cE : (4.39) Ee 2n£(1-a) +a(1-20) °g The void fraction can be determined from equation (4.39) by an iterative solution. For diabatic flows the right side of equation (4.36) is not equal to zero. In this case, no determinable solution is readily obtainable. The two-phase friction multiplier proposed by Levy is based on an annular flow model. ‘The two-phase multiplier given by equation (2.19) can be written as a ratio of shear stresses, (4.40) or48 o2 = fe. (4.41) £ (1-0) The two-phase friction factor is assumed to be related to an equivalent diameter of the film, = 46. (4.42) The film thickness is related to the tube diameter by 6 = (1-02. (4.43) The Reynolds number of the film can be written as (469¢U¢/i¢) which is equal to the Reynolds number where the liquid only is assumed to be flowing (G,D/u,). Consequently, friction factors written with these two Reynolds numbers are equal. Applying this result to equation (4.41) and substituting equation (2.25) gives 1.75 ) (1-x)? tf = : (4.44) which is in consonance with the Armand result given in equation (4.7) for the same flow regime model. Levy compared his pressure drop solution with the Lockhart-Martinelli and Martinelli-Nelson correlations. These agree rather closely when plotted as a function of the void fraction. This model was compared with pressure drop data of several experimenters and deviations of up to fifty percent were noted. The author attributes this to neglecting flow rate effects. 4.7 The Martinelli-Nelson-Jones Correlation [11] Jones devised an adjustment to the Martinelli-Nelson49 correlation. The two-phase friction multiplier for this correlation is the Martinelli-Nelson results for the flow of concern multiplied by a factor Q. The adjustment parameter Q is empirical and is a function of mass velocity and pressure, and is given by Q = 1.36 + .0005P + .1 { - -000724P( (4.45) for G < 700,000 lbm/hr-£t” and @ = 1.26 - .0004P + .119 aol -ooo2er (22° (4.46) for G > 700,000 1bm/hr-£e” Jones did not indicate the nature of the data used, nor the range of validity of this factor. 4.8 The Bankoff Variable Density Model [13] This model is based on the assumption that void fraction and velocity distributions within a flow can be described as (4.47) and < “4 (4.48) max where m and n are unknown constants. Using continuity con- sideration to determine the mass flow rate, equation (2.2), and determining the average void fraction for the distribu~ tion described by equation (4.48) leads to50 (4.49) where 2(m +n + mn) (m+n + 2mn) 1) (2n + 1) (m+ 1) (2m + (n By substituting equation (2.11) into equation (4.49) it is seen that a= KB. (4.50) This coefficient is identical to the coefficients used by Armand [4] and Armand-Treschev [9] as shown in equations (4.12) and (4.29). This result compares favorably to the Martinelli-Nelson void fraction if K is equal to .89. Baukoff found that K depended on pressure as did the afore- mentioned Russian investigators. By empirical means Bankoff determined that K = .71 + .0001 P. (4.51) Bankoff's pressure drop model is based on the ratio of the two-phase wall shear to that of the liquid phase. 1/4 p= \l/a 2 | (| : (4.52) £ ue} " OF The ratio of the densities is (4.53) and the ratio of the velocities is given as i e. get x =f). (4.54) £ PgiTable 4.7 values of B for Equation (4.66) 66 r G (Lbm/ne-£t?) B < 5 59.5 <3. 69X10 4.8 3.69X10°1.4x10® 1494/ og 7 9.5<1<28 <4. 426K10° 14123/pg-5 5 <4.426X10 2p 228 - 4.075x10°/ 9 5 r Gt51 The viscosity term is very near unity, consequently equation (4.52) becomes 3/4 qoeli-a sp h - 22 (4.55) Ce which is also the two phase friction multiplier as defined by equation (2.18) and can also be written as = vals] 7 rex (1-2! anna (4.56) fo Pe 2. . Bankoff's void fraction model, equation (4.51) with K = .89, agrees well with the Martinelli-Nelson void fraction correlation [6] in the pressure range of 100 to 2500 psia and for void fractions less than .85. The pressure drop model tends to fit the general pattern of the Martinelli- Nelson pressure drop correlation and is considered to agree according to the author even though there may be differences of up to seventy percent between the two. 4.9 The Sze-Foo Chien and Ibele Correlation [14] This correlation is based on the Lockhart-Martinelli [5] result (equation (4.23)) that the friction multipliers are dependent on the actual liquid and gas flow Reynolds numbers, Re, and Re,, respectively. This work gives a friction multiplier which is based on the actual gas flow, ¢3, which can be related to the two-phase friction malti- plier, 625, which is based on the entire flow being liquid by equation (2.25). This correlation is based on annular52 and annular-mist flow data for air-water vertical down-flow at pressures near one atmosphere. The friction multipliers are 22 7 “71 +725 og = 3-885 x 107 (Re,) (Reg) (4.57) for annular flows ani 2 -.34 +725 95 = 3-45 (Re,) (Ree) (4.58) for annular-mist flows. The transition from annular to annular-mist flow occurs when (Reg) (Re,)"79) = 1.199 x 10°, (4.59) No transition into the annular flow regime was determined. Figures 2.2 through 2.4 can be referred to for transition into annular flow. Equations (4.57), (4.58) and (4.59) were arrived at by empirical methods. The results are shown by the authors to compare favorably with the Lockhart-Martinelli correlation which is based on data taken at similar pressures. 4.10 The Thom Correlation [15] This correlation is based on steam-water data at pressures from 15-3000 psia. The mass velocities ranged form .3 x 10° to 1.4 x 10° 1bm/hr-£t?. The data was taken under both adiabatic and diabatic conditions and vertical and horizontal orientations. The data was taken in earlier boiler circulation studies [33] and provided an adequate data base to correlate friction and void fraction information53 over the same ranges that Martinelli-Nelson rather arbitrarily interpolated. The results of this work are presented in a similar fashion to the Martinelli-Nelson work. Thom found that the void fraction data correlated rather satisfactorily if a constant slip ratio § for a given pressure was assumed. Then equation (2.10) can be used to relate void fraction and quality. Table 4.3 gives the in- verse of the slip ratio multiplied by the density ratio as a function of pressure. ‘The resulting void fractions are tabulated in table 4.4 and plotted in figure 4.4, as a function of pressure and quality. The two-phase friction multipliers were based on data and presented in table 4.5, as functions of quality and pressure. 1_The arovzy Correlation [16] Raroczy noted that the generally accepted pressure drop correlations, namely the Lockhart-Martinelli [5] and the Martinelli-Nelson [6] correlations, do not account for mass velocity effects which are revealed by observed data. It is also noted that these correlations are also limited to either low pressure flows or steam-water flows. The Baroczy correlation was developed to take the mass velocity effects into account and also be applicable to other two-phase systems, as well as steamand water. This correlation makes the two-phase multiplier a function of a property index which is dependent on pressure alone, with quality as a parameter. This is in keeping with54 the results of many earlier correlations. To account for the mass velocity effects Baroczy introduced a correction factor for varying mass velocities which is a function of the property index and quality. ‘The mass velocity for the basic correlation is 1 x 10° 1bm/hr-ft?. ‘The property index used by Baroczy was the ratio of the liquid only pressure @rop to that of a gas only flow and as such is similar to the Lockhart-Martinelli parameter, X. The Baroczy prop- erty index is led, Gel BE This results in the upper limit of the friction multiplier being the reciprocal of the property index. The property index equals unity at the critical pressure. The friction multipliers are given in figure 4.5 and table 4.6 for the basic mass velocity. The correction factors are given in figure 4.6 for various mass velocities. The Baroczy correlation is based on the data of several combinations of liquids and gases. The steam-water data used ranges in pressure from 590 to 2000 psia and mass velocities from .7 x 10° to 5 x 10° 1bm/hr-£t?. This data amounts to about 130 points plus the Sher and Green [24] correlation for 2000 psia. The correlation was compared to the Martinelli-Nelson correlation and found to compare55 most favorably at low mass velocities. It was also checked against other steam-water data ranging in pressure from 139 to 1400 psia and found to compare favorably. 4.12 The Becker Correlation [17] Becker and associates conducted two-phase pressure drop experiments in vertical round ducts under diabatic conditions. The pressure profile was plotted and the two- phase multipliers were derived from the gradients of the plot. The results of this work is a correlation of the two-phase multiplier, +96 28 x fo = 1 + 32000(3] (4.60) The supporting data covers a range of pressures from 90 to 600 psia and mass velocites ranging 2 X 10° to 4x 10° 1bm/hr-ft?. ‘The authors compare their results with the Martinelli-Nelson and Lockhart-Martinelli correlations. This correlation gave results that were as much as 40 percent greater than Martinelli-Nelson's for comparable conditions. It compared favorably with the Lockhart-Martinelli correla- tion at pressures around 150 psia. 3__The Borishansky Correlation [18 Borishansky and associates concluded that the traditional method of correlating two-phase pressure drop data as a multiplier, being a function of both pressure and quality, may not be the best way to present such data. They chos2 to correlate pressure drops using56 ap ae (ge l- (#4 a (= - ae = F(x). (4.61) a "lL, baz le, F(x) = 0 and when x = 1, F(x) = 1. The experi- mental results of the authors indicates the data is suffi- ciently concentrated so as to be considered independent of heat flux, pressure, geometry and flow rates. F(x) is plotted in figure 4.7. To allow comparison of this correlation with others, substitution of equations (2.18) and (2.21) and a similar definition of the pressure gradient considering the entire flow to be vapor into equation (4.61) relates F(x) to the two-phase multiplier, Cruye eo me as 25 Pf T —|-1)4i. (4.62) £ °g The friction factor is assumed to be given by equation (2.24). F(x) is based on air-water and steam-water data at pressures up to 530 psia. 4.14 The Chisholm Correlation [19] Chisholm intended this result to be an easier to apply substitute for the Baroczy correlation. He employs a Property index similar to the Lockhart-Martinelli parameter X and the Baroczy property index. It is given as 5 -Sn Og yu ret a for smooth tubes (4.63) ®g Up57 and 5 G3 r= |= for rough tubes (4.64) Pg where n is governed by the appropriate friction factor relation. Chisholm reports that these relations when con~ bined with the approximation of the Lockhart-Martinelli correlation equation (4.25) results in oo =2+(r? - a] [sfx] (Qn) 2 (4.65) where and C is the same constant as in equation (4.25). Chisholm recommends using empirical values of B which are given in table 4.7. This coefficeint was determined by comparing equation (4.66) to the Baroczy and Lockhart-Martinelli correlations. As a result the significance of C which was used in equation (4.25) to approximate the Lockhart-Martinelli correlation is lost. The values of B determined by Chisholm were intended to give resulting calculations a degree of conservatism. 4.15 The CISE Correlation [20] During the early 1960's this laboratory conducted and extensive research program in two phase pressure drop. The correlation developed was based on using the homogeneous model to reduce data. The empirical correlation devised is58 Ee xo 786 ,0-4 ge, = kG ae (4.66) az pie K and n are given in table 4.8 and are geometry dependent. Equation (4.66) is written for SI or CGS unit systems. This correlation was obtained from and verified against two-phase vertical pressure drop data in both adia- batic and diabatic conditions. Data for a variety of fluids including steam-water was incorporated into this result. A large range of mass velocities, quality and geometry and pressures up to 1500 psia were covered by the steam-water- data. 4.16 A Summary of the Pressure Drop Correlations and Models It is noted that different correlations consider different variables. Several depend only on the pressure and quality. Some include the mass velocity and/or the equivaleat diameter, in addition. The effects of the mass velocity can be very substantial. A quick glance at figure 4.6, the mass velocity correction factors for the Baroczy correlation, shows that there is a large difference between the friction multipliers for high and low velocity flows. At the same pressure and quality the friction multipliers can vary by as much as a factor of three be- tween low and high velocity flows. Consequently, there can be considerable discrepancy between correlations based on narrow ranges of mass velocities. There is also a great deal of variety in the methods of the reduction of pressure drop data. Correla- tors have used data which has appeared inthe literature59 only in a reduced form as a friction multiplier. Table 4.9, which summarizes the correlations presented in this chapter, also shows that a variety of friction factor and void fraction calculations are used in the development and application of correlations. These effects can also cause variation in the pressure drop predictions made by the correlations. The general applicability of correlations is definitely limited by the data. Several correlations are based on data taken at low pressures. Their validity in cases of high pressure is dubious. The comparison with actual data as reviewed in subsequent chapters will give an indication of the limits of suitability of these works for steam-water systems.Table 4.1 Lockhart-Martinelli Correlation Constants 60 Liquid Turbulent Re_>2000 g Reynolds Number Flow Type Subscript Range Kj monic Gas Viscous w Rej2000 Liquid Viscous vt Re,<1000 16 -046 1 2 12 Gas Turbulent Re,>2000 Gas Turbulent tt Re,>2000 +046 .046 .2 .2 206L 00°T st‘z one 06°S a8 orst O-oE O€T sz7s 00°T 00°T Bere oz'b s6rL oveT eee O-ep otz oz 0670 00°T sevz st’ onre gr2t 6°28 91z $89 08-0 o0°T €z°z 96°€ sore Tet TZ Ohh 66T $z9 0470 00°T oz one peo Tt b°6T oron vLT Shs 0970 00°t L6°T Bere 6s°s 6°6 ovLT 6°bE SvT osp os-o 00°T €8°T zore zerb v8 BOT z6Z = STT ose ov-o 00°T 89°T zoe bore 879 9°IT ovez $8 soz o€'o 00°T TST 6tz sere ws 9°8 z-9T 9S ost ozo 00°T oet SLT shez ve ws 6°8 8z 69 oto o0°T “UT ent SL°T 4 ore €°S st of soo 00°T 00°T 00°T 00°T Ort ort ort T T 0 g0ze oo0g oosz o00z oost oo0t 00s 00T “pt KatTend (-e-T+s-d) emnssazg (OT] OgWYH UT pesn szeTTdT3TAN TeOOT UOSTEN-TTT>UTIIEW Zp eTaenTable 4.3 Slip Ratio Values Used By Thom [15] P 250 40.0 600 20.0 1250 9.80 2100 4.95 2000 2.15 3206 1Table 4.4 Thom Void Fraction Correlation [1] 63 Pressure (psia) Steam Quality 250 600 1250 2100 3000 % By Wt. a a a a a 1 +288 +168 +090 © .0476 0213 5 678 512 +340 207 102 10 +816 +690 521 +355 +193 20 +910 +833 .710 553 +350 30 +945 6895. 808 +679 +480 40 +964 +930 866 +767 589 50 +975 +952 -908 +832 +682 60 +984 -.967 +936 0.881 +763 70 +990 979.959 +920 +834 80 +994 +988 = .976 +952 +895 90 +997 +995 989 978 951 100Table 4.5 Values of 0, for the Separated Flow Model as given by Thom [15] 64 - Pressure (psia) Steam Quality 250 600 1250 2100 3000 % By Wt. tio | %o [fo | %o | *o L 2.12 1.46 1.10 - - 5 6.29 2.86 1.62 1.21 1.02 10 11. 4.78 2.39 1.48 1.08 20 20.6 8.42 3.77 2.02 1.24 30 30.2 12.1 5.17 2.57 1.40 40 39.8 15.8 6.59 3.12 1.57 50 49.4 19.5 R.03 3.69 1.73 60 59.1 23.2 9.49 4.27 1.88 70 68.8 26.9 10.19 4.86 2.03 | 80 78.7 30.7 12.4 5.45 2.18 90 88.6 34.5 13.8 6.05 2.33 100 98.86 | 38.30 | 15.33 6.664 | 2.48065 T 1 T 1 T T T T 1 T T T t] ot] ot T fee | oze Jose | evz | sez jeer | cert | ost |os-t |ge-t |ozt | ett | 90:1] zo-t| to-r £°0 rot | ot-6 |o0°e | oso j oss | oz-y |o9'e | ogs-z jos-z |so-z ject | yt | eet] ett | vot v0 eee | roe} ose) ztz| ert] zet| ovtt}oe-e eto | zy |syre |esez | tet] ss] att €0°0 oot | ose} 0-69) oes] Srey] see] ovze| o-0z] 0-97} v-zt]09°6 | oo | o8-y| o¢-e | 6s: To°0 | ose | oz | sst | ott | 0-98} ovea| sey) o-se| o-6z} e°zz] erot| ott} ose | o6'»| go-z 700°0 ooo'r | ov | ves | oe |ove | aot | gor | o-oc | z-ar| z-ve | g-zz| s-rt| o8-e | o9-s| ct-z 100°0 000*0r | 009°9 | ove *y | oso‘z | oo¢*t | o€9 | gze | €9t | 0-66| o-ey| s-9z| or9t| 02-6 | oss | oz-z To00"0 oor | og | 09 | oy | oe oz | st | or | se | s [se } z tT | so] to 4, | | S i Fl sam £q % Aaztend sodeq zo xepur Aqaedoaa Teotskua (otl z 33 2u/aTorxt=9 203 9% sertdratoR [eUOFADTAY Bseyg-oMy Jo Sa3eUTPI0-09 UoTIeTe1209 Kz00I0q ory eTaeLTable 4.7 Values of B for Equation (4.66) 66 Mi G(1bm/hr-£t?) B <9.5 <3.69X10° 4.8 3.69X10°1. 4x10 1494/ 9.5<1<28 <4.426x10° 14123/pg-5 5 <4.426X10 ay, 5 >28 - 4.075xX10 » ages67 Table 4.8 Constants for Equation (4.67) Geometry K n Round Tubes +83 (.087) 1.4 Rod Bundles and Annuli +213(.0354) 1.6 Two values of K are given. The values in parentheses are for use if the variables of equation (4.67) are in CGS units and the others are for use with SI units. n is independent of the unit system.68 a Pe wt ersd 000Z-6ET Sty *8Ta oT Azv01eg 2/990 eons. | ahYS aqny, y8noy etsd QOOe-sT s'y "9PL ST woul seies-weeIg é a etsd cy 2eeN iSty cuba 90 eTeqr x01e8-1Ty 3 UsTHD oog-225 awys é ¥sd QOOT 9sty ‘ubg | €T yyoqueg | xeqen-uwaag a aqny ysnoy a Soy cuba | TT sewop-uos Tax | -FTTOUrazen avs a ered oovt-09 | yyy cubz | zt adueyoxa roqeA-weeas unaueMoy £a97 ans I at etsd OOLz-0ST | Te*y cuby 6 asqosery, | re7eA-meeIs -pueuxy as Py =a esd QO0E-ST ty °8Td 9 WOSTON | seco Teaen-aas “srrouyasen avs -?Yn9-7F esd OS-ST Ty *8ta s TTTOUTaIen z spynbyT snoyrea-rty =312q9907 anys PY po 2d ersd ct | 6*y cuba | y puemry sz-"Y exo ee Worweag PYON | (Z) xoa9ey WOTIOTAT uozaesttddy 10/pue quaudoteasg uy pasa stepow eaeq Burazoddng | uopaeottddy | -30y Tepow 20 30 pourey, uotze {21109 suoTaeTezs09 aseyg-omy, Jo Areumns y ory eTaeL69 *sz03eTe1109 842 (%) Figure 4.2 Martinelli - Nelson, Two-Phase Friction Multiplier for Steam/Water as a Function of Quality and Pressure [25]0.8 Pressure psia ° ° s a Void Fraction @ ° in 0 0.001 001 Ol Mass Quality x Figure 4.3 THE MARTINELLI -NELSON FRICTION PRESSURE DROP CORRELATION C61 7273 Void Froction @ Oo Ol 02 03 04 05 06 07 O08 09 10 Quality x FIGURE 4.4 THOM VOID FRACTION CORRELATION [1511,000,000) 2 to TWO -PHASE MULTIPLIER, ¢, A Tt tt T (SOURCE: BAROCZY 1965) big NOTE? VALUES ARE FOR G= 1x 108 Ib/n-ft? pit 14.7 psia 4 100 psia i rut 200 psia 500 psia | 1000 psia | 2000 4 psia PROPERTY INDEX (py Mig)? (5 /py) WATER (°F) 212 328 382 467545 636 705 Figure 4.5 Baroczy's Two-Phase Friction Pressure Drop Correlation C252(1.0 x 108 ) 09 $55 $, TWO- PHASE MULTIPLIER RATIO: ce T T T T (SOURCE : BAROCZY 1965) 20 6= 0.5 x OP AND G = 3x10% tv [S=0.5x10% Ib/h-Ft2 | G=3x108) Ib /h-ft I @ ° o T 0.4 i t L | [s-a25x Ib/h-Ft G=2x10% | Ib/n-Ft2 [40 to 60 0.0001 0.001 0.01 oul 1 PROPERTY INDEX (pp/p9) ( pg /pg) Figure 4.6 Mass Flux Correction Versus Property Index C25]16 14 1.3 1.2 Lt 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 BORISHANSKY PARAMETER - F(x) O Ol 02 0304 05 06 07 08 09 10 MASS QUALITY -x Figure 4.7 Borishansky Correlation C18)Chapter 5 THE METHOD OF EVALUATION OF PRESSURE DROP CORRELATIONS 5.1 General The correlations reviewed in the previous chapter were assessed by comparing the correlation two-phase friction multiplier to that derived from data. A difference ratio 24) correlation - (62,) aata ts.) 3 : (e¢,] data was used to quantify the discrepancy between the correlation and the data. A similar ratio of the quality - averaged friction multipliers was used for diabatic data. The cal- culation of the terms of equation (5.1) is discussed in subsequent sections. Overall measures of merit of the correlations can be expressed as means, root-mean-square values or standard deviations of © as given in equation (5.1) for all data points. The minimum root-mean-square value of © for all data points is the primary figure “£ merit since it embodies the significance of both the mean error and the standard deviations. It is given as, € —. (5.2) 778 The mean and the standard deviation of all the values of are also presented. These same measures are used to assess data grouped in subsets of common physical properties. An uncertainty interval or range of possible error is calculated for each point as prescribed by Kline and McClintock [26]. The R.M.S. and mean uncertainty ranges were computed for various data sets Appendix A gives more details on the un- certainty analysis. Approximately 2,200 adiabatic data points were used in this evaluation. An overall merit was established using this data and it was verified against approximately 1200 diabatic data points. The adiabatic data was also grouped by similar physical properties and compared with the correla- tions to determine the impact of mass velocity, pressure and quality on the suitability of the correlations. 5.2 Pressure Drop Data The data used in this evaluation are identified in table 5.1. A primary objective was to use raw data so that common void correlations and friction factors could be employed in the manipulation of all data. It was also desirable to know the uncertainty in the two-phase friction multiplier derived from data. The adiabatic and diabatic data used ranges in pressure from 200 to 1500 psia and covers the entire quality range and a large spectrum of mass velocities and configura- tions. This data specifies the measured pressure drop or gradient, the flow conditions and geometry. Most data sets79 provide adequate uncertainty information for the measured variables. These error intervals are given in table 5.2. Some data is presented without the uncertainty range. A median value of the error interval for other experiments as is given in table 5.2 is applied to these points. There is a large amount of steam-water pressure drop data which has been presented in the literature in a graphical manner. For this sort of data the uncertainty information is often not presented. Then there is also the uncertainty in converting the plotted data to numerical quantities. Con- sequently, such data is not considered. Another limitation in the amount of data used is a function of the human re- sources in this effort. Consequently, adiabatic steam-water data at pressures less than 250 psia which is available was not used. The diabatic data is also limited to provide points covering a variety of geometries over the same range of conditions as that of the adiabatic data. 5.3 The Reduction of Pressure Drop Data As stated earlier the pressure drop data was reduced to the two-phase multiplier, a as defined by equation (2.18). Equation (3.14) was used to reduce the pressure drop to a friction pressure gradient. In the case of adiabatic data it was used as given. For diabatic data it was integrated in ten steps over the quality difference. In this case an average two-phase friction factor was cal- culated. It can be defined as80 (5.3) i # r| daz tn (2), The adiabatic raw data was reduced by several methods. Three methods of calculating the void fraction were used, the Thom correlation, the homogeneous model and the Martinelli-Nelson correlation. The Thom correlation was selected as the primary void fraction because of its reputed steam-water data base. The Martinelli-Nelson void fraction correlation and the slip ratio of unity (homogeneous model) were used to determine the effects of the different void fraction models on the results. The homogeneous void fraction model was used to provide data reduced by the seme method as that on which the CISE correlation was. It also provides for a completely homogeneous computation of data based multipliers for comparison with those homogeneous model friction multipliers given in equations (3.6) and (3.11) through (3.13). The effects of the friction factors were also examined. The adiabatic data was reduced using both approxim- ations (equations (2.26) and (2.27)) and the smooth tube friction factor given by equation (2.28). The diabatic data was reduced to an average multiplier using only the smooth tube friction factor and the Thom void fraction correlations.81 Much of the diabatic data was for flows having sub- cooled inlet conditions. The location of the point of zero quality was determined using equilibrium thermodynamics. The region from the inlet to this point the flow was treated as a single-phase flow with a friction factor of .0075 being used for the rod bundle data presented Lahey et al [35], which is consistent with their single-phase experimental results, and .005 for other ducts. The computer programs used to reduce data are dis- cussed and listed in detail in Appendix B. The output of these programs were the input to an evaluation program. This output described the geometry of the duct and the flow conditions as well as giving the friction multiplier base on the data and its uncertainty range. 4 The Evaluation of the Correlations For each data point the computer program assessing the correlations computed two phase friction multipliers for each of the correlations given in Chapter 4, as well as the four homogeneous two-phase multiplier given in Chapter 3. These multipliers were compared with experimental result using equation (5.1) and the appropriate mean differences, R.M.S. differences and standard deviations were computed. This output was given for each data set as well as the entire data collection. Each set of reduced adiabatic data was evaluated twice, once as sets based on the source of data and secondly as groupings of like properties. The property groupings combined data of similar pressure ranges,82 quality ranges and mass velocity ranges. The intention behind these groupings was to determine the flow conditions at which a correlation may be most or, for that matter, least effective. Table 5.3 gives the property ranges that were used. They were selected so as to provide a signif- icant number of points in each data grouping. In all, 42 subsets, each having a specific pressure, mass velocity and quality range were formed by mechnically sorting the output of the data. For diabatic data the correlation multipliers were determined and averaged over the quality range of the data point. The average multipliers based on the correlation were compared with the data average multiplier in the same manner as the adiabatic data. Appendix C provides further details on the programming of the correlations and a listing of the program.T90" | 90" LL "ZO" Tet OZvT-009 | 4c‘ | “ZFXOH |TauUeYD -3DeE/ 79 stv ert" tet’ |6er-zor]T2-2" OTOT-069 | 8£2'| unog |Teuueu -a00y) 97 aad sez: | 99t" J66r=zor]—T*Z-S* otet-005 | 82c° 4 |Tewueyp -a004y} ee-v Tez"; TST" |€s*-00" 6*Z-3" Ov0T-OTOT | 76T" aq snqnuuy} 77 wv 760° 180° |Zs*-T0" erz-s" ovot-orot | y6t’ do snqnuuy| ¢Z Te eso" | = €S0" | TS*- TO" 6°2-S* O8TT-0€L | 860° da sntnuuy| L€ oz-¥ 9s0° gso" |£8°-Z0" 6°-1'T 0ZOT-OOOT | ¢6T" aq edta "pul 92 6I-¥ 850° sso” |ze°-c0"| z"T-8* ogot-otot| g6t’| da edta “pul eT sT-¥ 6s0" oso" |te*-€0"} 6 "2-8" oeot-066 | cet’| dan adra “pul 99 uv 190° oso: |96°-zo-} 6*z-L" o€ot-oo0t | ooz"| da edta ‘Py) SST oI-v 799" zet* — |86*-z0" S*T-£* OTET-08Z | 86S" an adtq ‘pul 897 sT-¥ eso" Ls0" |86"~SZ" TiT-L" Ove -0€c | 86S" dq. edtd ‘Pal 77 yI-¥ zL0" 990° — |0"T-z0" 6°Z-€° oEot-0EL | T9E* do. edtd “Pu| 096 e1-¥ 190° 090° |09°-To" 6°c-8" OOEt-oTL | C6T da edta ‘pul ZZ zI-v 190 | = 090" Jos*-€o" | 6 "z-B* OzOT-oTOT | ste” do eda “pul Ts TI-v 990° 790° jv*-00"} 6 *Z~L O€OT-066 | LZT" da snqnuuy} TST or-v Sto° 490° |zz"-T0"}vre-8 ocot-066 | 92° da snqnuuy] 39 ov 190° 090° |j9z"-yo"} 972-8" ovot-066 | c6r" da snqneuy| 19 ev sto" 690° |sz*-€0"} 72-8 ozot-066 | gee'] da eqn “pul tz iW 6s0" 650" |S9°-ST™ Ore-L OTOT-066 | €ze" dq eqnL "pul LS ov £90" 190° |se°-€0"}OvE-L* ozot-066 | 8yz’] én ean “pul 9 sv 090" | 6so" fecr-zor}_ore-e* ozot-066 | soz’] da any “pul 8s al $90" | €90° — |y9"-TO"| EvE-L* otot-066 | cet} dn sang “pul 6” ew e90° | = €90° | TL"-To" 6°e-L" orzt-ogs | soz} da eany “pul Z/T w¥ 090° | 6s0*_—|g9"-s0" 67-8" otot-066 |soz’} dn pany “pul 9S T-v or Aaupeazooug [Aauyeazeoun| o8uey | 2 aazy| (FS) eBueyl (uy) luoravearq) worreaneFsu0g| sauyog|*3ou |(T 930N) eaeq sia | eaea uean |Aazrend| (HITED | oanssorg | ea | nora 298 Aaqo0TAA s8eH ee Apnag Stuy ur pasn e120 T's eTaeL,84 voTaeqerd ~ @ SoFaeqerPy — v :T 230N zt zy't — |s9*-90--] sce: ooct = eee dq freuueyp -a004} sz | ge zI-a ST'T ev’ |sy°-t8 wz" ooot = vén"] dn keay| Te | se Tr-a ont 97°T — |0€*-0°O IT SOOT-£66 £89" an edta ‘Pa s | ze ol-a sot" zor" |¥2*-90" oT ooot = azz} dn satnuuy | zt | zz 6-0 z20° TL0" | ”8"-TO" 6°z-8" OZOT-S66 S0z" an edta ‘pal ey | 82 e-a 060° 880" |€8"-90 6°-L" 009T-06S —S0z" dq edta ‘pu] 60¢ | 87 ia 280° ogo" |9g-zt ove-8" osot-066 c6T*} da adra “pul 12 | 62 9o-a 020" 890° |T6"-€T | o"e-8° OPOT-O00T L6T* do adta “pul 04 | 6z s-a Lo" s£o* |66°~"T | 6°7-8" OvET-OzZ Let" dq edta "Pu; 6ST | 6z ya s90° 990° |36°-90~ 8°Z-L" Ogot-oo0t o0z"} da edta “pul ox | 62 a tL0° €40° |96°-Te~ 62-2" | ovet-ozc 66t"} dn edta “pu| tz | 62 wa gyT" gery |tee-ct z't-2* | o¢ot-086 get} da edta ‘pal st | 6z T-a eet" ost: |sz°-so- 9'T-"T | o¢ot-og6 eg9"| da adra ‘pu} 9 | ze ee-v yor" so" |06*-60" Te" O0yT-009 zz] uno edra “pu; Le | 0€ Zev 290° 790" = |06*-60" LeT-8" 000T tye" | *2E20H edta ‘pu} yt | o€ Te-v 790° 290° |06*-60" 9° -e oot = -£z"t| *2Ta0H adta ‘pu) 9 | of oE-v z90° 790° |06"-60" TT-z" Oovt-009 ss6"| -2t20H edta ‘pu] op | of 6c-¥ Ter" 660° |G6"~60" Tt-2" | oovt-o09 ss6'| da edra “pu) 9¢ | o€ 8c-v 790° z90° — |06"-so* Tez-S* | otot-009 s€y°| *zr20H |rewueyy -a399y! gt | oF Lev 790° 90° |z6"-so° T'e-S* | OT#T-009 sey"| = dn [teuueyy -a00y} ez | of ov got Aquyeazeouq | 900 psia, G < 1x10°1bm/nr-£t?, 1x10® < ¢ < 2x10°1bm/nr-£t? G > 2x10°1bm/hr-£t?. QUALITY: o 9L9EETC 19380070- L eogtess ESbbEo Sezer ot 1R9€S°O €06L6°D ¥SLIB°O st 9980290 zee9z"0 2919T*9- oT yoeEz 0 eLI9Z"0 €L9CT*9- a LEwHorT yReovr Ee eaeour2 20 €BooL* €22z870 9ze%E°9 It Selcwr? 6Ee98"O Eon9L"9 ot 6SBL9°o SLES Sseez"o- 6 4926E°9 419095 sT6S¢°0 8 o9zser> o2usorl L622o°T L LO9ETS Z2Lyer0 6EETOTO 9 T60%L "1 s9sLore 620ETTT s SLLT?O €928€°9 6€0%E7O- » lztz2"9 69882°0 aysatto- € orsetrs oLeeeso 4% OL 2° 0- 2 1824279 zoe" e12otro- T 9LLLI"9 -Y9990"G GEzZ ce A3O GIS wOUsd SWY YUYYa Nw yous sous NOTAVISYHGI NOTLVISHYOD NOTLVIFVNOI NOTLVIRYNCD Sw¥ vivI NA Viv SINI3d SL3S vied99 oLbsz °C So99e ts zosuere’ dewoeru Lecle*o LSSEE TO SzEns*G Livse'o nB67Z °C zozazre LELHZ*O B@zez 70 bezneri uSEnere LLLSL70 caster nEeBE TD nbsLe a LL9Sn“O ezszsc SCEGE ‘0 seces‘o oceeary Leeoa'e Leese te 600SEtu €S60L"L nELZO°~ LmSL*o @LS6E"D 69027 "0 zzo0ero sone" Ssone "3d gosne'o OSEL7°9 Agu dus aOkes Sha xo}OeE UOTIOTIZ eseyd-eTSuts eqny, yR00ug PUY T®POW UOTIDeIZ PTOA snosuebowoH eyL U3TM PeonpSY eIed OTIEGeTPY JOJ S3TNSeX TTezeAO Ge9ZZ*u ozeeorc- bneoL te en6LL"9 Leszeto- 66SZL O- SOLzLte oonzerc o6ccL*c bnesz"c- eo9ene ESLOE TL ZELOd*G- BLeeC'L 9LASE TG mSE€0z*o- eLyut*c- CmEZLT O- eOdda NK KOILW1addCD NOTLV1S9d09 NOTLV1a0009 9°9 eTaeL ob “ut oh st mt et a ib aL ramsnernae NOTLVIS ded nZoLo “uy ious Shi WL¥G rL99o"C aouda Nk va seze SiNIOL ke Saas YING100 sumututu oy3 zo T-o uruara SHY suracy suorzetozz00 aayosery, a ~pueway SS —————— 8 asypsery, (Tt €) cubs -pueury Aysueystizog | Aysueystiog | snosueSouoy | Aysueystiog L (1T"€) cuba asyosexy, asyosexy, asyosexg, snosusbowoy | wroystyd =puewsy -puewty =puewsy 9 (tte) cuba} (tte) cuba | Tre) +ubg wToystyo snosuabhouoy | snosuebowog | Aysueystiog | snoausbowoy S @ 2 (ete) “uba | (2t-€) -uba (Te) uba x snosuebowoy | snosuebowoy Kzo0reg snosueSouon Kz00z¢@ » | 2 6 (97€) *uba (2t"€) *uba (zt-€) cuba | snosuebou0g Kzo0re¢ snosueSou0H Azoozeq snoatabouoy | € (97€) cuba wou, snosuebowoy woud wou, | wous, z (97€) cuba (g7¢) cuba (g*€) cuba Azoozeg wouL snosuehowoy | snoeuebouoy | snosuebowoy T Tepow uOSTON snosuebouog TTTeuT@zeW woug wouL wous, uotzORIZ PTOA eqns YrOOUS | eqnL YROOUS | Sqny YrooUIS | —-.eU/6L0°=F | 7.9N/9¥0"=F | ZORORE UOT OTIE pouren uorzonpey eqed uot eTTOD eFeq SITIU_ suL wats Aouedexostq yseeT e432 SutaeH ST@epow pue suCTJeTerI0D doiq einsseizq eseyd-om], 479 eTaRLTable 6.8 The Adiabatic Data Subsets Based On Physical Properties 101 Mass Velocity 2 Data Set (psia) 10 Quality Points Appendix E 250-900 0-1 20 1 42 7 29 13 34 19 28 25 53 31 48 37 1-2 30 2 37 8 28 14 31 20 17 26 23 32 17 38 2-3 13 3 8 9 9 15 9 21 9 27 9 33 900-1500 o-1 67 4 86 lo 79 16 68 22 54 28 110 34 94 39 1-2 107 5 143 ll 95 17 90 23 77 29 129 35 63 40102 Table 6.8 (continuted) Mass Velocity 2 Data Set Pressure don/hr-fe" Mass Number in (psia) 10) Quality Points Appendix 900-1500 2-3 84 6 90 12 76 18 63 24 57 30 69 36 27 41103 6BEzH 0 EpDLDO ZZE7S*O TP9T6 “0 D9STE*O OT8Ty*O v80LS*Z 0SZE6 "0 E0TS9“O O9PEE"O 6679 "0 SZEPT*T L9ESS*0 6508S °T 97167 "0 vOv9E"O LOOzE "0 6L¥zy"O qd dus LLWTSXYOD vlepyo seseyo s98ss"0 LEEeeT 6szLe"0 o6zzp0 s0667° 99SP0°T 4106670 68SSP°0 096£8°0 €6826°T Z16SS"0 6Z8E6°T ogser"o S9S6£°0 TSLOv"O vyszp'o YOuNT SHE NOILWISYYOD NOTLVIaHYOO zoyOeg uoTIOTIE eqns, y300us eseyg-aTbUTS puy uoT3eT91I09 UOTIOeIZ PTOA WOUL Os UAT paonpay e3eq OTIeGeTA TOF s3TNsey TTeIeA0 T8LZT°0 Tpz0T*0- 9LS6T°0 95896 "0 €086T"0- 8S€90"0- LOSbHE ZIeLy"o 9S6TL*O €960€"0- 9Lezs"0 E9€SS"T T6LLO*O zeTZT-T B98Te“0- L6bST"O- €zzsz"0- 08550 -0- YOuNa NW 6°9 eTAeL 8T LT oT ST oT &t et tT HaAmMsnonan 6za6z"0 zL9ZT “0 yous yours NOILWIGYYOD sWa WIVd NW Viva Tez SENIOd et sigs wiwaChapter 7 CONCLUSIONS In reviewing correlations it is seen that several of them are based on only small amounts of steam-water pressure drop data or data limited to certain flow conditions. It is not expected that these correlations would be very applicable for conditions extremely different from those upon which they are based. The Lockhart-Martinelli, Armand and Sze-Foo Chien and Ibele correlations are all based on very low pressure data, none of which was for steam and water. The Lockhart- Martinelli correlation compared with data most favorably (even though only marginally so) at the lower pressure, and lowest mass velocity subsets. The other two correlations compared marginally well with data having low quality. rt is obvious that these correlations are not applicable to the data covered in this study. The Martinelli-Nelson correlation, which has been generally accepted, shows unfavorable overall results. However, for the data sets with mass velocities less than 1x10°1bm/hr-£t? it compares very favorably. This should be expected since the data on which this correlation is based is within this mass velocity range. The Thom correlation, which is similar to Martinelli-Nelson correlation in forma: is based on data with higher mass velocities. Since the Thom correlation is based on and compares well with data 104105 near the center of the mass velocity spectrum, its deviation from data having higher and lower mass velocities is less than in the case of the Martinelli-Nelson correlation, which is centered on low mass velocity data. Different correlations will compare more favorably with different data sets, this all depends on the data, how it was reduced, the geometry and environment of the test. Any correlation can appear to be good if checked by selected data sets. However, as noted in the results of this study several compare more favorably than the others do with the entire data collection. The four which compare most favorably with all the data are the Thom correlation, the Baroczy correlation and the homogeneous model two-phase friction multipliers given in equations (3.11) and (3.12). These are recommended for yeneral application in the range of data covered in this work. The breakdown of these results by property groups offers the opportunity to identify that correlation which is most appropriate over a specific property range. This is not recommended for any sets based on a small number of points (for instance fewer than 50 since a few erroneous or " ad" points could have a noticeable effect with a small data set. Some reservation is also expressed if the method of calculation differs from the method used here to reduce the data. However, this study did show that overall the use of different reduction methods had only small effects.106 The mean correlation error for a particular correlation and property group and the correlation value obtained for a point within that property group substituted into equation 5.1 would yield in a friction multiplier more representative of the data studied here. As previously mentioned the results and recommendations of this study are only valid in the range of data studied. In terms of nuclear reactor technology, this indicates that the data is applicable to boiling water reactors. For the analysis of boiling water reactors the Armand-Treschev correlation is recommended for qualities below 0.3 and the Baroczy correlation is recommended for higher qualities. For applicability to pressurized water reacto.s a similar study should be conducted on steam-water pressure drop data at higher pressures than those examined here.lo. 1. 12. 13. REFERENCES J. G. Collier, Convecti McGraw-Hill, Londen, 1 iling and Condensation, G. B. Wallis, One-Dimensional Two-Phase Flow, McGraw- Hill, New York, 1969. %. 8. Tong, Boiling Heat Transfer and Two-Phase Flow, John Wile & Sons, New York, 1965. A. A. Armand, "The Resistance During the Movement of a Two-Phase System in Horizontal Pipes," AERE-‘rans. 828, March 1959. R. W. Lockhart and R. C. Martinelli, "Proposed Correlation of Data for Isothermal ‘Two-Phase, Two Component Flow in Pipes," Chemical Engineering Progress, Vol. 45, No. l, Jan. 1949. R. C. Martinelli and D. B. Nelson, "Prediction of Pressure Drop During Forced-Circulation Boiling of Water," Trans. ASME, Aug. 1918, D. S. Rowe, "COBRA-II: A Digital Computer Program for Thermal -Hydraulic Subchannel Analysis of Rod Bundle Nuclear Fuel Elements," BNWL-1229, Feb. 1970. D. S. Rowe, “COBRA-ITIC: A Digital Computer Program for Steaay State ard Transient Thermal-Hydraulic Analysis of Rud Bundle Nuclear Fuel Elements," BNWL-1695, Mar. 1973. A. A. Armand and G. G. Treschev, "Investigation of the Resistance During the Movement of Steam-Water Mixtures in a Heated Boiler Pipe at High Pressures," AERE-Lib/Trans. 816, Jan. 1959. R. B. Bowring, "HAMBO, A Computer Program for the Sub- channel Analysis of the Hydraulic and Burnout Character- istics of Rod Clusters," AEEW-R-582, Jan. 1968. A. B. Jones, “Hydrodynamic Stability of a Boiling Channel," KAPL-2170, Oct. 1961. S. Levy, "Steam Slip - Theoretical Prediction From Momentum Model," ASME paper 59-HT-15, 1959. §. G. Bankoff, "A Variable Density Single-Fluid Model for Two-Phase Flow with Particular Reference to Stcam-Water Flow," Journai uf Heat Transfer, Nov. 1960. 10714. 15. 16. lq. 1g. 19. 20. al. 22. 23. 24. 27. 28. 108 $ze-Foo Chien and N. Ibele, "Pressure Drop and Liquid Film Thickness of Two-Phase Annular and Annular-Mist Flows," ASME paper 62-WA-170, 1962. 3. R. S. Thom, "Prediction of Pressure Drop During Forced Circulation Boiling of Water," Int. J. Heat Mass fransfer, Vol. 7, 1964. Cc. J. Baroczy, "A Systematic Correlation for Two-Phase Pressure Drop," Chemical Engineering Progress Symposium Series, No. 64, 1968. K. M. Becker et al, "An Experimental Study of Pressure Gradients for Flow of Boiling Water in Vertical Round Ducts (Part 4)," AE-86, 1962. v. M. Borishansky et al, "Some Problems of Heat Transfer and Hydraulics in Two-Phase Flows," Int. J. Heat Mass Transfer, Vol. 16, 1973. D. Chisholm, Pressure Gradients Due to Friction Durirg the Flow of Evaporating Two-Phase Mixtures in Smooth Tubes and Channels," Int. J. Heat Mass Transfer, Vol. 16, 1973. rs Cc. Lombardi and in Two-Phase Flow, Feb. 1972. Peddrochi, "A Pressure Drop Correlation Energia Nucleare, Vol. 19, No. 2, N. M. Rohsenow and H. Y. Choi, Heat, Mass and Momentum Transfer, Prentice-Hall, Englewood Cliffs, N.J., 1961. R. V. Macbeth, "The Effect of 'Crud' Deposits on the Functional Pressure Drop in a Boiling Channel," AEEW-R 767, Apr. 1972. 0. Baker, “Design of Pipelines for Simultaneous Flow of Oil and Gas," Oil and Gas Journal, July 26, 1954. Sher, N. C. and Green, S. J.," Boiling Pressure Drop in Thin Rectangular Tubes," Chemical Engineering Progress Symposium Series, No. 23, 195). R. T. Lahey, Jr., "Two-Phase Flow in Boiling water Nuclear Reactors," NEDO-13388, Jul. 1974. S. J. Kiine and F. A. McClintock, "Describing Uncertainties in Single Sample Experiments” Mechanical Engineering, Jan. 1953. N. C. Sher, "Review of Martinelli and Nelson Pressure Drop Correlations," WAPD-TH-219, Jul. 1956. "A Research Program in Two-Phase Flow," CISE, Jan. 1961.29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 109 G. P. Gaspari et al, "Pressure Drops in Steam-Water Mixtures," CISE-R-83, Jan. 1964. E. Janssen and J. A. Kervinen, "Two-Phase Pressure Drop in Straight Pipes and Channels: Steam-Water Mixtures at 60C to 1400 psia," GEAP-4616, May 1964. A. Era et al, “Heat Transfer Data in the Liquid Deficient Region for Steam-Water Mixtures at 70 kg/cm? Flowing in Tubular and Annular Conduits," CISE-R-184, June i966. E. Jannsen and J. A. Kervinen, "Developing Two-Phase Flow in Tubes and Annuli," Part I, GEAP-10341, Feb. 1971. R. W. Haywood et al, "Experimental Study of Flow Conditions and Pressure Drop of Steam-Water Mixtures," Proc. Inst. Mechanical Engineering, Vol. 175, No. 13, 1961. W. F. Davidson et al, "Studies of Heat Transmission through Boiler Tubing at Pressures From 500 to 3300 Pounds," Trans. ASME, Vol, 65, 1943. R. T. Lahey, Jr., et al, "Two-Phase Flow and Heat Transfer in Multirod Geostries: Subchannel and Pressure Drop Measurements in a Nine-Rod Bundle for Diabatic and Adiabatic Conditions," GEAP-13049, Mar. 1970. N. Adorni et al, "Description of a Steani Loop for Heat Transfer Experiment with Steam-Water Mixtures (Part I) and of Pressure Drop Measurements in Adiabatic Condizions (Part II)," CISE-R-62, Dec. 1962. L. Berkowitz et al, "Results of Wet Steam Cooling Experi- ments: Pressure Drop, Heat Transfer and Burnout Measure- ments with Round Tubes," CISE-R-27, Oct. 1960. 0. Mendler et al, "Natural-Circulation Tests with Water at 800 to 2000 psia Under Non-Boiling, Local Boiling and Bulk Boiling Conditions," ASME paper 60-HT-36, June 1960. B. Carnahan et al, "Applied Numerical Methods," John Wiley and Sons, New York, 1969. U.K. Steam Tables in S$ 1970. Units, Edward Arnold, London,Appendix A TWO-PHASE FRICTION MULTIPLIER UNCERTAINTY General The uncertainty in the value of the two-phase friction multiplier is calculated using the method of Kline and McClintock [26]. The uncertainty is defined as the possible value of error that the data might have. For an observation the error is the actual difference between the true and observed values. The uncertainty in experimental data is a function of the measuring instruments, the apparatus, re- cording method and environment associated with the particular experiment. Most pressure drop data is considered the result of single-sample experiment, since a particular case is not repeated and if so, not sufficiently to analyze the data spread by statistical methods. The experimenter must, therefore, estimate an uncertainty interval instead of computing a frequency distribution. Consider the variable v, whose value in an experiment is the data point m, for which the estimated uncertainty interval is + w, or 27M Ey (a1) Now, let R be a function of n independent variables vj, R= R(Vy) Var Vgreeee Vg) « (a.2) 1101 The corresponding uncertainty interval of R is then given by (A.3) One unresolved matter concerning this result is the significance of this possible error range. No matter how minute the probability of an error exceeding a specified interval, the possibility always exists. The significance of the interval is the understood likelihood that it will not be exceeded. If the uncertainties w, equations (A.1) and (A.3) are not exceeded by more than one value in ten, for instance, then this is the significance of the interval wa: A significance on the order of one of ten or twenty is R suitable for engineering applications. Another connotation for the same ranges of significance is a confidence level of 90 or 95 percent in the interval. If a particular error is presumed to have a Gaussian or normal distribution it may be expressed as a standard deviation 9, in which case equation (A.3) can be rewritten The standard deviation is used by some experimenters [28, 24, 36, 37] to define the interval and significance of the uncertainty in their recordings. Others have specified a confidence level [30]. Many researchers have not denoted12 the significance of their interval. It can only be assumed that they are of engineering significance. ‘There is data published that does not even address the subject of un- certainty of error in measurements. The pressure drop in a given two-phase flow situation is compared to a correlation by computing the friction multiplier based on the data and the correlation. The value calculated from the correlation is assumed to have no error. Equation (A.3) is applied to the appropriate relations, which reduce the experimental data to a friction multiplier, to determine the uncertainty in that multiplier. A.2__The Uncertainty in Recorded Data Table 5.2 gives values of error intervals reported for various pressure drop experiments. The 1961 CISE report "A Research Program in Two-Phase Flow" [28] used the standard deviation as the measure of data uncertainty because of its recognized statistical significance. However, it indicated that the values of the standard deviation were maximum errors determined from nameplate data, tests ané estimates. (Maximum error is assumed to indicate an error of engineering significance) for each of the variables re- ported. Later reports [29, 36] from the same laboratory use the standard deviation also, but make no claim that the values used were maxima or not. A review of the assumed deviations indicates a strong likelihood that the values represented a range of accuracy that was of engineering significance. Thus, all the uncertainty ranges specified413 in table 5.2 are considered to be of a suitable level of confidence. In evaluating correlations the range of uncertainty of data must be known. Should the deviation from the data of more than one correlation be within the accuracy of that data the resulting evaluation must consider them to be of equal suitability over the range of the data. For those data sets for which the accuracy has been published, the uncertainty in the multiplier are evaluated using the given error ranges. All other data points (those without given uncertainty information) are evaluated using a median uncertainty. These values are also given in Table 5.2. A.3__Uncertaint Friction Multiplier For the adiabatic pressure drop the expression for the two phase friction multiplier is 2 _ AP = OP, ¢ = 2 (A.5) fo 2 2fe, CL Go Me D a [= = where AP, g, |Pa% * Pe a a and the acceleration pressure drop is presumed to be negligible in adiabatic flow. The pressure drop and dimensions are the only directly measured variables. The other variables are calculated from measurements of mass flow, pressure addition and heat losses. The uncertainties in the calculated114 variables of equation (A.5) are generated by applying equation (A.3) to the formulae used to compute those va: ables. One term of the expression can be written as 2 2 Mak 30f0 gap - | Lim Peo (APFOAP) - 6, (AP) bar o4P ~ |SAP+0 SAP . (a.6) Assuming that 6AP is sufficiently close to zero for numerical evaluation, equation (A.6) can be rewritten 2625 Sao SAP 2 2 eq (APHOAP) ~ $2, (AP). {A.7) Then for the computer solution, 862, = [2, captear) - #200]? + [02, (eese) - #251] + [20+ 60 - 200]? + [12, (cree) - #25(6)]? ‘ (A.8) The pressure P and quality x are in this expression because the properties and void fractions are functions of them. It is assumed there is no geometry uncertainty. Similar methodology is used to compute the uncertainty for diabatic pressure drop data. Figure A.1 gives the uncertainty in the multiplier computed from a set of adiabatic data [18] as a function of quality. Figure A.2 displays the major components of uncertainty which for this case were due to uncertainty in measurement of pressure drop and mass velocity. The pre- sentations of the two figures are related to each other by equation (A.8). ‘The behavior of these two components of the multiplier uncertainty is predictable. Consider first11s the uncertainty in the pressure drop measurement only. Equation (A.4) can be written SAP]. (a.9) 3AP Substituting equation (A.5) into equation (A.9) and dividing both sides by equation (a.9) gives (A. 10) as the uncertainty due to pressure drop alone. It can be seen from equation (A.5) that as the quality approaches zero followed by the void fraction, the gravity pressure drop in- creases to the value it would have if it were for a single phase liquid. At high qualities the gravity pressure drop decreases by a factor on the order of twenty. Equation (A.8) then reduces to (A.11) 2 £0 ap If the gravity pressure drop is considered negligible. Figure A.2 verifies these predicted limits. Similarly, for the case of the affect of uncertainty due to mass velocity alone : (a.12) This result is approximate because of the approximation of the liquid only friction factor is116 079 et ue (a.13) Equation (A.9) agrees with the results, based on data, which are given in Figure A.2. In that the uncertainty contribution of the mass velocity is independent of quality. In most experiments the variable recorded is not the mass velocity but the mass flow rate. Consequently, the errors cannot be applied strictly in the terms of equation (A.5). The range of uncertainty is known for independent variables such as, power to the boiler, and, inlet feed water temperature and yet the recorded value of thse variables are not given. The uncertainty effects of the latter two variables are, thus, more difficult to apply. For instance, the quality required to determine the void fraction is based on an energy balance, L 7) *= he [Bin tw 7 losses ~ Me ae fg * In most cases the losses can be presumed to be small. Applying equation (A.3) yields (A.15) Now since one (A.16)17 and assuming that h, fc tT. , (A.17) in ~ Sp Tin the uncertainty in quality can be expressed as 2 (%) 6%) W %. oi ceite \s 2) +S + tpt (a.is) In many adiabatic experiments the inlet temperature to the boiler heating the liquid to test conditions is low and if so (A.19) Then alt @ + or). (A.20) The components of the uncertainty in the quality due to each of the variables are {A.21) {A.22) And or, 6X, = oa (A. 23) in 5 When the error in mass flow rate is applied in equation (A.8), its effect on both the mass velocity and quality is simul- taneously evaluated in the term computing the error due to mass flow rate. There will be a term evaluating the error due to uncertainty in the boiler inlet temperature and11s another for the uncertainty in the boiler heat flux. Thus, the multiplier uncertainty, equation (A.8), can be expressed as 2 2 20 2 92 2 - 42 | 8b EQ [o2_ (apesaP) #25 (0P) | + [£0 (PSP) Geo (P) | 2 2 - + [42, extn +80, - #o(s-s)] (A.24) A.4 Uncertainty in a Diabatic Two-Phase tion Multiplier The uncertainty in the diabatic result can be computed in exactly the same manner as the uncertainty in the adiabatic multiplier. The behavior of the error range is much different though. The error range in diabatic data is strongly in- fluenced by the inlet subcooling or quality and the change as the flow travels through the test section. The inlet condition of the flow dictates the upper limit of the elevation pressure drop of the flow. The un- certainty in the multiplier will approach the limiting value of the ediabatic condition as the heat transferred to the flow decreases to zero. As the heat flux is increased the elevation term decreases in conjunction with the mean density. This change in quality also gives rise to an acceleration pressure drop. The acceleration has the same effect on the129 multiplier uncertainty due to 4P as does the gravity pressure drop, namely 32 E 5 . (A.25) %£0 As the acceleration term increases the multiplier uncertainty does also. The other effect that the acceleration term reflects strongly is the error range of the mass velocity. The uncertainty of the multiplier due to the mass velocity for a diabatic case can be expressed as (A. 26) AP-aP AP, = (A.27) 2FeoG7L ToPsP Assuming the friction factor given by equation (A.13) applies, equation (A.27) can be written as <2 _ AP=AP_-6'K, %Fq = (A. 28) for convenience. Then 32 ag 1.75(AP-AP_) .25 K 6° - aos a ta.29) G Ky G kK, Substituting equations (A.28) and (A.29) into equation (A.27 gives120 =2 . _ . [8¥£0 . 1.75(AP-4P,) ~ .25 apy 6 em 3 aP-aP_-OP c . \ 32, 2 AP c The friction pressure drop can be written as 4p, = 4P-AP -AP.. (a.31) This equation substituted into equation (A.30) gives 632) 1.75 OP, + 2AP fo : £ a | 8G 8, (a.32) \e IP, G fol ¢ which also shows an increasing multiplier uncertainty caused by an increasing acceleration term. Figure A.3 is a plot of multiplier uncertainty which slows the increase in incer- trainty due to larger quality changes, hence, acceleration Pressure drops. Figure A.4 breaks down a segment of the data of the previous figure into uncertainties in the mult:- plier due to the error ranges in mass velocity and pressure drop. Inlet subcooling complicates the error analysis immensely. If the subcooling is large or the outlet quality very small the error range in the multiplier due to uncer- tainty in the pressure drop measurement may be very large. The uncertainty in the multiplier due to AP uncertainty is q2 2% ¢0\_ bap me 52 AP-AP_-AP “AP Ec . % Fo. lap Zz a If the error in measuring pressure drop could amount to 2.5 percent and the subcooled length amounts to 90 percent ofqa. the tube, the two-phase friction pressure drop would amount to perhaps five percent of the total. For such a case the error in the multiplier is 50 percent. There could cer- tainly be even more extreme cases. The uncertainty of all errors will have greater effects for subcooled inlet conditions. Intuitively, this could be expected. Since the two-phase friction pressure drop is small relative to the overall all pressure drop. Small changes in these other components of the pressure drop would be relatively large with respect to the two-phase friction drop.392, $ MULTIPLIER UNCERTAINTY- 2 fo P= 1000 psia ob BAP | BAe = 025 G=.8x10° Ibm/hr-ft? 28 = 021 06 122 G6 =.8x108 Ibm/hr-ft? 3° & G=2.9x10°lbm/hr-ft? 02 -- ° 12 4 6 8 10 QUALITY - x FIGURE A.| MULTIPLIER UNCERTAINTY FOR ADIABATIC DATA.842 $ fo MULTIPLIER UNCERTAINTY - o @ z fo. 06 04 02 UNCERTAINTY IN MULTIPLIER G =.8x108 Ibm /hr-ft? P= 1000 psia BAP _ “ap 7-025 3G. S702! DUE TO UNCERTAINTY IN THE MASS VELOCITY ,G. UNCERTAINTY IN MULTIPLIER DUE TO UNCERTAINTY IN THE PRESSURE DROP, AP. FIGURE A.2 4 6 8 1.0 QUALITY-x MAJOR COMPONENTS OF MULTIPLIER UNCERTAINTY FOR A TYPICAL ADIABATIC CONDITION.MULTIPLIER UNCERTAINTY- 124 G=.8x108 Ibm/hr- #12 P= 1000 psia BAP. coat Bay = .025 3G. S =.021 06 oak ADIABATIC LIMIT .02|- A | ! | l ° 2 4 6 8 lo EXIT QUALITY - xe FIGURE A.3 TYPICAL MULTIPLIER UNCERTAINTY FOR DIABATIC DATA.32, ¢ MULTIPLIER UNCERTAINTY ~ P= 1000 psia SAP . a Ap 7-025 06 |- — A ore ao i 2 7 oe 05;- ie w Za oe oy 00 o4ab we °. 03 |- yeneee erpinty IN TO EY oO | | | | 0 2 4 6 8 1.0 07 2 fo. 12 EXIT QUALITY - xe FIGURE A.4 COMPONENTS OF MULTIPLIER UNCERTAINTY FOR A TYPICAL DIABATIC CASE.Appendix B DATA REDUCING PROGRAMS B.l The Programs Pressure drop data presented in the literature is found in different formats and expressed in different systems of units. To facilitate evaluation of correlations and models, the data was reduced to a common form for adiabatic and diabatic data. A basic program to reduce data was written for each of these two types of flows. Program statements were altered to apply the appropriate unit conversions, convert gradients to pressure drops, mass flow rate to mass velocity, and enter the data un- certainty as required hy each particular data set. In adiabatic data the quality may be expressed as a mean or else the inlet and outlet values are given. For subcooled diabatic data the subcooling may be indicated by the temperature, the specific enthalpy, or by a negative quality. These were accounted for by adjusting the program for each data source. The output data of the data reduction programs is the input to the correlation evaluation program. The English system of units is used for the output. The output for adiabatic data identifies the data set and point, gives the geometry and flow conditions and expresses the pressure drop as a two-phase friction multiplier, as aefined by equation (2.18). The uncertainty in the 126127 multiplier is also given. For diabatic data the output is similar, except that inlet and outlet qualities, the heat flux and an average multiplier is given. The main program reads the number of data sets, the uncertainty intervals of the independent variables, the number of points in the set and geometry. It then reads the data points expressed as a pressure, mass velocity, quality and pressure drop. Unit conversion is handled by the main program, It calls subroutine PHI to calculate the two phase multiplier and the terms required to calcualte the uncertainty interval by equation (A.6). Finally it punches the output and proceeds to the next data point. Subroutine PHT calculates the two-phase friction multipliers. It calculates densities and liquid viscosities by the method used in the reactor code HAMBO [10]. The acceleration and static head pressure drops are calculated and subtracted from the total pressure drop. Subroutine FRICT gives the friction factor which in the case of the sample program is the smooth tube, liquid only, friction factor. This value is used in subroutine PHI to calcualte the two-phase multiplier assuming the entire flow is liquid. For diabatic data the program is essentially the same as for diabatic data. The average static head loss is calculated by integration as opposed to use of the mean of the inlet and outlet qualities and the resulting multi- plier is the quality average multiplier.128 In the following section a sample data reduction program is given. This program shows a typical input and gives the output. 2_A Sample Program A sample data reduction program is listed in this section. This program is used to reduce the data of reference 29. The sample data set consists of 27 points. Both the input and output are listed. For this example the homogeneous void fraction model and the smooth tube friction factor are used to reduce the data.129 TX+1IWND=TIX d0+d=1d 00%d130+d730=1d0 ‘94H/LO=€X (944/44 WD) edd = 2X (944 /4H+ WA) #OG=1x dx2ZZ*-"088=94H d#€91" +" 09€=5H @ea(G0+"1)/9=29 WO#9+9=19 GAHO * GG + GAHO=10 B31Td111NW NI ALNIV1839NN 31v1NIIVI (LOVIHd! 1X *0AHG'd1304d 43 IVND AT IWND SDH TIV2 vaTId111NW 3LvINdIWd 22° led 13G=4130 22" oled=d *ELEL9=9 *2/(11WwAG+3 ND) =1VND d130‘d*3WwNO*lIwnD+9 (eOT*s)avau SANTOd viv NI Gv3u N¢T=7 2 OG 476° Z/QAHO=GAHG (82 bees" 2) /1X27K @oo*9/vauv=vaUy NOISH3ANO9 4 INN AX *V3NV SGAHG *N44NO91 (ZO *S)0V38 A¥LSWO39 GNY SINIOd 40 ¥3SWNN NI GV3¥ si3SI* =x 2 00 16‘DU ‘ddd *dd*aa ‘Wa eoT*s)ava" S31@VIwVA LNSON3d30NI YO4 S3TANIVIY3INN NI Vay S13S1 (1Ol*s)avau S13S Viv 4O y3aWnN NI Ovay *WV¥9OUd NOTLVAIVAR NOTLV13¥¥O9 3H1 OL ANdNI YOS SOYVI G3HINNd SI LNdiNO 3HL *SY3T Id] 11 NOTL1NS 3SVHd-OML OL €8-¥-3S19 NI N3ATO SV VIVG dO¥G 3YNSSIYd SLYZANOD WYYIOYd STHL ° eouee o130 ana (OTOTS9)LVWYOS OT (OTS4E*ET*ZILVWYOS = ZOT (€1)ivayOs TOT IS "OTHE 41 "84 *T OT 44S *odz “SIE VAVWEOS OT IHdG* LOVIHd* Twn 4d *9*V3¥VGAHGS INODI 414 (6OT4L)SLT UM L 4NdinO HINNd 3 Gta (Ze (1 IVIHd-91Hd 424% ( LOVIHd-SI Hd) +244 ( LOVIHd—9IHd )4T @ax(LIVIHd-€ Id ) +2%% (1 VIHd-2 Hd 14244 (1 9VIHd-TIHd) )=1Hd0 (91Hd*1xX*0AHO 44730 ta “eax*e1x* OVIHd 11V2 (SIHd*1X4 —Ta4d73G 4d *3TWND FT IWNO*ZOVIHd VIVI (y1Hd 1X4 0AHO4d 130 4d * 23x*ZIX*TONIHd WWI (ETHd #1X*GAHO 4d 130 4d 4 TAX TIX*9)IHd TW (Z1Hd $1X GAO +d 730 1d 43 WD 41 VAD “OVINE VWI (TIHd # 1X *GAHG ‘Td 4d #3 WNO ST WWNO49DIHd 119 2x-31WNO=Z3x (1°39°Z3X) dT 2x-11WNO=ZIX (4 1°399°ZIXIAT Zx#3 WAD=Z3X exe qwnd=z1x €X-FIWND=€3X (71739 EINDAT EX-1WAO=EIX CT TTEUT EI XD AT Ex#3 WND=E IX EX4#11VND=EIX Tx-3WOD=13x (7 1939" T3X) ST TX-11VMO=1IX (91939 TIXDST 1X#3 WND=1 3x*2/CIveayd =v (Vx86C36 919°) /*Z449% ( SOHU/ CCT V-8 19/2 288 (1 WNO-* TT (30-91) /* 2a (3 TWNO- 9) 4OGHE/ CLV /* 21 WO) -(3V/°2"*3 TNO) =Vda ((31vN0/(3 WO" 1) (SOHN /90HU D4? TD /* 1230 CCDWAO/ C1 WO~* 1) (40HE/90H8 149 TD 72 T=1¥ 4130-d=3d (ANWX#* ZT) /OAHO#9=0439 TT ((H-*06 1452" +H) /*81 14800" =4NWX TT OL 09 (*°06°39°H) AT A/*8114800°=4NWX 02 8° 694Ns( Z0aZEEE* +2 Me (TO3EZZZ* +N LLEZ* HN ( 1C-3ECL 9" +Mx( 10-38Z99°-N#(ZOIGGET “+1 Nx (19-3402 T* +e ( Z0-JEEES* Ne (€O~-3Z948" +e 4O-31LL4*-PPIDIIDIIZH OT oz O1 09 Z°LGS +Nx(€OIBO9T* +e (ZO3GHLET+T Ne (ZOIBET* +a ZOIGOT* +Mel TOTS 41" +N ( 1OF6HL I" +N¥ELBS*~1) 990) =H *2-n=n 41 OL 09 (*S92°31°d) AT (d90IW=N Zz Ad/d=90HY A/*1=4OHY 2 EDIVELES +e ZOI9ZET * #N#( 1034¥OT T° +e ( 10-39% +e (20-3 ITED "=1 M#( 2O-3TLIE*~N¥(€9-3Z2L9* +x (€O~31 OZT* -MeSO-3981°-19)99)) =Ad TO-39 191 * +x (€0-F6EOT * +x ( 40-FE TEE Ne ( SO-394L9° 44 (50-36402 °-T Nx (90-3569 * -Ne (90-3696E" +x (20-31 41° 80-3894) DIIDII=A ET 21 01 09 €03%4° +e l Z03TS19*—Ne(ZO3TOES*-T Mie ( ZO3LELZ* Ne ( ZOFEHS?*— Nx ( TOATOS9* —N* 10994149) 2) D=Ad 10-39612* #Nx(Z0-3L 409" +e (Z0-3HLOT° #1 Me (ZO-aZOTT* +e 20-352 12" +e (ZO-3LZ4T * +N¥EO-3BEIT*— ID =A *L-nen €I OL 09 (*OSH*37*dd ST (d)907V=n 22/(11WNd+3 NO d=TWND MIT IILINW NOILIIYS 3SVHd OM S3LVINIW 9 (19V THd* 1x *GAHG*d1304d437¥ND 41 WWNO49) THd 3NTANONANS132 (289% S#L0-39S15°I/6 ana NYNLaY QAHO*4SOHU) *( ZdO-Vd-1X/d 13904" 441 )=L9VIHd (944 4 NWX *GAHG*4OHY)LII8S TIV9 (¥-" TL) ed 0H¥+¥e90H¥=Zd0133 ana NunLay z 1 O41 09 4 (OZ*O3 TST WISI dd /d=4 2 O41 09 (S0000°*37°(44/4-4)S8V) dT J¥e9*~(3Y*OITY) O10 Westar’ b= 4d (Ad LUDS=48 t soo° =4 ISI C4OWX#* 21) /OAHO#9=04538 WOLIVA NOILITY4 JNA HLOOWS S3LVINIWI 2 (9¢4*4NWX*GAHO*SOHY)L9IYs 3NTLNONENS134 nCELnT 89Sus° LzLgs* 8096S" nLLe9* Soave: €eszn* €nnos* 9LLSS* eezes* Loseg* 6L9SL* 8SeLe* 95986" ensore oz nel" 8€ 89" nLene* nLbone neLzn* SLozs* SLLLO* B9LeL* S908L* Lnese* 6L286° Sholch zu ZezLe S9Z"LLE mete LoEtLL BzrLL nha @sstte Zante SbmrLe EomLe LmerLe 9BETLL Lentee €Z9° LL 6S9"LL bzerbe ones ohhh ene BmrLe meniLe Beetle BZELE MOETLL L9ZLL EOL 2meLe szo* Sas ViWd 40 ¥aawON duis? LogLs* 86499" ones. zsn6g° 48sso° bnoLes zosstt nzoz* zoeez* socez* Sonse* g0zznt ssen* 4695" 9ezzo* 89680" €66S1° n9zzz" nSLz* €LLse* nSeen* LLges* LUgLg* 497269" ssoe.* noLes* “Ss zuis* 6ouis* 82499" z9e6L* Leones? zZenso* 6L60L* S9ESL* n0Loz* Lezez* ZLeLz" Lzse* €L6n* 9zn6n* nLegs* 9zzzo° 68880" Loest* Lizz? enhlz" zzgse* eozen* 9nSEs* LLigt €nz69° n908L° Beles" s*49 is? bor stL2z eLttez L6°L2Z £678 Lz £S7 02 Lor Lee en Zoe. Bn L6e €S706€ €S*06E En" Z6€ ES*06€ LO Lee €s*06e €S* 068 LO" LGE SE*O0E us" Loe 46° LOE 46" LOE nO LOE LOT LIE 8s" O0€ 18° Oce L2° Loe Le*o0e L6° LCE L0BtLz Lb 900° L SLINO 1¥NIDINO NI SUNIOd YoEa SEN1Od dO UAGWON GNV AULaWOED +, SAIINIGLEGONO ONIGHOOaa Wa¥C 2a GUYS GHIBL AYO dNOIas qavo Lsala WaD0%d NOTLOOGaY Wid OL WL¥C GoaNT135 9484570 = LoLZ* OL ongL97do © SZL 86" OL O€€69-0 666LETZL oosnL70 §©— S9ZDET EL e€zse°0 = LBB9L "SL 9Lezi"0 = nheEL*Z 9SLBL°O © OS9LLTE zgez*0 «= ne6 ESTE BeEnz"O —EBZEZ*H L9ssz"o | ZLOS*h 6LLLz°0 = S6988"h Lnnee"c —99906°S €Lese"o oenge"o BZLEn"O = HOT HLL 4949870 © E8167" 8 989EL"0 © 9hELETL anzerrG = ELOLLTE N6ONZ*0 —-LLZHZ*h 9S647°0 6 E9E6* b066z"0 = LS98z"S 6089E"O = ZL6LS*9 N6OEN"O = SEZHOTL €9Lzs70 -Lez9z"6 €€9ss"0 gegse'6 80L09°0 9666L" OL nLiGL*O = HZSON*ZL B9EBL"O —76096°EL SHLLsto eeLes*o 8849970 SS86L"0 nLn68"0 6ESso"o ogearto SENSL*O zLL0z*0 n6ZEZ*O Be6Lz"0 Leese G OnLznro Ben6n"o 2z69S"0 9szzor0 6268070 Ln6St*0 L1zzz70 n60LZ"0 L99SE°0 LLEEnO Leses*o €6Lt9°0 S$z769°0 6S0BL°0 tzLes-c QreLoL fT ELOL S7mboL etn QTELOL E"SLOL S*LLOL 6°SLoL S*SLOL nrSLOL S*mLoL b'stor “tStoL S*@LoL or6toL Z*hLoL 9°9LoL or 9Lon 9°StoL mr OLGL e*SLoL b7SLOL ET mLOL 67 ELOL mele S*ZLOL O*LLOL OTeLLEESL orsteneot O7MBS9ESL OrOLt HSL orz9z9zoL 0° 9L6zeaz G*SeEe6az O*LeE geez Or LLeei ez O*LLE6LEz O"SeEeeez O-LLE GL Ez O-9L6ze8z O*LLE6LEz O° LLE6LEZ Or Hzngzzz OTLemmtzz O*EEDEZZZ ornzngzzz ornengzzz oregselzz Ornzngzzz O°9LL912z O°ZLBL1 22 Or EgzLZzz O*ZLeLizz Or nengz7z 4964070 4964070 1961070 £96L0°0 £96L0°0 4964070 L96L0°C £96L0°0 L96L0°0 £96L0°0 L96L0°O 496L0°0 £96L0°0 L96L0°C L96L0°C £96L0°C L96L0°G L96L0°0 £96L0°0 4964076 £96L0°0 4964070 £96L0°0 4964070 4964070 £96L0°0 £96400 Zeevero ao ZeLteto Lb oz ok ZLLie*o Lb st ob ZeLteso L nz ob Zeetero b e@ ob 2ectero b 7. ZLLLE7O L wok Zecte*o L oz ot Zeetero b 6b ok ZeLterG ek ook ZeLtero Lb abo ZLLLETO Lb gt ot ZeLLETG Lb shook ZLLLETO Lt mob ZLeLero b eb ob ZeLLErO Lb zoo Zeetero bk book ZLLLETO Lb oro o4 ZeLteso Lb 6 t ZeLLETO L 8 t ZLLLE‘O Lb L t ZLLLETO L 9 t ZLLLETG L s L ZLLLE*O L c) L ZLeLErG Lb € L ZeLtero L z L Zectero Lb L L ALNINLUZINA BALTEILTOW “(HING) BATTAL OW *ALITWOO DAY ‘EENSSAed ‘ALTIOIEA SSTH ‘Wau ‘ALAWVIG INATWAINOG *NOTLWEAOIANOD “BARWON LNIOd ‘UdaWON JES VLVG WOd SNWOTOD ‘GaHONOd SGYYD coaLnoAppendix C THE CORRELATION EVALUATION PROGRAMS ¢.1 The Program for Adiabatic Data This program computes the multipliers for the eighteen models and correlations and then compares them with the data. The difference value, ¢, is computed for each correlation. Cumulative mean values, R.M.S. values, and standard deviations of © are successively calculated for each point. The results for each data set and the entire data collection are printed at the end of each data set. The main program reads in tabular correlation and property matrices and the data points, calculates correla- tions and the evaluation measures and prints the output. Those correlations which have been presented in Chapter 4 in a closed form are calculated directly. For the Lockhart-Martinelli correlation $2, is computed using the approximation given as equation (4.25). The Martinelli- Nelson correlation is interpolated using the log of the pressure, as was done by Bowring in HAMBO [10], and a third order Lagrangian polynomial. Similar interpolation was used to find the multiplier for the Thom, and Baroczy correlations. ‘The Borishansky correlation was directly interpolated using third order Lagrangian polynomial. The 136137 Levy momentum exchange model void fraction computation is accomplished by an iterative method. The main program also compares each of the correlation results with the data and computes the mean, RMS and standard deviation values for the difference value. Subroutine PROP calculates the densities and viscosities. The saturated vapor viscosity is interpolated by a third order Lagrangian polynomial from the 1963 International Skeleton Table [40]. The remainder are cal- cualted as in the reactor code HAMBO. Subroutine TERP sets up two-dimensional correlation matrices for inter- polation using the log of one axis, such as pressure for the Thom and Martinelli-Nelson correlations and the property index for the Baroczy correlation. Function FLAGR is given in Carnahan et al (39] and was used to interpolate using Lagrangian polynomials. Subroutine FRICT iterates as solution to the smooth tube friction factor. Should this program be used to compare diabatic pressure drop data, it would have to be altered to calculate a mean multiplier for each correlation. The multiplier would be averaged over the quality range from inlet to outlet conditions. A sample program is given in the following chapter. ¢.2 A Sample Program A sample adiabatic correlation evaluation program is listed in this section. ‘This example evaluates the output of the previous chapter's sample program. All ofthe tabular correlation and property data is listed. The data points are not listed since they are given in the previous section.Veubss ef UG *O=ZIHdS9 *O=Z1HdS *O=1HdS9 *O=1HdS O=NN O=N T=WW 8L=9N S31GVIYVA 3ZIIVILINI 9 (02 IOTS#(0Z)9ISd*( TT )O8 *C11) 08x (S*OT* 8) avOIE “COLI DEX! (STIX! (8IDIGA! (TT Td 4(ST#TTIGTHd *(OZ)8Ya £5) 89Z *(0Z)A90 4 (02) dday! (OZ) YYBNX * (OZ INNS *(0Z)DSNST * CET OINWIHd $C GNA *CETINHO SC ETINWX NOISN3AIG (02) YYIYD 4 (OZ 1 BBX D* (OZ) WNS9*(0Z IOSNS94(0Z1AFGD NOI SN3WIG CETILX#(ETS9ILIHdS(9)1d NOTSN3WIG (OZ)IHd NOTSN3NIG (62)L 994 (621 9NWX NOWWOD ¥3HL3901 G3dWN7 38 4SNW 13S ViVG HOV} YOI SGI 2 Quv) WNV1@ 1@ G3MO710I 38 ISMN GUVD Viva 1SV¥7 9 ZO ANSW34VIS 33S LVWYDS INdNI 9 S31@VIYVA INdNI WV¥9O%d NOISYIANOD vivd WOU4 VivO *VLVG QILVEVIGV YO4 WV8DOUd NOLLVNIWA NOI LV 1a¥Y0d YaTTdILINW BOs ZONVE ALNIVLYRINA THAD 2 W3ITdIL IW NOILIIYS 3SVHd-OML AIND GIND1T L9VIHd 9 aLTqwn — 1WwAd 2 VISd JUNSSaud 4 2 ZxxL4-YH/WIT ALTIOIIA SSVW 9 2 @eeNI VaUV | VU 2 NI -¥3L3WVIG IN3TWAINYS = DAH 2 G3¥IS30 SV S¥3HLO 2 SMINNNY 2 2 NOLLvaNdT4NOI 4NOIT 2 YaSWAN INIOd 1 2 ¥aQNNN LASVIVG w 2 2 2 21)A309, (CN) AVO14/ CT WAS= (1) YBNX ( ee ((NDLVOTS/ CI) WAS )-(NDLVOTS/ (1 )OSNS) LOS=(1)A30 ON‘I=1 LT OG 4 OL OD (W*03°WHDAT ANTYd GNV SYALSNVEVd ONTILVY 3LVNIVAZ N3HL 13S VIVO 40 GN3 4I 93H) 9 (S°OTSE*T*84*T*O14*S*8424*SIE)IvWYOd ZOT THdd #1 VIHA # Wd 4449 ¢vauyGAHO44NOIT 44H Z0T#S) Jvay € viva NI Gvay 9 T4NN=NN T4N=N 2 € 01 09 (O*S4STT)LVWYOS LOT T)084)4 (114 T=14¢1)08x) (LOT 4S) OVaE (11°S38)/T° OTSS/T* S4OT/T*S4B)LVAYOS SIT (S*T=y8 (OT *T=P4 (8 STH x80 4TYYOIIIIT *(S*T=n*(0189) (OTT =P 416) 98x) (84 T=14( 1) 91d (STL *S)OVae (CU SaT TD /T°SSST/T° SATII LVAYOS = YTT (ST4T=rt TEA THT (G4 TD GTHd DOCS TE T=P 8 (POX DE CLTAT HLT 1d 8) (VTL 4S OVE ((O"S4 69/0°S46)1VAYOS 90T (ETAT =P e(64T =I (PT NWIHGI) (6 *T=1 S(T) NWA) (90T*S) V3 ((0°S49)/0°S49) LV8Od SOT CELT =P (98 TST ACPA TD LTHd) 14094 T=14 (1) Ld) (5014S) 0V38 (CO*SSETILVAYOS OT (ELST =1 C1 NW) (HOT *S)0V3E SS3ITYLVW Y3TIdIL1Nw NOTLV13exOI SNOTYVA NI Ovay 2 5 (O*S49TILVNYOS = EOT (OZ *T=1 41 91S) *C024T=1 (1) 91 Sd) (EOT #9) VEN (6Z*T=1*( T)L9NWd) (62 *T=1 41) LOMWX)(EOT *S VEN SBIIMLVW NOISN3L 39V4YNS ANY ALISOISIA YOdVA NI ava¥ 9 *O=C1)OSAS = 2T *O=(1)WNS, *O=(1)0SNS9 *O=t1)aNso CUt‘t=rseme ee eee BSE ASE DONT G PE SUN OR STINA TS IVNOT TPP SAL PI Ad=( 7) 1d SZ" ee (* T-ANWX/ OW) TVNO+? T) (TD THd=(€) Hd SZ" ee ( (*T-ONWX/INWX De TNO? 1) / CT THd=( 2) Td (*T-90H¥/40HU)* TWNOF* T=(T) THe $13G0W SNO3N390WOH 2 (aniixe® 20) C100" 1 #94G AH) = 430 (OMnxX#* ZT) / CWA I*DAHC) =939N (ONWX *4NWX *4OHY*OCHY dd dO¥d 11V9 3NNILNOD dOLS (OOA*HADST WeWW (T)OSNS IT *O=(1IWNS oz‘t=1 tt og 0: *O=ZIHdS *O=1HdS ((S* ETSE £61 $X0S) /5*64Z*61Z*XOT) AVAYOS TOT a ON*T=T (CT) A909 4 (1) BHAUD* C1 YHIXOST) $Y3dXYO* IGN ANNE WW (TOTO ISL TUM (70A3D G4LS MOUNT SHE BOUND NWe*x8T 4 uz a3 BOUT *XZE/sNOTLVIFYYOI NOLLVISYZOI NOTAVI9YYOI NOT LV 13T BBCI SWYWEV NW VIVO SINIOd SLAS VLVG eo *XOT/////// a1 e)AVWYOS OTOT COTOT#9) 311 aM CONS T=14 (1)N304 (1) MaI YS CL MYINK #1) 4Y9dXY!VTGX §NAWW TOT 9 )SLT UK (/ea30 Gs YOUNT SHY WOUNS NWe XBT edz owas WOUITe *XZE/eNOLLVI9YYOD NOTLVISYOI NOILV13¥¥OI NOTLVI3T WHOD SWY VAVG NW ViVG SINIOd 13S VIVO «*XOT///////el es) LVHYOS BOT (80149311 UM (NI AWO14/ Id S=¥939 dX (( NDAVOIS/ZTHdS)1YDS=839d XY CONNDAVO14/Z1H4dS9) LYOS=¥39dXYD (NN)1LVO14/ IHdS9=¥39dX9. COND 4VOTS/CIOSNSDLUOS=CT) NTN LT (ONNDLVO74/(1)OSNSO)LYOS=C1 BUI YD.142 s9 O1 09 *2I=19 940° =99 wT "9 *T=D €9 99*€9*E9(* 0002-934 19. 29*T9*T9(*000Z-43¥)51 09 NOILV13¥YOI 1773NILYVW-LYVHNIOT 9 SL* Tea ((VL39-° T/C 1WND-* 1) (G00 * +d*92T000°) =(9) THd (6° 719° V1398) 41 (d¥701090°46°T Dex (VHA TW-"T)/(S2* Tex ( WNO-" T)e 8% )=(9)IHd 6S 09 OL o9 2 Tae (VHdW-"TI/(SL" Leet TWNO-"T))=(9)IHd 8s 6S*8S*8S(G*—VHd TW) dT VL3Gx((ZZ*YT/dIOTIOWsesO°+EER")=VHd IV 9S NOIAV73NHOD AZHISAYL-GNVWYY 9 99° Lee ( VHAW—" 1) /SL* Lee (WWNO=* Te EL T=(SDIHd = LS 9s 01 09 2° Zee (VHA TW" 1/52" Lee (WVNO-" i beBLYe=IGDIHd SS 9S O14 09 Zo" Tee (VHd W-" TI /SL° Tee (WWND-"T=(S)IHd 9S SS *S *¥S (59° -WHd Wd 4S OL 09 (6°*39°VHdTWIdT ES (Oro *T4(WL 38-91 /VL 3G eZ 94° SIA Ze —HT? L4* 9)-* TeV G *#&( SOHY/OOHUDESZE* #44 IUS IIe Y=Zd CCUWNO-* 1) #VaaVeI"yZLOOO® +°4)a(VL3E-"1)469°=19 25 €$ 01 09 VLIGeees*=vddIV TS 25*1S*1S(6*-Vi3a) JI (1 T+ TWAD/(WWNO-* 1) «(40H / 9048) 1/* T= W138143 O=4NON *O=1VV *o=TIvN0X lvnd=vv 88 TAGOW 3ONVHIXA WNLNSWOW AAZT 2 (9/d&*08Z49/*0006T1+d%4000"-92"T)#(BIIHG=(OTIIHd LB 88 O1 09 (d#9%60-391L "9490-31 * +deG000"+9E* Tet GIIHd=(OTIIHd © 98 28*98498(*000001-9)41 NO11V73¥Y99 S3NOP-NOST3N-1773NILYYW OD SL" Lee WAD =" T)eSL" Leet ((90HY/SOHY—* Te TVNO-* Te SL ax ( ( JOHB/9OHY~" TeV W-* 1) =(6 0TH CCF T1007" 1) (4OHY /90HU 4° 11 /19=VHd dxT000"+TL°=19 1300W 440%NVa 3 CONSINE (811d ‘NWTHd *1¥NOSNND 4d *NHdIdYdL TV €1=tn 6=IN CUINWX=(T NWO 1 €lti=I T og NOI4V13Y4OI NOST3N-I773NILYvW 2 SL* Tee (TWND~* Te (82 HRW IX/* THWIX/T 94° =(L) THe CZ aad *T-CVVNO/* 1) De (4CHY/9OHY) # (99/39) ((OeedTY) /(44#99Y)DLYOS=WNIX SO *Oz=19 940°=99 990°=49 atau z=0 19 $9 O01 09 OT=19 *91=99 990° =49 ao zd 99 29499*99(*000z-934) 4I 29144 ZLU*2L1*eLt*O0VD00E-9)41 LT TAT TZ SOLT4* 0000SZ-9) JI 1=a CONSING COT) IHd BIHd* WNOaxX*11da*1de) d¥al TIW9 ST=tN TIsIN 2 ex (OMWX/ SNWX) /(QOHY/4OHY)=11 dd NOTLVI3¥xOD AZIOUNVa 2 CONSTNECETIIHASLIHd TAD? IX4d44d) d¥dL TIWI €I=tNn 9=IN HL)NWK=C1) AX s €t't=1 $ OG NOILV73¥¥09 WOHL 9 GL° Tae WNOx( 90H /4OHY) SZ" #4 (INWX/OANX) #DIHD=(ZTITHd = 26 GE *#RDITU/LIS" #e4dURGZYE=DIHd = T6 26 OL 09 GZL° ee4IVHTL**#9TNHIO-3SBB*E=DIHd — 06 16406 *06(* 000661 T-10¢* ¥#( 434) ¥99N) dT NOILV13¥xOI 37381 GNY N3IHD O04-37S 9 Zaa(VW-*1LI/SL°Tee(TNO-"TI=(TT IHG = 08 €g O1 09 ‘tev 18 Te*ee*est tI -v)al (10 xX-TWND )#3d01S+V VEN WO x= TIvNOX Wel (1TvNOx-TWNOX)/ CIvW-vW)=3d07S zB Ze*08*0B(100*-CIvndx-TwnbISavdal = SB 08 01 09 "O=VV 48 78*S8*S8(TZ-INON) JT 1#1NON=LNON Cvvae 22 a" TeV t Zee VV" TD (SOHU/ SONY DTZ) / CCC CVU Re 2-9 DVT OV DT #(OOHY/ SOHY ITZ HVVS Zea (V VEZ" TDD LUOSRV UEC VUE Z~"T)eVYD=TYNOX = EB145 T#NIW=NIW YLT HLT AOL T SEL TUCNIWIOGX=TWNO) ST SLT €=NIW TI=@N €=9301 . NOILV13¥YOI ANSNVHSI¥OG 9 96" ##(d/WND)#* OOGZE4* T= ST) IHd NOI1v13¥¥09 ¥aNI3e 2 ZYYOI*( YTV EHd=( HT) THA ZAYYOI# ( ZAVYOI-TAYVYOI) 1 INWZ=ZYIOI OX BYOI+ OXYYOI-EXYIOD DL INWA=ZABYOD 2KYYOI+(ZXYYOI-TXYYOI) el MWA=TAYYOD (1-44 1-248 1-0) a0 FC (T= WE TZ 4 TTA) VON CT HE T=Z HOTA) OD DL NW X= XVOD (TW 2H T= THI BAO D+ CCT ZH TTD OI (T= ZH TH) YYOD DL TWX =E XYIOD (MTZ HT 1) BHO I+ COME TZ 4 T- D9) HYOI~ O44 1-24 THD MOD) wh TW X=ZXYIOI CAE SHET HTD BROIE CONE SH*T THI YAOI~C HZ THD YYOD) LTA X= TXBIOI, (1-89-0109 (1-9) 89-9) =170WZ (1-24 98X=€2NI9GXI/ CC T-Z9) X= WN I=L INNA (CCT-18991d8)901V-T (CTA) 91d9)90W I/ (L CT-T9Td8 90 W-C11dGIIOW) =LIAWX =—-9L. 82 O1 09 T+2n=24 Lk LLALL9LC(Z9V9GX-WNDIAT BL sty EL st O41 09 T4TNs Tok heh EL (CTNIDIAG-11d)4T SL TsId OL zz 01 09 Tess TL T2tte4oL COGI-9) 3D 2k 02 o1 09 S=y 221 04 01 09 z= Olt146 4yS€0°=¥X 9° T=NX €S1 OL 09 (2°3N°4NOIIDST 7° T=NX 280°=)X (SN41 49301 *vd 91S 491 Sd) uOV Td =VAdIS @-1=1 @T=1(61°99° INST OST zst 01 09 isl IST ISL*ISTOSTC(1)9ISd-vd) 41 25T €0S*4I/d=vd Oz=SN €=9301 €=1 NO11V73¥H09 *3°S°I°9 9 (SL* Tee 1VNO04SL8 "4 ( CTVNO-* TD TWD Dad De (91 -Zee VHD 4° T=( LTD TH (G*#DeZ ee VHWVSI/°00SL0% =8 (*8Z°99"VWHVO)ST BT WHNVO/*1Z=8 98T 81 O01 09 (VHHV9eS *489)/°EZTHT=9 SBT 9eT*S8T*set(*0092%%-OI41 TST “81 OL 09 S*#e9/*%6¥T=8 (903%°1°99°9) ST 9/903LL°T=8 EBT +81 01 09 Bto=8 Z8T eet 'zet*ze1("00069€-9)41 O8T T8T +081 *08T(S*6-VAWVD) dT GZ * #4 ( INWX /ONWX)#S*¥e( D0HY /4OHU)= VAAVD. NOI1V 734409 WIOHSIHD 2 14 (7 T= (90HU/4OHY) #52 8 ANWX/ONWK DD ¥BI=(9T) TH (@N*NIW*93GI * WO *O84 *O8x) YOVI4=84 2-NIW=NIW OI=NIW (OT*S9°NIW)SI ELT Sit O1 09147 ana z 01 09 (8839s O89 + OOSAS=O10SNS C00 889+ OUWNS= OWNS (0) 883e ON BYa+ OOSNSI=()0SNSI ON 883# OUWASI=O NSD SQWIHd/( LOVIHd-Q0) Hd = Onea2 IN‘I=¥ 9T OG @ee1Hddd+Z1HdS9=Z1HdS9 IHdQd+IHdS9=IHdS9. @ee1Hd0d+ Z1HdS=ZIHdS THddd#IHdS=IHdS ADWIHd/1HdG=1HdGd or BALIGIAING VIVO HLIM 3¥vdWOD 9 WIdd/YYOIdG=( BT)TH (LI*2€%* 9 Te GAHO#IOHY) /( Ze ( °009E/9 Dede" Z)=9 Idd (943 440AX4OAHGS4OHY)L91US T1V9 2° Tae ( GAHO8YS °2 1/9" ee VWOIS#98* ee (1° T-9OHY/4CHUIT TNO + T)# (4OHB/Z7° 29) DENXHe( TELEL/OD#(OHTLZ/YXI=YYOIdD EST148 € 01 09 T+1sI 1 T*T#2(¢1)19Nwd-Wd) ST € €-1 ez=r €=9301 €0S*4I/d=Vd IT ((H=906 DeSZ*° 4H) /° 8114800" =3OWX TI OL 09 (*06°39°H) ST H/7@T1+800°=4NWX OZ 8° 694Nx( ZO3ZECET+Z Mx (TOJEZSZZ* +e LL8Z* HM ( 10 -FEOTY* +e (1 0-38299* -e (ZORERET "HT Nx (10-3402 T* +M¥( Z0-I6EES* —N#( €O-3Z99B* +e O-FTLL IID EH Z oz o1 09 Z°LSS +Ma( £OISO9T* +Ne(ZOIYLE*+T Me (ZOIEGET* +Mx( ZOIGOT *+M#l TORS T¥L* +e ( LO364TT* +e ELBS°~1 0) =H tu-N=n 2 OL 09 (*S92°37°d) AT (4) 90 = Ad/d=90HY A/*t=40HY2T EOIIEEE™ Ha (ZOIIZET * +x ( TOBVOT I * Mx ( 10-39" +e (Z0-3TTE9* -T Mal 20-311 0€* —Me( €O-3ZZ2L9* +H €O-3LOZT*—M¥SO-3981°-109)) D=Ad TO-3vI9T * +x (€O-F6EOT * +Ne ( ¥O-JE TEE +x (SO-397L9° +e (S0-36407*-T Nx (90-356 9€ *-Mx (90-3696E "4 (LO-4L4L°—NeBO-389H" IDIIIIZA 21 OL 09 €03%Y* +Me( TOATST9* —N#( 203 TOES *— Ne (20 3LOL2°—N4 (20 3E 922 *—Me (T03T6S9*—MeTO 39%L4* 9) =Ad 10-39612° +Nx( 20-32 40%" +N (Z0-39L61 * +1 Me (Z0-32OT 1 +e ( Z0-3SZTZ* Me ( ZO-ILSHT * + M#EO-IBEIZT™— IDI =A . *4-n=n 6 OL 09 (*0Sy*3aT"d) ST (d)90TV=N (6z)19NWd* (6z LONWX NOWdOD SSILISOISIA GNY SAI4ISN3G S3ivINIW) 9 (ONWX! 4WX* SOHYSOOHY4d) dod JNIUNOYENS \ —.149 ona NunLay *0090901/6*81%Z*9NWX=9NWX (P41 49301 *¥d*19NWX* LONWd YOVTS=ONWX @-1=1 @z=1(ez"39°1) 41o-CN=tr BT eLtet*et(r—Z-eNIdT oT st 01 09 t2rr £1 oLETEL(e-PdsT 27 @-Ie1l Lt 21 01 09 o-IN=II OT LUSQTAOT(1-Z-INDAT UT et O1 09 T=Il OT TL OT *Otle-1) dT T=NIW €=9301 42 (x 901W=1x (ON XK 190 W200 Vx 8 IN‘i=x 8 0G 22 2 01 09 X=IX OUXXSO0VX EZ IN‘I=¥ €z OG 2Z OL O9 CHTt3TAIST © 0Z 22 01 09 OZ O1 OD (ST*O3°PNDAT z 9 01 09 Teese S4S*Z ((PIAA-ADST 1st % O01 09 I+1=1 € EOL CCLIXX-XDST ” I (y OIHd* (4 THA Cy IETHd #1 VIHAT “CDVBTHd (708A 407) 9X4 COZIVXECPNFINITHd 4 (PNDAASCINDXX NOTSNAWIG NOLLVIOdYSLNI YOd SIDTYLVW BAI IdT41MW dN SLaS 9 (ENS INGE THA STHd *ASAASXEXX) dda. 3NTLNOVENS aon151 ana NUnL3y (VI‘NIN‘93GT* ACBIHd*GADYOVIS= — ETHd (VI SNIWS93GI*TX*G1Hd *8X) YOVIS=( 9) 3TH (VI ‘NIWS9301 1X *9THd 48x) WOU IS=( E) STH (VI'NIWS93GT Sx *@1Hd $8x )YOVI14=( 2) aTHd (VIINIWSO3GI 1X *VIHd 8x) YOVI4= (1) 3THd Ceser *t-v1 el DTHd=(¥ 1) QTHd (2400 T-VIOTT IHd=(VI) 9TH (1400 ST -VI41 TD THd=(V1)8T Hd (PP ST-WIF TT THd=(V 1) SHA CIV 1400) ADEA C1-VIFT TD vx=(V1) 8X o*tsvi L0G z-r=er st 01 09 st 6t152 ona Nunday ASBA=¥9OVTS WHIL+1S3A=1S 9A COPD X-CLIXD/WUSLEWYIL CPTINTTD ST XVWSNIWEP & OG CUD X-98VX) 801 9V44 1 AEHUL XVW ‘NIW=I S OG *O=1S3A (CP) X-98WxX)# BOLIVI=BOLIVS NunLay (Py AsvovId Z OL 09 (UP dXTENTOWXDST XVWNIW=E Z OG ‘9301 +NIN=XVH *T=¥0L2v4 (NDASCNDX NOT SNAWIG (N *NIW*93G1 49N¥xX*ASX) BOVIS NOTLONNS153 ana NUN ay z 1 O1 09 *0=4 (0Z*03°1) 41 T4151 j4/a=4 2 OL G9 (S0000"* 31° (44/4-4) SAV) ST Ue (3840438) 0190 Waddie’ b= 5d Cad LaDdS=3u t soo*=4 1 (ANWX«*Z1)/GAHO#9=0434 wOLIV4S NOIAITYS FNL HLOOWS S31VINIIWI 9 (943 *4NWX*GAHO44OHY)L91¥S INTLNOVANS154 TL Betz *L serz Lh Eze tL Lez tL Leth gt TL egtL Zore TL eoth zz TL EStL ebtZ szte TL Eth SLTL Gntz TL cLTL Emth Such TL SOTL LTL Zh Sere Thoth th te tk en Anne amma *90ZE*99£"C0SZ*9007" 0951" *L sntcng9* OEE SL tL €etz TL Bee L €0° “h gete tL OELTL Leste tL ontk “e nzee ork vorL Lo zore . a chook ete tan 6°@Z gtan nTLZ 9th nteL tn TLL OTnE norh Z6e Orbe t€z gts Z79L ns 678 gre €°¢ otk oath thotk 0901 *90¢ GO°9 HEL Shes n°Zb ge" hel “OL Ler 6n°6 69re £0°8 ZLte 6579 LStZ LETS zorz “Lee BH°L GETZ bzth Zork thoouth tht "90ZE" DOE *OULZ*OSzL “Let gt Ltt Gt ot ete ite sor 7G Lh* €0°% LTE LLTS SLTL GEMLLL* OLLOT GLH’ LZZB°EZEL*ITYS "BZ _GTCEES*SESC"OMOS “9MBL*AGZ“LZZS “OLZL*9ELNSOL LT ONLO"BZL6°TLLE9"° C676 "SASH" RLEL“N9SO"SSHE“IhBL"GEBH"EERS*ZLBO"GLSS*ZL SL“SELO*L TLS *OLZL"SBLH "SILL" OHLY"BTLO"ZLLO9"B6Z6 “SASH” HL SL“H9GL *GSH6“ONLL"GEBH EERE LZ Z°EZBOTOLSS"SLSS*ZLEU TOL ZE"L BLO OL"HHLO-ELCL*Z SOG*LEEMLELTTL Lh S"OESH*OZSL*HZ L°ZZ L*LZGO"GZ ETOZEL*OLLZ*OLML*BLLE “SL OLE ES*LEDLLLBL*OLZH*OLLO“SLZL*SLLE“SLZU"SLOI*HL ETMLHETELLSTEL Z*ELES*TLSH*ZL9O*ZL WIVG LOGNI NOTLVIZHAOD GNW ALYadCRa 2 Tobe toe 7912 "G89 766L *S29 THLE *SS "Shh 73S “SLL *OSE tes tShz 79S OSL ez 769 “Skt ce ste 9tg Thoth 700k L* ht £7898 "86 Sthe 9°98 L70€ 278k 6°9z 8789 zrez br 6s Stel béh G°Si 87 6E L‘Zb Z70E Zhe 9°07 Blin Leb 98° 6279 Ontk ZL°Z th otk "009 *osz tor to eetelis zt155 STL OZTL Heth HZTL ZL Zeth Guth out SZtL ETL LZTL GLTL Zh ZeTh bth OTL Thoth th tk Thoth th th ELTL €OTL BOTL EOTL ETL SLTL MLTL SETL TEL HZTL GZTL GZTL STL tmth GETL SETL StL tmth getL SETL Erk OTL Zbrh ezre mh LETL GLTL ECT mL ZUTL ZETL ObTL tLotk th tk "090009 "9090002 2000001 thet gt ont zt te "D00L *O0E "COL TIL “OOOL 7929S *9SZL"99L *IOLL9T99 *OS "QOL *@6E “ENT “LEL "SB TLS OTEN "GES "6LZ “SSL "BOL °69 ‘on Z°LE TOEE OGL TOLL "C8 ES BtLE NTEZ ‘one "6EL 798 "29 STEM BTLE "SZ tml 796 T£9 t6n n7sz ste "BOL 7OL St6n "BE Oo mse *OL 10S "9 RZ Sthl arab Zen SLE "6Z EZ £ ett 2°38 @*n€ 9°BZ e°2Z S*EL H°ZL BBB IL*9 erZZ E707 ETOL BTEL 976 9B°9 99°S e78 71'S BTL 98S Bth NTE HZ 9°S GZS 6th LHth ETE 6S°Z GEL GL°Z €LTZ BOTT BSL OSL ZETL LZTL Thoth th th th th th rr e* oz" sit 4* seo so" Seo *O0IL 99S "OSZL"I9L *OILL9T99 "0S “LGitt LE 6TS 9°8 uztL Leth zh sete Sere Zbrb 6O°L BOTL Thoth thootk Bork gore SLTL £ttk HEL BEL ETL 9Sth EntL 9Sth LETL Once 6zth EZtL @erb cute vhotk 000095 so* bor "OE OL eee OL "OE b°6 ‘oz *8 zLz sto EtLL Sts trek eth “LL ove 6th 9% bro g*z ith Sorz snr BLrL igre zzrb SSTL ZLTk Zee Hore th otk tot s90° etee TOL "GL TOE Enb th zeth tL 6orL tk goth tk hth th tk govk otk 6lh tk the k meth tk meek tk mStL tk StvL tk beth oth thot *o000sz toot to th tk eee tk ze tk 987% tk ehez tk szez tk eel tb eenk th est tk sth tk gevh tk 9th tL 907L tb Zork tk bork th Thoth Loot *0 cere tk “OEL "Sz156 L i ae ger Sut ast ist Lor “st ss* bee oh L zr est eu* nL? nee zu te e 1 L h oh ci tL ca tL oh Lh Lh oh gork SLtL zs* ce uL* of Of 16" 18t a° see ze" go* L9* cst rh v1 0 *L tL Lh of ap tL “L oh tL Lor rt s a e c s9° €or gg" ast 9° a a7 oh 6° te? Let sL* Set €or €L" a8" sy L “L tL oh 1 vl nh oh 1 1 L Lore bre a ou° not O° zor no ‘9° ns? ier a nh os° “Lt fet ne Lut ne LL 68° 1 L L nL L “L ai oh i *t L Lh gork ete ZL SETL LbTe LL 9eTL eZrL Eetb were 89° set z L e9° st ist 6s" 13° 69° esr 68° nL Lh “ee so* so" tee ng elt 6g° m6 cu 1 a 1 L oi vL L L 1 al av Bork SUTL 9zrb eer ‘" *L s* Lh tne 9s so ie ee* ne" cy L tee es? as* 89° bet el Ler 96° 1h L L as tL nL tL an vL 1 oh tL Lote “eth zere sctL 70 L ct a 1 L L nL a 1 L 1 L “L Lh L oy L a oi ce *L 1 1 ci 1 Lh L a tt L oh Oy a 1157 “LOEETLSLETLLSZ*LEML“L Ze6* GOB" 9E9° EhHe LZz*