UNIVERSITY OF MICHIGAN DEPARTMENT OF MECHANICAL ENGINEERING CAVITATION AND MULTIPHASE FLOW LABORATORY UMICH 014571-4-T STUDY OF LIQUID FILMS, FINGERS AND DROPLET MOTION FOR STEAM TURBINE BLADING EROSION PROBLEM by Wontaik Kim Approved by: F. G. Hammitt Financial Support provided by: National Science Foundation Grant No. ENG 75-2315 and University of Michigan (SEP) funds. September, 1978

ABSTRACT STUDY OF LIQUID, FILMS FINGERS AND DROPLET MOTION FOR STEAM TURBINE BLADING EROSION PROBLEM by Wontaik Kim Chairman; Frederick G. Hammitt Some phenomena leading to large steam turbine blading erosion have been investigated experimentally. Primarily these are the behavior of thin liquid films, and their subsequent break-up into droplets and fingers at the trailing edge of a stationary blade in the University of Michigan low-pressure wet-steam tunnel. Both high-speed (15000 Hz) cinematography and still photography have been used to visualize the pertinent phenomena and measure the important parameters. Also electrical conductivity microgages were used to measure liquid film thickness. A film "Transition Map" for various flow regimes has been constructed, based on the author's results. Three film flow regimes were delineated: rivulet, wavy-film and an entrainment regime. Critical conditions for rivulet flow were obtained experimentally. Comparison with existing theoretical models shows that the wetting contact angles required in these models are much smaller than those measured here and elsewhere for static conditions. "Bulges" at the film-rivulet stagnation point, and waviness in the film, have been

observed and may be responsible for discrepancies between the author's observations and earlier theoretical models. Film wave characteristics have been measured, showing some waviness always. These wavelets are either twodimensional or three-dimensional depending on the flow conditions. A shear perturbation instability mechanism for wave generation seems pertinent for the present thin film case, while pressure perturbation is pertinent for thick-film cases. The author's wave propagation speed measurements show that for two-dimensional waves, the classical capillarygravity wave propagation model is suitable, while for the high-subsonic waves Wurz's model(Wl4) correlates the results well. For the entrainment regime, a correlation has been found based on. the author's film thickness measurements. This correlation explains horizontal film flow data well (including the author's), but not vertical passage data. At the trailing edge, time-dependent liquid finger behavior has been observed. It is found that Karman vortex shedding in the steam is not a controlling factor, rather surface tension and other dynamic forces of the steam play more important roles for finger formation. The author's measured wake droplet trajectories show two.important phenomena which were not reported previously in the literature. These are periodic deceleration of wake droplets and break-up time-dependence upon droplet Weber number. The primary cause of periodic deceleration of drops

may be steair vortex-shedding from the blade into the wake, Direct measurements of such vortices were not attempted in this study, Droplet break-~up time dependence on inverse Weber number is explained well by classical wave growth theory.

STUDY OF LIQUID FILMS, FINGERS, AND DROPLET MOTION FOR STEAM TURBINE BLADING EROSION PROBLEM by Wontaik Kim A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Mechanical Engineering) in The University of Michigan 1978 Doctoral Committee; Professor Frederick G. Hammitt Chairman Professor Vedat S. Arpaci Professor James W. Daily Professor Jerzy R. Moszynski, University of Delaware Professor Charles M. Vest

ACKNOWLEDGEMENTS The author is very grateful for the support and assistance of many individuals towards the completion of this work. In particular, he would like to express his appreciation to Professor Frederick G. Hammitt for his careful and consistent advice during this thesis work; Dr. S. Krzeczkowski for his useful suggestions and assistance while he was here as a Visiting Scientist; Professor Vedat S. Arpaci for his many discussions and encouragement during the author's most difficult time in the graduate work; to Professors James W. Daily, Jerzy R. Moszynski and Charles M. Vest for their serving as a doctoral committee members. Thanks are also due to large number of student assistants for their time and effort in the test operation and data reduction, The assistance of Ms, Ruth Howard in the preparation of the thesis manuscript is most greatly appreciated. The author also gratefully acknowledges the financial support by the National Science Foundation, Grant ENG 752315, internal University of Michigan Funds (SEP) and Univers ity Fellowship. And finally the author extends his appreciation to his wife, Songmin, and daughter, Christina, who provided shelter and comfort throughout the entire graduate work, ii

Dedicated to my father Mr. Kyung-Suk Ma iii

TABLE OF CONTENTS Page ACKNOWLEDGEMENT....................................... ii DEDICATION....................................... iii LIST OF TABLES........,..........................S..vii LIST OF FIGURES.....................,* viii NOMENCLATURE xiv I. GENERAL PROBLEMS ASSOCIATED WITH STEAM WETNESS................. 1 A. Condensation of Steam............ B, Deposition of Primary Droplets on the Blade Surfaces 8 C. Film Flow and Secondary Drop Formation. 9 D. Velocity of Sound in Wet Steam........... 11 II. INTRODUCTION.15 A. Film Flow,...,.......... 16 1. Rivulet Flow Regime................... 20 2. Wavy Film Regime........... 22 3. Entrainment Regime....... 26 B. Liquid Fingers at Blade Trailing Edge and Secondary Droplet Formation....... 27 1. Formation of Liquid Fingers......... 8 28 2. Acceleration of Secondary Droplets.,..* 29 3. Break-up of Droplets................. 30 III, EXPERIMENTAL FACILITY AND MEASUREMENT CONDITIONS 32 A. Steam Tunnel General 32 B, Blading Profile and Electrical Conductance Microgages...................... iv

Page C, Photographic Facility................... 41 D. Experimental Measurement Conditions......43 IV. BEHAVIOR OF LIQUID FILM.............. 45 A, Visual Observation of Wavy Film.........4 B. Measurement of Film Thickness and Estimate of Interfacial Shear Stress..............77 C. Rivulet Film Flow Regime................. 97 D. Wavy Film Regime.........................113 E. Film Entrainment..........................146 V. FORMATION OF LIQUID FINGER AND SECONDARY DROPLETS A. Characteristics of Finger Formation.......160 1. Visual Observation....................160 2. Reduction of Experimental Photographic Results............................. 182 B. SECONDARY DROPLET MOTION.................. 210 1. Droplet Acceleration............... 210 2. Observed Droplet Decelerations........218 3. Disintegration Time for Droplets..... 225 VI. CONCLUSIONS............................ 236 REFERENCES......................2...241 APPENDICES A. Design, Calibration and Generalized Characterics of Electrical Conductane Film Thickness Gages............................ 256 B. Calculation of Normal and Tangential Perturbation Stresses Upon Wavy Film............ 265 C. Computer Program for Data Reduction on Acceleration and Break-up Time of Droplets 273 V

Page D. Statistical Linear Regression.........2.... 293 E. Comparison of Experimental Droplet Acclerations With Those Using Serafini, Abraham and Churchill's Empirical Correlations for the Drag Coefficients, CD.................... 341 F. Computer Program and Output of Droplet Break-up Time Data Reduction............... 359 G. Wave Growth Theory on Liquid Droplet Surface. 369 vi

LIST OF TABLES Table Page 3.1 Thermodynamic Operating Conditions of University of Michigan Steam Tunnel Facility..,........ 35 3.2 Experimental Parameters.............. 43 4.1 List of Flow Conditions and Oscillosgram Picture Set-up for Plates in Fig, 4.1............... 46 4.2 Estimate of Interfacial Friction Factor, Equivalent Sand Roughness and Shear Stress. 90 4.3 Comparison of Average Measured Film Thickness and Predicted Film Thickness (Smooth Wall Assumption... 4.4 Criteria of Rivulet Formation (Comparison of Author's Data with M-M and H-M Force Balance Models)........................................105 4.5 Comparison of Author's Data with H-M Minimum Power Model....................... 110 4.6 List of Wave Characteristic Parameters (TwoDimensional Waves)....... ~'... 120 4.7 List of Wave Characteristic Parameters (ThreeDimensional Waves).................'...'~*...... 130 4.8 Comparison of Critical Weber Number for Film Entrainment (Woodmansee-Hanratty(W2), Van Rossum(V2) and Kim-Hammitt)....................57 5.1 List of Flow Conditions for Plates in Fig. 5.1 — 161 5.2 Characteristic Parameters at the Trailing Edge...208 5.3 Comparison of Experimental Am and Serafin As Droplet Accelerations, R-Time Average Acceleration Ratio of Am to As.- All Values are Average Values for the Conditions Shown.................... 212 5.4. Characteristic Parameters Concerning the Periodic Deceleration of Droplet in the Wake Region.. 224 A,l Error of Thickness Gage Calibration. 24 264 vii

LIST OF FIGURES Figure Page 1.1 Temperature-Entropy Diagrams of Conventional Steam Turbine Cycles. 1.2 Efficiency Loss Due to Moisture From Laboratory Turbines 411.........................2 1.3 Diagram of Phenomena Leading to Steam Turbine Erosion.............................. 5 1.4 Diagram of the Forms of Water in Wet-Steam Turbine Stage.......................... 1.5 Approximate Fractional Deposition of Fog Drops on Blading (C2).............................10 1.6 Vector Diagram of Steam Flow in Turbine.......12 1.7 Notching of Moving Blade at Inlet Edge by Water CC19..1............................ 3 1.8 Acoustic Velocity Ratio, a/aF, of Steam vs. Wave Propagation Angular Frequency, C2 (P2) Including the Effects of Drop Size, Moisture Content and Pressure.......................14 2.1 Categorization (Family) of Liquid Film Flow Studies.......................................18 2.2 Liquid Film Measurement Data by Several Investigators (K10)...........................O19 2.3 Schematic of Velocity and Pressure Changes in Flow over Wavy Surface.......................23 3.1 Facility Schematic....................... 33 3.2 Photograph of the Overall Steam Tunnel Facility....................................34 3.3 Schematic of Blade with Electrical Conductivity Microgages................................... 39 3.4 Schematic of Characteristic Wayy Film Parameters....................... 44 3,5 Schematic of Characteristic Parameters for Finger Formation and Secondary Droplet Motion 44 4.1 Photographs of Wavy Films under Various Flow viii * * VI11

Figure Page Conditions (Plates 1,1 to 6.7 and Flow Conditions are listed in Table 4.1),........ 48 4,2.a Transition Map of Wavy Films, Us vs q.. 75 4.2,b Transition Map of Wavy Films, Ti vs q..... 76 4.3 Schematic Portion of Test Section with Liquid Films ~ ~~~e............ *............... 80 4,4.a Film Thickness, h, vs, Steam Velocity, Us, with Equivalent Sand Roughness. q = 0.010 cm3/cm sec..,.... —. 83 4.4.b Film Thickness, h, vs Steam Velocity, U, with Equivalent Sand Roughness. q = 0.021 cm3/cm.sec................ 84 4.4.c Film Thickness, h, vs Steam Velocity, Us, with Equivalent Sand Roughness, q = 0.043 cm3/cm.sec....................... 85 4.4.d Film Thickness, h, vs Steam Velocity, U, with Equivalent Sand Roughness. q = 0.083 cm3/cm.sec................86 4.4.e Film Thickness, h, vs Steam Velocity, Us, with Equivalent Sand Roughness. q = 0.104 cm3/cm.sec...................... 87 44.4f Film Thickness, h, vs Steam Velocity, Us, with Equivalent Sand Roughness. q = 0.125 cm3/cm.sec.................*..o.. 88 4.4.g Film Thickness, h, vs Steam Velocity, Us, with Sand Roughness. q = 0.208 cm3/cm.sec.................... 89 4.5 Friction Factor, fi vs Steam Reynolds Number, Res..................................... 91 4,6 Ratio of Measured Film Thickness to Predicted Thickness CSmooth Wall Assumptionl, hm/h Vs Film Reynolds Number, Ref............. 93 4.7 Nondimensional Film Thickness, h+ vs Film Reynolds Number, Ref (With Laminar, Hughknark (H6) and Kim-Hammitt Correlations).. 96 4.8 Nondimensional Film Thickness, h+ vs Henstock Hanratty (H7) Parameter, F............... 98 ix

Fisgure Page 4.9 Photograph of Rivulet Formation on Exper.imental Turbine Blade, w - 50QQ rpm, d - 380 mm. Msteam = 0.7, p = Q b.6 bar Csupplied by Visiting Professor Povarov, Moscow Power Institute)..101 4,10a Schematic of Hartley-Murgatroyd (H-M) Force Balance Miodel *******~....................... 103 4.10b Schematic of H-M Minimum Power Model........ 108 4.11 Photograph of Wetting Angle Between Water Droplets and Stainless Steel at Room Temperature (taken by H. Hamed and W. Kim, 1977)......... 106 4.12 Dimensionless Minimum Film Thickness vs Wetting Angle (from Mvikielewicz-Moszynski's Results (M3)112 4.13 Dimensional Wave Length,, vs Film Reynolds Number, Ref. (Two-Dimensional Waves)......... 115 4.14 Dimensional Wave Length, \, vs Film Reynolds Number, Ref. (Three-Dimensional Waves)........ 116 4.15a Dimensional Wave Speed, Cvs. Film Reynolds Number, Ref. (Two-Dimensional Waves)........ 117 4.15b Non-Dimensional Wave Speed, C* = C/7i/Sf vs Ref. (with Wurz (envelope) model)............ 129 4.16 Dimensional Wave Speed, C vs Ref (ThreeDimensional Waves)................... 118 4.17 Non-Dimensional Wave Speed, C' = c/Vf vs Ref. (Two-Dimensional Waves)................. -.- 137 4.18 Non-Dimensional Wave Speed, C' vs Ref. (Three-Dimensional Waves).................. 138 4.19 Disturbance Wave Growth Ratef o(Ci Ref vs o<, (Frron Craik' s ReSults (Cl10................... 145 4.2Q Non-Dimensional Waye Number, o(, vs Ref, ( Two-Dimensional Waves)..................... 147 4.21 Comparison of Various: Entrainrjent Inception Criteria, Us vs Ref...................... 150 4.22 Schematic of Commencement of Entrainment Inception................................... 154 4.23 Comparison of Various Entrainment Inception Criteria............... * ^8 158 x

Figure Page 5.1 Photographs of Finger Formation and Secondary Droplet Migration Taken From X = 5 and 10 cm Down Stream of the Trailing Edge (Plates 1.1 to 9,4, Flow conditions are listed in Table 5.1).........................:............. 5.1). 163 5.2 Maximum Finger Length, L vs Time, t for q = 0, 01 cm3/cm.sec................... 184 5.3 Maximum Finger Length, L vs Time, t for q = 0.043 cm3/cm.sec..................* 185 5.4 Maximum Finger Length, L vs Time, t for q = 0,104 cm3/cm.sec............. *..... 186 5.5 Maximum Finger Length Distribution (% vs L) for q = 0,010 cm3/cm.sec........... 187 5.6 Maximum Finger Length Distribution for q = 0,021 cm3/cm,sec.............. 188 5.7 Maximum Finger Length Distribution for q = 0.043 cm3/cm.sec................... 189 5.8 Maximum Finger Length Distribution for q = 0.063 cm3/cm.sec.................. 190 5.9 Maximum Finger Length Distribution for q = 0.104 cm3/cm.sec................... 191 5.10 Time Averaged Finger Length, L vs U....... 193 5.11 Larger Finger Appearance Period, T vs Us.. 195 5.12 S 1 = vs Re. 197 L 5.13 Sf2 = T U s s..198 UsAT s 1/2 5.14 Nondimensional Periodt T l -C ) vs U 2 - f Weber Number, Wes s L........ 201 5,15 Number of Fingers, N vs time, t for q = 0,010 cm /cm.sec.. 0 202 5.16 Number of Fingers, N vs. time, t for q 0.043 cm3/cmsec......................... 203 xi

Figure Page 5.17 Number of Fingers, N vs time, t for q = Q.104 cm3/cm. sec..... 204 5.18 Finger Wave Length'Xf vs U.. 205 5,19 Nondimensional Finger Length, L* vs Ref.. 207 5,20 Measured Droplet Acceleration, Am vs Us... 213 5,21 Measured Droplet Acceleration, Am vs Re... 215 5,22 Comparison of Measured Acceleration, Am with Thoseusing other Correlations for Drag Coefficient, C_. (Serafini (S6), Churchill (C14) and Abraham (:Al)..................... 219 5.23 Acceleration Ratio, R = Am/As vs t for Us = 93 m/s *. * O * 221 5.24 Acceleration Ratio, R vs t for Us = 159 m/s 222 5.25 Schematic of Vortex Generation at the Trailing Edge............................. 223 5.26 Blade Strouhal Number, SB = vs D Droplet Reynolds Number, Us Red s= (B is blade thickness, At periodicity of droplet deceleration and D, droplet diameter) 226 1/2 5.27 Nondimensional Break-Up Time, TbQm vs Droplet Break-Up Weber Number, We for Us = 93 and 159 m/s (M = 0.23 and 0.3A) 2. 30 52Tb1/2 5.28 Nondimensional Break-Up Time, TbQm vs Droplet Break-Up Weber Number, Web for Us = 250 m/s (M Q.61) 231 5.29 Nondimensional Break-Up Time, TbQm/ vs Droplet Break-Up Weber Number, Web for U = 297 m/s (lM Q,72....... 232 5.30 Nondimensional BreakrUp Time, TbQ /2 ys Droplet Break-Up Weber Nubier, m Web for Us = 335 m/s (M = Q.81). 233 5.31 TbQr1/2 vs Web Comparison of Author's Data with Shock Tube Experijnent Data......35 235 xii

Figure Page A,I Schematic of Electrical Conductivity Microgages...7 A,2 Design Drawing of Gages............... 257 A.3 Wheatstone Bridge Circuit and Related Equipment........................ 257 A, 4 Schematic of Calibration Method.......... 257 A,5 Typical Film Thickness Calibration Curves.. 260 A.6 Dependence of f(h) on Film Thickness, h,. 261 A. 7 Dependence of Product of Geometric Factor and Water Conductivity on Temperature..... 262 xiii * * Xlll

NOMENCLATURE Symbol Description Unit aF Frozen speed of sound m/sec AC Cross sectional area of the test 2 section m Am Measured droplet acceleration m/sec2 As Droplet acceleration using Serafini 2 drag coefficient m/sec a,b Widths of test section walls cm A Wave amplitude cm AD Projected area of droplet in the 2 axial direction (L2 cm 4 B Blade thickness cm C Wave propagation speed cm/sec C' Non-dimensional wave speed (= C/Vf) ~C* Non-dimensional wave speed (= C/VT/; Wurz) CD Drag coefficient D Liquid droplet diameter cm Do Initial liquid droplet diameter cm DH Hydraulic diameter (= 4 x rea y) cm periphery f Film friction coefficient f; Interfacial film friction coefficient F Henstock-Hanratty Parameter T Ref )/f f 1/2 Re0.9 Ps tS 5 g Gravitational acceleration cm/sec2 xiv

Symbol Description Unit G Reciprocal Froude number 2 ( 2 vf h Film thickness Fm h Film thickness of wavy crest pm Hhg Mercury manometer head cm or inch h+ Non-dimensional film thickness = = h (xi/f)1/2 Ki Isentropic expansion, index Kn Knudsen number mean free path of vapor molecule ( =diameter of droplet K Von Karman' s constant 2 TT -l k Wave number ( = 2 ) cm1 L Maximum finger length cm L Time averaged maximum finger length cm L* Non-dimensional finger (= M Mach number N Number of fingers 2 PO Pressure N/m AP Pressure drop N/m2 Pk Negative pressure differential due to curvature of wave N/m 2 A (Surfa c tension stress N/) Ps Surface tension stress N/m2'( Kc E'^ A ( oe _ xv

Symbol q Q Qm Red Ref R R sf 1 Sf2 S sB Description Specific liquid flow rate per unit width cm / Non-dimensional flow, i.e. ratio of dynamic steam pressure on the droplet surface to upstream dynamic steam pressure U 2 sU2 ( = - 5 q Maximum Q U D Droplet Reynolds number (= — ) vs Liquid film Reynolds number (= V4 h ) Ratio of measured to Serafini's acceleration As Time averaged acceleration ratio Finger Strouhal number (= AT ) AT Vf L Finger Strouhal number ( L= U ) s Strouhal number for droplet deceleration B U At S Steam temperature Break-up time of droplet Reciprocal Weber number 0(= 2) fVfh Finger formation period Droplet deceleration period Non-dimensional break-up time (=Ur tb 1/2 r p.. s 1. -. - - % EJnit cm. sec To tb T AT t Tb oc sec sec sec xvi

Symbol Description Unit U Droplet Velocity cm/sec Ur Relative steam velocity (= Us-U) cm/sec U Bulk average steam velocity in the test section m/s u* Friction velocity (= CTi/ 7f ) m/s Specific volume of steam cm3/gm Volume of a liquid droplet Tr D3 3 6 ) cm Vf Film surface velocity cm/sec V Capillary - Gravity propagation speed 1/2cm/sec (= + Tk 1/2 Ww Mass flow rate of steam Kg/hr W(y) Velocity profile in rivulet cm/sec Weh Weber number (= f h Vf h h Vf2 Wep Weber number f p 2 2 Wes Finger Weber number Ss LUs Web Droplet break-up Weber number 2 D (s Ur Do x Steam quality X Transverse span in finger formation cm tg Density of gas or density of steam g far ahead of droplet gm/cm f, ~ Density of steam gm/cm3 s 3f Density of liquid gm/cm3 \ Film wave length cm xvii

Symbol \f T Ti;)s A (Xt) II Description Finger wave length ( - X/N) Kinematic viscosity of steam Kinematic viscosity of liquid Non-dimensional finger formation period U AT s 1/2 (. — () ) L f Interfacial shear stress Shear stress due to smooth wall Dynamic viscosity Sand roughness Wetting angle Surface tension Non-dimensional wave number (= k.h) Non-dimensional sinusoidal perturbation profile Non-dimensional normal perturbation stress Non-dimensional tangential perturbation stress Unit cm 2 cm /sec 2 cm /sec dyne/cm2 dyne/cm2 dyne/cm2 gm/cm. sec cm dyne/cm xviii

CHAPTER I GENERAL PROBLEMS ASSOCIATED WITH STEAM WETNESS One might ask the question, "Now that steam turbines have been used and improved since the late 17th century, why do we still need further research thereon?" That may seem a plausible question because the first substance discussed in elementary thermodynamics courses is steam, and all power engineers must have thoroughly studied the properties and behavior of steam, concepts of steam turbine efficiencies, etc. Nevertheless, there continue to be many on-going research programs, particularly relating to wet steam flow in steam turbines which, being multiphase flow, is itself a very complex subject. Therefore, it is of benefit to review the history of the various problems concerning steam turbines, specifically in the area of the author's research. From the early days of steam turbines, designers have made substantial efforts to improve the efficiency of the steam cycle. The most promising way of accomplishing this was to raise the temperature and pressure of steam to as high a value as would be compatible with the best construction materials available at an acceptable price. It became apparent soon that the limits set to the temperature of highly stressed parts were rather strict, while the limits concerning the pressure in the boiler and turbine inlet could be raised more easily by improved design. As seen in Fig, 1.1 an increase of steam pressure Po, 1

2 7 7' -S (a) Low P; Low Carnot efficiency Low exit wetness (1 - XE ) (b) High P; High earnot efficiency Higher exit wetness (1 - XE ) Fig 1.1: Steam Cycle Characteristics (Temperature and Entropy Diagram) Fig 1.2: Wetness Loss Data From Tests on Laboratory Turbines ( Miller, 1972 ) 1.00.98 EFFICIENCY RATIO /wet / /dry.96.94.92.90 ~0.02.04.06.08. MEAN WETNESS

3 at constant steam temperature, T ^. raises the mean temperature, T, of heat addition to the cycle, which means an improved Carnot efficiency, However, it was also necessary to limit pressure, Practical pressure limits were set both by stress considerations and limitations on wetness fraction at turbine exit, For increased exit wetness, two problems are noted. First, severe erosion damage to the low-pressure blading occurs. After a few months or years of service, poorly designed low-pressure rotor blades can be significantly damaged, i.e., seriously roughened, pitted or even broken. To prevent grave erosion damage as well a' substantial loss of efficiency, limitation of exit wetness to A 12% (by mass) was found necessary. Thus wetness became as severe a limiting factor to the attainable efficiency of steam turbine plants as the temperature and pressure resistance of materials. Second, it was found that the aerodynamic efficiency of turbine stages operating in the wet steam region was considerably lower than that of dry stages. One of the earliest empirical correlations of turbine efficiency to account for wetness loss was that of Baumann(Bl), He concluded that 1% of mean wetness reduces efficiency by 1%. It was also experimentally verified (Ml) that supersaturation produces a significant part of the total loss in efficiency CFig. 1.2). Sources for this efficiency loss include the thin wavy film of deposited droplets on the casing wall and blades (Cl) which results in higher friction loss than would a dry wall.

4 Also a part of the vapor energy is spent in accelerating the liquid droplets, This energy term does not add to turbine power output, The fact that present day large water-cooled nuclear reactors produce only saturated high pressure steam forces the study of moisture effects in steam flow. For typical plants (such as PWR) the high pressure (45-65 bar) saturated or near-saturated steam enters and expands in the high-pressure turbine with ^-15% or more of wetness developing at a pressure of *5 bar. Continued expansion would give ~ 25% wetness at turbine exit which is clearly not acceptable. Figures 1.3 and 1.4 show the events leading to steam turbine erosion in such plants. A, CONDENSATION OF STEAM In many technical processes, such as expansion, compression or heating and cooling, the wet steam is subjected to rapid changes of state. In such processes, time is not sufficient for full phase equilibrium. The necessary evaporation and condensation is limited by many factors. Therefore, non-equilibrium processes take place, which are characterized by finite temperature differences between and within phases, Equality of pressure between vapor and liquid phase is usually maintained, because pressure disturbances propagate at acoustic speed, Thus supersaturated (or subcooled) steam and superheated liquid may coexist, However, these states are very unstable. They occur only

5 EROSION PROCESS WATER DROP FORMATION BY CONDENSA TION DEPOSIT OF DROPLETS ON THE SOLID SURFACES (I.E. STATOR BLADE AND CASING WALL) AND WATER FILM IS FORMED WATER FILM DISINTEGRA TION DROPLETS IMPACT ON THE ROTOR BLADE 4.. _~~~~~~~~~~~~~~~~~~~~~~~~~~ Figure 1.3: Summary of Erosion Process.

6 Deposition of part of fog and coarse water Re-entrained coarse water Coarse water spray centriguged from tip Coarse water strav.1. * * —w _ g- -'-.{ ^Impact and splashing of coarse water'og and ing of | Re-entrained water Centrifuging of deposited fog and coarse water 5304 Fig 1.4: Diagram of the Forms of Water in a Wet-Steam Turbine Stage.

7 during dynamic processes, and disappear when equilibrium is recovered. Thus in the fast steam expansion in a turbine, steam condenses in the so-called "Wilson zone" W6 Numerous investigators such as Yellot Yl'Y2) Gyarmathy GlG2) and Ryley(Rl) found that the "Wilson zone" lies at ^5 %o equilibrium wetness. The size of droplets in spontaneous condensation is known to be in the range of 0.01^ 0.1 um. (G Vl) Also the number density of droplets per unit volume has been estimated to be 1015/m3, assuming a uniform size of 0.1 pm (G2), Droplets formed at the Wilson zone will grow in the following downstream expansion region. The growth rate of droplets has attracted the attention of many researchers. (H1) Hill developed a growth law from kinetic theory considering the droplet growth rate to be the difference between the molecular impingement rate of the vapor at the droplet surface and the evaporation rate of the droplet. GyarmathyG2 used Fick's diffusion law and Fourier's law of heat conduction to develop a growth rate for spherical droplets in a continuum environment of arbitrary supersaturation and moisture fraction. Moses and Stein ) compared their experiment with several growth laws and concluded that Gyarmathy's model is more accurate than the others, Whether the vapor behaves with regard to a droplet as a continuum or as a free-molecular gas depends on the value of the Knudsen Numberr

8 K mean free path of vapor molecule n diameter of droplet i.e., Kn - 0.01; continuum flow Kn? 4.45; free molecular condition prevals. Most primary droplets studied in steam turbines fall between these extremes. B. DEPOSITION OF PRIMARY DROPLETS ON THE BLADE SURFACES Primary steam droplets are not in general large enough to cause erosion damage by direct impact. Rather water from deposition of primary drops is collected on blade surfaces, and then torn from the stationary blades in larger "droplets" which then cause erosion of the next rotating row. Traupel(T) has shown that the thermal boundary layer will maintain blading surfaces in a superheated condition. Thus evaporation rather than condensation is expected, so that direct condensation on the blade is unlikely. Gardner() proposed two possible mechanisms of collection, "Brownian Diffusion" for drops smaller than 0.1 pm and "Eddy Impaction" for those larger than 0.1 Pm. For the Brownian diffusion model, molecular diffusion controls near the surface and eddy diffusion at a distance. The efficiency of collection decreases as drop size increases. On the other hand, for the eddy impaction mechanism, particles approach the wall by eddy diffusion, and proceed the remaining distance by already gathered inertia, dissipated by viscous resistance. Contrary to Brownian diffusion, the

9 collection efficiency increases as drop size increases. (C2) Crane carried out experiments for the "collection efficiency" of the eddy impaction mechanism, Some typical results for HP and LP turbine blades are shown in Fig. (1.5). (G5) Recently Ganic and Rohsenow proposed a modified deposition mechanism, considering lift forces, due to drop rotation within the laminar sublayer caused by high vapor velocity gradients, and reaction forces due to asymmetrical drop evaporation inside the laminar sublayer. In addition inertia, drag, bouyancy, and gravitational forces were included. C. FILM FLOW AND SECONDARY DROP FORMATION The deposited drops form a very thin water film, or sometimes separate rivulets on the stator blades. This water will flow along the blades under the influence of steam "drag" (interfacial shear). Finally the film is torn away at the trailing edge or at the wave crests of a wavy film upstream of this edge, depending on the prevailing conditions. Droplets detached from such films are in fact much larger than the primary droplets, and cause the observed erosion damage to downstream rotating blades. Thus the behavior of these thin films, and the size of the secondary drops, their acceleration and their break-up and time of break-up into smaller though still large droplets, must provide the design bases to avoid steam turbine blade erosion. This is the major subjects of author's present research. Information about the droplet acceleration is very

10 1 -- a 1 - J I I I 4''/1r / /' - -/ Fog Deposited on Blade /7).2 FRACTIONAL DEPOSITION OF MASS FLOW RATE 0.1.2.5 1.0 2. 0 DROP DIAMETER 5305 Fig 1. 5: Approximate Fractional Deposition of Fog Drops ( c2) on Blading.

11 important,since the closer the velocity of the droplets approaches the main steam velocity,the less will be the normal impact velocity between the droplet and the blade surface,thus causing less damage. (see Fig.1.6). It is generally agreed that only the normal component of droplet velocity contributes to the erosion damage. Example of damaged blade is shown in Fig. 1.7. D. VELOCITY OF SOUND IN WET STEAM Mach number is an important design parameter in steam turbines, and the steam flow is generally in a compressible flow regime. For this reason,steam velocities are here usually presented in terms of Mach number. Theoretical and experimental studies of acoustic velocity are due to many researchers lC3,P2G6) (P2) Petr's results are summarized graphically. The speed of sound(Fig.1.8) was expressed in terms of the "frozen" 2 K P speed of sound, which is defined as: aF = where Ki is isentropic expansion index, aF is the gas phase speed of sound, inferring that there is no interaction with the liquid phase( which is frozen). Petr considered the effect of frequency on the sound wave speed. As seen in Fig.1.8, acoustic velocity approaches aF as frequency is increased. For reduced wetness fraction and increased density of the gas phase, the mixture tends towards the single-phase acoustic speed aF. This is used in this thesis.

12 Fixed Nozzle Blades Moving Blades Steam Flow B r Lf A C AC - Moving Blade Velocity AB - Steam Velocity Leaving Fixed Blade CB - Steam Velocity Relative to Moving Blade AD - Velocity of Entrained Water Drops at Entry to Moving Blades CD - Velocity of Drops Relative to Moving Blade at Impact Point E Where Erosion Takes Place FD - Normal Component of Drop Velocity Relative to Moving Blade Fig. 1.6 Vector Diagram of Steam Flow in Turbine

Fig. 1.7 Notching of Moving Blade at Inlet Edge by Water ( Taken by Christie and etc., 1966 ( c19) )

14 1. 0 r5 0.95-. 0.90- 0u 0.8 - U * 0.80,1,,! 1 102 103 104 105 106 107 O 5.~ a) Angular Frequency Q 1.0 - 0.950~ f0.90' U 0 1.0 -'~) >.0.95' = 0.85- 0.1 o 102 103 104 105 106 10 bc) Angular Frequency n =, F requency of Acoustic Wave Propagation > 0.90u o 1.0 (Z 0.85- 0.1 u 102 103 104 105 106 107 Fig. 1. o c) Angular Frequency n Effects of a) Drop Size, b) Moisture Content, and c) Pressure on Acoustic Velocity Ratio, a/a P where P F a ki-n,; Frequency of Acoustic Wave propagation ( From Petr's results ))

CHAPTER TI. INTRODUCTIQON The behavior of a horizontal thin liquid film driven by interfacial shear from a cocurrent high-velocity subsonic, low-pressure wet steam flow and its subsequent break-up in the wake of a simulated turbine blade into relatively large droplets have been researched-in detail. Conditions are nominally adiabatic. The results include the identification of several wavy film regimes under various flow conditions, and the construction of a flow-regime transition "map". Important parameters governing the different map regimes, such as film thickness and the detailed wave characteristics like amplitude, form, and propagation celerity, have been measured using electrical conductivity microgages and high speed photography including both still and motion picture. Author's experimental results concerning the critical parameters controlling the transitions between regimes, and the wave characteristics have been compared to other studies in the literature. Where important discrepancies exist between analytical and experimental results of other investigators and author's data, new theoretical analyses or plausible sources of the discrepancies have been suggested. Liquid phase behavior in the blade wake is highly important, having obvious direct application to the steam 15

16 turbine blade erosion problem, Measurements of the "liquid finger" characteristics and the observed periodic decelera-. tions of large droplets are'"firsts" from this study, as are the detailed high-speed motion pictures. Due to the overall complexity of the problem and non-existence of previous comparable experimental or analytical results in many cases, much of the experimental data have been reduced using dimensional analysis, rather than direct solution of the required differential equations. The following sections are descriptions of author's overall results, including pertinent literature surveys. They consider primarily two subjects; 1) the behavior of the liquid film flow and 2) the formation of "liquid fingers" and secondary "droplets", and their subsequent trajectories. These phenomena are of major importance for the erosionresistant design of the subsequent rotating turbine rows. Some results are also pertinent to the "loss of coolant accident" of liquid-cooled nuclear powerplants such as PWR. A. FILM FLOW As explained in previous chapter, understanding of the thin liquid film behavior on turbine stator blades, and its subsequent break-up into liquid droplets, are essential to improved steam turbine blade designf concerning particularly erosion, However, besides the steam turbine application, the study of such liquid film flows has many other important applications. Conveying of liquids by co-current gas

17 streams in oil pipelines, "loss of coolant accident" and emergency core cooling in liquidrcooled nuclear reactors, internal boiler tube flow, transpiration cooling systems for re-entry vehicle nose conesare pertinent examples. The family of liquid film flows is categorized in Fig, 2.1 To allow a study of this range of phenomena and applications, a high-velocity, low-pressure, wet steam tunnel had been designed here and constructed, A simulated turbine blade (in this study, a thin flat plate, Fig. 3.3 ) is inserted parallel to the flow direction along the axis of a transparent rectangular test section. For film thickness measurement, electrical conductivity microgages were develop(M1 0) ed, calibrated and installed at various suitable locations flush with the blade surface. The various attainable film flow regimes, together with wavelet characteristics, were observed using both high-speed cinematography and a still polaroid camera with a 1,2 us flash and flood lights. Most previous investigations involving film thickness measurements were for relatively large film Reynolds numbers, because of their interest in relatively thick liquid films occurring in pipe flow applications,(Fig. 2.2). Author's film Reynolds nunber (pertinent to the steam turbine case) is very low CRef < 56.* As injected liquid film flow rate is increased from zero at fixed steam velocity, various film-flow regimes have been observed here, A film-flow'"map" from this study is thus *Based on surface film velocity and mean film thickness.

FILM FLOW | i R I VU LET I I GRAVITY DRIVEN FILM -I 0-J SHEAR DRIVEN FILM - - - WAVY FILM WITH HEAT TRANSFER ISO'lTERMAL hartley & MFhrgat royd (1964) Hewvtt & i.acey (1965) Hobler (1968) Bankoff (1971) Mikielevicz & Moszynaki (1976) Contact Angle, Critical Liquid Flow Rate, Hinitm.im Power I Norma Mc In Zuber Staub Orell Bankc Brown Marti Ihnat Gumko Hikie Tempe Deper Surfa (.arr Cr ti Flux NON-ADIABATIC UNIFORM FILM WITH HEAT TRANSFER n & & CONDENS4TIOtN HEAT TRANSFER (1966) OR EVAPORATION L & - I I If (1971) Nusmelt (1916) Duckler (1960) iL In (1971) Gater L Miii1 & 1'ecuyer(1970) Chung (1973),owi' & Denny,Milla & Brumfield & levicz(1977) Juaionts(1971) Theoftnous (1976) - I~~Chun & -- Seban (1972) >ratureI ident ice Tension Film Convection ongoni Force), Coefficient Lcal Hleat -- — I I WAVY ISOTHERMAL FILM FtLH7 ISOTHERMAL WAVY FILM I STATISTICAL LINEAR FINIrE AMPLI TUIDE.I - -I Yih(1954- Levy(1966) 1960 ) Reynolds & Lin(1967) Porter(1967) Krantz & Lln(1969etc,(1976) 1971) IHughmark (1973) T:. 1: Telles & Duckler(1970) Chu L Duckler 19741975) WITH SUBSONIC GAS Benjamin (1959) Van Ro8Num(1959) Miles (19601961) Hanratty, Woodnansee L Cohen (1965-1971) Craig (1966) Tait (1966) Ishl & Gromles(1975) Neutral Stability, Phase Lag, Equivalent Roughneal I1- WITH SUPERSONIC GAS Chang & Rumael (1965) Nachtaheim (1970) Nayfeh, saric, Lekoudi & Bordner(19711976) Superitabtility due to Non-linear effect 7~~~~ Orr-Sommerfeld Equation, Tollmien-Schlichting Wave i 0C 00) I I AP'PI,ICATION Local hBUcnout In Nuclear Power Plant Core Cooling, Ablation of Reentry Vehicle and so on I APPLICATION Cooling Tover and other Chemical Process APPLICATION LP Steam Turbine Erosion, Gas Turbine Cooling, Intake Manifold System in Automotive Engine 5308 Fig. 2.1 Categorization of Liquid Film Flow Study

19 s+=0,.0504 Re 7 100- o o > +Y -1+ -++fi~+ Investigator ~~~~~10 ~ rQ S~~f^Kutatelad.ze - __ _ ^ —-3-O O + - McLeod O - Charvonia ~ 0.^^"^,,_ rJF 1/2 A - McManus x - Belkin 00o - Collier & Hewi V - Chien 1. I I', I I......... 1 I I I I _ I I, I 10 100 1000 1000 ReL Fig. 2.2 Liquid Film Thickness Measurement Data ( Kosky, 1971 ( K 10 ) ) )0

20 presented here. The film flow regimes, are; i. Rivulet flow (partly dry and partly wet} ii. Two-dimensional wavy film ('small wave) iii. Three-dimensional wavy film (large wave) iv. Entrainment regime Cdroplet formation at the crest of waves) 1. Rivulet Flow Regimes The transition between regimes Ci) and (ii. ) is of particular interest for safety considerations of nuclear powerplant core cooling (dry-spot formation) as well as to turbine stator blades at the low pressure end, (H2) Hartley and Murgatroyd's analysis ( in 1964 was the first attempt to predict a critical liquid flow rate and thickness corresponding to continuous film break-down into rivulets. The mechanism initiating such dry patches was not discussed. The study was concerned rather with whether once a dry patch is formed, will it remain dry or become rewetted. They suggested two criteria. One was based on a force balance at the upstream stagnation point of a dry patch; the other on a minimum total power criterion based on kinetic and interfacial energy in a transversely unrestrained stream. Their minimum power criterion did not include any influence of the contact angle, which since has appeared to be very important in this regard. (JH3) Based on Hobler's analysis(-, where a mnimum power criterion allowed for the effect of contact angle, Mikielewicz and Moszynski M3) suggested an improved criterion.

21 Both Hartley-Murgatroyd"s force balance (HEM) and the Mikielewicz-Moszynski models (M-M) included contact angle criteria. Comparing with authar's experimental minimum flow rate criterion, contact angle, 9, needs be/ 60 and v10~ respectively for these models to be compatible with author's data, The static contact angle between a stainless-steel plate and a water droplet (in air) is believed to be 49~4~. However for a single curvature such as in the case of a rivulet, the measured angle(M) was found approximately half of that static angle. Anyway the wetting angles necessary to make the H-M and M-M models realistic are too small to be exactly valid. Of course contact angle is dependent on surface conditions, i.e., roughness, cleanliness, etc. and also on gas phase. Such aspects were not considered in the theoretical analyses. Author's photographic studies (Fig. 4.1), show an abrupt liquid bulge just upstream of the stagnation point for a dry-patch. Some analyses (M2'Ilmentioned this particular geometry. This may be partially responsible for the large discrepancies in contact angle between observation(M5) and requirement to match the H-M and M-M criticality models. NonrAdiabatic Dry Patch Formatign For cases with strong heat Flux, Zuber and Staub considered surface tension variation with temperature. Orell and BankoffO01 added the liquid pressure due to the thrust of evaporating molecules. Both. these analyses are based on the Hartleyurgatroyd model2) the Hartley-Murgatroyd model.,2 These studies are not direct

22 ly applicable to author's case which was nominally adiabatic. 2, Wavy Film Regime If liquid flow rate is further increased at fixed steam velocity, the blade surface is completely wetted and the dry patch or rivulet regime no longer exists; rather a wavy film regime appears. This is shown from author's photographic results confirmed by film thickness gage output. The interaction of the external driving vapor flow with the liquid interface produces two important effects. The first is the establishment of a velocity profile in the liquid due to the shear stress at the liquid-vapor interface. The second is the action of perturbations in the normal and tangential stresses exerted by the vapor due to the appearance of waves. The main steam flow may be considered inviscid because of the high Re ( >10 ), In our case it is also subsonic (M < 0.8). The pressure perturbations in the liquid are then in phase with the amplitude of the surface waves;. In fact the pressure perturbation from the vapor would be 180~ out of phase with the wave amplitude due to the resultant vapor streamline curvature. Thus there is an energy transfer from the vapor flow to and from the wavelets in the liquid layer (Fig. 2-3a), This is the classical Kelvin-Helmholtz instability mechanism for wave formation, In real cases, however, the pressure distribution in the liquid is not, in general, precisely in phase with the surface wave amplitude, As seen in Fig, C2-3b) the perturbation distribution upon the profile slopes leads to a resultant stress in the

23 Velocity change Pressure change in vapor Surface profile a) Kelvin-Helmholtz pressure changes Mechanism lag surface profile by T + Velocity change Pres sur change in vapor e I I I I I - 1 I i I i I I i iI I - i zo i I I - )f. \A Z L - I I I. -I. 4 Surface profile b) Pressure changes lag surface profile by TT/z Fig 2-3: Velocity and pressure changes in flow over wavy surface.

24 flow direction, and does work on the waeves. This interfacial stress, derived from the normal pressure distribution near the boundary, is sometimes called "form drag". Since the behavior of the liquid film is largely governed by stresses exerted upon its surface by the vapor flow, good estimates of these stresses are obviously desirable. In general, the vapor stream is turbulent, and consequently random fluctuations of stresses exist at the liquid interface. Although the resonant response of a liquid surface to random pressure fluctuations is important in the generation of (P5) ocean waves, such a resonance mechanism is not important in the author's present problem, since the convected velocities of the vapor flow fluctuation are always much greater than the wave velocities in the liquid film. Thus only a very weak response of the surface is possible, The surface stresses produced by interaction of the mean steam flow with small periodic perturbations of the boundary surface have been evaluated by Miles(M4 M5 M6) and Benjamin(B4) for mean vapor-velocity profiles of the boundary-layer type. However, their approaches lead to great mathematical complexity, MilesM4 proposed the "Tollmien-Schlichting" instability mechanism for large film Reynolds number. In this mechanism, the unstable disturbances receive their energy from the mean flow of the liquid layer, rather than from the vapor, by way of the Reynolds stresses in the liquid near the wall, The disturbances receive their energy either from the mean flow

25 of liquid or vapor depending on whether the wave speed is less than or greater thn th.he intePrace velocity, Miles coneluded that a necessary condition for instability,, assuming a Tollmien-Schlichitin mechanism, is Ref 203, Saric and Marshall [S) verified this mechanism experimentally for supersonic gas flow, and concluded Ref must be >100 for this mechanism to occur. The analysis and experiments of CraikC) with an incompressible air flow showed that thinner films may be more unstable than thickner ones, His expression for shear perturbation in phase with wave slope is inversely proportional to film thickness; the thinner the film the larger the shear perturbation. Because of this thin film consideration, author's experimental results can best be compared to Craik's model. The analysis of Nayfeh and Saric(N3 extended Craik's work to include an arbitrary body force. External flow was supersonic and a long-wave analysis was used. A simple relation between the wave speed and growth rate was derived. This author's vapor velocity is subsonic. However an interesting result was reported for the supersonic vapor flow case. Nayfeh and Saric's non-linear analysisN45) showed that stable linear disturbances die faster whereas unstable disturbances achieve steady-state axplitude instead of growing indefinitely if disturbance houndary layer is larger than subsonic boundary layer, This was yerified experimentallyS3 by observing no liquid removal eyen for supersonic gas flow,

26 3, Entrainment Regime From author"s photographic study, for higher steam velocities, as liquid flow rate is further increased, the waves grow and crests are broken, much like ocean "breakers". Droplets are then shed from the wave crests into the vapor flow. From the viewpoint of turbine blade erosion, this entrainment mechanism, at least the parameters controlling its inception, is important. Since the entrainment takes place along the stator blading or casing wall, the flight time of such entrained droplets is different from those generated at the trailing edge. Experimental inception criteria for droplet entrainment in horizontal two-phase film flow have been measured by many (V2,W2,W4) and numerouis inceptTSn criteria for vertical-pipe flow have also been proposed by others. (C16,S4,G7,G8) All of these used air-water, relatively high film Reynolds number (Ref > 100) and low gas velocity. The tests were generally performed in low velocity wind tunnels. Attempts(W 6) have been made to correlate the various data available. However, these were empirical and fit only the limited range of data used. In general these do not fit the high-velocity, thin-film data of present interest. Results are conflicting due to effects of flow orientation, i.e., horizontal, or vertically u? or down. ZhiVaikin's data2) shows that the inception of droplet entrainment depends on flow direction at low liquid Reynolds number, while Wallis 7 indicates that onset is independent of flow

27 direction, An extensive entrainment inception model was proposed CI2 1 Zl by Ishii and GrolmesI Based on Zuber's analysis, they considered the wave amplitude and flow within the wave crest, They assumed that the interfacial shear force at the top of the wave induces an internal velocity of same order as film velocity, and that the velocity of the wave crest with respect to the film is proportional to film (H6) velocity. Using Hughmark's correlation(H6) for the effective interfacial friction factor, they concluded that critical non-dimensional vapor velocity for entrainment inception is proportional to (- -) power of film Reynolds number for relatively thick films (Ref > 160). For low film Reynolds number flow they speculated upon the similarity between (H8 ) Hinze's critical droplet disintegration mechanism() and their entrainment inception, They concluded that nondimensional vapor velocity for entrainment is proportional to ( - ~) power of the film Reynolds number. Compared to author's experimental results, their correlation for low film Reynolds number criterion is high. B. LIQUID FINGERS AT BLADE TRAILING EDGE AND SECONDARY DROPLET FORMATION Liquid films formed on the stator blade can be partially removed by suitable separators or by the entrainment mechanisms already discussed, However^ the remaining film is entrained by shear force of the driving vapor from the blade trailing edge, forming into "secondary droplets" in the steam wake.

28 The liquid droplet motion in the wake near the trailing edge is very important to the erosion problem, controlling the upper size limit for drops reaching the rotor blade. Since the erosion potential of droplet impact increases strongly with droplet size CP6fS9), this is an extremely important question for blading erosion. 1. Formation of Liquid Fingers In author's experiment, liquid film detachment from the trailing edge is never continuous, but appears as discrete "fingers". An extensive study of the time-averaged shape of such fingers, together with the time parameters involved in the frequency of finger formation has been made, It is important to secondary droplet formation, directly pertinent to blade erosion potential. Surface tension, together with shear, inertial forces and film flow conditions at the trailing edge are governing parameters for finger behavior. Axial finger length and finger wave length (distance between fingers) as function of time have been measured from numerous high-speed motion pictures for the various steam and liquid flow conditions. The frequency for the formation of longer fingers was especially studied, since these are of particular importance from the erosion viewpoint, transporting large. slugs of liquid into the path of the next rotating row. Dimensional analyses were used to study the effects of surface tension, inertial and viscous forces of the vapor

29 and liquid flow on the frequency of formation of the longer fingers and wave lengths of fingers, It is shown that Strouhal number based on formation frequency, finger length and liquid film velocity decreases with increasing steam Reynolds number, 2, Acceleration of Secondary Droplets When droplets are detached from the fingers, they are accelerated by drag forces due to vapor flow in the wake region. Information concerning droplet acceleration is most important, since the closer the droplet velocity appoaches steam velocity, the less their erosion potential, since (P6) impact damage is a strong function of normal velocity alone. Many studies of liquid droplet acceleration have been motivated by aircraft and missile rain erosion problems. In such cases the velocity is supersonic, so that droplet desintegration is influenced by shock-wave patterns. Thus the trajectories and acceleration of droplets just after passing through a shock wave is of interest S ) Nevertheless, an important, partially unknown parameter is the drag coefficient for droplets as a function of Mach number and shape. This of course is also a function of drop Reynolds number. Empirical or theoretical drop or bubble drag coefficients have been reported because of their importance in heat and mass transfer, (Al,C14 All of these drag coefficient correlations considered single droplets with external gas or vapor flow, Also of interest is the drag coefficient for a drop in the presence of other drops

30 (T21. of varying sizes. Tam r deriyed a forma f for the drag exerted on a cloud of spherical particles of a given particle size distribution at low Reynolds number, Based on the statistical concept of distributed particles, he concluded that the drag experienced for a particle depends only on the first three moments of the distribution function. None of these correlations can be directly applied to author's particular case, because author's flow regime is the wake region of a tapered flat blade. Author's experimental droplet accelerations, when plotted as a function of time, show periodic abrupt changes, ie., an accelerating droplet suddenly decelerates, sometimes reverses direction, and then accelerates again. This observed phenomenon can best be explained by the trailing vortices shed from the blade, similar to a Karman vortex street. The presence of a discrete vortex near a droplet would produce an induced velocity field which may be responsile for the observed changes in droplet acceleration. The periodic vortices shed from the droplets themselves may also be important. A very similar observation is reported by Tyler and Salt(T4)for solid spheres in freeflight. The periodic behavior is assumed due to Karman vortices shed from the spheres. 3. Break-Up of Droplets In the higher steam velocity ranges, the disintegration of the droplets was found to occur more rapidly. Author defined the break'up time, tb, as the duration from the time when discrete droplet forms from fingers, to the time when

31 the droplet becomes amorphous or disintegrates. Nondimensional break-up time based on initial drop diameter and vapor flow conditions was found to be approximately inversely proportional to the droplet Weber number.

CHAPTER III EXPERIMENTAL FACILITY AND MEASUPEMENT CONDITIONS A, STEA4 TUNNEL GENERAL Figures 3,1 and 3,2 are a schematic and a photograph of the low pressure wet-steam tunnel facility, "Wet Steam Tunnel," at the University of Michigan, This wet steam tunnel was designed (K5,H9'Hl0to produce flow parameters in the test section which would generally match the average steam parameters in the last stages of large steam turbines. Mean static steam conditions in the test section are near saturation, probably never superheated, showing varying degrees of moisture depending primarily upon test section velocity,( Table 3.1).* Of course, wetness increases with velocity head for fixed supply conditions. Actually in many cases, upstream of the throttle valve has rather low quality (almost hot water according to pressure and temperature measurements at the upstream orifice under low flow conditions). However, after traversing the valve, water is trapped in the "stilling tank" so that the quality of steam reaching the test section is still nearly dry. This is confirred by the fact that water build-up is not observed on the blade or in the test section. The test section temperature (which is almost stagnation temperature). does cover a certain temperature range (usually decreasing as steam flow rate is reduced). It is assumed for present purposes that supply conditions are constant, though this is not precisely true. *These are "typical" conditions, not entirely accurate. 32

33 Fig. 3.1 Facility Schematic 1-.KO I I i I Gcw - ~ PCW CW I I I ----—. tpw MIXING CONDENSER MIXING CONDENSER

SillinLg Ta. (J3 U-. ul Fig. 3.2 Photograph of the Overall Steam Tunnel Facility

35 STEAM PROPERTIES: steam T3 P3 3 S X3 T1 P1 X1 h S ft/sec 0F lb BTU BTU 0F lb U Bm U.2 Ibm lbm F 2 lbm lbm F in. -; in 200 130 2.225 1L07.456 1.894.99 227.96 20.95 1108.225 1.6621 600 1101.065 1.877.98 1500 1063.316 1.825.95 Not.e: Conditions at "3" and "4"-approx. same \-A =ZOO ft/ses _st 1200 E H 1100 + H 1000, Z'I / P=2. Z24psia T=130 F line 1:Z 1.4 ENTROPY I.'6 BTU/lbm~F 1.8 Z.0 Table3.1 Thermodynamic Operating Conditions of Steam Tunnel

36 The supply is from the building heating boiler. No actual measurements of steam quality in the test section have been made, but only conventional thermodynamic calculations. However, precise knowledge of test sectiqni steam wetness is not of great importance for present purpose, where liquid film is injected and conditions are nominally adiabatic. Test section steam velocity can be measured either by a Pitot tube installed in the test section or by the upstream orifice. Once agreement between these devices was established, only the flow orifice was used, to avoid the flow perturbations caused by the presence of the Pitot tube. The relationship used for measurement of steam velocity is as follows: _Ww 7 Us 3600A (3.1) where: U = bulk average steam velocity, in the test section (ft/sec) W = mass flow rate of steam (lbm/hr) w = specific volume of the steam (ft3/lbm) 2 Ac= cross-sectional area of the test section (ft2). For the present geometry A<= 0.069 ft2. And from the handbook, "Principles of Flow Meter Engineering" (S10). Ww = 2781,72 (hgP )/2 W V 2 where 2 Hhg = manometer head, inches of mercury T2 steam line temperature C~RI P2 saturation pressure at the steam line temperature (lb/in2-abs.)

37 More details of the steam flow measurement and related data reductions are available in referenceM9 Static pressure was measured along the side wall of the test section. Three pressure taps were installed; the first opposite the leading edge of the blade, the second 11.5 cm downstream with respect to the leading edge, and the third 25.5 cm downstream from the leading edge (5.5 cm downstream of the trailing edge), These static pressures were read using a precision Heise gage as negative gage pressures. Ann Arbor atmospheric pressure of course varies from day to day. The 4 2 average pressure, however, is 14,3 psia (9.86 x 10 N/m ). Thus the average absolute test section static pressure ranged between 1.2 and 1.3 psia (0.83 x 104 and 0.89 x 104 N/m2). One thermocouple was embedded flush with the side wall of the test section. The temperature reading obtained there was always higher than the saturation temperature corresponding to the test section pressure. This positive temperature differential is of course due to the boundary layer recovery of part of the kinetic temperature in this high velocity steam flow near the wall. (S8) Average parameters in the test section, i.e. position #4 of Fig. 3.1 are; P4 0.Q9 bar C1,25 psial test section pressure T4 43QC C109,4~FI saturation temprature at P4 X4 = 0,95 estimated steam quality 4 = 16 n3/Kg (256.47 ft /lbm) steam specific volume

38 U max 335 m/s CllQO ft/s) maximum obtainable steam velocity The test section C8 cm x 8 cm x 100 cml is constructed of clear transparent plastic CLexan) which allows the necessary photographic studies. A simulated turbine blade, i.e., essentially a thin flat plate with tapered ends.( Fig. 3.3),with electrical conductance microgages on the blade to measure liquid film thickness, including time variations, is inserted along the centerline and parallel to the flow axis. The design and calibration of these gages is discussed in Appendix A. The "stilling tank" (Fig. 3.1) provides a honeycomb flow straightener to produce uniform flow in the test section. The condenser is of the jet type (rather than conventional shell-and-tube design). Jets of tap water are used to condense the steam in a counterflow vertical tank. The bottom of the tank serves as a "hot-well" from which a centrifugal condensate pump takes suction. The pump then exhausts the combined flow into the laboratory drain system. To improve condensing~and hence obtain increased vacuum and steam velocity, perforated trays were installed in the condenser to produce more efficient mixing. In addition, a mechanical vacuum pump (Kinney Hi-Vacuum pump, Model KC-15) was used to increase the test section-condenser vacuum still further, B. BLADING PROFILE AND ELECTRICAL CONDUCTANCE MICROGAGES As shown in Fig 3,3., a thin flat plate used to simulate the steam turbine blade was designed, fabricated

II II II 11 II II I - I I.N N -. I - i - I.-, I —--— ~ —-.1.-'. II II II II I I U. M. Gage Mod. 1 IF'FM Gage Mod. 3 U. M. Gage Mod. 2 U. M. Gage Mod. 2 0.152 I I Fig. 3.3 Test Section Blade

40 and installed in the test section for these tests, In one version, a liquid film is injected through a small slit normal to the tunnel axis, to simulate the liquid film flow condition encountered naturally in the turbine stator blades. Minimum film injection rate depends upon the measuring and throttling capability of the flow meter, and is 5 cm3/min (0.625 cm3/min-cm, "specific" liquid flow rate per unit blade width). The maximum rate used was 100 cm /min (12.5 cm3/mincm). The pressure differential for film injection from the open reservoir used is due to the vacuum in the test section (-1. 25 psia) plus an elevation head between liquid reservoir and the test section. Electrical conductance microgages were installed flush with the blade surface to measure the liquid film thickness variation with respect to time. Since detailed decriptions of gage design and calibration are given in Appendix A, only a few important aspects are discussed here. Essentially, the electrical conductance of a thin liquid film connecting the central pole of the gages with an external annulus (separated by an electrical insulator), all installed flush with the blade surface (Fig. 3.3),is measured by an a.c. circuit including a Wheatstone bridge. The output is connected to a dual-beam oscilloscope, so that the simultaneous output of two gages can be compared at a given instant of time. An a.c, circuit is used to avoid problems of electrostatic charge,. While results depend somewhat on circuit frequency(C15,'H19 it was found that the response is flat frequency,~

41 with a frequency in the range of 1 kHz CH1QF1) which was thereafter utilized, The output in terms of circuit conductivity was found to be approximately linear for small film thickness, but gage output becomes"saturated" at higher thickness ( >300 pm. (M10,F1,H10,Hll,H18) Thus the useful operating range of the gages is limited to ~ 300 pm. (See Appendix A and references Fl, M10). This is sufficient for present purposes, C. PHOTOGRAPHIC FACILITIES Photographs were made using both a still polaroid camera (4x5 Busch precision camera, Model D) and a modified high speed Wollensac Fastax camera (WF-1) for the investigation of wavy films, finger formation, and droplet trajectories. In the Fastax WF-1 camera, film is transported continuously past the aperture while a rotating prism, geared to film motion, produces an image which is stationary with respect to time. Individual frames (16 mm x 8 mm) constitute the 100 ft. rolls of film used. (Kodak Tri-X reversal 7278 film). A "Wollensac-Goose" control regulates the camera motor speed. The motor continuously accelerates the film, which is initially at rest. The maximum acceleration, and hence framing rateis determined by the film strength. Since the film is initially at rest, the framing rate changes continuously, so that the maximum framing rate occurs only near the end. The exact framing rate at a given point is determined from timing marks placed along the edge of the film using a 60 Hz timing light,

42 For present work, the framing mechanism was removed and a magnetic coil installed near the drive sprocket, Framing was achieved using the magnetic pickup coil, which is triggered by the drive sprocket. Pulses are then delivered to a strobo-trigger circuit CEG&G high-speed stroboscope Type 501), which fires the flash lamp (EG&G-FX2). Synchronization of camera and strobe is thus ensured at all film speeds. An EG&G stroboscopic light source which produces high intensity 1.2 ps light pulses at a maximum rate of 6,000 flashes/sec was used to reduce the exposure time per frame. Since the Fastax camera has a rated top speed of - 14,000 frames per sec., framing rate was limited in the present case by the flash rate of the light source, which was in fact sufficient. The Fastax cinematography was sufficient to "freeze" all droplet and wave motions, as was the still photography used. By using the timing marks on the film the time increment between frames could be determined accurately for the purpose of establishing wave propagation speed and droplet accelerations and break-up times. For the still photography, two sets of flood lights were employed (Color Trans Quartz-King dual 650, Model no. 160-011). The lens settings were selected as required. Film used was Polaroid 4x5 land-film/Type 57. Typical exposure time. was 5 ms with two flood lights, This relatively long exposure time was sufficient, even though much shorter exposure times wereused for the Fastax. These different exposure time requirements are due to the combined effects

43 of light intensities and film speeds used. D. EXPERIMENTAL MEASUREMENT CONDITIONS The range of experi"mental parameters. and measurements is listed in Table 3.2 below-\ TABLE 3,2 - Experimental Parameters Independent Variables Steam Velocity, Us, Specific Liquid Flow Rate, q Dependent Variables Wavy Films Film Thickness, h, Film Wave Length,\ 20 to 335 m/s CMach No. = 0.05 to 0.81) 0.010 to 0,208 cm3/cm-sec CRe = 1.4 to 56.2) (see Fig. 3.4) - 1 to 10 mils (25.4 to 254 urn) -5 to 12 mm (small waves) 10 to 20 mm (large waves) Wave Propagation Speed, C, -20 to 200 cm/s (small waves) 100 to 1.500 cm/s (large waves) Finger Formation and Droplet Motion (see Fig. 3.5) Finger Length, L, - 3 - 30 mm Finger Wave Length, \, - 2 to 10 mm (average distance between fingers) Frequency of Large Finger Formation, f, - 8 to 16 sec. Initial Size of Largest Drop in Wake, Do, - 0.5 to 2.0 mm Droplet Acceleration, A, 300 to 400 m.sec."2 m ( 30 to 40, gl Duration of Large Droplets, - 0,Q01 to 0.02 sec.

44 Us -.0 Vf_ _7 _C k.. I h Fig 3.4: Schematic of Wavy Film Characteristics D2 O ( a V v~S x~ N ( N = 9 in this case ) TRAILING EDGE Fig 3. 5: Schematic of Finger Formation and Secondary Droplet Motion

CHAPTER IV BEHAVIOR OF LIQUID FILM A. VISUAL OBSERVATION OF WAVY FILM One of the most important features of film flow are the wave patterns which appear at the free surface or interface between the liquid and vapor phases. Such wavy film flows are highly complex. Thus no satisfactory general theory has yet been evolveda rather only specialized and simplified models are available. Further, the complete experimental determination of the parameters characterizing very thin wavy film patterns, as in the present case, is extremely difficult. To study the wavelet patterns in detail, about one thousand still photographs have been taken at various steam and liquid flow conditions, as well as ^ 100 rolls of 100 feet length of Fastax high speed motion pictures. Oscilloscope photos of film thickness conductivity gage voltage output were taken, simultaneously with the wavelent pattern photos. The photographs of Fig. 4.1 show the various wavy film patterns for different flow conditions. For different steam velocities, the transition points between various wavy pattern regimes, as the liquid flow rate is increased, have been observed. Plate Series #1 of Fig. 4.1 show the various wavy patterns for steam velocity, Us=30 m/s, which is the minimum steam velocity used for this investigation. Instrumentation limitations prevent the use of still lower steam velocities. As can be seen in plate 1.1 for minimum 45

46 Table 4A. FLOW CONDITIONS AND OSCILLOSCOPE SET UP FOR FIGURE 4.1 Plate Steam No Velocity Us(m/s) Liquid Flow Rate q, cm/cm. s Oscilloscope Set Up X Y ms/unit V/ unit Features Pl.1 P. 2 P1.3 P1. 4 P1. 5 P1.6 P1. 7 P2.1 P2. 2 P2.3 P2.4 P2. 5 P2 6 P2. 7 P3.1 P3. 2 P3.3 P3.4 P3.5 P3.6 30 30 30 30 30 30 30 50 50 50 50 50 50 50 80 80 80 80 80 80 0. 010 0.021 0. 043 0.083 0.125 0. 167 0. 208 0.010 0. 021 0. 043 0.083 0.125 0.167 0.208 0. 010 0.021 0.083 0. 125 0.167 0. 208 10 0.5 10 0.5 10 0.5 10 0.5 10 0.5 10 0.5 10 0.5 10 0.5 10 0, 5 10 0.5 10 0.5 10 0.5 20 1.0 1C 1.0 10 0.5 10 0.5 10 0. 5 10 0.5 10 1.0 Rivulet Flow Ripples Smooth and 2-D wave Ripples and 2-D wave 2-D and 3-D wave 3-D waves 3-D waves Dry patch and abrupt bulge at the stagnation point Dramatic dry patch and apparent wave appearance 2-D waves 2-D waves 3-D waves 3-D waves 3-D waves Rivulet Flow 2-D wave and dry patch 2-D and complicated ripples 3-D waves 3-D waves 3-D waves

47 Table 4.1 P4.1 P4.2 P4.3 P4.4 P4.5 P4.6 P4. 7 P5.1 P5. 2 P5. 3 P5.4 P5.5 P5.6 P5. 7 P6.1 P6.2 P6.3 P6.4 P6.5 P6.6 P6.7 144 144 144 144 144 144 144 227 227 227 227 227 227 227 305 305 305 305 305 305 305 0..010 0.021 0. 043 0.083 0. 125 0.167 0.208 0. 010 0.021 Q. 043 0.083 0. 125 0.167 0. 208 0.010 0.021 0. 043 0.083 0.125 0.167 0.208 10 0; 1 10 0.5 10 0.5 10 0.5 20 0.5 10 0.5 10 0.5 10 0.5 10 0.5 10 0.5 20 1.0 20 1.0 20 1.0 20 1.0 10 0.5 10 0.5 10 0.5 10 0.5 10 0.5 10 0.5 10 0.5 Wavy rivulet 2-D waves 2-D and 3-D waves 3-D waves Entrainment, Entrainment. Entrainment 2-D waves 2-D waves 3-D waves Entrainment Entrainment Entrainment Entrainment 2-D wave 3-D wave 3-D wave Entrainment Entrainment Entrainment Entrainment

s T p 1p 3 I0l-. ro A jXpa n;D, ~I SX A A- IA 0o s qd -e ao o d I jST,: 00s VE PIN z-fi2>B ~-x~~zz-,>~wS's~,2zgx 39.&>,.z<69%9532.g,,A3NN' f,-W,'m Adz'a >1 S,:.B a >-.R IMMO ME,~~~~~~~~~~UT~I~J OSTd~ _M _ -AdZ~K> M ~~~ 4 ~~~~~~~ A 4K:A~~~~~~~~~~~~~ ~,gg

Fig. 4 1 Continued

0 Fig. 4.1 Continued

51 0 U 5379

l:: Syl ISM, co~~~~~~~~~~~~~~~~~~~~~~ij \, i ol~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Ml Mrm~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~a3 ~ ~ ~ ~ ~ ~ K sx~~\~':~~ <>~AM. SHM,, 0 >, umj M%4 Ko~ 0 ~ ~ >(. Is M,~~~ N~~~~~~~~~~~~EI

00 C i ~1~,i,~1~!!R 21 -' - 4 l Co.n..e. Fig. 4.1 Continued

Fig 4o1 Continued

~~~~00oFig 4 Cotin Fig. 4.1 Continued

00 Fig 4 1 Continued

57 13 <y 9~. P-4 s^ 4J) 0 U bo).r4 N

'o mi;W'"< M m FBI Y1I1diLb~ Ion~~~~~~~~~~~~~~~~o ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i- F el C I C on-tinu ed

00 ~~88""~~~~7~~"3 ~~~~"-$ " ~ ~ ~ ~ ~ ~ c Ln

co o ~: - C

61 0 ~ -A b~ 5389

~'~'" ""q" "''' n i'~'"~~~~ I..i 11, I rl-1l-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~!1i ~~~~~~~~~~~~~~~~~~~~~~~~~~~ Fig,, 4,1t Continued

Fig. 4.1 Continued~ii

U1.1~ ~~~ 11g., 4. 1 Continued

Fig. 4.1 Ctiued

66 rd C1) oS bf) eb4 han 5394

si~~~::: -C;B; Ii~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0.~~~~~~~~~~~~~~~~~~~~~~I~~~~~~~~~I -:i::i, a~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~iiij Fuji~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,,",:E Li"; ~~~~~-V —--------— ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ — ----------- F~~~~F~~~3 ~:

00 Fig. 4.1 Continued

69 film flow (q =.QQ1 cm3/cm.sec), the liquid film on the blade is not continuous but contains "dry spots", i.e., one region is dry, and contiguous regions are covered with a thin liquid film. To be consistent with existing literature, (MM) this flow pattern is called "rivulet flow". The existence of rivulet flow requires a balance of the various applicable forces at the junction of wet and dry regions. These are due to surface tension, shear, inertia and gravity. Surface tension and shear(viscosity)are strong functions of temperature. Thus for strong heat flux (not present case) temperature dependence of these properties need be considered. The films are never waveless in author's experiments (Plate 1.1). Thus a fully realistic analysis is very difficult. If the liquid flow rate is further increased, as in plates 1.2 and 1.3 (q = 0.02 cm /cm.sec), the blade surface becomes completely covered with a thin liquid film which remains relatively smooth. Plate 1.3 includes portions of both wavy and relatively smooth films. Further increasing liquid flow rate, q, one finds more complicated wave patterns of the liquid on the blade. Plate 1.5 (q = 0.125 cm3/cm.sec) shows relatively symmetrical wave patterns about a plane parallel to the axis, while plates 1.6 and 1.7 show patterns which are non.symmetrical about this axis. These symmetrical waves will be called "two dimensional waves", and the non-symmetrical waves "three-dimensional waves", The important differences

70 between these wave forms will be discussed later. Plates of series #2 of Fig, 4.1 for a somewhat higher steam velocity (Us = 50 m/s) are quite similar to Series #1 in showing the wavy film pattern transitions. There are, however, a few peculiar features not found in the plate #1 series. In plate #2.1, a more irregularly shaped rivulet forms than that of plate #1,1. Also, at the junction of dry and wet regions, a bulge in film thickness is observed. The bulge geometry is of interest, because of its possible influence on vapor flow patterns near the stagnation point. Hartley and Murgatroyd(H2) assumed the liquid film thickness to decrease smoothly from the up-stream value to zero at the stagnation point. The forces acting on the stagnation point due to changes in vapor flow pattern could then be relatively easily estimated. However, with an abrupt liquid bulge at the stagnation point as in author's case, the vapor flow pattern around the stagnation point becomes more complicated and difficult to estimate, Some (1 I) considered this geometry but did not get fully acceptable results. The significance of this particular flow geometry will be discussed in more detail in section C of this chapter. 3 Plate #2.2 (q = 0.021 cm /cm.sec) shows a relatively unique rivulet formation. There are only two dry spots, one at the center, and the other a little further downstream near blade axis. In most "rivulet" photos, there are many rivulets along with parallel dry strips. Here also

71 a film thickness bulge is seen, t the stagnation point, Again, there are waves in the liquid film ahead of the stagnation point bulge, Plates #2,3 (q = 0,043 cm /cm) and #2,4 Cq = 0.083 cm /cm sec) show typical "twordimensional waves" formed on the blade. In plate #2,3 only relatively faint wave outlines are observed, while those in plate #2.4 are more clearly defined. In both photos, the wave fronts are nearly straight, and most are perpendicular to the flow direction. Since the wave length in plate #2.3 is greater than in plate #2.4, it appears that wave length decreases as liquid flow rate is increased. As liquid flow rate is further increased, the shape of waves becomes less regular and their cross-sections attain a relatively steep front with long tail (See plates #2.5,q= 0.125 cm /cm sec), #2.6 (q = 0.157) and #2.7 (q = 0.208)). Usually each such three-dimensional wave is preceded by a number of small satellite wavelets. The wave lengths from plate #2.7 are smaller than those from plate #2.6 or 2.5. Also the transverse length of the three-dimensional wave front is reduced, and consequently a more "pebbly" shape appears. as the film flow rate is further increased. From inspection of the oscilloscope photos, the wave heights can be estimated, In a particular case Cplate #2.4) the wave amplitude is /20% of the average film thickness, It is ^100% of average film thickness in plate #2,7. The amplitude of the "three dimensional" waves is typically as

72 large as the mean film thicknesa itself, Plates #3 are photos of wavy films for a larger steam velocity (Us = 80 m/sl, These show the same transition of wavy films as do the #1 and #2 series. Plates #3.4, 3.5 and 3.6 indicate very typical "three-dimensional" waves, with pebbly wave fronts. Plates #4 are for steam velocity, U = 144 m/s. In plate #4.1 (q = 0,01 cm /cm sec, wavy liquid filaments are in good contrast with the adjacent dry strips, Plate #4.2 (q = 0.021 cm /cm sec) shows two-dimensional waves, parallel to each other and with wave fronts perpendicular to the flow direction. Further, the wave lengths are relatively uniform. Plate #4.3 Cq = 0.043 cm3/cm sec) shows threedimensional and two-dimensional waves at the same time with relatively irregular wave lengths; i.e., waves with longer wave lengths are accompained by three or four small satellite waves. Plate #4.4 (q = 0.083 cm /cm sec) also shows threedimensional waves (with steep fronts) as well as a dry spot at the upper right of the picture. Dry spots in the vicinity of three-dimensional waves are not common. Plate #4.4 is one of a very few pictures obtained showing such dry spots near three-dimensional waves in author's entire experiment, Such a complicated wave shape in front of the stagnation point associated with the dry spot nakes the analysis of this particular film breakdown phenomenon difficult. A different phenomenon can be observed in plate #4.5

73 (q = 0,125 cm3/cm secLf ie,^ a white blur apparently above the film appears in the picture, The blurred portion, though somewhat out of focus appears to be a spray of liquid. It can thus be presumed that the crest of the three-dimensional wave has been entrained into the main stream flow. This type of liquid entrainment also is observed in plates#4.6 (q = 0.167 cm /cm sec) and#4.7 (q = 0.208 cm3/cm sec). This entrainment mechanism is of considerable interest having application to many mass transfer problems in various chemical and aerospace processes, besides that of steam turbine erosion. If relatively large slugs of liquid, formed on the stator blade, are removed by this entrainment mechanism, the probability of forming large droplets at the trailing edge of the blade may be less than without entrainment beforehand. The largest droplets from the (K4) trailing edge of the blade are of order 2 mmK4 However droplets entrained from the wave crest are much smaller, and thus have much less erosion potential. Also they are formed further from the next rotating row. Thus they have more time for acceleration to main steam velocity, losing some erosive potential (Fig. 1.6). Plates #5 are for steam velocity of 227 m/s. Even 3 for very small film flow rates, (plate #5.1, q = 0,010 cm / cm secl no rivulet flow regime was observed. Because of the limitations of the instrumentation, liquid flow rates less than q. = 0,010 cm3/cm sec were not investigated, but presumably a rivulet regimes would have been found there.

74 Plates #5,1 and 5,2 Cq = 0.02) show two-dimensional waves with relatively small wave lengths. Plate #5,3 shows three-dimensional waves, while all the plates between #5.4 Cq = 0,083) and #5.7 Cq = 0.208) show the entrainment regimes. Plate #6 series is for the steam velocity of 305 m/s. Trends concerning wavy film transition are the same as found in the plates #5. Plate #6.1 (q = 0,01 cm3/cm sec) shows two-dimensional waves. Plates #6.2 Cq = 0.021 cm3 cm sec) and 6. 3 (q = 0.043 cm3/cm sec) show three-dimensional waves, and all the plates between #6.4 (q = 0.083 cm /cm-sec) and #6,7 (q = 0.208 cm3 cm sec) show the entrainment regimes. The entrainment regime pictures do not show any peculiar differences for the various steam and liquid flow conditions. Rate of entrainment is of considerable interest in light of mass transfer, but no measurements of entrainment rate were possible. Flow conditions and oscillogram scales are listed in Table 4.1. Based on all the above, a liquid film "Transition Map" (Fig. 4.2) for the experimentally attainable steam and liquid flow conditions was made.(K6) From this map (Fig. 4.2) the wavy film regimes can be identified for given steam and liquid flow conditions, Physical interpretation of the transition zone will be discussed next in as simple a manner as possible, without losing the basic physics.

Fig. 4.2.a Transition Map, U vs q S 330 300 270 240 1 al I I (n N a) >1 U a) 4J En 2101 1801 150 I I I Film Breakup I'-3-diriensional wave wave - \ ^"^\ -2dimens ional.._ _ wave __ Rivulet _Flow > and Droplets Shedding (Entrainment) 1201 901 60 30 (dry patch)'. I I I I I I I U q / 3 cm / cfll, 0.021 0,042 0.063 0,084 0.104 0.125 0'3 Specific Liquid Flow Rate, q, cm/cm.s 5 6 11. 26 16.9 22.5 28.2 33.8 I,iquid Film'Reynolds number, Ref- f.146 0.167 0.188 39 4 45. 1 56. 3

Fig. 4.2.b Transition Ma p, Ti vs q 600 - 550 5315 500M ~~~~~~450L-~~,|~~ ^ ~0.74 450 to 400 P 350 F ilm eak N 300 Q) M Film m 250 I r-A n~.4 200 \ O; \ __ V —.s —-- M N 150 x 0.46 100...... 0 100 _ ^ ^ ^ ^ ________ M = 0.32 ~ M50 -.^-^< ~* ----- * —-- M = 0.18 ~_____-.._ I _ ,. / 60.021 0.042 0.063 0.084 0.104 0.125 0.146 0.1167 0`188 dyne/ q Specific Liquid Flow Rate,q, cm/crr, s f *s' I I I I I I 5.6 11.26 16.9 22.5 28.2 33.8 39.4 45.1 56.3 Ref, Liquid Film Reynolds No, ef _= _h

77 B. MEASUREMENT OF FILM THICKNESS AND ESTIMATE OF INTERFACIAL SHEAR STRESS Liquid film thickness measurements have been made using electrical conductance microgages previously discussed. The use of a "generalized characteristic curve" to reduce the data is explained in Appendix A. Some initial film thickness measurements were already reported 10'HllK7) Among the most important parameters determining the thin-film flow regime are the film thickness itself, since it is necessary to compute the interfacial shear caused by the high velocity steam flow. The vapor flow establishes an approximate Couette-type velocity profile in the liquid, i.e. constant shear imposed at the interface. Wavelets are also initiated due to perturbations in the normal and tangential stresses exerted by the vapor. Increased flow resistance (axial pressure drop) is found when waves are present, due to their effect on "roughness". Consequently, the velocity profiles in both phases may be affected according to the different interfacial wave structures, It is convenient to describe the effect of the interface structure by comparing the drag on the liquid surfaces with that on roughened solid surfaces. Cohen and Hanratty (C9) were perhaps the first to measure the increased drag due to wavy liquid films in terms of an equivalent sand roughness. Utilizing the "universal velocity profile" for the vapor phase, U = 2303 log (41) K Z

78 where u* is "friction velocity", y+ = y u*/y, K = 0,4 is Von Karman's constant. From Nikurads.e's measurements (SSof the friction factor and their own experiments, they concluded the ratio between equivalent sand roughness and root-meansquare wave amplitude of the liquid film (C 2 mm of mean thickness) to be'4.2, Their velocity profiles were also found to be different from those reported for flows over sand-roughened surfaces, (K9) Kordyban obtained similar results from his experiments for thick liquid films (min. water depth - 30 mm) in an air tunnel (max, air vel, = 6 m/s). Numerous empirical two-phase pipe flow studies are reported L5,W7'Hl3) which generally present the interfacial shear stresses in the form p V ti fi v2 (4.2) where Ti is interfacial shear stress, fi is interfacial friction factor, 0 is density of the gas or vapor phase, and Vr is relative velocity between liquid and gas. For the present study, a simple but physically defensible analysis has been made. The film thickness data from the electrical conductivity gages has been used to compute gas phase Vf T4 <..h liquid phase Y t /ug=-Ly

79 the shear, assuming a classical constant shear, Couette flow i.e., a thin laminar film with smooth and flat vapor-liquid interface, and linear velocity profile, A film Reynolds number, Ref is defined, Vh f h Ref = (4.3) f Yf where Vf is the nominal average film surface velocity, h, film thickness, Vf kinematic liquid viscosity. The velocity profile can then be expressed as'i u- -- Y Pf (4.4) where t. is the effective shear stress, jf absolute liquid viscosity. The effective shear stress Ti is discussed later in detail. Specific flow rate, q, i.e., volume flow rate per unit blade width, is obtained by integrating the velocity profile q =J o u dy = Tih2/2jf = Vfh/2 (4.5) so Re = 2 q ff Then from equations (4.5) and (4.6). 2.= Re 1/2 -- V -- f (4.6) (4.7) Figure 4.3 shows schematically the flow regime in the test section. The cross-section of the active portion of the steam flow channel consists of three "Lexan" plates, and the liquid film surface on the blade. Flow is from

80 wavy liquid film of thickness h Cs: shear stress from smooth wall -T:: shear stress from wavy film L: axial length of the section h: average film thickness p: pressure Fig. 4.3: Shear stress schematic of test section with liquid film

81 left to right. Shear stresses exist with the three (hydraullically smooth) Lexan plates and the liquid film, which has an effective roughness due to surface wavelets in most cases. Applying an axial force balance to this geometry, one obtains the following relation: a bAp = (a+2b) L Ts + a L Ci (4.8) s 1 where a and b are the widths of the Lexan walls, L is the test section axial length considered, ie., blade length, and AP is the axial pressure drop for distance L. - is s the shear stress due to the smooth plastic wall, and T. that due to the roughened liquid film. Also the pressure drop Ap Can be expressed as: L _sUs Ap= f L = DH 2 (4.9) where f is friction factor, DH is hydraulic diameter, 5s the steam density, and U is steam velocity. 5 If it were assumed that approximately the interfacial shear stress C. is the same as the smooth wall shear stress, -s, then Xi = s'= (4,10) Combining Eq. (4.8),(4.9), and (4.10) one obtains L us 2 -C 2 (a+b) L f s s (4.11) 0 D 2 (4.11) Substituting the actual dimensions and conversion factors, Eq. (4.11) becomes: 2 L = 5.4 x 1Q-3(f S ) (4.12) wheeimson o 2 where dimension of o0 is lbf/in2 ls Ibm/ft3 U ft/sec. f~ ~ ~ U I -ec

82 To improve the precision of Eq- C4,101 separate friction factors can be considered for the walls and the film. Then, weighting f according to wetted perimeters one obtains: T= 5.4 X A'%3 1a+22b a \ Ps s To 54b x 10 (T) + 2 (a+b) i) 2 (4.13) where fs is the friction factor due to solid walls, and Si to the film, f is a function of steam Reynolds number, Re, 5 s only butfi is a function of Res, and film equivalent sand roughness, e D - e., (4.14) DH 4i - % (Res,' ) (4.15) 1 D Using Eqs. (4.5) and (4,15), one obtains the equivalent sand roughness as a function of film thickness, h; i.e. = E (h).. An iteration method was used, involving liquid flow.rate q, measured film thickness h, steam velocity U, and the density and viscosity of each fluid. (K7) First results using this method have been already reported. Figures 4.4a through.4.4g show the measured film thickness as functions of steam velocity and liquid flow rate, and equivalent sand roughness as a function of average measured film thickness. Interfacial shear stress, Ti, equivalent sand roughness, EChI, and friction factor, fi, as functions of steam velocity and liquid flow rate are listed in Table 4.2. In Fig, 4,5 the variation of friction factor, fi, is

83 1, ci) c).-4 P-4 -H;7 250 200 150 100 50 pmn I.0. f 0 9.C 8.C 7.(.1.0 I I (.* == Specific q - 0.01 - I Flow Rate cm3/cm sec 0 - measured (Gage 2) X - smooth wall ( = 0 ) A - rough wall. (6 h) 5317 5.r J _4.C - 3.( 2.( m i = 4.2 h 1. mi: 3.8! 3.; 3.8 I. I I -- ---------- 60 120 180 Steam Velocity, U (m/sec) Fig. 4.4.a Flim Thickness vs. Steam Velocity 240 300

84 250 200 Specific Flow Rate q m 0.021 cm3/cm-sec 10. I - 9.4 O - Measured (Gage 2) X - smooth wall(.= 0 ) A - rough wall i= h) 8.C 7.C = 150 u le 100 f10 5318'5.t - 4'.o 3.0h 50 F.0t %% %A 1% x A.2-44 - g..4... 5.2h 3.8 3.8 3.0 1.0 mil.., - d-A& 1 I 1 I 60 120 180 240 Steam Velocity, Us (m/sec) Fig. 4,4.b Film Thickness vs. Steam Velocity 300

85 250 200 150 U, c.) c r-4 I LO.0 8.C 7.C 11.0 9.( L Specific Flow Rate q - 0.043 cm3/cm-sec O- Measured (Gage 2) X- Smooth wall ( =-0 ) A- rough wall ( = h ) m A m, m 5319 5.( m \0 \ 100 _4.< m. 3. m 50 2 5.0' 1.4 t mi] _ = 5.5h I 6.2 6. I I I I 60 120 180 240 300 Steam Velocity, Fig. 4.4.c Film Thickness vs. Us (m/sec) Steam Velocity

86 250 200 150 0. V&" 0; w a V-4 -H;L4 I 9.0o X_ Smooth wall (6= 0 ) A- Rough wall (e= h ) LI.Or 3.0t 7.0 Specific Flow Rate q - 0.083 cm3/cm-sec 0- Measured (Gage 2) J6. 0 LO -r 5.0o X 5320 \ 3.0I 50 pm -2.0 "ft _. 0 3.0 1.0 I e - 6.Oh 5.2 4.8 mil I I I 60 120 180 240 300 Steam Velocity, Us (m/sec) Fig. 4.4.d Film Thickness vs. Steam Velocity

87 EL Or Specific Flow Rate q = 0.104 cm3/cm sec 10. 0 250 1 9.0 o 8.01. 200 _ O- Measured (Gage 2) X - Smooth wall (C = 0 ) - Rough wall (6=h) 5321 7.0 A 150 -6.o 0. m u -PE- aP-4 5.0o 1o00 p 3.0 %% - T'.I la 0*% %b A=m ww ~ M~-0 -2.0 50 I pfm 1$0 mi 4.4 4.4 = 5.2h 1 w 4.6 I I 60 120 180 Steam Velocity, Us (m/sec) Thickness vs. Steam Velocity 240 300 Fig. 4.4.e

88 250 200 150 l.O1I 0.O 9.0 8.C 7. C.6.C I A.,,. m.,. o- Measured (Gage 2) X Smooth wall (- = 0 ) A- Rough wall (6= h ) 5322 X_ ^^^X A x6~ Specific Flow Rate q - 0.125 cm3/cm-sec 5.( i 4 r. E-4 4.C tm 100 \ 3.C m om-M. mm A M s~~f __ _, An+_~C 50 Ij pm -2.1 1.C [mi e = 5.Oh I 5.0 5.0 3.1 I 1. I I 60 120 180 240 Steam Velocity, U (m/sec) Fig. 4.4.f Film Thickness vso Steam Velocity 300

89 ll.0 o Specific Flow Rate q - 0,208 cm3/cm'sec 250 10.0 9.-0 200 8.0 O- Measured (Gage 2) X - Smooth wall (E o0 ) - Rough wall ( =h ) 5323 7.0 ci C; E r-4 -r (w 150 6.0 \ \ 5.0 100oo P-0 3. O A-i-,.. ~ 4fttm I alm'nwslft 50op e = 5.2 h 8.2 6.2 4.2 m ufm 1.0 mil 1 1 1 I 60 120 180 Steam Velocity, Us (M/sec) Fig. 4.4.g Film Thickness vs Steam Velocity 240 300

90 Table 4. 2: ESTIMATE OF INTERFA CA L FRICTION FACTOR Z. 1 V 2 i s E: equivalent sand roughness dyne/cm = 18.1 |: =47.09'C = 81. 87 Z = 196.48 s L s s Q Re4 Re s V (fps) 240 (73. 2rrs) 435 (133n/s) 625 (191m/s) 1000 (305=') (c fRes 4.2x10 7.6x104 1. 1. 75x105 q(cm / ) ^ M q(c ) 0.18 0. 32 0. 46 0. 74 Ref _ 5 f. 0. 006. 0059 0.005o9 0.0059 0.010 4.2 h 3.8 h 3.5 h 3.5 h 2.8 -i 17.68 57.16 118.00 302.08.....i D, 1... 1 " -0.0065 U0. 0065'.006Z 0. o061 0.021 5.2 h 3.8h 3.8h 3.0 h 5.6 r.i 19.6 62.97 124.00 312.32 20 f. 0. 0073 0.0075 0.0068 1 0. 0067 0.043 s 5.5h 6.2h 6 2h.5.0 h 11.26 -i 21.53 72. 66 136.00 343.04 30 0076. 00767. 007 7 0. 0077 0.063 4.5 h 4.5 h 4. Oh 4.0 h 16.9 -. Z2. 41 74.6 154.00 394.24 40 f. 0.0075 0.0079 0.0078 0.0077 0.083 5.22 h 4.5 h 4.5 h 4 0 h 22.05 22.12 76.53 156.0 394.24.-Ci 39. 24 0.104! 5.2h j 4.6h 4.4 h 28.2 ti 22.12 77.51 166.0 435.20 60 fi 0.0078 1 0.0088 0.0090 I 0.0120 0.125 h50h 5.0 h 5.0 h 3.1h 33.8 2i 3.00 85.26 180.00 1 614.4 80 i. 0.0090 0. 0090 0. 0088 0.0095 0.167 s 8. Oh 62 h 6.Oh 2. 5h.45. 1 Zbi 26.54 87.19 176.00 486.40 100 i 0 0 10 0. 0 12 T 0 015. 0098 0.208 5s 5.2 h 8.2 h 4h 4.2h 56.3 X 2i 9.49 116.25 300.00 501.76 ~~~~~~~..........,.?......

* Liquid Rate Film. 3 /msReynolds ( cm /cm's) No. 0.010 0.009 0 v 0 a q o 0.010 A 0.021 0 0.043 V0.063 Ref 2.8 5.6 11.26 16.9 s-I 0 4iU (I4 N 0 4-H s.rA kA rl4 0.008 0.007 V 0 V T 0 v 0 v7 t3 0 V 0 V vl 9 0 00.083 22.5 *0.104 28.2 l0.125 33.8 10.167 45.1 *0.208 56.3 V2. 1. -;. = _ J "' A A 0.006 h 0 0 0 0 0.005 I I Steam Reynolds Number, Res Fig. 4.5 Friction Factor vs. Steam Reynolds Number

92 shown to be more dependent on the liquid film Reynolds number than on the steam Reynolds number. In general, friction factor increases for an increasing film Reynolds number, Ref. This is as expected, since film waviness and thus equivalent roughness, increase with Re f For high Ref, fi increases for an increase in steam Reynolds number, Res, indicating that increased steam turbulence increases waviness, as expected. However, no direct analytical confirmation is available. Author's photographs show large three-dimensional waves and the entrainment of wave crests at these conditions. Both phenomena would likely increase effective friction factor. In Fig. 4.6 and Table 4.3 ratios of measured to predicted film thickness, based on the smooth wall assumption, are shown. These are always smaller than 1, indicating an effective film roughness for the film, and generally increase with Re. The average ratio of the measured film s thickness to the predicted film thickness in the author's experimental range is: h m - 0.55 (4.16) ht for or 2.8<Ref < 56.2 4.0 x 104 < Re < 1.75 x 1Q5 There are numerous thickness measurements for high film Reynolds numberf(Ref > 10Q) particularly for pipe flow. These are well summarized in reference.K However, for the low film Reynolds number range Cauthor's case), such

0 4) -,'U *H rd 04 0 o4 Q) 0 / -4 u r-i *d r-4 * (U o - 3 0 (d (0 2,-I 0) 4 0 <U 0*d0 4J1-, rd JT (^ 3 1.0 01 0 o Res = A Res = o Res = 0 0 o 0 4.2x104 7.6x104 1.1x105 1.75x105 5325 0.8C 0 0 0.60 O 0 0 0 0 0 O 0 8 A 0.401 0 0 aC 0 0.201 0 I I I I I I I I /Rf 8 16 24 32 40 48 56 64 Film Reynolds Number Fig. 4.6 Ratio of Measured Film Thickness to Predicted Thickness (smooth wall)

94 Table 4. 3: COMPARISON OF AVERAGE MEASURED FILM THICKNESS AND PREDICTED FILM THICKNESS (SMOOTH WALL) h: Measured film thickness (mils) m ht: Predicted film thickness (mils) h /h: Ratio of above two thickness mn t V (fps) 240 (73. 2rr/s) 435 (133mn/s) 62Z5 (191m/s) 1000 (305nvrr/s) CQ S Re 4.<x104 4 5 5 Average cc/C,, Re 4. 2x 7. 6x10 1. x0 1. 75x10 h 2 s h q(cm/s) M 0.18 0.32 0.46 0. 74 h Ref t 5 h 2. 67 0. 90 0.67 0. 55 0.010 h t 2.92 1.69 1.24 0. 79 2.8 hm/h 2.8 th/h 0.91 0. 53 0.54 0.70 0.67 10 h 1.00 0. 87 1.00 0775 m 0.021 h 4.12 2. 39 1 1.74 1.11 5.6 h /ht I 0.24 0.36 1 0.58 0.68 0.47 20 h2 2 67 1 22 1. 07 1.25 0. 043 hr' m t0 0.46 0.46 0.43 0.8 0.51 30.26 h.82 j 2.8 2.73. 3 1.5O 7.. 0.063 ht 7.14 4.14 3.03 1.93 16.9 ^hT/1t| 0.54 0.6 0.57 0. 52 0.56 40 h 3.40 2.88 1.63 1.52 0.083 h 8.24 4.78 3.49 2.23 Z2.5 hr At 0.41 0. 60 0. 47 0.68 0.54 50 h -4 4.5 j 3.03 2.27 2.35 m 0.104 ht 9.21 5.35 3.90 2.49 28.2 h /h 0,48 0.57 0.58 0.94 0.64 60 h 4.65 - 2. 37 2.00 m 0.125 h 10.10 5.86 4.28 2.73 33.8 h ht 0.46 55 - 7 80 h 5.57 2.85 2.25 1 2.45 0.167 h 11.65 6.76 4.94 3.15 45.1 h/ / M 4.t 0.48 0.42.0.46 0.78 |. 54 100 hm 5.43 2.88 3.00 2.10 0.208 h 13.03 7. 56 5.52 3.52 56.3 h /Ht 0.42 0.38 j 0.54 0.60 0.49 Avg.. h l 0.49 0.48 0.52 0.71 0.55 Avg hm/t 0.49 0.48 0.52 0. 71 I I 0.55

95 thickness measurements hardly exist, even for the pipe flow case. In Fig, 4,7 a non-dinensional film thickness, hC i 1/2 h P f ) h+ = is plotted as a function of Ref. The form of h+,generally accepted in most thickness measurement literature, is based on the previously estimated shear stress, T. The only available data, in this low Ref range is from the measurements of Charvonia (C12) and Collier-Hewitt. (C3) They both studied air-water flow in a vertical pipe. This data is graphed using solid symbols, while the author's data is plotted with open symbols. The first solid line represents the ideal case, with the shear stress of Eq. (4.7) found assuming a smooth wall. The second solid line is the empirical correlation of Hughmark,(H16) based on data obtained by others for upward flow in vertical pipes. The shear stress used in the friction velocity term for the dimensionless film thickness was not clearly specified. It appears that he used wall (rather than liquid film) shear stress. The author's data can be correlated by the following expression, using the previously estimated interfacial shear stress, 1/2 h+ = 0.6 Ref (4.17) This curve is the bestrfitting curve for all the data points in author's entire experimental range, Comparing with the Charvonia and Collier-Hewitt data, one finds that their nonrdimensional film thicknesses are always higher than the author's data for the same Re,

96 Fig. 4.7 Film Thickness Factor as a Function of Film (g Reynolds Number (Comparison with Hughmark Correlation, 1973) 10 8 a 6 - to:from smooth wall assumption 4. 1/2 1/2 ho ( ) = Ref1/2 4 3 2 A i O~ h += o. 6 (R f)/2 Kim and Hammitt correlation (1977) -IN 14 ".-cm Hughmark correlation (1973) 2 < R' < 10o 0 - A - 0 - 0 - a - Res= 4.2x10 Res 7.6x104 Res= 1.1x105 Res= 1.75x105 Charvonia, vertical Collier & Hewitt, vertical 0 3 4 5 6 78 10 L0 Ref, Film Reynolds Number

97 These deviations can be explained in that their data is for a relatively thicker film and lower gas velocity, while the author' s data is for very thin films with very high steam velocity, sometimes accompanying the entrainment of liquid film. In Fig. 4.8 author's data is plotted in terms of (H7 ) the Henstock-Hanratty ) parameter F= y Ref Vf 1/2 (4.18) Re —0.9 V (4s18) (H7) which is a Martinelli-type flow parameter. Here. = [(0,707 Ref 05) 5+ (0.379 Re 0.9) 2.5] 0.4 Their correlation h 6.95 F (4.19) DH (1+850 F) ti/2 was only tentatively suggested due to the lack of sufficient data. This correlation over-estimates the author's data. C. RIVULET FLOW REGIME As already discussed, a thin liquid film flowing over a solid surface under an applied shear stress due to a high speed steam or gas velocity, may separate into a number of filaments (Figs. 4.1 and 4.2)., Such rivulets may then occur on turbine stator blade surfaces, In such cases, the spectrum of detached droplets impinging on the next rotating row depends on the rivulet characteristics. This type of film break-up is also related to boiling'"dry-spots" in

98 Fig. 4.8 Nondimensional Film Thickness as a Function of F (Comparison with Henstock and Hanratty Prediction, 1976) h -6.59F DH (1+850F)1/2 Henstock & Hanratty 0o 0 0 0 oo 0 00 o — W. Kim & F. G. Hammitt (1977) 5327 0 0 10' 00 O O 0 0 0 10-5 F = r(Ref) Res0.9 r (0.707 2 0 Ref 2.5+ (0.9 2.5 0.4 Ref ) + (0o0379 Ref I

99 emergency cooling of nuclear powerplants, as well as to various chemical engineering applications, Because of applications to numerous important engineering problems, several models describing such dry-patch or filament stability have been suggested CH2,H3,B2,M3 These analyses are generally based on force or power balances at the junction between rivulets or dry patches and the continous film. Since the surface tension plays an important role along with the dynamic pressures of liquid and vapor flow at the stagnation junction, the relations for critical (or minimum) flow conditions for rivulet stability should include the contact (or "wetting") angle, e, between liquid and solid. Unfortunately, e depends heavily upon other parameters such as surface condition. Experimental studies on the formation of rivulet flow were also conducted by Hewitt and Lacey(H) for air-water flow in a vertical pipe. Even though their experimental geometry and flow conditions are quite different and thus cannot be directly compared with the author's experimental results, some pertinent points should be made. The upward air velocity used in the Hewitt-Lacey work ranged from 18 to 95 m/s, while the downward flow critical liquid film thickness from 30 to 210 pm, This thickness was not direct(H14) ly measured, but estimated using the Hartley-Robert's model. The film breakdown was studied by artificially producing a dry patch using an air jet directed perpendicularly toward the tube wall. it would no doubt have been more

100 realistic had they produced the rivulet or dry patches without using such an air jet. They concluded that for the higher gas flow regime, (H2) minimum wetting rate (which Hartley and Murgatroyd used to express the minimum liquid flow rate required to re-wet the surface after the formation of a dry patch) decreases with an increase in the gas flow rate. At lower gas flow rates, however, the minimum wetting rate passes through a maximum, and then begins to decrease with a decreasing gas flow rate. They ascribed these results to a flow regime change between the climbing film regime and "churn flow", which occurs at higher gas flows in this vertical annular pipe flow. Such a maximum critical wetting rate was not found by present author. They also observed that at liquid flow rates above the minimum wetting rate, the patch was more or less circular. Rewetting of the surface occurred in a complex manner, usually by splitting of the dry patch by filaments, followed by the disappearance of the ends and sides of the resultant secondary patches. This kind of rewetting process, Fig. 4.9, also was observed on a laboratory turbine blade. (F2) For cases with strong wall heat flux, additional effects need to be considered. Examples are the thermocapillary effect(Zl) due to the temperature dependent surface tension and vapor-thrust effect (Nl' caused by surface evaporation. These models show that such thermal effects become dominant for the strong wall heat flux cases in determining

101 5397 Fig. 4.9 Rivulet Formation on Experimental Turbine Blade = 5000 rpm, d= 380 mm, M =0.7, p=0.6 bar steam ( supplied by Visiting Scientist, Dr 0. Povarov, Moscow Power Institute )

102 the stability of dry patches. These thermally dominated situations are most directly applicable to the nuclear power plant "loss of coolant accident" (LOCA,. while the isothermal considerationsof stability of dry patches are directly pertinent to the steam turbine problem here emphasized. No new analytical model was developed by author since many-.models already exist. The need for better experimental data seemed more pressing. However, some modifications of existing models are suggested in the following discussions considering author's experimental results. (H2) Hartley and Murgatroyd (H-M model) present the first relatively comprehensive theory for the stability of dry patches. Mikielewicz and Moszynski (M-M model) involves a more complete analysis, which can more usefully be compared to author's experimental criteria. H-M Model(H2) Hartley and Murgatroyd suggested two models for the stability of dry patches. The first model was based on the balance of forces at the liquid flow stagnation point between dry and wetted regions (points G in Fig. 4.10a). If the dry patch is stable, the surface tension forces must balance the fluid pressures over the stagnation line GsGp in Fig, 4,10a. Surface tension and pressure terms can be expressed respectively as, Fr = dX, Cl-cos Q) f AB 2 Fp = dXJ Of [W(y)] /2 dy O

103 z --- x W(y) D F J H C F = dx j W() /2 dy F =dx.60. (1-os) F_ = dx.r. (1 - cosS) By Fp = Fr h= 1.82 [cr(1 - cose)/] qc = 33' 3 ( ^/ ) /3 1/3 2/3 c3 ( 1 - cos 2/ [<r ( I - cos &)' 9 should be 4v8~ in comparison with our data Fig 4.10.a: RIVULET FLOW MODEL HARTLEY and MURGATROYD (1964) 1. FORCE BALANCE MODEL

104 where Q is the contact angle between the liquid and solid T liquid surface tension! dX the unit length in X direction, W(y) the velocity profile, 6AB the fluid thickness before the stagnation point, Assuming a linear velocity profile and Fg = F for neutral equilibrium, the minimum (or critical) film thickness and flow rate become, h = 1.82 [Cr(l-cos 8)/9]1/3 / 2/3 (4.20) q = 3.3 [r (l-cos e)]2/3 ( / )1/3 (4.21) where h is the critical film thickness and qc the critical flow rate, pf the liquid viscosity, Ti the interfacial shear, and ~ the liquid density. A comparison between author's experimental criterion observed from the "map", Fig. 4.2, and Eq. (4.21) discloses that the contact angle, 9, would need to be between 4 8~ (see Table 4.4). However, the static contact angle of a water drop on stainless steel is known to be about 50~ (see Fig. 4.11 and Ref. H4) at room temperature and atmospheric pressure. Thus the values required by the H-M model are not realistic. Of course, the contact angle is dependent on many parameters such as temperature and surface conditions Cwet or dry, cleanliness, roughness, time of exposure*, etcl, A possible source of this apparent discrepancy is the geometric consideration of the stagnation point in their model, (See point G of Fig. 4,10a). *Shown by author's work measuring changes in 6 as droplet size decreased due to evaporation CFig, 4.11).

105 Table 4.4 CRITERIA OF RIVULET FORMATION COMPARISON WITH MIKIELEWICZ-MOSZYNSKI (M-M) AND HARTLEY-MURGATROYD (H-M) THEORIES U (m/s) 73.2 133 191 305 s Q( cm3/min) 10 7.5 2.5 2.5 Re 5.6 4.2 1.4 1.4 -C.(dyne/cm ) 22 60 110 280 x 10 cm 8.13 4.57 2.69 1.78 h mils 3. 1.8 1.06 0.70 h 0.0060 0.0058 0.0060 0.0070 e (M-M) 10 9.5 10 11 - deg 9 (H-M) 8.4 5.5 4.7 3.8 -deg f = 1 gm/cm3 /{ =7.4x 103 dyne sec/cm - = 72.8 dyne/cm h-fo h /6 2.(T

106 5398 Fig. 4.11 Photograph of Contact Angle Between Water Droplets and Stainless Steel Plate at Room Temperature. (taken by H. Hamed and W. Kim, 1977)

107 In the H-~M model CFig, 4,10al, the stagnation line GSG is assumed to decrease smoothly from the upstream thickness GABA However, author s rivulet pictures (plates #2.1, 22 and 3,2 of Fig, 4.1) show an abrupt bulge in the liquid at the stagnation point ahead of a dry patch. There is no doubt that the bulge geometry may change the force balance considerably. Another factor, which complicates an adequate modeling, is the existence of waves in the upstream liquid film. Such waves always exist in the author's rivulet pictures. They suggest that a modification of the H-M force balance model to include their effect is necessary. The other H-M model, called "minimum power model", is based on the assumption that the film will attain a condition such that the sum of surface and kinetic energy across a rivulet cross-section will be a minimum. For such a stable rivulet (Fig. 4.10b), the kinetic energy flux Ek is k PQ 3 Ek x 9 [W(y)] 0o.- 2 dy where W(y) is the velocity at element PyQy. Also surface energy flux E6. is EC- X,, WSpl Then the total energy flux, ET is The minimum for ET is obtained by setting ET/bs = 0

108 A E F B M z EK = 3 L H Y N ip r ------ Q-R =x.O. W(SpQ) DK J C E = E + E = 1.59 / / Fig 4.10.b xRIV T FW MT I PsHARTLE Qsand MURATROUD (1964) dy... x Pp By 3ET/ 6 =o hc = l- 59 (6 Y3 /% In/t, ) qc = 2. 5 (, ~ /t /r1/3 2/3 Fig 4. 10.b: RIVULET FLOW MODEL HARTLEY and MURGATROUD (1964) 2. MINIMUM ENERGY MODEL

109 Assuming a linear velocity profile in the liquid as in the first H-M model and author s analyses, one obtains the critical flow rate as h = 159 tCr/?1/3 lf/ Ti23 (4.22) qc = 2,52 C fPf/il/3 o 2/3 (4.23) Equations (4.22) and (4.23) do not include the contact angle, rather assuming the rivulet cross section to be rectilinear, i.e. 6 = 90~. Thus a portion of the surface energy of the rivulet was neglected, Comparison between author's direct measurement of the critical film thickness and the H-M power model (Eq. 4.22) shows that the predicted critical film thickness is -4x that measured by present author (Table 4.5). A partial explanation of the difference may be their assumption of a rectilinear rivulet cross"section. In light of the minimum surface energy, the cross-section should not be flat but rather it approaches a circular segment. This shape requires less energy than a flat form. Present author observed from his rivulet photographs (plates #1.1 and 3.1 of Fig. 4.1) that a stable approximately flat cross sectional rivulet continued to be stable. Of course, the exact rivulet shape is not known at this time. Another possible reason for the difference in measured and predicted critical stability parameters is the negligence of wave effects in the rivulets for the HrM minimum power model. Two-dimensional waves moving in the rivulet flow

110 h. ] ~73.2 133 191 305 x10 3(cm) Measured,-(W. Kim) 8.13 4. 57 2. 1. 78 (W. Kirn) Predicted 2 16 11 | (H&cM) \ 32 16 11 6_ (Hl&M ),,~~~~, I,,=, ---,, I,, I Predicted Critical Film Thickness is - 4 times larger than Measured Film Thickness. Table 4. 5: COMPARISON OF THE MEASURED AND PREDICTED MINIMUM POWER CRITERION OF H-M MODEL

111 direction were observed Csee plates in Fig, 4.1) and should be considered in any rivulet film flow stability model. M-M Model More recently Mikielewicz and Moszynski M3proposed a modified H-M minimum power model based on Hobler's (3) basic approach. As an improvement over the H-M model, they assumed the rivulet cross-section to be a circular segment. Assuming that the rivulet is stable when the total energy flux (kinetic plus surface) is minimum., they obtained the following equation for the critical film thickness for critically stable rivulet flow, h+ 3.2-2/3( cos e(1/3 2/3 2/3 s ( 1-cos 3.2 (sin.' Cosee g(9) 1-CO(/ sn (4.24) where 2 3 h+ = f Ti h is the non-dimensional thickness, 3 13 15 8 3 g(6) = -1/4 cos3 e - cos + 8 sin, 8 8 sin 28, (6) = s sin 3 - es cos 6. In this analysis, the velocity profile of the liquid film was assumed linear, They presented a graphical solution for the above expressions (Fig, 4.121 Comparing the author's critical film thickness measurement with the predicted M-IM model, the contact angle 8 should be about lQq for the model to be compatible with the experiments. (see Table 4.4). This presents substantial improvement

112 Q so 20 30 __ o 50 __ __ I3 l o 10 20 30'o 50 6o 70 a0 9 where pr2h3 Fig. 4.12 Dimensionless Minimum Film Thickness as a function of the Contact Angle ( Mikielewicz-Moszynski, 1975 ) 5330

113 over the H-M minimum power model, but still the required contact angle, 8, is less than half that of measured contact angle(^' 25~) (15) in such a single curvature flow as rivulet. An important reason for the discrepancy with author's experiment may well be the neglect of rivulet film waviness. D. WAVY FILM REGIME When the liquid flow rate is increased at fixed steam velocity, the liquid film covering the blade surface becomes continuous, and rivulets and dry patches disappear. For the range of steam velocities here used, the surface of the liquid film is only rarely entirely smooth. Rather it is nearly always somewhat wavy (see plates in Fig. 4.1). The shape of these wavelets for low steam velocities and liquid flow rates is two-dimensional with relatively small wave lengths (^0.5 cm) in direction of flow. With further increase of liquid flow rate or steam velocity, the wavelet shape becomes more complicated with much larger wave lengths (, 2 cm). The author has defined this wave form as "threedimensional." These larger waves are usually preceded by a parasitic group of much smaller wavelets. In the "film map" Fig. 4.2, the division between two dimensional and threedimensional wave regimes is indicated by the broken line, This division is not "sharp", so that in the three dimensional wave region (Fig. 4.2), two dimensional waves sometimes exist, appearing and disappearing. Extensive measurements of film thickness, wave length, and wave propagation speed have been performed for the

114 different steam and liquid flow conditions attainable. Test points (Tables 4,6, 4.7) were selected to cover with adequate detail the full range of steam velocity (steam Mach number) and film flow (film Reynolds number). Calculated results include interfacial shear stress (section B of Chapter IV) wave number, capillary wave celerities and other important parameters. The measured wave length data for the two-dimensional and three-dimensional waves are shown in Figs. 4.13 and 4.14. Figure 4.13 shows that the wave length, A, of twodimensional waves decreases as the liquid film Reynolds number, Ref increases, for lower steam velocitieswhile it slowly decreases with increasing Ref for very high steam velocity. The Mach number is then high subsonic (M=0.61 and 0.72), so that compressibility may importantly influence the flow. In general the wave length,\, decreases with increasing steam velocity or Mach number for a given Ref. Also the data scatter decreases with increasing steam velocity, perhaps because the relative precision of the measurement of these larger parameters improves. For three-dimensional large waves, X, deceases with increasing Ref at lower steam velocities, but increases with Ref for higher velocities. This trend is not found for the two-dimensional waves. Figures 4.15a and 4.16 show measured data on propaga_ tion speed of two and three dimensional waves for various Ref and steam Mach numbers, M. These were directly measured

1. 1.2 1.0 V s (m/s) + 21 0 37 v 61 0 93 A159 0252 X298 M Ti (dyne/Cm ) [.05] 2.5 [.09] 6.2 [.15116.2 [.23]32 [.38]80.2 [. 611190 [.72]340 4J t( CI 0 0.,.~ Q)., 0.9[ 0.8L 0.7 H 0.6 - 0,51 0.4 0.3 0.2:1 1 0.1 L t I! 1 10 Film Reynolds No., Ref Fig. 4.13 Wave Length,, vs Film Reynolds No. Ref (2-D waves) 100 5331

3.S 2 to 0, 30 U< 1( C2 0 3 t + 21 0 37 V 61 0 93 A 159 0 252 X 298 (3-D waves).05.09.15.23.38.61.78 5332 sa M (m/s) Fig. 4.14 Wave Length, X, vs. Film Reynolds No. Ref (3-D Waves) } f 1t i it If I I I I I a 9 I I A A a a A. I.0 —-------—' ----------' -.- - --. X u *L - * xX- I - I - _ I. _ - L - - I Film Reynolds No., Re 10 100 t

117 [V (m/s)] s 200r o - M =.05[21] * 0 - M= [ O - M.2 =~9 i Average A - M = 3 ] A Values XX 0 - M =.61 252 8 X - M =.72 [298 10 small waves /,, M-0.72 / A 0 / I x 150 A A M=0.61 M=0.39 -) CIJ 13 U 3 100l,- M=0. 23 -— M=0.15 0 0 M=0.09 M=0.05 50 OO a7 o _ ^ O O 1 10 20 30 Film Reynolds Number, Ref 5333 Fig. 4.15.3 Wave Propagation Speed, C, vs Film Reynolds No. Ref (2-D waves)

30 I I Iu u 20 10 to o.4 c: 1s 0 EO 10 1 t 0 V _ O 0 X V s (m/s) 21 37 61 93 159 252 298 M.05.09.15.23.38.61.72 (3-D waves) Fig. 4.1!-Wave PropagationSpeed C vs Film Reynolds No, Ref (3-D Waves) I. 00 I'II As I I I I I I I I I I I a a a a a I~~~~~~~ taII I. I I i I I 10 Film Reynolds No., Ref 100

119 from the high-speed motion picture frames. The wave propagation speed C, increases with both M and Ref for two-dimensional waves (.ig. 4.15.a). Solid symbols represent average values, while open symbols are individual points. To help clarify the operative wavepropagation mechanisms, the celerity of capillary-gravity wave Vc has been computed (Table 4.6). The classical solution for combined capillary and (C17,L6) gravity waves (long waves), Vc, is expressed as V [ + 6k 1/2 (4.25) c k 5k where g is gravitational acceleration and k = 2 T/X is wave number, Comparing measured wave speed, C, with capillary wave speed, Vc, it is found that the measured speed is 1.86~ 0.36 x that calculated from the conventional capillary-gravity model for author's two-dimensional waves. Thus the celerity of the two-dimensional waves is in reasonably good agreement with the classical capillary gravity model. Recently Wurz(Wl4) proposed a two-dimensional film wave propagation speed model, compared with his own experimental data. Using'universal" velocity profiles (laminar and laminarturbulent) for the film, two lixtiting correlations are suggested; C*= 2 32 Re1/2 and C* - /4 91/3 1/3 C* = 29/ 9 Ref

Table 4.6 FILM WAVE CHARACTERISTICS ( 2-D WAVE) Vs q h| V C VcI W W.; k o< C/V W G o< C/V C (m/)) e (cm, (cm) (cm/)ml (cm/ ) Weber Weber /dyne\ f -nre on- ee C = ~~~~Stea.~m Spei N~~(on- Wem be nr Cut- cd-m7 Steam Specd- Film Inter- Wave'Wave Capil- Numbe umbe 2 Film Dimen- F roude si ona C NonLiqud nesV Film at Speedton Wave ) 1 al nold Wave Wave isVf hr Numb speed Wave FwVciedeloty Speed Seed / Cr umbe iumbe e fa.010 3.5 2.34.061 2.0 2.8 6.71 160 0.070.021 4.7 3.49.082 2. 1 5.6 20.0 96.5 0.092 1onal. 2h. 3. Wav.923 49.0 22.3 6.34.113 2.4 16.9 6.81.112 6.47 130.4 28.4 0.156 22 31.0 21.4 1.23 38.0 19.3 8.45' " 5.11.084 5.02 " " Z.0 24.5 (M=0.05.063 |6.5 |7.57 1.46 45.6 |17. 7 10. 0 4.30.071 6.02 " 6 2. 29. 4.769 31.1 24.4 5.28 " 8.17.135 4.11 " " 1.3 20.1 2.15 27.4 14.6 14.8 " 2.92.048 3.62 " " 1.9 17.7.615 62.0 27.3 4.23.156 2.8 28.2 10.2.231 6.73 264 26.2 0,174 2.3 37. 1.577 34.2 28.2 3.96 " 10.9.246 3.71 " m e 1.2 20."4.185 76.0 49.8 1.27' " 34.0.769 8.25 " 1.5 45.4 38.0 R a te n 4.13 * 22.7. 0I0 3.5.34 I I0I1 1.0. 8 6.6-0 —-— I- -_-1_ - I r0o o I. I -- I - I - - I - I - I I - - - - - I - I - l l 4 l l, —-. lI.l. l l I-..

Table 4.6 FILM WAVE CHARACTERISTICS ( 2 - D WAVE) Vs (m/s) Steam Velocity I - q/ (cm /s) Specific Liquid Flow Rate I h (mils) Film Thickness r I I I.1 4 V v' f, (cm/a) Interfacial Film Velocity - I I x (cm) Wave Lengtl C (cm/s)'.ave Propa gation Speed r 1( - Vc I WA (cm/a) Weber Capil- INumbe lary Wave Speed'rA I v2 Its a1 ~f T I }1 W.h Weber Numbel 2 - tc /dyne\ tcmr anterfa cial 1 - Ref Film Reynolds Numbe -Vf k (cm'1) Wave itumbe! 0< NonDimensional Wave Number (kh) - C/Vf NonDimen sional Wave Speed I - Wf Weber Numbe'PVCrh G nverse Froude Numbe: /-h f 0<c C CutOff Wave Numbe r C/v ty.on - dime - sional Wave speed I Shear Stress * C = c/f- 7 N.onaimensiona Wave perz14 Wurz) xlOt4 xlO31 4 4. 4 I. I. I I.010.021 2.4 3. 7 3.42 4.43 I.385 34.2 34.5 I 7.74 I.189.123 4.9 5.6 I 5.6 2.8 1_ 16.3 1.153 7.72 I 9.8 51.3.10 25.4 47. 1.115 I 4 231 36.0 44. 5 4.64 II II II 27.2.256 8.13 It It 1.0 0.8.385 38.0 34.5 7.74 *" " I 3 16.3.153 18.58, " " 1.1 16.1.231 22.6 44.5 4.64 ", " 27.2.256 5.10. " 05 9.6.....,............ -.....~~~~~~~~~~~~~~~~~~~~~~~~~. t 14. 5 15. 2 I..308 22.8 38.6 6.19 I I 20.4.192 5.15 I II II t 0. 6 I 9. b 36.6 ( M= 0.089.462 46.3 31.5 9.28.327 7.0 16.9 13.6.221 6.02 132.5 27.1 0.176 1.5 117.6.385 52.5 34.5 7.74 " "t " 16.3.265 6.83 " " 15 19. 8.385 45.6 34.5 7.74 " " 16.3.265 5.93 " " 1.3 17.2 0.063 6.4 7.69 —. —'._.462 36.0 31.5 9.28 1 " 13.6 221 4.68 1 " 11 15.6,' " ".221 4.,,,,,, i. ~~~~~~~~~~~~~~~~~~~~~I i.104 I 8.5 I 9.65 I.385 52. 5 I 34.5 I 7.74 I 434 4.4 I 28.2 16.3 I.352 5.44 I 277 I! 0.182 1. 5 ti - --—... -.. I i...-.. 1 1 1 61.0 ( M= 0.15).005.010 i I 2.56 r I I.227 19 I 1.4 I I 3.7 61.0 0. 088 I -_ _t -t. 1. _1 -.- ------ ----- -i i -t i i i i i i -1~~~~~~~~~~~~~~~~~~~~~~~~ 1.6 5.13 21.4 I -I _ 12 I 2.8 4.17 14. 7 15.2 0.097....... I. I _._I I..., - i --- i i 6.2 <

Table 4.6 FILM WAVE CHARACTERISTICS ( 2 - D WAVE) i. I f - I I I - I I -— r —------ _ I Vs (m/s) Steam Velocity I q2I (cm ~] Specific Liquid Flow Rate h (mils] Film V I InterInter A * (cm) Wave C (cm/a)'.Wave Propagation Speed Vc I Wx (cm/a) Weber Capil- Numbe W. Weber Numbe] B V2_ (ou rhick- facial Lengtl ness Film Velocity lary Wave Speed ^wT V2 dycm cm/ Interfacial Shear Stress Ref Film Reynolds Numbe v fh/ _vf_ k (cm1 ) Wave hIumbei o< NonDimensional Wave C/vf NonDimensional Wf Weber Numbe 4'Vy h\ K27 o-/ Yl+ G nverse Froude Numbe h f wint 3 oc7 C/Vc c y.onCut- aimestonal Off Wave Wave speed Numbe I c* C =c/zf/f N.ondimers iona Wave sPe l14) Wurz) IurF.... u wave Numbe - J (k h) bpeed I......... ~.... 4 4,, I.........I.._...I.'v.v. _. 114.227 12 2.8 22. 2 14. 7 15.2 0.097 i 32. 9.010 1.6 2.4 5.13 r I._+.1... -.. I I —- I I ~- _ - --.539 76.0 29.2 30. 1 II 11. 7.923 44.2 22.3 51. 5.340 15 5.6 6.81.042 6.47 38.8 12.9.198 2.6 2.0 1. 1.021 6.83.385 37.0 34.5 21.5 " 16.3.099 5.42 I' I I~~~~~~~~~~~~~~~ 42 " - - 21. 9 11.4 9.6 -.385 34.2 34. 5 21.5.. 16. 3.099 5.01 If 1.0 8.8 ttN I 1 -- -I..1.- _.. I _. I _.539 91.3 29.2 30.1.567 20 16. 9 11. 7.119 7.42 212 6.61 312 3. 1 6 1. _ 1_.. i_ I —-- _ -I-. _. 13. 7 I.385 87. 7 34. 5 21.5.I II 16.3.166 7.13 II II 2. 5 19.6.063 4.0 112.3.539 57.0 29.2 30. 1 " "11.7.119 4.63 ", " 2.0 12.8.308 72.0 39.6 17.2 " " 20.4.207 5.85 I " 1. 8 16.1.577 46.6 28.2 32.2 ",1 " 10. 9.111 3.79 " 1.7 10.4.539 59.5 29.2 30.1 " 11.7.119 4. 83 " " " 2.0 13.3 61.0 ( M=0.15 ).539 53. 3 29. 2 30.1 11.7.119 4.35 II 1.8 I 11. 9 It. I. _._.._., 1 I I._ -- _1 I,: _I I. 385 137 39. 5 21. 5, 780 22 16. 3. 228 9. 2 427 6.19 561 3.5 -- _. - 1 1 4 -. I..104 1 5.5 114.9.462 114 31. 5 25.8 13.6.190 7.65 --..-..-...i.4 __.__.___F__I _ -e 3 8 5 6 8. 4 3 5 2 1. 5l b.- 3 — 8 4-.5')..385 68.4 39.5 21. 5 It 16. 3. ZZ8 4. 59 3.6 i., 24.3 7ayT. < — 29.2

Table 4.6FILM WAVE CHARACTERISTICS ( 2 - D WAVE) -Ve q h V C Vc W W i k < C/V W G <C/V C (cmR-) (cnG/sc ch R (m/8) (cm/) (mnilB) (cm/e (cm) (cm/u) y(cm/s) Weber Weber A/dyne I -1 on- nverse Cc U.on- c C/ (mis)'on-(cm ) line a - Steam Specd- Film Inter- Wave'.Wave Capil- NumbeNumbe I 2 Film m imen- b Froud atona-s l ionVelocity, fic..,,, cm Wave.,Dimen- Numbe F r Off -lrroud tonVelocity fic thick- facial Lengtl Propa lary 2 2 / Rey- nal n Numbe Wave solPa *g *. V W\ V hnterfa- umbe Omb sional i Wave Liquid ness Film gation Wave' a s s nolde Wave fonal 2 W speed W Flow city-her(h speed Wave aVelo- SpeedSeedial nmbe umbe Wave V f h Numbe speed Waxve Velo - cty Speed Speed \r /t hear Numbe Numbe d r Spedl) Flow City (A V Sp14) Rate f tress f | Wurz. _ -.f. x_04 x103.269 68.4 41. 2 35.1.497 29 5.6 23.4.089 6.28 62.4 3.17 372 1.7 I 12. 7 10.9.021 1. 5.231 68.4 44.5 30.1 " " " 27. 2.104 6.28 " 1.5 12.7.269 62.2 41.2 35.1 " " " 23.4.089 5.71.. 1.5 11.6.308 48.9 38.6 40.2 1, 20.4.078 4.50 1.3 9.1.269 48. 9 41.2 35.1 23.4.089 4.50 1.2 _1 9.1.192 85.6 48.8 25.1 697 31 11.3 32. 7.174 5.45 181 2.13 534 1.8 15.4 93.3 ( M=0.23).042 2.1 15. 7.308 114 38.6 40.2 " " " 20.4.109 7.26 " " " 3.0 20.5.231 68.4 44.5 30.1 " " 27.2.145 4.36 " " I. 12 3' — _]....1 I I Nt ro).308 114 44.5 40.2,I 20.4.109 7.26 I, I -i 1 1. --. II - -- -, -.192 68.4 48.8 25.1 I. i i i 1 i i - I - - 1, 32. 7 19.5.174.129 4.36 5.45 32.5 1. 1.82 _.. I, 55 55 2.6 1.4 2.3 20.4 12.4 14.9.323 85.6 37.6 42. 2 863 32.7 16.9.....1 4 1.. 4 I... _. ___.308 85.6 38.6 40.2 I, II 20.4.135 5.45 Il I,.063 2.6 18.9.269 76.0 41.2 35.1 " I 1 23.4.155 4.021 " 1 i 2.2 [ Z.S 1.8 2.5.308 95.1 38.6 40.2 II 20.4.135 5.03 14.9 13.3 16. 6 II I t 4 I + 1 _-_ I_ _. _ - _...I i I I - i i -1.1~113 5.3 2..104 3.4 24. 1 1. 13 35.3 28. 2 691 1.46,70

Table 4.6 FILM WAVE CHARACTERISTICS ( 2 - D WAVE) Vs| q ho Vn C Vc | W W.t k | < C/V W G C< C/V C (m/a) (cm/s) (mll) (cm/)| (cm) (cm/ m/s) Weber Weber /dyne\ f ( 1 on- Non-nverse c on- c Steam Specl- Film Inter- Wave'Wave Capil- Numbe umbe 2/Film Dmen- Wbe Froude Slonl.onLelocity fiqd s ia Wv V Vh e - Wa be D imenaI 2 Numbe Wave si d ona 1 Velocity fic rhick- facial Lengtt Propa lary 2 2 cm/ 2 ey- a sional eionae ave iq 1_ Fil io Wav Vi V hn nterfa- n embe Wave onal Liquid nesa Film aton Wave NV speed Wave Venold- Wave Ve lo- Seed Speed Spee T a C h Numbe Numbe - Numbe pe l4 4 _____________' kh._ Rate e f4 3Wurz IF —-l1 —-oIw I I I I I I 1\:-ToI I I ~I7 (hX10) _ 1. 77 114 671 674 76.3 2.8 3.55.006 9.74 33.5 1. 28.478 5.0..i..385 114 146 II II 16.3.029 9.74 II I..010 7 11.7 3.4 2.9 i 13. 1 II. 2 13. 1.385 97.8 146 16.3.029 8.36 i.385 li4 146 16.3.029 9.74 II 3.4 I 13. 1 r -1 -I - -I - 1 -4 1 - I I I ~ I - -- I I I I —-).269 132 41.2 102 963 79. 5 5.6 23.4.059 8.05 94.1.930.560 3. 2 r 14.8 N.021 1.0 116.4.192 85.6 48.8 72.8 " " 32.7 083 5.22 " 1.8 9.6.192 97.8 48.8 72.8 " 32.7.083 5 96 " " 2.0 10. 9.177 68.4 50.8 67.1 " " " 32.7.083 4.17 " 1.4 7.7 _............................ _ 159 ( M=0.39).169 97.8 52.0 64.1 32.7.083 5.96 I -1- I- *-! -- -- -!- I I *l —I 79 794 1.9 1.8 11.0 10. 7 58.4 4.18.154 97. 8 54. 5 1.35 83.2 11. 26 40.8.145 268.639.154 85.6 54.5 58.4 " " " 40.8.145 3.66 " " " 4 1i. 6. 9. 4.185 72.0 49.8 70 1 " 1 34 0 121 3.08,, —. 1L 5 r_ o C I AF: n n o 7n I if A I all% I ft I t~ 4 1 -.042 1.4 23.4. 103 03. 0 4V. 0 (U. I 34.0. IZ1 3. bb 1. I I e n r - r - __. r. -_ __.. I I.139 91. 3 57. 5 5Z. 7 II 45. Z 161 3. 90. I,_I I.- 1 I. _ 1 13 68. I 8 57, 5 52.I 45 2 16 I.94 9,.139 68.8 57. 5 52.7 1.6 1.2 10.0 7. 5'. 4q 45.2 161 2. 94.I,I

Table 4.6 FILM WAVE CHARACTERISTICS ( 2 - D WAVE) vs q, h V V, C Vc W w ti R k ( C/Vf Wf G. C/VV C 21WA I i Re k f f,.c c m/H) (cm/)l (milf ) l(cmnt/ (cm) (c m/a) {(cm/a) Weber W ebher /dyne\ Rf -1n_ N on- nverse - c C/o-/ Steam Speci- Film Inter- Wave'Wave Capil- Numbe umbel 2Film (cm Non- ut- Wmet - on Velocity fic Thick- facial Lengtl Propa lary 2 Rey- ona imen-Numbe Numbe O Wave t Liquid ness Film qation Wave hX V nterfa- umbeWave sional 2 Wave ve V ILiqud n hests Film *atlon Wave psa a; n enolds Wave Wave _ V/ h speed Wave Velo- Speed Spehea Speed a V1 spe'4) Rate |r7 tress ( I) 1 x4|x.115 143 63.0 43.6 1.64 86.1 16.9 54.6.236 4.95 268.639 2..794 23 15.4.123 114 61.0 46.6 " " " 51.1.221 3.95 1 " 1.9 12.3.063 1.7 28.9 _____.192 97.8 48.8 72.8 ". 32. 7.141 3.38 " " i 2.0 10.5.215 79.6 46.1 81.5 " " " 29.2.126 2.75 n " * 1.7 8.6.104 2. 2 37.3. 2.12 90.9 28.2 1070 1.58 1.084 Ra t e / Vih86.092 124 70.4 87.6.726 182.0 2. 8 68.3.052 9.05 19. 7.400.282 1.8 I 9.2.010 0.3 13.7.077 97.8 77.1 73.3.726 1 ".81.6.062 7.14 " " " 1.3 7.3.108 91.3 65.2 103 58. 2.044 6.66 I. 1.4 I 6.8 252 ( M=. 61 ).092 137 70.4 87.6.967 184.4 5.6 68.3.069 6.68 58.8 2.38.420 2.0 10.1.021 0.4 20.5.077 85.6 77.1 73.3 " 81.6.083 4.18 "." " 1.1 6.3.069 114 81.3 65. 7 " "91.1.093 5.56 "' " I14. T I I I F-.069 163 81. 3 65. 7 1.57 190. 5 11.26 91.1.150 6.47 145 256 516 2.0 I 11.8.042 0.65 25.2.077 132 77. 1 73.3 It I, 81.6.135 5.24.I It t - I 1 1-4 —--.1. _____ ____ I__ I. i___ I.,,.I.I II 1.7 I I I _ _ _ _ t — - - - --- - _- _ __ I_ | _I_ *_ _.100 97.8 67.6. 95.3 62.8.104 3.88.I 1. 5' 9.6 7. 1 i II ---- I I I - I I I I. 0 9 1 1 8 - - ~ ~ 7 0 4 7. 6 8.4 3 1 1 4. 6 8 I.092 118 70.4 87.6 68.3.113 4.68

Table 4.6 FILM WAVE CHARACTERISTICS ( 2 - D WAVE) Vs (m/s) Steam Velocity (cm2e) Specific Liquid Flow Rate h (mile) Film Thickness V'' (c m/) Interfacial Film Velocity (cm) Wave Lengtl C (cm/a)'.Wave Propa gation Speed (cm/s) Vc I W W.:m/s) Weber Weber Capil- INumbejNumber lary Wave Speed (A 9..) V, a~ / Vi tdyne\ 2) \cm [nterfa-:ial Shear Stress Ref Film Reynolds Numbe -fh -v-J k (cm 1) Wave!tumbe o( NonDimensional Wave Numbei (kh) C/Vf NonDimen. sional Wave Speed Wf Weber Numbe If Va f x10+4 G tnverse Froude Numbe / ah \ xl0+3 Lb e I11, I I 0< OCe c CutOff Wave Numbe C/V ~ on-, djmeQslional I Wave speed 1.5 1.6 c* C =c/m7F onsiona Wave sPe f 14) Wurz) 7.3 8.2.100 101 67.6 95.3 1.82 193.0 16.9 62.8.120 3.08 282.174 556.063 0.75 32.8.092 114 70.4 87.6 II II 68.3.130 3.48,t.9 It 87.6.092 137 70.4 68.3.130 4.18 I. 2.0 9.9 I 1~ -1 - 1- i - I I - I I - - I i j.104 1.0 41.0 2.43 200. 0 28.2 588.149.696 1 1 1 1 + 1 - 1 1 4 1 I. I 1. I 1 _..L........064 85.6 84. 5 82.3 1.96 310 5.6 98.2.150 3.14 157.201 f 1. 122 1.0 2.1 i 4.9 9. 7,. ran Us.021 0.6 27.3.067 171 82.4 89.3 II II!I 93.8.143.103 91.3 89.0 137 " " " 61.0.093 3.34 " I " o 1.0o 5.2.116 114 62.9 155 o " 54.2.083 4.18 { i j1.8 5 I. 1.8 6. 5 I.. -t~ I.............................. 6.26 I,.064 95.1 84. 5 82.3 II II 98.2.150 3.48 II II I oI t -— t $ —-I I I It I I I I I I -- - ~I —-. —- 674 1. 1.077 172 77.0 103 2. 72 340 11.26 81.6.166 4.20 471 119.042 0.8 41.0 97. 8.I It.I 2.39 I. I, 298 ( M=0.72) I I I I.-..I..-..092 105 70.4 123 II II II 68.3.139 2.56 I1 II I I - - I - - I ----.051 196 94.4 67.9 I II 1.674 7 2.2 9.3, - =,... _..... II 1. > I 5. ( 5.3 5.4 3.38 380 16.9 123.312 3.98 847 103 i - I i.. -- -- 1 - - - I. - -- I - I-.062.063 1.0 49.2 171 86.2 82.6 101.257 3.48 II 1.670 II 2. 1 10.1 2.0 8.8.077 127 1 77. 1 103 Be~~ -I~~ I....- I1 ~ ~ I. ~ e-I I A I I -I --- I U - - I i I - -.. I - I ( I (. 1 1 0. -I8. 6I2 0 2. 5 8 SI I I. 7V I. I. 81.6 207 2. 58 I,.I II 1. 7 b. 5

Table 4.6 FILM WAVE CHARACTERISTICS ( 2 - D WAVE ) I,,.,-,, ----- -. _ _ r r. r - l. -r _. T d * _ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ <W V8 (m/8) Steam Velocity I (cm /s) Specific Liquid Flow Rate h (mi)s) Film hickness Vk I (cm/se Interfacial Film Velocity I A (cm) Wave Lengt) C (cm/8)'.Wave Propa gation Speed I Vc I W; (cm/s) Weber Capil- Numbei lary ^ 2 Waveed8 a ASeed 0- / (A.i W. Weber Numbet s V2h' 0 t; t.dyne\ 2) \cm /:nterfa-:ial Shear Stres I Ref Film Reynolds Numbe /Vfh \_v/ I k (cm)1} Wave gumbei I I I I. I o( NonDimensional Wave NumbeI (kh) I C/Vf NonDimensional Wave Speed Wf Weber Numbe f 2h ca / x10+4 I G nverse Froude Numbel V/ xII Ec CutOff Wave Numbe C/ Vc l.on - ~ime - slonai Wave speed 1 CL. -c/T7s tNondimeps ona Wave sper14) [Wurz) I I.063 1.0 49.2.092 196 70.4 123 3.38 380 16.9 68.3.173 3.98 847 1.03 1.670.104 1. 2 68.3 4.06 430 28.2 ____ 1960.064 2.316.,. _ I.. _.- I - _ i _. I _-I I - - ):i,Z.8 1 10.1 I t t i t I I I I I I I I I f I i 4 I I I_ I I I I I I I _ _ I N -. I I i I 1 I i_ I_ jI.i~i 7II. __ r.I. I I I I I I { } I }

128 where C* =C/U- = C Most of author's high subsonic data (M = 0.61) falls into his limiting "envelope" (Fig. 4.15,b), However Wurz's model does not adequately explain the lower subsonic portion of author's data. For this range of data, a more accurate estimate or measurement of the interfacial shear stress appears necessary. Alternatively a more sophisticated model may be required. However, the propagation speed of three-dimensional waves (Fig. 4.16 and Table 4.7) is much greater (C/Vc = 32.2+20.7) and the data more scattered, especially for high Mach numbers. Their speed is the order of A 10 m/s, while that of two dimensional waves is * 1 m/s. The basic wave initiation and growth mechanisms are yet in question. No adequate calculating model for the three dimensional wave celerity is yet available. As shown in Fig. 4.17 and 4.18, dimensionless wave speeds C'=C/Vf = 1, i.e., the wave velocity is always greater than the liquid surface velocity, so that as expected, the waves are always progressing downstream. Since wave propagation speed is always much smaller than the average steam velocity, a resonance mechanism of (P5) the type suggested by Phillips P5)is not applicable to the present case. Turbulent steam pressure fluctuations with wave lengths comparable to those here observed would have velocities much larger than the wave velocities observed.

28 26 24 22 20 18 16 T~lt U 11 CB 0>,. N.) %.0 14 12 10 8 6 4 2 1 Fig. 4.15.b Non-Dimensional Wave Propogation Velocity, C*, vs. Film Reynolds Number, Ref.

Table 4.7 FILM WAVE CHARACTERISTICS ( 3 - D WAVE) V6 qz h Vf C Vc W, W t k C C/V W G 0( C/V C (m/s,) (cmr/e (mile)( (efRf (/) (cm/) (mil) c (cm) (cm/n) (cm/a) We ber Weberdyne f on- ebernvere oCut n-C (cm ) Non- Weber CutSteam Speci- Film Inter- Wave'Wave Capil- Numbe umbe 2 Film imen- Froud O siona oncm Wave Dimen - Numbe Off Liquid neas Film gation Wave BV) terfa-d Wave n V hm Wtgh bave peed Wave Veloaa Speed Speed, Numbeod Numbe Wave af NumbeV u e Velo - iSpeed S~Pr opeed bl Se V h heary Spe Waed 4 _______Rate__ __ x ___ __^_" __ _j}__________ x104 xl03_______T_______ 1.23 342 19.3 2. 7 131 1. 1 3}.''. t 1.13 35.4 277 Z7.. t.1 S 17.7 163.0 1.16 254 19 9 23.3 " ".42 1.i2'. 3 " 12.8 121.0 36.6 1.08 244 20.6 21.7 " " " 5.82 1.63 25.3 " " 11.8 116.3 (M=.08.104 8 5 9. 65 1. 16 171 19.9 23.3 1 " " 5.42 1.52 17.7 " " 8.6 81.5.923 228 22.3 18.6 " " 6.81 1.91 23.6 1 " " " 1. 08.7 1.16 274 19.9 23.3 " " " 5.42 1, 52 28.4 " " " 13.8 130.6 SpeLSed (t 0.005.010.021 1.6 1.6 2. 4 2.56 5.13 6.83 P-t-.227 10 1.41 3.7 I 61.0.088 I.227 12 1.41 14.7 15.2.097.............. _......340 15 5.63 38.8 I 12.9.198 I 61.0 ( M=.15 ) 2.31 298 14.1 129.567 20 16.9 2.72.358.24.2 212 6.61.312 21.1 66 6.063 4.0 12.3 1. 1. —---- ___ 152 " " I *' *5 ( 1 34.0 1. 54 342 17. 2 86.0 780 22 28.1 4.08.739 23.0 427 6.19 561 1.08 342 20.6 60.3 j " f " "_ 5.82 1.05 23.0 " I" t 2.16 274 14.6 121 " I' 2.91.527 18.4 " o " 20.0 16.6 18.8 12. 5 72.9 72.9 58.5 48.6.104 5.5 114.9 1.39 228 18. 2 77.6 4.52.819 15. 3 10

Table 4.7 FILM WAVE CHARACTERISTICS ( 3 D WA VE)) Vs q,/ h V 1) C Vc WI W ti R k G o< i C/V C (m/s)..,ys~l (miie~l~och c ck (m/s) (cm/s) (mils) (cm/) (cm) (cm/s) (cm/s) Weber Weber dyne f -1 Non- Weber nver s c onc = C/R Steam Speci- Film Inter- Wave'Wave Capil- Numbe umbe 2 Flm F Dlmen- beFroud siona[ NonVelocity fic Thick- facial Lengt Propa lary V 2 V2 hnterfa Rey- umbe ionl ional WNumbe av Wave s iona Liquid ness Film a~tlon Wave Dim\, h aterfa- Humumbe Flow city, e /V h (kh) e ) IFlo O| Velo- Speed Speed ar umWbe umbe Svd 2 rNumb Rate c i, tress ( f 1 3 Wurz).4 5 1 104+5.5 14.9 3.85 228 10.9 215.780 22 28.! 1.63.296 15.3 427 6.19.561.3 48.61.104 5.5 14. 9 3.85 228 10.9 215.780 22 28. 1 1.63.296 15. 3 427 6.19.561 2, 3 48.6 010.0 21 1.0 1. 5 8.2 10.9.333 2.82 I. 500 29 5.6. 62.4 3. 17. 72' i.042 2.1 15. 7.700 31 11.3 181 2. 13.534 r -_1.. - - - -- 4- - I I I.- I...923 274 24.4 121.866 32. 7 16.9 6.81.583 14. 5 32. 5 1. 82.550 11.2 r 47.9 18.9.063 2.6 30.0.846 214 111 " I I 1 7.43.636 11.3 7.1 37.4 1.15 342 23.2 151 " 5.46.468 18.1 " 14.7 59.8.923 342 24.4 121 " " " 6.81.583 18.1 " 14.6 59.8 I —. 93.3 M=. 23 ) 1.08 342 23. 5 141 II I! 5.82.498 18.1 I.. I. I _ I,.700 1.00 171 23.9 131 1.13 35.3 28. 1 6.28. 704 7.10 691 1.46.104 3.4 24.1.846 228 25.0 111 0 ( " 7.43.832 9.46 ".923 456 24.4 121 " " " 6.81.763 18. 9" 1.15 342 232 1. ". " 5.46.612 14. 2 r - 2.08 342 272 3" " " 3.02.338 14.2 " " 22_. 63. 14.6 7.2 9. 1: I 59.8 28.8 38.4 -6* 38. 4 228 9.46 I. I,

Table 4.7 FILM WAVE CHARACTERISTICS ( 3 - D WAVE) V. q2 h V \ C Vc W, W. Re k C( C/Vf W - G oX C/V C (m/s) (cm/i)l (mil) (cnm/s (cm) (cm/) (cm/i) Weber Weber | dyne\ I 1 Non- nverse.out- - Steam Speci- Film Inter- W -Wa.Wave Capil- Numbe umbe cm Film e Dimen- umbe Froude s Ial.oM - Velocity fic thick- facial Lengt Propa lary 2. Z Rey- siaal |ionalNuZ berv Wave Wav na ~Liqiuid nee o Film at ion Wave a t a nold Wave h hm h speed Wave i:al Wavm e t f NumeWe peedV1 Velo- Sp ee d, hear umbe N be Speed - J 14 low,city. V h\ (kh) f / wu ) Ratexl~0te x(kh)3Wurz ) Rate 4 h ( x 104 _ 3. 4 _ 4. 1.=.923 274 I 24.4 I 121 I 1.13 35.3 I 28.1 6.81 I.763 11.4 691 I 1.46 700 I.1I-I___a. -..... 1. 1 I - 11. 2 I 46. 2 1.29 228 *1 22.9 I 168 I II II.010.021 0. 7 1.0 11. 7 16.4.675 76.3 I It 4.87 2.81:.546 9.46 33. 5 I II 1.28.478 10.0.....4 _1 38.4 30.7 274 I.964 79.5 5.63 16.7 94.1.930.560 159 ( M=.39 ).539 342 29.4 204 1.35 83.2 11.3 11.7.538 14.6 268.639.794 11.6 37.5 1.23 342 23.0 466 1 t 5.11.236 1 14. 149 37.5,042 1.4 23,4 1.23 456 23.0 466 1 " " 1 5.11.236 19.5 " " " 19.8 50.0.616 342 27.7 234 " " " 10.2.471 14.6 " " 12.4 37.5 1.23 456 23.0 466 1.64 86.1 16.9 5.11.286 15.8 268.639.794 19.8 49.1,769 428 25.7 292 " " 8.17.458 14.8 s " 16.7 46.7.063 1.7 28.9 9 ___ ___. -.923 342 24.4 351 " " 6.81.381 11.8 " " 14.0 36.9.615 326 27.7 233 " " 10.2.572 11.3 " " 1 11.8 35.1 t\).104 i 2.2 37.3 1.85 684 I 22. 6 _ 702 _1 2.12 90. 9 I 28. 2 - 3.40 i.246 18.3 1070 F 1.58 1.084 1 30.3 1 1. 54 428 22.6 I 584 I!.. 4.08.296 11. 5 It II 18.9 44. 9 71.7 - I _ I I ----- I - - I -1 I I1 - -— t I1 I I I I I I, 1I3I I.1 1 1 6 191 55~~~~~~~~~~~~~~~~~~~~~~~~. 18.3 28,6 71.7~~ 2.69( 684 I 23.9 I 1020 I I I{I I 2.34.169 18. 3 I I{I. 28.6 I 71. 7

Table 4.7 FILM WAVE CHARACTERISTICS ( 3 - D WAVE) Vs c/ h V. C Vc WeI Re VB | q2- h Vf I | C Vc I W W.h t{ R' k o( C/V[ Wf G occ (cm/i) (c)Non- Webe r Cut(m/s) |(cm/e) (mle) (cnrm/) (cm) (cm/a) (cm/) Weber Weer Weberdne Non Non Webernv e CutSteam Speci- Film Inter- Wave'Wave Capil- Numbe Numbec 2 FilmWe Dimen Numbe Fud O'M ~ lalW ave mef Off Velocity fic rhick- facial Lengti Propat lary 2 2 era ey- onaubeionl Lqine FlatoWaerV V h nterfa- umbe sional Wave Liquid nese Film g Igation Wave ls tal nolds Wave tv Vf h \ umb Velo- Speed Speed /hear Numbe Numbe S Flow city \(kh Speed a Rate, Ai If +4 +3........4...... x10 x.10 C/V t.,onstolnal Wave speed 30.1 30.1 37.2 II C = C//T7Nondimens tona Wave spe 1l4) Wurzl 71. 7.104 1. 92 684 1.39 684 22. 7 I 22. 7 527 728 I 2.12 # i 90. 9 I 4.52 28.2 3.27 I.328 18.3 # #.237 18.3 1070 1.084 23.0 1I.'1 ItA # a.11.270 92^0 If # 2.2 37.3 71. 7 89.8 1.69 684 22.6 641 " 3.72 270 18.3 * * 30.1 71. 7 2.15 684 23.0 815 " 1 2.92.212 18.3 29.7 71.7,846 570 25.0 321 " 7,43.538 15.3 " 2.8 59.8 *923 684 24.4 351 # " 6.81.493 18.3 " I 28.0 71.7 1.15 570 23.2 436 " _" 5.46 396 15.3 24.2 59",...............,., ~::....010.021 0.3 0.4 13.7 20.5 1.62 22.6 _' 614 1.725 182.0.966 184.4 II: 0i 3.88 1.41 - 2.79 i.281 19.7 58.8 2.38 ia iN.400.282.420 I. 1.00 684 23.9 953 1.57 190.5 5.62 6.28.135 28.6 27.1 145.256.516 252 (M-.61) 0.65 25.2 1.08 684 23. 5 iN iN # 5.82 1030 -.125 27.1 N Mi " 29.1 49.6 49.6 49.6,846 684 25.0 806 # if N 7.43.159 27.1 27.4 1 27.4 I I - I 1 — 1 -- I I I I I..-. I I I I

Table 4-7 FILM WAVE CHARACTERISTICS' (3D WAE ) ( 3 - D WAVE ) V8 ql h V C 1i W h i R k 0< C/V W G o< C/V C (m/a) (cm/) (mile) (cm/a( (cm) (m/s)(cm/a) Weberer Wef oere f nverse c o n- c C/ d c- Fim Iter Non- Weber Cut- =C//'['I"e Steam Speci- Film Inter- WveWave'.ave Capil- Numbe umbe 2) Film Dimen- Froud a.on xcm Wave Dimen-Numbe Off lNon Velocity tie rthick- facial Lengtl Propa lary 2 2 Rey onal Numbe Wave Wave sona V hV H erfa umbolave s.ha Veo I|Liquld I nehss Film | hggation W|ave 3 a to J ah nold a Wave V h\/ speed Wave La edai Wave tV/h Velo- Speed Speed -). r al umbe Wv r umbe( Wave Numbe ape Flow citdy,Shear umbe umbe Speed - Vr %14) Flow city FV h (kh) Rate _...._ ------------ ---- ---- — ate- ----- --- - ---- - tress (-e,)- { k W urz ) 1,00 1370 23. 9 953 1.81 193.0 8.44 6.28.155 41.8 282.174.556 57.3 I 98.6 I.1 12. z 2 I I I 6 1.15 1370 3 1100 " i 5.46.135 41.8 * 59.1 98.6.769 684 25.7 733 " - " 8.17.202 20.9 i " 26.6 49.2 25.7 26.6 49.2.769684 _ _ 733 " 8.17.202 20.9 " " "__ 1,15 684 23.2 1 100 5*46.135 20.9 29.5 49.2.539 570 29.4 513 " " 11.7.288 17.4 " 19. 41.0 1,08 570 20.6 1030 " 5.82.144 17.4 " 27.7 41.0..I........................ I.063 0.75 32.8 252 (M. 61 I.923 684 22.3 879 If f f 6.81.168 20.9 I If f 30. 7 _ 49.2 / I. _- - - -... —-I- -- -f 1- t I I,.. _ —I _- 1 _____ I i.769 1360 25. 7 733 200.0 8.17,269 33.2 52. 9 I 96.2 2,42 14.1 588.149.696.104 1.0 41,0 1.15 856 23.2 1100 2.42 " " 5.46.180 20.9 " " 36.9. 60.5,846 1140 5.o 806 " " 7.43.245 27.8 " 45.6 80. 6 23.0 29.7 48.4 2.1r5 684 1 ~ 2050 " " 2.92.096 16.7 48.769 684 25.7 733 " 8.17.269 16.7 " " 26.6 48.4.923 684 22.2 879 " I 6.81.224 16.7 " 307 48 2... — 1....16.7..,, -' ~ ~ ~ ~ ~ ~~~~~~. 2 8 3 1 6. 7~....731 684 - - 0. 696 8.60.283 16,7 f0 #f G~.. I 4o. 4 I

Table 4.7 FILM WAVE CI-ARACTERISTICS ( 3 D WAVB ) - I 1 - 1 -:.. I..... ----.. * - Vs (m/e) Steam Velocity q( h v (cf (cmq/) (mile) (cm/8) (cm) Specl- Film Inter- IWave fic rhickLiquid ness facial Film Velocity C (cm/a).Wave Propa gation Speed Lengtl Vc I W. (cm/s) Weber Capil- Numbe lary 2I Wave A) Seedf I' W. Ih Weber Number a v^) 8 s fdyne 2) \cm/ nterfa-:ial Shear Strea s Ref Film Reynolds Numbe "PR k (cm ) Wave Mumbe] o< NonDimensional Wave Numbel (kh) c/Vf NonDimen. sional Wave Speed Wf Weber Numbe ( 2 er / +4 xlO G Inverse Froude Numbe xlO 0< C c CutOff Wave Numb5 C/v l.ondimensional Wave speed C = c/f-7Y N.ondimens tona Wave sPeW 14) Wurz) Flow Rate t. _. - I I 1- 1.. -I 1-.. _ _ -: 2.08 684 22.9 1980 2.42 200.C 14.1 3.02.100 16.7 588.149.696 29.9 48.4.104 1.0 41.0 1_31 684 22.8 1250 I" " " 4.,80.158 16.7 " " " 1.31 760 22.8 1250 " " 14.80.158 18.5 I I I i..-. -i R i i -i -8 18.5 "a 30.0 55.3 37.6 I 48.4 53.0 48.8 1.39 684 18.2 1320 # 0 U 4.52.149 16.7 # I U I- - - 1 i - 1 1 1 - i 4 - l l l l l l- I.021 0.6 27.3 2.03 310 5.62 i UI. (A) 157.201 1.121 i I I -1 1 4 I -- 1 - I I.I _.I I i i ~~-'' I...,....:...... l~.615 684 27.7 819 2.71 340 11.3 10.2.269 16.7 471.119 1.674 24. 7 I 37. 1.042 0.8 1 41.0 *577 1370 28.5 770 " 10.9.287 33.4 48.1 74.3,846 684 30.0 1130 " " 7.43.196 16.7 22.8 37.1 _.846 684 30.0 1130 " " " 7.43.196 16.7 1 " 22.8 37.1.846 684 30.0 1130 " " 7.43.196 16.7 22.8 37.1 1.23 1370 19.3 1640 I 5I11.135 33.4 71.0 74.3 298 (M.72) 61, 1 684 27. 2 Rio # U M in_9 I- 20;0 1 A r, 25.2 I 37.1 i # ItIt_" - - I-," ----. r h-__,,.__ v --- I - - - - - — I I.1 ( —. I I I - I I I I.I..769 1370 25. 7 1020 3.37 380 16.9 8.17.269 22.7 847.103 1.670 - l-. l -,- 4 _- I I 53.3 41.4 I 70.3 51. 8.063 1.0 49.2.923 1010 24.4 1230 N 1 6.81.224 20.5 U U U 1.5 41-I I 1I- I I I-. 2 0 —— 0 --- I 1-3-4 0 - - - -1......................................... I - - r ~ ~ - 1.54 2200 1 22. 6 2050 4.08.1341 40 Is 100. ( N # i 117.0

Table 4 7 FILM WAVE CHARACTERISTICS ( 3 - D WAVE ) Vo q, h V, l C WVc W to k |< C/V W G o< C/V C (m/a) (cm%) (mile) (c/ (cm) c (cm~) Weber Weber dyne R (_I onf nverfe~ =c c / Steam Speci- Film Inter- Wave.Wave Capil- Numbe umbel 2 Film' Dimenn-e Frud Cut- onal pncm Wave Dimen -Numbe oalOff oVelocity fid.. sional.,_ Numbs Wave seonaVelocity fc. hick- facial Lengt. Propa lary 2. s_2 -/s- WaeLiquid ness V nterfa umbe alional 2 a Wave spedWv qud e Film aton Wave a a nolds Wave Wa V h speed Wave Velo- Speed Speedl r N;;;' Numb er spapt) RtFlow city A1V tress (Y Speed-'- CTl Wur14) _ Rate __trees f f Wurz)xl R"]T,10 /.846 3420 25.0 1130 3.37 380 16.9 7.43.245 69.5 I.103 1.670 136.8 I 175.4 I. a I...... _, —_. — ----—. ----.063.846 3420 25.0 1130 # # tf 7.43.245 69,5 1.0 49,2 1,00 1140 __. 23.9 1330 W # f 6.28.207 23.2 i # # M a 136. 8 47.7 140, 2 1 175.4 58. 8 175.4 _923 3420 24.4 1230 I 6.81.224 69.5 M a # # I.769 1370 25. 7 1020 N # f 8.17.269 27.9 M i 53.3 I 70. 3 I I I.- -. I - I I. I _ 1,t 1.46 1710 22.6 1950 4.07 430 28.1.170 25.o 1960.046 2.316 75.7 82. 5 298 (M-.72).104 1.2 168,3 23.2 1.15 1370 3 1530 " " 5.46.216 20,1 " I 159.1 66.1 1.15 1370 23.2 1530 " 5I 6.216 20.1 " 59.1 66.1,, - 759.1 82. l 1.31 1710 22.8 1750 i 4.80.190 25.0 70 825 1.15 1710 23.2 1530 " 5.46.216 20.1 I " N 73.7 82.5 1.08 1370 23.5 144o 5.82.230 20.1 * 58. 3 66.1.769 1370 25.7 1020 8.17 323 20.1 " 53.3 66.1.846 1370 25.0 1130 " 7.43.294 20.1 54.8 66.1.846 1370 25.0 1130 " 7.43.294 20.1" " 6.1 " 1,15 1710 23.2 1530 M I N 5.46.216 25.0 c I r rr 73.7 I 82. 5

s IMl]; 14 (m/s) (dyne/Cm2) + 21.05] 2.5' 0 37 [.09] 6.2 12 V 61 [.15] 16.2 iz - ^'2~ ~0 93 [.231 32 A 159 [.38] 80.2 0 252 [.611190:4' X 298 [.721340 10 _ UQ)t ~ ~ ~T I |(2-D waves) c) 8 3V) -- ^ j- <> T TT 5336 6 U) 4.r 1 2 1 1 10 100 Film Reynolds Number, Ref Fig. 4.17 Dimensionless Wave Speed vs Film Reynolds No. (2-D aves)

138 701 V s M (m/s) + 21.05 0 37.09 V 61.15 0 93.23 A 159.38 O 252.61 X 298.72 (3-D Waves) 6C 50 40 - 0) U) w U, C 03 c O C -.4 im 30 20 r t r. * it ii 10 - 0 5 10 Film Reynolds No., Ref Fig. 4.18 Dimensionless Wave Speed vs Film Reynolds No. (3-D waves) 15

139 Thus only a very weak coupling to this type is possible. The mechanism responsible for wave generation in thick film liquid flows (author: s case is thin film) is understood as follows. Disturbance in the steam velocity field caused by the waves creates pressure and shear variations over the wavy surface. These transfer energy from the steam to the liquid film. If the rate at which energy is transmitted to the waves by these mechanisms is not larger than the rate of viscous dissipation in the liquid, the waves will decay, In thick film cases, shear perturbations are very small compared to pressure perturbations (C7C) Therefore, the contribution of the shear perturbation as an instability factor can be neglected. Based upon the pressure perturba(C7) (N6) tion mechanism alone, Cohen-Hanratty, Nachtsheim, (N3) and Nayfeh-Saric formulated stability analyses for thick liquid films below a horizontal cocurrent gas flow. To evaluate the pressure perturbation magnitudes, CohenHanratty used the Jeffreys2) and Miles(M6) models, One of the important findings of the Cohen-Hanratty analysis is that viscous dissipation at the interface is predominant for short wavelengths (compared to their height,while viscous dissipation at the solid wall is. important for long waves, Because of their interest in a transpiration cooling system for missile nose-conesa the Nachtsheim and NayfehSaric analyses were. focused on highly supersonic gas flows over relatively thick liquid films, Their linear analyses neglected shear perturbations, and derived the rate of wave

140 growth and wave propagation velocitv- in terms of body force, surface tension and pressure perturbations. As normally expected they conclude that while surface tension is stabilizing, pressure perturbations are destabilizing, and body force can be either,depending on its direction (downward is stabilizing). They also found that films adjacent to supersonic streams are more unstable than those adjacent to subsonic streams. (N4,N5) In later non-linear analyses, Nayfeh-SaricNN5) and (S3) Saric-Nayfeh-LeKoudis S3)modified these conclusions. According to their non-linear theory, (N4N)linear unstable disturbancies do not grow indefinitely, but achieve steady-state amplitudes in the supersonic case, provided disturbance layer is larger than the subsonic layer, For very thin films, a quite different stability mechanism was suggested by Craik(C1). According to his analysis, small wave-like disturbances become unstable when the normal stress component in phase with the wave slope is sufficient to overcome the resistance to motion of the liquid surface due to gravity and surface tension. He also held that the destabilizing role of the tangential stress component became important for very thin films, and that instability may occur whatever the velocity of the gas stream is, provided the film is sufficiently thin. It is appropriate to examine the wave behavior in author's experiment using the Craik model. (C He assumed a two dimensional wave of the sinusoidal formwhere iX = k h =

141 Y _____ ^x~ ^ ^~- 777(x, t) 7(x, t)= i (x- c t) = 1+y h is the (real) non-dimensional wave number and C is the non-dimensional wave velocity, which may be complex. The non dimensional wave amplitude 8o was assumed to be sufficiently small for the problem to be linearized, Using a perturbation stream function with the linearized Navier-Stokes equations, the Orr-Sommerfeld equation is obtained. This is a fourth-order differential equation for the perturbation stream function. To solve it Craik assumed a perturbation solution in powers of dimensionless wave nurber A<. To specify the required 4 boundary conditions he used the BenjaminB4) expressions for shear and pressure perturbation stresses at the wavy interface, together with no-slip, nopenetration conditions at the solid wall. In Benjamin's derivation for perturbation stresses over a wavy wall, it was assumed that the disturbances introduced into the gas flow by the waves do not interact with the turbulence in the gas. Substituting a power solution for the perturbation stream function into the four boundary conditions, he obtained four homogeneous equations, whose determinant should vanish in order for the equations to be consistent. Neglecting the higher order terms, he obtained, 2 3i+'3 6 7 27 2 TCX2 + G - II + 4|- _ (l=)(ie^ + I (cl-c- i V Ref'2 -Ref (4.26) (4.26)

142 whe re T = -P h- reciprocal Weber number Pfh'' G = 9-h reciprocal Froude number Vf 2 IT ( = k h = h dimensionless wave number n= 31 + irll pressure perturbation ~~~r 1 (Benjamin, 1959(B4); = S + i2. shear perturbation I Ir 1 (Benjamin, 1959 (B4)) C = C + iC. dimensionless wave speed normalized by film velocity, V Re = Vfh film Reynolds number Yf On separating the real and imaginary parts, 6 27 27 2 2 5(Cr) + 8 (Cr-l)-(Ci/o(Ref) (3+-D( +-5 oRefCi)=T O2+G- r -(3 ji/2CW) (4.27) C 12 (C- rl)] + [(C -l)/o<Re (3+27 o2)= 3/2 ( ) (4.28) From Eqs. (4.27) and (4.28), the rate of wave amplification O(CiRef and wave velocity, Cr, can be found. (( C.Re f)2 r 2 3i 3 3 < CiRef = ( Re_(C + 3Ri e2 e e.29) Cr- 1 = - ((Re.F -0Re i + 2 ef r30 27 2 12 3 + -( + d Ci Ref It is seen from Eq. C4.29)_ that instability occurs when E~r+3. i > T 2 + G ie. when the effect of the surface stresses is sufficient to overcome the restoring forces of surface tension and

143 gravity, Although the pressure perturbation component ]r is generally much greater than the shear perturbation component ij, the effect of tie variable shear stress cannot be neglected when C is small. The term in.i becomes dominant for sufficiently thin films.. Because of this, instability is more likely to occur for sufficiently thin films, The present author defines this as the "shear perturbation mechanism". These results are in good qualitative agreement with the author's experimental observations reported in section 4 of Chapter IV, that waves always exist over the entire range of the experiment including even the rivulet film flow.regime. One can explain this wave formation mechanism in the following way: The component of normal stress, rI, is in phase with the wave displacement, and it attempts to deform the liquid surface by exerting an upward force upon the crests and a downward force upon the troughs of small periodic disturbances. Also, the tangential stress component,., is in phase with the wave slope, and it has the effect of accelerating the liquid on the windward slopes while retarding that on the leeward slopes. This mechanism tends to displace liquid toward the crests and away from the troughs of existing wavelets, thereby increasing their amplitude. The thinner the mean film compared to the wave amplitudes, the more effective is this mechanism, Thus it should be highly effective in author's case.

144 Compared with the pressure perturbation instability mechanism for thicker films, the shear perturbation mechanism is a entirely different one from the pressure perturbation mechanism. In the pressure perturbation mechanism instability is caused by an irreversible energy transfer from the vapor flow to the disturbance through the stress components represented by i and r'* The flow becomes unstable when film viscous dissipation is less than energy input. This energy transfer takes place through non-conservative forces, provided by the normal stress component in phase with the wave slope, and the tangential stress component in phase with the wave displacement (articles 348-350 in Lamb's "Hydrodynamics" (L6 - However, the stress components causing the instability in the shear perturbation mechanism are ]r and 2i (Appendix B) both of which act conservatively on the wave. The disturbance growth rate o CiR, calculated from Eq. (4.29), is shown in Fig. 4.19. It is essentially a frequency-response curve for infinitesimal initial conditions. Very large wave number disturbances (greater than the critical wave number, (c) remain stable because of the dominant effect of surface tension. The smaller wave number disturbances are unstable because of the controlling effect of the tangential and normal perturbation stresses. A precise estimate of perturbation stresses is essential. Author's calculations of the normal (U), and tangential (2)

145 Unstle Unstable Q * C) 4J C: Q) I Wave number Stable Fig. 4.1.9 Disturbance Wave Growth Rate as a Function of Wave Number

146 (B4) perturbation stresses. based on Benjamin's analysis, with listing of related computer programs, are given in Appendix B. Figure 4.20 compares author's observed wave numbers with the critical wave numbers calcualted from Eq. (4.29). Since the analysis is based on a linear small perturbation theory, it is applicable only to the small two-dimensional wave regime. Most of the observed wave numbers for high steam velocity are smaller than the calculated critical wave numbers, and hence are in the unstable region, (Fig. 4.20). However, the observed wave numbers for low steam velocities (21 and 37 m/s) are either larger than, or in the range of, the calculated critical values. These are in the relatively stable region indicated in Fig. 4.19. Comparing with the flow regimes in author's "map" (Fig. 4.2), most of the measured wave numbers are in the unstable region. Obviously there are inherent uncertainties in calculating the critical wave numbers. First the analysis is based on a small perturbation linear analysis and a "universal" turbulent steam velocity profile. Second the estimate of interfacial shear (.see Section B, Chapter IV)may involve errors. Nevertheless, the results of Fig. 4.20 seem reasonably acceptable for such a highly complex problem. E. FILM ENTRAINMENT For very large waves, as already mentioned, entrainment from the wave crests occurs. Good understanding of all

147 Fig. 4. 20 Comparison of Measured Wave Number., with Analytic Critical Wave Number, oc (2-E wave) 10 3 z a) 3 (0 I= in a) *E 1.0.8.6. 4 14 I I 4 i3 1 m * 0 5338 0 so so 0.].04.0(.0'.0-. 0:.O. * a r o o A 0 0 s O(c ov (m/s) * o 21 A A 37 0 93 o 252 298 A 8 i (dyne/Cm ).05 2.5.10 6.2.23 32.61 190,.72 340 J 0.0! I II I I I _ — I -1 I I -- 2 4 6 8 10 1214 16 18 202224 28 Film Reynolds Number

148 pertinent liquid entrainment mechanisms is important for both steam turbine blade erosion and moisture performance effects, as well as numerous other applications. These include annular duct flows important in many chemical process applications. Thus the critical condition for the onset of such entrainment has been studied by many researchers in both vertical 4,G7,G8) and horizontal ducts. (2 VISUAL OBSERVATION Plates #5.4 to 5.7 and 6,4 to 6.7 of Fig. 4.1 reveal such liquid film entrainment. The interfacial structure is characterized by a series of liquid flow surges, i.e. large waves. The surface becomes then highly agitated, with sudden increases in liquid film thickness. Imposed on this main wave structure are complex smaller wavelets resulting from a chaotic tumbling at the wave front, like an ocean breaker. The entrained liquid droplets usually appear frothy (i.e., not well-defined). Sometimes several clear large droplets are seen together. Author's high speed motion pictures show some features of the surface entrainment mechanisms. They include "splashing" of the liquid interface, after the impingement of a large liquid droplet. Most of the surface droplet entrainment, however, was. observed to occur following rapid acceleration, lifting, and then shattering of small complex ripples present at the top of large waves, The bursting of air bubbles trapped in the liquid, as suggested by other (rB7,N7,W14) researchers,, was not observed here,

149 In authorts. "mmap'! (Fig. 4,2), the transition between wavy-film and entrainment regimes is shown. As liquid flow rate decreases, the steam velocity for surface entrainment increases. For high steam velocities, or for thin films, a limiting behavior was observed for which the critical conditions are insensitive to changes in steam velocity, i.e., there is a limiting liquid flow rate below which no surface entrainment occurs (' 0.07 cm3/sec cm). Similarly, a minimum steam velocity for entrainment even for large liquid flow rates is. expected. However, it was not possible to test this possibility due to facility limitations on minimum steam velocity. COMPARISON WITH OTHER RESULTS Author's critical steam velocity criterion for surface entrainment is compared to other results in Fig. 4.21. There are numerous studies concerning such entrainment inception for two-phase annular flow in vertical pipes (Z2,S4,C18,U1). Much less experimental data is available for horizontal channels. It is of course necessary that experiments be reasonably similar for meaningful comparison. (V2) The channel heights of Van Rossum ()- and Woodmansee-Hanratty (W2) were 10 cm and 2.5 cm respectively, while author's is 4 cm. Thus they are of the same general magnitude. The ratio of liquid film thickness to channel height has an important effect on critical gas phase velocity for entrainment. How ever, for relatively very thin films, as in author's case, precise value of this ratio becomes less important, The

150 u 1 E 10 I — 1/2. ~' Res a) (0 E Ct1'J:> 0 a) in 0 a $. CL V 102 10 10 0- Kutaleladze (1972) Expmt. Molten S.S-sodium vapor 00Q0S - Steen and Wallis (1964) Van Rossum (1959 Woodmansee & Hanratty (1969) Ishii & Grolmes (1972) Kim and Hammitt (1.977 ) i Expmt. Air-Water Analytic Expmt. Water Film - Steam I I, s I 1,, I,, t I,1. I I 11, t 1 10 103 Liquid Film Reynolds Number Ref Fig. 4.22 Comparison of Various Entrainment Inception Criteria Ug vs Ref

151 experimental data O (y2) and S (W2) in Fig, 4,21 can thus be properly compared with author'as results 6 for which the critical steam velocity is much higher. Except for Zhivaikin - Z2 ), (using sodium vapor and molten stainless-steel) and the author i, fluids were air and water. The analytical correlation ~ (I2) is discussed in the next section. The air-water experiments O (4) for annular upward pipe flow indicate that the entrainment air velocity is constant with liquid flow rate. The cause for this discrepancy with horizontal geometry results is unknown. ENTRAINMENT MECHANI SMS Wicks and DucklerW5) proposed an empirical entrainment correlation for upward vertical annular flow as observed by Collier and HewittC1) Hughmark 6) proposed a completely empirical correlation based on the experiments of various (C15,A3,C18) other researchers.( A318) More recently, a theoretical correlation was proposed (12) (Z3) by Ishii and Grolmes. Based on Zuber's analysis they analyzed a smooth roll-wave geometry with an internal flow in the roll-wave crest. They then assumed that interfacial shear stresses at the wave crests induce a velocity within the wave of the same general magnitude as the film velocity, with a linear velocity profile within the wave-crest. To express the roll wave amplitude, they used a modified HughmarkH6 correlation for the friction coefficient, and concluded that the critical nonrdimensional gas velocity for entrainment was proportional to Ref/ for a relatively

152 thick film (Ref > 160), For low film Reynolds number, based on assumed similarity between droplet disintegration (critical droplet Weber number conceptl and liquid film entrainment, they derived an expression showing that critical non-dimensional gas velocity is proportional to Ref/2 As discussed in Section IV-D, the shear perturbation instability predominates for very thin films. However, the analysis applies for wave generation inception rather than film entrainment. Author observed entrainment to occur from the wave crests, i.e., from the thickest part of the liquid film. Therefore, the pressure-perturbation mechanism, important for thick films, is here the most probable entrainment mechanism. From author's photographic studies, the crests of the large waves, when entrainment is observed, are usually not smooth, but composed of many small wavelets. Thus Ishii 1(12) and Grolmes' assumption of smooth wave crests and linear velocity profiles within the waves is questionable. In any studies of critical conditions, it can be concluded that film entrainment will occur whenever the destabilizing components of hydrodynamic forces applied to the surface (pressure and shear) are not balanced by the stabilizing components, such as pximnarily surface tension and gravity, A completely comprehensive analysis is unfortunately not yet available, and probably impossible in the present state of the art,

153 Figure 4.22 is a schematic of wave profiles before entrainment commences. For large wave amplitudes, the steam streamline network around the waves is contracted, so that a pressure reduction at the wave crests will occur. This negative pressure differential, Pk. at the crest is propor2 tional to ps(Us-C). Pk should also increase, in general, with increasing wave amplitude, A, (due to increased streamline curvature), and decrease with increasing wave length,. Then for correlation purposes, it might be assumed that A and X effects are linear. Then: Pk P s (Us-C)2 () (4-31) where Us is the average vapor velocity, C wave propagation velocity, A amplitude and X wave-length. The surface tension stress, Ps can be expressed as s S o< ( ) (4.32) x A where C is surface tension. In cases with high velocities and thin films (such as authors),the gravity term is negligible compared to P, if X is small. s Following these arguments, entrainment should occur whenever, k > Ps or 2 Ps (Us - c) 1 > const (4.33) According to this model the critical condition is affected by steam or liquid viscosity only, as they affect wavelength,,. Author's observations and others4 l )

154 /,// / /.////IP/,~f f/_ / Vapor flow _/ \ \ ^^ ^.A H I1 h I Liquid I //?//'/////')/'"r / ff^/// ////,, —-- ---- l I h: mean film thickness X: wave length a: wave amplitude H channel height Fig. 4.21 Schematic of 3-D Wave Formation in Entrainment Regime 5341

155 show that X is. a strong function of film thickness, h, Due to viscous damping at the Qo.lid plate, the wavelength X, cannot be sustained,if it is too long compared to the film thickness, h. Assuming that approximately, X < h, Eq (4, 33) becomes C$ (Us-C) 2h -.- > K (4.34) Combining Eq. (4.17) for h with Eq. (4.34), a critical non-dimensional relationship is obtained: (.U-C) FJf fAe -1/2 (Us-C) 1f -? K I0.6 e f (4q35) where f. is friction factor, The constant in Eq. (4.34) is similar to a critical Weber number for film entrainment (V2,W2) Two different critical Weber numbers were defined, i.e., Weh based on average film thickness h, and Wep based on the maximum wcve height hp. Data from Van Rossum and Woodmansee-Hanratty (W2) together with author's results are listed in Table 4.8. Their critical Weber number, Weh, are of the same order as author's. The critical air velocities found by Van Rossum and Woodmansee-Hanratty, however, are much less than the critical steam velocities of present author, while their (V2) Ref values are much higher. From Van Rossum-, the maximum limiting critical Weber number was found to be 17, This is not comparable, for causes unknown, to either W-H(W2 data or author's. The critical Weber numbers, Weh. for entrainment agree

156 reasonably well in all cases, in spite of the use of different gases and geomet.ries, Using an average K = 2.5 CTable 4.8 friction coefficient f. = 0.005 (Table 4.2 and Ref, W8), the critical condition of Eq- C4,35) becomes (Us C) f > 0,21 Ref (4.36) Although the wave propagation velocity C is relatively variable, it is always much smaller than steam velocity, Us (Table 4.7). Therefore, the steam velocity Us itself is used rather than the relative steam velocity, U -C, in calculating the critical non-dimensional gas phase velocities. Eq. (4.36) then becomes Us - >0.21 Re12 S 9 d9f f (4.37) This relation is plotted in Fig. 4.23. For given Ref author's critical non-dimensional steam velocity is much smaller than that predicted from Ishii-Grolmes (2) shown as () in Fig.4.23. Their results are primarily for data from vertical pipe flows, with little data for horizontal flows. They based their analysis on the similarity between droplet disintegration (.i.e., critical Weber number Wle, = 22 from Hinze (1115) We = 22 from Hinze's H5 and film entrainment in this Ref range. Thereforef the critical gas velocity for entrainment in Eg 4q37) is appropriate for the horizontal duct data (curve (-,, () in Fig, 4.23), However, a comparison of the measured critical data and predictions based on the proposed

157 i ~,,,,,.... U (m/s) Ref VEh Wp I! 45 68 3. 7 38 69 2.9 30 83 2.1 27 89 1 1.9 9 24 99 1.7 5. 5 21 117 1.5 5.3 12 305 j 11 5.5 8 2065 1.8 4.6 2 82 6 - 2.5 78 3.8 3.0 60 3.0 above 25 below 20 i 17.0 1 66 52 93.3 34 i 159 28 2.12 - 252 1 22 1.52 8.2 298 17 3.38 15.2 s (U,- c) h2 s (us- c) h Woodmansee ( Hanratty (1969) Van Rossum (1959) W. Kim A F.G. Hammitt (1977) Wh = a- 2 fs (Us- c) hp 2q Ref = - f-Y Table 4.8 COMPARISON OF CRITICAL WEBER NUMBER FOR ENTRAINMENT OR ROLLWAVE FORMATION

158 - I'.4 re 1 I -~ _ % I %% **% - 1/2 - % Ref I V "P, I_ i ~~~~~~~~~~~~~~ R~~~~~~~~~~f1% I 14 I *& #. 4).1-, u 0,-4 U, tn (0 0 O 0l c 2 Q0 E i-n 10-1 10-2 "lef -& I I -1/2 1 21 Re I a, I ~~~~I. Steen and Wallis (1964) Van Rossum ( 1959 ) Woodmanse and Hanratty(1969 ) Ishii and Grolmes (1972 ) Kim and Hamrnitt(1977) t I I1 t. I I I I 111 10-3 1 I L I I It!l 10 10 2 Liquid Film Reynolds Number Ref Fig. 4.23 Comparison of Various Entrainment Inception Criteria VS Ug r[g vs Ref 6 L 9f 10 3

159 mechanism is limited not only by the lack of detailed information on the large wave shape, but also by an inability to adequately calculate the pressure variations over waves of arbitrary shape and size.

CHAPTER V FORMATION OF LIQUID FINGERS AND SECONDARY DROPLETS The characteristics of the liquid phase motion in the wake behind the trailing edge are very important from the erosion viewpoint, since the upper size limits and velocities of droplets reaching downstream rotor blades are directly influenced. Such information is crucial for the prediction of droplet impact blading erosion rates. Liquid shed from the stator blades enters a highly complex flow region. Author's high-speed motion pictures of the blade wake region show a region of liquid "fingers." These grow relatively slowly, but extend deep into the wake before disintegrating into droplets, which then are rapidly accelerated by the high-velocity steam. Larger droplets then rapidly become unstable, disintegrating into a spray of much finer droplets. These processes are obviously closely involved with the blade erosion problem. A. CHARACTERISTICS OF FINGER FORMATION 1. Visual Observations Still photographs (Fig. 5.1) show typical finger shapes, as well as droplet distributions down stream of the finger region. Stream conditions and liquid flow rates, as well as location of each photograph, are listed in Table 5.1. The two concentric circles in some of the photos indicate a pressure tap in the far wall of the test section. The diameter of the inner circle(0.95 cm) provides a useful scale, since the lens focal length is long compared to the test section 160

161 Table 5.'1: FLOW CONDITIONS AND LOCATION OF LEFT END OF EACH PHOTOGRAPHS FOR FIGURE 5.1 Plate Steam Mach Liquid Location Features No Velocity No Flow Rate of Left End Us, m/s q of each cm3/s. cm Photo X, cm P1.1 P1. 2 P1 3 P1. 4 P2.1 144 144 144 144 227 P2.2 227 P2. 3 P2. 4 P3.1 P3.2 P3.3 P3.4 P4. 1 P4. 2 P4.3 P4.4 P5.1 P5.2 P5.3 P5.4 P6.1 P6. 2 P6.3 P6.4 P7. 1 P7. 2 P7.3 P7.4 P8.1 227 227 309 309 309 309 144 144 144 144 227 227 227 227 309 309 309 309 144 144 144 144 227 0.35 0.35 0.35 0.35 0.55 0.55 0.55 0.55 (. 75 0.75 0.75 0. 75 0.35 0-.35 0.35 0.35 0.55 0.55 0.55 0.55 0.75 0.75 0.75 0. 75 0.35 0.35 0.35 0.35 0.55 0. 043 0. 043 0. 043 0. 043 0. 043 0. 043 0. 043 0. 043 0. 043 0. 043 0. 043 0.043 0. 083 0.083 0. 083 0. 083 0.083 0. 083 0. 083 0.083 0. 083 0. 083 0. 083 0. 083 0. 167 0.167 0. 167 0.167 0. 167 0 Bubbly Fingers 0 Wide Base Fingers 5 Disc Shape Droplet 10 Large Droplet 0 Finger and Droplet Shed 0 Finger and Droplet Shed 5 Disc Shape Droplet 10 Small Droplet 0 Liquid Sheet And Fingers 0 Finger 5 Cluster of Droplet 10 Droplet 0 Complicated Fingers 0 Fingers 5 Clear Droplet 10 Droplet 0 Fingers 0 Finger and Sheet 5 Droplet 10 Droplet 0 Liquid Sheet with Holes 0 Finger and Sheet 5 Droplet 10 Droplet 0 Fingers 0 Fingers 5 Large Droplet 10 Droplet 0 Finger and Sheet

162 Table 5.1: cont'd Plate Steam Mach Liquid Location Features No Velocity No Flow Rate of Left End Us, m/s q of each cm3/s. cm Photo X,. cm... 11_ L: -...,... P8. 2 P8.3 P8.4 P9.1 227 227 227 309 0.55 0.55 0.55 0.75 0. 167 0. 167 0.167 0. 167 0 Finger and Sheet 5 Droplet 10 Droplet 0 Sheet and Fingers Disc Droplet 0 S heet and Fingers 5 Cluster of Droplets 10 Droplets P9.2 P9. 3 P9.4 309 309 309 0.75 0. 75 0.75 0. 167 0.167 0.167

163 5399 Figo 5, 1 Phtographs of Finger Formation and Secondary Droplet Formation

164 Fig. 5. 1 Continued 5400

165;:;;'; I;; Fig. 5 1 Continued 5401

166 PFig 5,1 Continued 5402

167 Fig, 5K i Continued 5403

168 Fig. 5 1 Continued 5404

169 Fig. 5 1 Continued 5405

1 70 ^^;:::l: 000000:::/c:;:;.*::;;:,^^,;'.,',:.;.: \ * *: ^..: *; *:': /::: f-:f:F::< X -::::::i;:f.:'::;;:;:^ i M -::-_:-:i-:.' *:.WN,:. —:::':::^:::~: :i ^ *' ~:::^.:!i c m: C. "':?-:-t:.;';':*::: -:'. *.' * j~ >. a t~;: C: C -X-S; iX;,:,G:,.'0 0 If 0 00. 0:::^ ^ E 0 L L.0;00 0; 0;: v:P: 4 o:'::::::::;:: 4 o 2 Fig^; by Al 00 c 5.1 C ft S 540 Fig. 5 1 Continued 5406 V~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-::::.''~:::i_:~:.:::i~ V~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1:::__::::i::-:-::::::::~i ~4~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~: /K (0.,: —-:~::~~::: ~ ~ ~ 4:::I 1 <::':: i; —-.j:,'i,-::: ~~~~ig S I Continued 5406~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~::-:~:::::.:~

1 71 Fig. 5 1 Co ntiinued 5407

172 Fig. 5.1 Continued 5408

173 Fig. 5 l Continued 5409

174 Fig. 5.1 Continued 5410

175 a::::::f::::;:: d;:0: fXW::;0; _ _~~~~~~~~~~~~~~~~~~~~::::; "i::;:- ~`:nm:0:fff::f:;:::::;0::0:::::D- F) 7~~~~~~~~~~~~~~~~~~~~~~~:e En:::: ~:-i I e:;.W..:::;: Fig, 5.1 Continued 5411

1 76 Figo 5 I1 Con'tinued 541 2

1 77 Fig, 5 1 Continued 5413

1 78 Fig. 5.1 Continued 5414

179 Fig 5 1 Continued

180 FigF 5 1 Continued 5416

181 width, so that little optical distortion is involved. Flow is from left to right. Finger shape varies with steam velocities at fixed liquid flow rate. One finger in plate #1.1 of Fig. 5.1 appears, e.g., as an approximate conical tube, with wide base commencing at the trailing edge. Other fingers appear more as cylindrical spikes of smaller diameters aligned with the flow axis. Plate #1.2 (Fig. 5.1), with the same flow conditions as Plate #1.1, shows one long spike-like finger and others of irregular shape (upper part of photo). The liquid sheet from which the fingers are formed, is wavy in the transverse direction, no doubt affecting the location, formation frequency, and size of these fingers. As steam velocity is increased, more and smaller fingers are observed. (Plates #2.1, 2.2, 3.1, and 3.2 of Fig. 5.1). Detailed study of the fingers is necessary to understand resultant droplet formation. Finger lengths, L, and the distances between them, Af, vary with steam velocity and film flow rate. (Plates #2.2 and 3.2 of Fig. 5.1). As shown by the photos, fingers of differing lengths exist at any given time. The maximum observed finger length, L, is here called the "characteristic" finger length. One finds that maximum finger length, L, and finger wave length, \f, (average distance between fingers) decreases as steam velocity increases (see plates #1.2, 2.2, and 3.2) at fixed liquid flow rate. With higher liquid flow rates, more complex finger

182 patterns appear along with more complicated liquid film patterns (i.e, waves, etc). As steam velocity is increased at fixed liquid flow rate, more fingers appear. Larger fingers are very unstable, and do not extend so far axially as do the smaller fingers, appearing at lower liquid flow rates. Sometimes several holes appear in such fingers. At inception, they may be of triangular shape with thick cylindrical boundaries (plate #6.1). These grow to more nearly circular shapes probably because of surface tension, before the finger disintegrates. Plates #5.2 and 6.1 show this intermediate process and the final shattering. Author defines maximum finger length, L, as axial distance from blade trailing edge to downstream end of the finger. For the highest liquid flow rate (plates #7, 8 and 9), the blade trailing-edge flow appears in total chaos, and of extreme complexity, so that L and Af cannot be measured. However, they are still well discernable for lower steam velocities, even at higher steam velocities, except for the exact instant of finger disruption. 2. Reduction of Experimental Photographic Results Finger formation at the blade trailing edge has been recorded (as already discussed) using both a still polaroid camera, and Fastax high-speed movie camera. The better optics of the still camera enables better observation of the forms occurring at the trailing edge, but cannot provide any information on the flow kinetics. However, the timedependent processes at the trailing edge were measured

183 by the high-speed motion picture camlera at 5000 Hz. In most cases, every 80th frame (every 16 ms of real time) was examined with a calibrated viewer-screen. Observer error was estimated to be + 15%, for the same film comparing 5 different observers. Finger formation dependent parameters are maximum finger length, L(t), and finger wave length, Af(t). The number of fingers, N, was counted for a given transversal length, X. Then Af = X/N. Typical measurements of maximum finger length, L(t), as a function of time, are shown in Figs. 5.2 to 5.4. For all liquid flow rates, upper limit of maximum finger length, L, decreases as U is increased. For all U average of L s s increases as liquid flow rate is increased. Also there exist somes "dead" periods before a new finger is formed after one is shed. The number distribution of maximum finger length, L, was investigated for all flow conditions. Finger length was measured in 2100 frames (70 frames on each of 30 rolls of 100 ft. length). Thus samples used are statistically meaningful and reproducible. The number of maximum fingers in each length increment was normalized by the total number of observations for a given flow condition. Typical maximum finger size distributions are shown in Figs. 5.5 to 5.9. The largest number of finger lengths (number peak) occurs between 2.5 and 4 mm for all steam velocities for

rif 3 $& Maximum Finger Length. L. mm 0 3 _ s v~~~~~~~~~~~l:% 71 m, 2) a c 7pi r, I "IO oD.e.a rr I L-. --,! i M * a I 0 DI < a > 0 f78T

o amum Finger L h, L ~~ - - ~0 0 as 0t 79 in X 5> 0 x YC0 b~g I 1 tl'30 (h Q I.l 0 I (I -q 0 9 C0 1 01 2 0~~~~~~~~( O. 01 b ~~~~~ m o~ ~ ~ ~ ~ ~~r NO - ~ ~ LI A ~ ~ ~ ~ %@ ~~~x,~~E ~~~~ ~ ~ V i: ~ ~ _ I 0, K 0 S8T

5345 I J j6 a r a1 g4 Ik I hCO 160 t - AS 320 480 64U FIg. 5.4 KFxlusrm Flger.ength, 1., vs Tlow

70 60 50 40 q =.01 cm /cm sec U (m/s).8....~0.... 93 -- -A —- 159 252..0-.. — 297 v —- 335 5346 I I i. *. II<Uk If I / I "'\ )' ~0 I/I,,,,/ \ I! 10.!/ *.\,% 0 5 10 15 20 25 30 35 Finger Length (mm) Fig. 5.5 Finger Length Distribution

3 q =.021 cm /cm sec 70_ U8 (m/s) - 110 == — - 93 -— A —- 159 60C- / -.-0. — 252 / {**0~~~~~~~~-^~....^0.. 297 /I9 — -7335 5 4 30 I I= Ii \ rdar~~~~~~ 30 ^~~ l5347 PI / I 0 5 10 15 Finger Length (mm) Fig. 5.6 Finger Length Distribution 20

70 q =.043 cm3/cm am - r - - 60 50 40 sec u (m/s) 93 159 252 297 335 -. "_ a... -V- --- Qi 4.) C-) 4' GJ 30 20 -30 f/ 5348 10 0 0 5 10 15 Finger Length (mm) Fig. 5.7 Finger Length Distribution 20 25 30

701 3 q a.06 cm /cm sec u (m/s) 8 -0 —-* 93 _ A — - 159 -. * O. — - 252..m -O.- - 297 IV+^... 335 a) bO tl u a) P 5349 tD 0 21 0 I%-6..0 OW A'-% A...p. 01%\ 0 5 Finger Length (rmm) Fig. 5.8 Finger Length Distribution

70 3 F ~~~~~~~~70 -- q =.104 cm /cm sec U (m/s) 8 — 0 --- 93 60 - - -- --- 159 -. -0- -- 252 \.. -0-.- 297 50 v 335 40 I bO \ mI 30\ 5350 04 20,' Ato.,,.\ -.o\ i A20 \ —- 0 I 5 10 15 20 25 30 Finger Length (Irmn) Fig. 5.9 Finger Length Distribution Fig. 5.,9 Finger Length Distribution 35

192 liquid flow rate, q = 0.010 cm3/cm.sec (Fig. 5,5). The number peak density increases with increase of steam velocity. This is presumably due to increased finger instability as steam velocity is increased. An increase in number peak length was also observed at higher liquid flow rate and lower steam velocity (Fig. 5.7 for q = 0.043 cm3/cm.sec). For higher liquid flow rate (Fig. 5.9, q = 0.104 cm3/cm.sec), an increase of number peak L at higher steam velocity was found. For two higher steam velocities (297 and 335 m/s), no peak shift was noted. Time-average maximum finger length, L was calculated from the distribution curves in Figs. 5.5-5.9. Figure 5.10 shows the dependence of L on the various flow parameters. At a given steam velocity, L always increases with increase of liquid flow rate, q. An increase of Us at fixed q gives a decrease in L in the low U region. However, all L's are strongly increased by increasing steam velocity from 297 to 335 m/s. Minima of L for all liquid flow rates are observed (Fig. 5.10) at Us = 297 m/s. The reason for the occurence of these minima is unknown at this time. it cannot involve shock-wave phenomena since the steam velocity is well subsonic (M = 0.72). The data scatter (Fig. 5.10) is reduced for increasing Us and decreasing q. The bar indicates the full observed range of data. The frequency of appearance of fingers longer than L has been computed from the L vs. t curves (Figs. 5.2-5.4), dividing the time increment for a particular data

193 Nomenclature q (cm /cm sec) O 0.010 a 0.02 O 0.043 A 0.063 a 0.104 10 9 8 AJ co C 0 0) 0) 4 7 6 5 4 3 2 1 U 1. (mm) / S (uM/S) 100 200 300 Steam Velocity Fig. 5.10 Average Finger Length, L, vs Steam Velocity, U S

194 set by the total number of peaks found therein. This gives the average interval, AT, between the appearance of fingers larger than the time averaged finger length, L. Figure 5.11 shows AT vs. Us. In general, AT increases with increasing Us, and decreases with an increase of q. The general dependencyof AT on Us for fixed q can be explained as follows. An increase of Us causes a thinning of the vapor boundary layer along the blade, and at the trailing edge. It also leads to an increase of interfacial shear (Table 4.2). Both vapor boundary layer thinning and increased shear stress enhances liquid removal at the trailing edge, due to increased entrainment, because of the closer approach of high velocity steam with consequent greater shear stress. It thus requires a longer time for sufficient liquid to be accumulated for the next such removal. For given Us, a given quantity of liquid of course accumulates more quickly for increased q, as shown by data (Fig. 5.11)..An unexplained devia3 tion from the general trends occurs for q = 0.021 cm /cm'sec (Fig. 5.11). AT increases with increasing Us up to Us = 250 m/s and decreases thereafter, while for all other q, AT always increases with increasing Us. Two finger Strouhal numbers have been here defined, based on film and steam velocities respectively, Sf = L (5.1) AT Vf and

195 Liquid Flow Rate q (cm3/sec cm) 0 0.010 A 0.021 O 0.043 A 0.063 * 0.104 5352 Cl 600 cu 0 <: X 400 a) 00 E o 300,-4 X 100 3o Zc,.V w1 I I I 100 T (ms)/ s (ads) Fig. 5.11 200 Steam Velocity 300 Appearance,LT, Average Time Interval for Large Finger vs Steam Velocity, Us

196 SL= C5.2) f2 AT U s to correlate finger formation frequencies. Figures 5.12 and 5.13 show the relation between Sf, Sf and steam Reynolds Number, Res, In general S and Sf decrease as 1 2 Res increases. Also for fixed Res, both Sf and S increase s s1 2 with film Reynoldsnumber, Ref and S differ by the 1 2 ratio of film to steam velocity. Sf, then, varies from 0.04 to 0.93,thus being in the same range as Strouhal numbers measured in other experiments, On the other hand, Sf -4 is ~10, and thus not consistent with usual S data. The finger shedding phenomenon seems obviously closely related to vortex-shedding from a circular cylinder. Such (R3) results by Roshko show a Strouhal number increase with Re up to S = 0.21 at Re = 103, No futher increase of S occurs for a further Re increase. Here, nD VD S = and Re = v where n = vortex shedding frequency D = cylinder diameter. Author's Strouhal number trends with Re (Figs, 5.12 s and 5.13) differ considerably from those for cylinders. Both S and S decrease with increasing Re. Thus the similarf f2 i1 2 ity between these processes is limited.

s?; aa'o spTou~aU m:a3S A' S'ON aiqmnNl spTouXa1 uSealS gOTXg gOT TvqnoajS aa~uTl ZT'g'S~a to7 xS OT1 /s — I'O -1 r o ftl H1 0 zt 0 ~ I' 0 r Ca!9,0.^ it M^4 ESES 6'91 V 9ZITT 0 9'S V 8'Z 0 *oN sp-IonXa{ mIT! -- 9-0 L'O 6'0 L6T

198 ~T 0 3 E-3 (4 Eo z O Q) 0 f& 16.( 15.( 14.( 13.( 12.1 11.1 10. 9. 8. Film Reynolds No. Ref 0 2.8 A 5.6 O 11.26 16.9 * 28.2 5354 2. f2 104 5x104 10 5105 Re s Fig. 5.13 Finger Strouhal No. Sf vs Steam Reynolds No. Re f2 S

199 Reference (B.91 provides a good summary of Strouhal number and Reynolds number correlations for various geometries. There, Strouhal number for flow past a thin flat plate (as author's case) is given as 0.156, which falls within author's range for S C(Fig. 5.12). S-ince S is based f f 1 1 on liquid rather than steam velocity, it appears that the vortex shedding in the steam in not a major factor for finger formation frequency, Presumably liquid finger shedding is related to a balance between the driving forces and the stabilizing effect of surface tension. To study these effects, two new dimensionless parameters have been selected, -the first being a non-dimensional time, T: Ur AT (.A 1/2 L \Pf Physically, T is the finger formation "period", normalized by a parareter having units of time formed from maximum finger length and densities of vapor and liquid. A rearrangement as below shows the physics more clearly 1 2 \ 1/2 (L/AT)pf / L/AT can be interpreted as a "deformation velocity" for the fingers. The numerator is the square root of dynamic pressure of the vapor flow acting on the finger, and the denominator is a measure of the deformation stress upon the

200 liquid finger, The other nondimensional parameter here used is Weber number where 2 - gsUs L We = -- (5.5) s C In this case We is the ratio of the driving vapor dynamic pressure to the stabilizing surface tension stress for the finger. Plots of C vs. We for various film Reynolds numbers, are shown in Fig. 5.14. The peak in T is reduced, and the corresponding value of Wes also increases, as Ref is increased to Re. - 8.4. For higher Ref, no peak in r occurred. Rather it increased smoothly with increasing Ref. Thus, surface tension appears more important for low than high Ref. as would be expected. Another important length parameter, mentioned earlier, is the finger wave-length, /ff, i.e., the average distance between fingers. It was measured by counting the number of fingers, N observed in individual Fastax frames, and dividing the transversal distance along the trailing edge by N. Plots of N vs. time are shown in Figs. 5.15-5.17. In general, the number of fingers was found to increase with steam velocity. Corresponding finger wave lengths, Xf, are plotted in Fig. 5.18 as a function of steam velocity. >f ranges between 3 and 13 mm for the whole experimental range. Finger wave-lengths decreases for increasing steam velocity, except for highest steam velocity, indicating a minimum

201 0 - 0- A, Nomenclature Film Reynolds No. 2.8 5.6 8.4 11.26 16.9 28.2 scattering of data I, -- 5295 200 400 600 800 2 - We Ps Us L Fig. 5.14 Non-Dinsional Te and Webe Fig. 5.14 Non-Dimensional Time and Weber 1000 1200 1400 Number for Finger

q " 0.01 cm /cm sec Steam Vel (m/s) ----- O 93.... 159 - - - 0 252 ~-^ 7 335 20 z 15 1S 0 10.. _ I 160 320 80 640 LI S I t q, c; 1C 7n N n TF M ut Tc - rig. D. i3 Numoeu ut r I ~`IIrt!Lb, II 9 Vb -I[t L

q = 0.043 cm3/cm sec Steam Vel. (m/s) 093.-... -a 159 ---— o 0 252 V -- 335 5357 20 15 z 0 O Pr4 10 4-4 0 It 5 0 N/ Ntime T 0) w 160 320 480 640 ms -- Fig. 5.16 Number of Fingers, N, vs Time, t

q = 0.104 cm 3/cm sec. Steam Vel. (m/8) — 0 93 N._... 159 ---- 0 252 ----- tV 335 20 15 10' 5 N/ time T 5358 i\ iA It. I 0 4 b~ OZ 480 640 ms --- Fig. 5.17 Number of Fingers, N, vs Time, t

1.3 \ 1.2 1.2 \.\ 1.0 0.9_ \. \.\ \\ 0. 60 ^^ ^ i \ ~ / is'4 ^ o-6-...., o. / r< 0.4 0 = 0.625 A = 1.25 Q3 o= 0 1.875 = 2.50 A = 3.75 0.1 5294 100 200 300 (O Steam Velocity, m/s Fig. 5.18 Finger Wavelength vs Steam Velocity

206 with steam velocity, It is likely that the fingers become increasingly unstable, attaining smaller widths, as steam velocity is increased. This may result primarily from steam turbulence. A non-dimensional finger length, L* is here defined as: L* =(5.6) Af i.e., a ratio between average finger length and transversal distance between fingers. In general L* increases with Ref at fixed Re (Fig. 5.19). Fluctuations with increasing Res are similar at each Ref. These may be due to overall facility resonances depending upon steam velocity. All finger formation parameters are listed in Table 5.2. Complete raw data and parameters are also available in our previous reports, (N9,T6,Bl0Bll) all done under author's supervision. No results similar to author's current observations appear to exist in the literature, so that no comparisons with other data are possible. Thus author has made a "first attempt" which has obtained comprehensive experimental data. Later investigators will hopefully provide more complete theoretical understanding than obtained here, using this data as a basis.

207 2.0 1.9 1.8 1.7 1.6 1.5 1.4 < 1.3 " 1.2 -K q(.C; 1.1 6) I 1.C QI 0.9 c m 0.7 S 0.6 0 - z 0.5 0.4 0.3 0.2 0.1 L*/ Re Film Reynolds No. Ref 0 2.8 a 5.6 O 11.26 A 16.9 * 28.2 1 5x104 105 5x105 Steam Reynolds Number Fig. 5.19 Nondimensional Finger Length, L*, vs Steam Reynolds No. Re' s

Table 5.2: TRAILING EDGE PARAMETERS V q Vh \ LRf L fh L We Wef Si SZ L m/s cm /s cm/s mils cm cm xlO x 93.010 7.45 1.1 1.25.418 2.8 53.3.098 60.3.33.572 4.58.33 " 021 10.93 1.5 1.09.697 5.6 " 068 101 1.19.936 11.00.64.043 16. 40 2.0, 93.900 11.26.061 129.8 3,46.900 15.87.97 -158.5.010 13.67.6,74.320 2.8 90.9 ~089 134.3 85.26 2.24 43 " 021 18.22.9.65.440 5.6.072 184.5 2.09.34 3.91.68,031 22.36 11 57 84 6 1 9 4280.4 056 3.05.34 4.80.72.043 25, 23 1.3.66.468 11.26 ".067 197.18 4.26.28 4.46.71,063 30.75 1.6.66.643 16.,9 ".057 268.5 8.69.37 7.18.97.104t 39.10 2.1.66,643 28.2 ".073 268.5 14,04.23 5.67.97 4-. - ___- -. - 1 252,010 16.40,5'34.202 2.8 144 300 211,1 78.04 0.26.59.021 27.33.6.41.364 5.6 " 245 380.2 3.'88.05 0.54.89,031 35.14,7.41.382 8.4 " 103 401.3 6.74.11 1. 53,93,043 - 36.40 9 40 495 11 26 118 517 3 9 37.11 1 59 1.24,063 49. 20 1. 0.,43.508 16 9.087 528.0 17 57.12 2.34 1. 18.104 58.60. 1.4.37,823 28.2 " 067 866.0 40.37.21 4.88 2.22 I __ _ _ _ ~ _ __ I I 297 II I, II It.010 20.5.4.83.162 2.8 170.393 238.9.97.02 0.14.20,021 32. 8. 5.425.180 5.6,188 265.4 2.77.03 0.33.42. 0 31 41.0. 6.44. 243 8.4.159 353.9 5.84.04 0. 55.55,043 46. 9. 7. 50.360 11.26.104 531.0 11.31.07 1.11.72 063 61.5. 8.50.300 16.9.131 442.5 16.21.04 0. 83.60.104.46.64 28.2 943.9 1.58 _ __ ___ _ _ _ _ _ _.. __ __ __ _ __ - _ -_ - -_ - -_ - -__- __ - __ - _ _ -

Table 5.2: TRAILING EDGE PARAMETERS V-f -q V h ~ f a a f |V q f h.f L Ref Res t We. Wef S1 S2 L In1/s cm /s cm/s,l/s cm /a cm,/a nrila cm cm xlO s 335.010 27.3.3.64.387 2.8 192.506 732.0 4.14.03 0.24.60.021 32.8.5.55.468 5.6 ".152 882.4 7.19.09 0.88.85.031 41.0. 6.46.481 8.4." 152 901.1 11. 55.08 0.98 1.05.041 46.9.7.37.504 11.26 ".282 938.8 15.84.04 0.56 1.36.063 61. 5. 8.46.639 16.9.200 1201.4 20.08.05 0 92 1.39.104 82.0 1.0.45.724 28.2 220 1351.7 69.55,04 0.78 1.61 0 ED

210 B, SECONDARY DROPLET MOTIOON Due to drag forcea from the. vapor flow in the trailing edge wake., the "secondary" droplets detached from the fingers are rapidly accelerated. From the viewpoint of turbine blade erosion, direct measurement of droplet acceleration in the wake region is most important. As the droplets approach steam velocity, their normal impact velocity, important for erosion, upon the next rotating blade row will be reduced. (See Fig. 1.6 ). For higher steam velocities, many of the larger drops were observed shattered into much smaller droplets, within a distance of 4 cm downstream from the trailing edge. Droplet size distribution data as a function of position is also very important for erosion, defining the probable upper limit of droplet size reaching the next rotating row. Droplet break-up time, tb, is here defined as the timespan between droplet formation from the liquid fingers and their disintegration. 1. DROPLET ACCELERATION Using frame by frame studies of the Fastax pictures (as for the finger formation data) the trajectories of large liquid droplets were recorded under various steam and liquid flow conditions, Droplet size and position from the trailing edge were recorded every 5 frames (every 1 ms). Values of the second time derivative of position were determined using the following equation,

211 (2x x CS 7) 2> xi+ t 2x; - Xil 57) At2/ (Atl2 The "second divided difference" (here used) is the best approximation for the second derivative, if differences of (D6,W13) higher order are not known. This procedure allows identification of those position values containing significant errors. The second divided difference can then be recalculated neglecting the questionable data points. When position values are thus neglected, the time intervals are not equal. The second divided difference can then be written: Xi+l - xi xi - Xi-1'x2 b=. -... A, tzi+l ti.. t.. -1 ii\ i-l til "" ^ ^ ^ ^- (5.8) at i ti+ + t-1 The steam and liquid flow conditions for which data on accelerations and break-up time were obtained are listed in Table 5.3. All data on droplet size and position were stored on the computer tape. Details of computer programs used and output obtained are found in Appendix C. The time-averaged accelerations of droplets under various steam and liquid flow condit.ons are listed in Table 5.3. and plotted in Fig. 5.20. Observed droplet accelerations range from \3xlQ4 to "3.5xQ5 c/ssec2 (X30 to 350 "g"), and increase with:team velocity, as would be expected. Considering the nature of the phenomena observed, a sta

U ___ 300 158s50 o 2i 5150 _ 29720 i. 335 30 o qcrrys -Am. As. R Am As R Am As- Am As R Am A R lnrmcm / cr/s2 c m/s2 cm/s2 cm/s2 cm/s cm /s c/s2 cm/s m/ 010 25550 37855 782 75144 109893 699 287500 244372 1.418 - - 9039 36320.910 — ______ —-- ------— 1- — t t - I, -___t~1.021 28780 30893 1.010 104906 117845.983 94750 159862 2. 059 12187 310066 1.152 0340 46545.700.032 -1 - 82940 88814.966 460001215681 1.26233750311230 1.2704500 528665.690 - ~ - -~ j ---------- ---- --- i _1 1.043 23454 42556.775 69542 87936.791 700001161763 1.820 09117 312999 770 5662E 512930 699 t —--— C —-- t —- t. t t P r 1t -— ~~ —l.063 80275 102841 828 193593198627.953 189140 332318.625 5537 509915.606 - 1 I_ _ L..____ ----- - - t - - _____.083 22117 30593.727 - - _.............. —t1-1t1 1 - --- -— t —.104 - - - 65203 90400.746 1471871185582. 964 - - - 9583 478515.680 1 1....... --... __(S) TABLE 5.3;COMPARISON OF EXPERIMENTAL, Am, AND SERAFINI, As, DROPLET ACCELERATIONS. R = (Am/As) ALL VALUES. RE AVERAGE VALUES FOR THE CONDITIONS SHOWN. N

213 q (cm2/s) 0.010 A.021 0.043 V.063 0.104 Ref 2.8 5.6 11.26 16.9 28.2 T" 4) 3 10 E <( r= 0 4o 0) $. u u < 0Q t 4 I 60 120 180 240 Steam Velocity, 300 8, m/s 360 Fig. 5.20 Measured Droplet Acceleration as a Function of Steam Velocity

214 tistical method oQf data reduction was necessary and desirable. Thus a linear regressiSon analysis was conducted. Full details of the analyses are given elsewhere C The full output of the statistical analysis is given in Appendix D. An important result of this analysis is that droplet accelerations are largely unaffected by changes of liquid flow rate for all steam velocities. This is particularly true for low velocities. (Fig. 5.21) THEORETICAL CONS IDERATIONS Three important assumptions were needed to obtain an analytical expression for droplet acceleration. First, spherical drops were necessarily assumed, though many are severely distorted. An equivalent droplet diameter was defined as that of a disc of the same area perpendicular to the flow direction. Next, the local vapor velocity was assumed to be the mean steam velocity at the blade exit plane. Finally, the only force upon the droplets was assumed to be steam drag. One then obtains the following equation, pTdU = CDA s Ur Pf.dU = CDA Ps U2' (5,9) dt 2 where ^ Fp = density of liquid droplet ps = density of vapor = volume of droplet C= HeD volume of droplet C )6

215 CN S o ~- - | X~FX 93 0- o_ O-X 159 uE's Co~0 —o0 252 l0-0-_ 297'mo.. 335 O (_' o - o c.. -.. 0. 021 0. 043 0. 063 0. 083 0.1 04 3 53 cm /cm sec Fig. 5.21 Average Observed Droplet Acceleration versus Liquid Flow Rate at Various Steam Velocities. Dotted lines show regression analysis results. Note: Slopes of regression lines were not found to be significantly different than zero. See Statistical Analysis.

216 dU dU = drQplet aQceleration dt CD = Drag Coefficient AD = Projected area of droplet in the axial direction C= r D2) U = Relative velocity of droplet(= U -U ) r s Rearranging Eq. (.5.9), droplet acceleration for drops of equivalent diameter, D, is dU 3 Ps D U 2 (5.10) dt 4 pf D r Many empirical results are available for CD. As stated in Chapter II, all calculations of CD were based on droplet acceleration due to uniform flow around a single droplet. No direct measurements of droplet acceleration in the wake region (as in the author's case) were found in the literature. Author's wake-droplet regime is complicated by the presence of many droplets together. Because of spontaneous condensation in the "Wilson zone" combined with secondary droplets formation, a droplet size distribution exists in the wake. The drag for a given droplet will be influenced by other droplets. Such droplet'dShielding"' should thus be considered for a more precise analysis. The drag coefficient shield effect was apparently first considered by TamT2) He derived a statistical formula for the drag exerted on a cloud of spherical particles of given

217 size distri;bution. He concluded that the. individual particle drag depends s igniifcantly only on the first three moments of the distribution function. However, no experimental verificatiQn yet exists. Tamn's solution is not used by present author, because it is applicable only to very low Reynolds numbers. (C14,A1,S6) Instead three different already published empirical correlations were compared with author's data. These are:`294 9/5 1) CD ReD (+ (6 ) (5.11 0 < Re < 3000 (Churchill & Usagi(C14)) 2) C = 0.292 (1 + 9.06/vReD) (5.12) (Al 0 < ReD < 5000 (Abraham ), and 3) CD = 24 (1 + 0.17 Re 2/3) (5.13) ReD 0 < ReD < 6000 (Serafini (6)), where droplet Reynolds number, ReD is defined as ReD = U-D (5.14) D IVs These three empirical correlations were substituted into Eq. (5.9), and comparisons with author's data made.

218 (Appendix El, Calculated accelerations, using Eqs. (5.11 - 5.13) were compared with those measured (Fig. 5,22). For low acceleration, i.e., low steam velocity, the differences between measured and calcualted accelerations, using any of the CD correlations are small. However, for higher steam velocities, the calculated accelerations using any of the CD correlations are larger than those measured. The Serafini correlation fits best the present data. Therefore it alone has been used in the following. 2. Observed Droplet Decelerations Author defines an acceleration ratio, R, A R = (5.15) s where A is measured acceleration and A that calculated m s from Serafini's CD correlation (Eqn. 5.13). Time averaged results (R) are listed in Table 5.3. One finds R < 1 for high liquid flow rates,-(q' 0.043) and for the highest steam velocity (Us = 335 m/s) for any value of q. For fixed flow conditions, abrupt oscillations in R with respect to time are found (Appendix C and Figs. 5.23 and 5.24). The wake droplets are not continuously accelerated, but actually decelerate at times and exhibit "back-flow" moving against the mean vapor flow. Several of our high-speed movies show this intermittent droplet directional change. To be more explicit, droplets detaching from a liquid finger move a short distance ( - 2 cm) down stream, then reverse

219 0.8 0.7 / / / / /... --.-.... */ _ I* /' 4- - 0.61 CM U u 0 -F -c U, U, o $U U, 0.51 / * 0.4. ~~~~~~ s~~~~~~~~~~~~~Itf s&O. Is4~~~~~~~~~~~~uf ME _ A_. — -9-. 90-. — 0.3. 0. 2. Measured Serafini Churchill Abraham U Us Trend 93 m/s 335 m/s 0 * ---- A A --- O *. v w__ 0.1_ I I I _ _ I 0.021 0. 043 0. 063 0.083 0.104 3 q, cm/cm s 5.6 11.26 16.9 22.5 28.2 Ref, Film Reynolds No. Fig. 5.22 Comparison of Average Measured Acceleration and those using other Correlations of CD

220 direction, moving a short distance (a2 cm) upstream, These negative accelerations are shown; in Figs. 5.23 and 5.24 The only plaus.ible mechanism for this droplet deceleration and motion reversal appears to be periodic vortex shedding from the relatively blunt blade trailing edge. Figure 5,25 is a schematic of the hypothesized flow pattern. If an individual vortex is not aligned precisely with a droplet center (see lower portion of Fig. 5.25), then such a droplet will experience an additional (upstream or downstream) relative velocity and drag due to the velocity field of that particular vortex, For such upstream relative velocity, it would be momentarily decelerated. Even though this hypothesized mechanism is consistent with author's data, vortex measurements have not been possible. Nevertheless it does appear the only plausible explanation for the observations. Some related parameters can be deduced. Average time interval A t between decelerations, calculated from the R vs. t curves (Figs. 5.23 and 5.24), is a measure of the vortex shedding frequency. A calculation of At was possible only for low steam velocities (U = 93 and 159 m/s). For higher steam velocity, the droplets disintegrated too rapidly for accurate measurements of this type, Values of At and related parameters are listed in Table 5.4. Two non-dimensional parameters ($ and ReD) were selected to correlate the droplet deceleration frequency. These numbers are defined below.

221 4.0 3.0 2.0 1.0 2 m$ l U - 93 m/s q 0.010 cm/cm.s q - 0.010 cm3/cm.s 0 -F-4 Ai m'4 C 0 -W4 U u 0.4 S 0 t W'-4 W Cu ha 0) C-F 0) U -P<m -1.0 -2.0 -3.0 -4.0 U - 93 m/s s q - 0.083 cm3/cm.s 4.0 3.0 2.0 1.0 2 ms I I -1.0 -2.0 -3.0 -4.0 Fig. 5.23 Acceleration Ratio, R, versus time R= A /A m s

222 2 ns U - 159 m/s 4.0 Lq - 0.010 cm3/cm.s 4.0 3.0 2.0 1.0 3.0 2 ms Us = 159 m/s 0 o q t 00104 cm.s 0.04 -2. g -3.0-4.0 4.0 3.0 2 ms Us -159 n/s b. 1 q - 0.104 cm /cm..s 2.0 i' 1.0 S.ix aix A/\IV t 0 -3.0 -1.0 u -2.0 -3.0 -4.0 Fig. 5.24 Acceleration Ratio, R, versus time. R= A /A m s

223 Vapor Velocity Flow Vortex Ot o t Droplet 0 C) 0 Blade 5367 Induced Velocity Field An Ideal Discrete Vortex Liquid Droplet Fig. 5.25 Schematic of Vortex Generation from the Trailing Edge of the Blade

Us q D Am n, Vi Fr At t Vf S, ReD SB S n c m/s m/s 10 cm/sf r/s cm/h/c cm cm/s n M/s m2/s 10-3s cm/s. sec" -- ---— /s - cl/m c ms 7'............ 93.010 93.021 93.043 93.083.11.12.09. 12 27550 28780 23454 22116.78 1. 01.78.73 44187 55.20 51. 16 43.23 159 159 159 159 159 159.010.021.031.043.063.104.08.08.09.07.09.09 75144 104906 82940 69542 79275 65203. 70.98.97.79.83.75 74.34 63.11 70. 76 44. 24 63.30 58.68 7. 573 2.669 9.293 3.248 8. 594 4. 556 7. 262 2.397 12.489 2.333 10.602 3.500 11.887 3.942 7.432 4.083 10. 634 4. 200 9. 859 4. 306 7.45 10. 93 16.4 25.75 13.67 18. 22 22.36 25.23 30. 75 39.05.00403.00331,00236.00538.00216.00144.00144. 00140.00.135.00131 6.03 2.82 2.54 1. 94 2.51 1. 25 1. 02..874.697.535 958. 2 1036. 5 970.4 1156. 0 1364. 7 1316.3 1541. 6 1526.1 1443.3 1506. 3. 01611;.01324.00944.01794.0108.00719.00638.00616.00599.00584 20. 110 11. 266 5.353 6.479 12. 522 6. 255 4.522 3.870 5. 088 2. 370 U: steam velocity s q.specific liquid flow rate D average drop diameter a: measured acceleration.mn a s: acceleration ratio a m Vi: induced velocity r' vortex strength; h u 5S B blade thickness =.4cm AT: periodicity of deceleration Vf: velocity of film D Sv: Strouhal number ( f ) 5 8 Sv: Stroulal number ( - ) vf VfAt UsD ReD: ( -p ) B S ( B B At U B SB. B AT SBv ATV ) Vt f N) N' tJ Table 5, 4: Parameters Concerning the Periodic Deceleration of Droplet in the Wake Region.

225 B S,1B = Uf-'t- Strouhal number (5.16) D U ReD= Droplet Reynolds Number D Vs (5.17) where B is blade thickness, At periodicity of deceleration, and D effective droplet diameter. SB is plotted against ReD in Fig. 5.26. Although data scatter is large, the solid lines show the general trends. SB decreases as Us increases from 93 to 159 m/s. This is due to the increase in both Us and At. (Average at = 3.1 ms for Us = 9.3 m/s and 3.8 ms for Us = 159 m/s). For a given steam velocity, SB increases with increasing ReD, remaining roughly constant at a higher ReD. The present observation of droplet wake deceleration is apparently the first such reported. Thus no comparison with other experimental or theoretical results is possible. 3. Disintegration Time for Droplets For high steam velocities, many of the large droplets were observed to undergo initial acceleration and then disintegrate into a mist of much finer particles. Break-up time, tb, is here defined as the time between droplet formation (from the liquid fingers) and droplet disintegration into either an amorphous mass or numerous smaller droplets. One of these disintegration processes is "disc" disintegration. Analogous behavior has been observed in shock tubes. (L8, H8, H15, J1, N8, W8) An intially almost

226 0 X m Re f 5.3x104 9.0x104 U (/ s (m/s) 93 159 0.02C 0 0 00 0 0.,018 0 0 0 O 0 0 0.016 - 0 0.014 - 0 0 0 00 0 0 X X X X 0.012t 0 0 4) 03 o Ui u 0.010 - 0 0.008t 0 x x x X x y x 0.0061 X X 0.004 0,002j I 1 t 1 J) ~ I* I'I II SB Red 600 Droplet 800 Reynolds 1000 Number, 1200 1400 1600 1800 Re. US 0 Ps Fig. 5.26 Strouhal Number, Sg, versus Droplet Reynolds Number, Re a

227 spherical droplet assumes a disc shape with axis parallel to flow direction. Upon further acceleration, it breaks into several smaller droplets. Another disintegration mode here observed is the "club" or "claviform" process (H16 Kll) A droplet is formed around and behind a long liquid jet ("club" or "claviforn"). Break-up of the "club" then occurs with the residual droplets also quickly disintegrating. Further details are provided elsewhere (H16 Kll) In the case of supersonic aircraft rain-erosion, two additional modes of droplet break-up have been suggested(A4' R4, El, K12 W9, W10, R5 B8)involving shock effects. Passing through a shock wave, droplets are either flattened (normal to the axis), with smaller droplets stripped from the circumference, or almost instantly torn apart. Generally these other investigators use geometries different from author's. Liquid droplets of known size are often injected perpendicular to the flow. Shock waves may or may not occur. Author's droplets are formed from the liquid fingers previously discussed and shock waves are not involved (M<1). Author's droplet disintegration datahave been described using the following non-dimensional variables; (1) Weber number, Web, i.e. the ratio between steam dynamic liquid surface tension effects, here defined as, eb P r2 Do We = tr a b ar

228 Steam density, ps,. and velocity,.Ur, relative to the droplet, are used, D is the original droplet diameter and 6 the liquid surface tension. (2) A non-dimensional break-up time, Tb, ie,, droplet lifetime before break-up, normalized to consider droplet size and both droplet and steam densities. Tb, used previously(A4' W9,W10, R5,B8) is defined: U th p 1/2 Tb = -r (5.19) Do f where tb is the pre-disintegration droplet lifetime. pI and p are liquid and steam densities. The square of Tb is the ratio of dynamic steam pressure to a deformation droplet stress. This concept was discussed for finger formation. (3) Q is non-dimensional flow around a droplet, al(A4,B8,R5,W91WlO) ready defined elsewhere A4It is the ratio of dynamic steam pressure on the droplet surface to upstream dynamic steam pressure, and is defined: 2 ~g Ug es U 2 (5.20) Us s2 Here pg and U are local steam density and velocity and ps and Us the corresponding upstream values. Here pg ~ ps Then Q becomes: U 2 Q = cU (5.21) Since Q is a locally varying quantity, a characteristic. is its maximum.. For low to moderate subsonic flows over a sphere, potential flow gives for a maximum, Q = 9/4 Csee Arts, 92-96 in Ref, L6 and Section 15.30-15.53 in Ref. M13).

229 However, for high subsonic velocities as in some of authorts cases^ empirical results are required. A simple CR5) is empirical expression for CR5) is IM = 0,78 + 23'4; OM3.5 (5.2. 1 + 2,1. where M is steam Mach number. Combining Eqs. 5,19 and 5,21, non-dimensional time, Tb, can be modified (Wl0R5BS8 as in the following: b Q /2 - u I - (5.2 Tb Qm D (5.2. o \Qf1 s Then with the assumption that U r U, valid for the present r s tests, Eq. (5.23) becomes: 2) 3) /~*) U t, /s1/2 Tb Q1/2 = 9 (5.24) The modified non-dimensional break-up time,(Eq. 5.2) can be physically interpreted as the square root of the ratio of maximum local dynamic pressure on the droplet surface to the accompanying droplet deformation stress. TbQm1/2 is plotted against Web in Figs. 5.27 -5.30, and is found to be roughly inversely proportional to Web. The proportionality constants increase with increasing steam velocity. Experimental data obtained elsewhere in shock tubes CN8,H16,H17,A4,R5B8), B ) ( H 6 / 1 shows the non-dimensional break-up time, Eq. (5.19) or C5,24), to be approximately proportional -1/4 to Web

230 *V tm= 93 m/S (M = 0.23) steam 159 rn/ (M = 0.38) s 45 team D = 0.45' 1.35 mm o 102 0,4 E 0. 0 10 1 0 0 T l/2 _ 300 10 Web Fig. 5.27 Nondimensional Drop Breakup Time vs Droplet Weber No.

231 V = 250 m/S (M = 0.61) steam Do = 0.45 ~ 1.35 mm 102 \- O 8 0 0 O 0 10 31 __. _ I L t.-. - I t I I I 1/2 700 b m Web 10l Web eb Fig. 5.28 Nondimensional Drop Breakup Time vs Droplet Weber No.

232 steam = 297 m/S (M = 0.72) Do = 0.45 N 1.35 mm 102 10 1 T Qm1/2 = 1000 b m We b 10 102 Web Fig. 5.29 Nondimensional Drop Breakup Time vs Droplet Weber No.

233 V = 335 m/S steam D = 0.45 p 1.35 o (M = 0.81) mm 10 p I 0 Tb Q1/2 1500 Tb Qm Wei 9 "I, 0 0 0 P-4i (5 tSI 1 I I I I I I I. I I I I I, I 10 102 o03 Web Fig. 5.30 Nondimensional Drop Breakup Time vs Droplet Weber No.

234 Figure 5,31 compares the shock tube data with that of present tests, The solid line is a previously proposed(RS) empirical correlation of the shock tube data. Author's data are shown as circles, The exponent difference in droplet Weber number dependence can be explained by considering the wave growth model on the (B8, R5) droplet surface proposed by previous investigators.(B It is therefore concluded that an inverse Web dependence (as found by author) is primarily due to a Kelvin-Helmholtz instability (C17,L6) near the equator of the droplet. A We 1/4 dependency (from shock tube data) is due to the prevailing Rayleigh-Taylor instability (C17,L6) near the pole of the droplet, Details of the wave growth analyses originally suggested C617,L6jR5jB8) by previous researchers (617L6,R5B) are given in Appendix G.

Breakup Time vs. Web 0 M = 0.23 S M a 82 ma * ~ t resent lcrodrop 28. 26 - x A00+ v AOO+V x Hasler (steam-water) (HI) Nicholson (air-water) (" ) Hanson, Domich & Adams (air-water)(HIq) Anderson & Wolfe (air-water) (A4) 24 22 1 20 18 16 TbQm /2=23(We-8)-1/4 Reinecke,Waldman Correlation (R5) x~ ~ Y A v 0 RI0 0 X Hasler Pdyne (N/M2) A 1.558x103 1. 492x103 O 0.971x103 V 0.615x103 1.382x103 Tsteam0C Do(m) Pstat 67.15 4x10-3 1.4x104 65.41 4x10-3 1.4x104 106.50 4x10-3 9.6x104 66.6 1.2x10-3 1.4x104 198.9 4.0x10-3 9.6x104 N rE-0 14 A rr A tn tn ul 0 X 12 - + +O x I n _ n lU r 8 6 4 2 - K x X'0 A A O_ A -~Ot Al:wft %EMu-.V. or 0 0 I I _ --- -- — ~~~~~~~~~~~ - -- -- I~~~~~~~ --- ] 0 102 10 We e, Fig. 5.31 Nondimensional Breakup Time vs Droplet Weber No. (Comparison between shock tube experimental data and author's data)

CHAPTER VI CONCLUS IONS Phenomena leading to large steam turbine blading erosion have been investigated experimentally, These include the behavior of thin liquid films, and their subsequent break-up into droplets and fingers at the trailing edge of a stationary blade, here modelled by a thin flat plate. Based primarily on photographic results, a film "Transition Map" for various flow regimes has been made. Three different film flow regimes were observed: rivulet, wavy films, and an entrainment regime corresponding to more vigorous wave action. Liquid film thicknesses were measured as a function of time and position using electrical conductivity microgages. A best empirical correlation for author's film thickness data is: h' = 0.6 Ref 1/2, where h+ is a non-dimensional film thickness and Ref film Reynolds number. Few previous correlations for this very thin film range exist. However, an extrapolation of Hughmark's thick-film correlation(H6) along with some data from Collier-Hewitt(C3) give a prediction higher than author's correlation for given Ref. Film entrainment at author's high steam velocities may be partially responsible for this discrepancy. The author obtained the critical conditions for rivulet flow experimentally, Comparison with existing theoretical analyses-shows- that-the -wetting contact — angles -required in the analyses are much smaller than measured for static 236

237 or one curvature condition. Modifications of these analyses are suggested based on author's observations. First, the flow geometry around the filament or dry spot stagnation points should be considered in more detail. Author's photographs show that there is an abrupt bulge in the film near the stagnation point, instead of a smooth decrease in film thickness from upstream as had been previously assumed. Also, film wave characteristics should be more carefully considered. Photographs from the present study show that some waviness always exists in the rivulet flow regime. These wavelets will certainly affect the dynamic forces causing film separation. For the usual thick film analysis, the contribution of pressure perturbations to film instability with respect to wave generation is usually dominant, and that of shear perturbations negligible. For thin films (author's case) shear perturbations(C become relatively more important as film thickness decreases. Since author's photos always show some waviness even in the rivulet flow regime (which corresponds to the thinnest films), it is believed that the thin film shear perturbation mechanism is most important for author's case. Measured two-dimensional wave propagation speed has been compared with Wurz's 4) model, which well correlates author's high subsonic data. However this is not true for low subsonic data. All the two-dimensional wave celerity data is better explained by the classiciai capillary-gravity

238 wave model. For the film entrainment regime, both experimental and theoretical criteria are found. Liquid film entrairnment was photographically observed to involve a rapid acceleration, lifting, and then shattering of small complex "wavelets" at the crests of larger waves. There is a limiting liquid flow rate below which no surface entrainment occurs ( 0.07 cm3/sec.cm). Based on author's thickness measurements, a correlation for critical entrainment conditions is found. This correlates the horizontal duct data better, including author's, than previous correlations. (2) However, confidence in author's correlation is limited by the present lack of detailed information on actual shapes of larger waves, and the present inability to accurately calculate pressure variations over waves of arbitrary shape and size. At the trailing edge, liquid finger length, L, and finger wavelength, Af, have been measured photographically. The upper limit of L for all time-spans decreases as steam velocity is increased. Also "dead" periods, while no finger is forming, are observed. Finger shedding frequency decreases with increasing steam velocity, Us, and increases with an increase of liquid flow rate. The dependence of shedding frequency on Us is explained by boundary layer thinning with increasing shear stress. It has been found that Karman vortex-shedding in the steam is not a controlling factor in finger formation.

239 A peak in nondimensional finger shedding period, T, was found for lower film Reynolds number ranges. The peak value of t is reduced, and the corresponding values of Wes is increased as Ref is increased up to an intermediate value, Ref = 8.4. For higher Ref, no such peaks were observed. Rather C then increases smoothly with increasing Wes. Non-dimensional finger length, L*, normalized by transverse finger wavelength, Af, was found to increase with Ref at fixed Ref. These fluctuations could be due to various overall system resonances, which may themselves depend upon steam velocity. Since author's finger formation observations are apparently the first reported, no comparison with other related studies is possible. Measured droplet accelerations were compared with those calculated using Serafinils empirical drag coefficient However, author's wake-droplet behavior is complicated by the presence of many droplets together, i.e., significant "shielding" may exist, Deceleration of droplets which sometimes reverse direction completely and move against the main stream, was observed. This phenomenon can be explained by the reverse velocity field induced by vortex shedding in the steam. A Strouhal number, SB, considering the periodicity of deceleration, increases with increasing drop Reynolds number, Red, remaining roughly constant at higher Red. This general behavior is consistent with most vortex-shedding observa tions.

240 Disintegration time for the larger range of drops observed was measured along with characteristic droplet disintegration shapes Cdisc and "club'" formsl. Non-dimensional break-up time, Tb, is found to be proportional to Web -1/4 However, shock-tube data shows Tbc < We. These dependencies are explained using surface wave growth theory.

REFERENCES A Al, Abraham, F. F., "Functional Dependence of Drag Coefficient of a Sphere on Reynolds Number," Physics of Fluids, Vol. 13, pp. 2194-2195, 1970. A2. Allen, J. M., "Evaluation of Compressible-Flow Preston Tube Calibrations," NASA TN D-7190, 1973. A3, Alia, P., Cravanolo, L., Hassid, A. and Petrocchi, E., "Liquid Volume Fraction in Adiabatic Two-Phase Vertical Upflow - Round Conduit," CISE Rept, R-105, 1965. A4. Anderson, W. and Wolfe, H,, "Aerodynamic Breakup of Liquid Drops," Proceedings of 5th International Shock Tube Symposium, sponsored by the Div. of Fluid Dynamics of the Amer. Phy. Soc., April 1965. A5. Austrow, J., "Trailing Edge Observations of the Steam Tunnel Experiments," DRDA Rept. No. UMICH 014571-19-I, April, 1978. B Bl. Baumann, K., "Some Recent Developments in Large Steam Turbine Practice," Eng., vol. III, p. 435, 1921. B2. Bankoff, G. S., "Minimum Thickness of a Draining Liquid Film," Int. J. Heat Mass Transfer, 14, 2143, 1971. B3. Brown, C. and Matin, S. A., "The Effect of Finite Metal Conductivity on the Condensation Heat Transfer to Falling Water Rivulets on Vertical Heat-Transfer, Trans. ASME, J. of Heat Transfer, pp. 69-76, 1971. B4. Benjamin, T. B., "Wave Formation in Laminar Flow Down an Inclined Plane," J. of Fluid Mech., 2, p. 554-575, 1957. B5. Brumfield, L. K. and Theofanous, T. G., "On the Prediction of Heat Transfer Across Turbulent Liquid Films," J. of Heat Trans., Trans. ASME, pp. 496-502, 1976. B6, Benjamin, T. B., "'Shearing Flow Over a Wavy Boundary," J, Fluid Mech., vol. 6, pp. 161-205, 1959, B7. Blanchard, D. C. and Woodcock, A, H., Tellus, yol. 9, pp. 2,.1957. B8. Barber, J. P., "Water Drop Break Up/Impact Damage Thresholds," AFML-TR 76-126. 241

242 B9. Bleyins, R. D., "Flow Induced Vibration," Van Nostrand Reinhold Co., 1977. B10. Blome, S., "-Wet Steam Tunnel Observations of Fingers at Trailing Edge," DRDA Rept. No. UMICH 014571-8-I, April 1977. C Cl. Cohen, L. S. Hanratty, T. J., "Effects of Waves at a Gas-Liquid Interface on a Turbulent Air Flow," J. Fluid Mech., vol. 31, p. 467-479, 1968. C2. Crane, R. I., "Deposition of Fog Drops on Low Pressure Steam Turbine Blades," Inst. J, Mech. Sci., vol. 15, pp, 613-631, 1973. C3. Cole, J. E., and Dobbins, R. A., "Propagation of Sound Through Atmospheric Fog," J. Atmos. Sci., vol. 27, May 1970. C4, Chu, K, J., Dukler, A. E., "Statistical Characteristics of Thin Wavy Films," A.I.Ch.E. J,, vol. 20, no. 4, pp. 695-1706, July 1974, and vol. 21, no. 3, pp. 583-59.3, May 1975. C5. Chun, K, R. and Seban, R. A., "Performance Prediction of Falling Film Evaporators," J. Heat Trans., Trans. ASME, pp. 391-396, 1971. C6. Chun, K. R. and Seban, R. A., "Heat Transfer to Evaporating Liquid Films," J. Heat Trans., Trans. ASME, pp. 391-396, 1971. C7. Cohen, L. S. and Hanratty, T. J., "Generation of Waves in the Concurrent Flow of Air and a Liquid," A.I.Ch.E. J., pp. 138-144, 1965. C8. Cohen, L. S. and Hanratty, T., "Height of a Liquid Film in a Horizontal Concurrent Gas-Liquid Flow," A.I.Ch.E. J., pp. 290-291, 1966. C9, Cohen, L. S. and Hanratty, T,, "Effect of Waves at a Gas Liquid Interface on a Turbulent Air Flow," J. Fluid Mech., vol, 31, pp, 467-479, 1968, C10, Craik, A.D. D., "Wind-Generated Waves in Thin Liquid Films," J. Fluid Mech,, yol, 26, pp, 369-392, 1966. Cll. Chang, I, D. and Russell, P, E., "Stability of a Liquid Layer Adjacent of a High-Speed Gas Stream," The Physics of Fluids, vol, 8, no, 6, pp. 1018-1026, 1965.

243 C12. Charvonia, D. A., "A Study of the Mean Thickness of the Liquid Film and the Characteristics of the Interfacial Surface in Annular Two-Phase Flow in a Vertical Pipe," Purdue Rept. 1-59-1, l95,. C13, Collier, J. G, and Hewitt, G, F., "Data on the Vertical Flow of Air-Water Mixtures in the Annular and Dispersed Flow Regions-Part B; Film Thickness and Entrainment Data and Analysis of Pressure Drop Measurements," UKAEA, AERE, 3455, 1960, C14, Churchill, S. W. and Usagai, R,,'A General Expression for the Correlation of Rates of Transfer and Other Phenomena," A.I.Ch.E. J., vol. 18, no. 6, pp. 1121-1128, 1972. C15. Collier, J. G. and Hewitt, G. F., "Film Thickness Measurements," ASME paper no. 64-WA-HT-41. C16. Chien, S. F. and Ibele, W., "Pressure Drop and Liquid Film Thickness of Two-Phase Annular and Annular-Mist Flows," ASME paper no, 62-WA-170, 1960, C17. Chandrasekhar, S., "Hydrodynamic and Hydromagnetic Stability," Oxford Univ. Press, 1961. C18. Cousins, L. B., Denton, W. H., and Hewitt, G. F., "Liquid Mass Transfer in Annular Two-Phase Flow," Symp. on Two-Phase Flow," Exter, England, 1965, C19. Christie, D. G., Hayward, G. W., Lowe, H. J., MacDonald, A. N. and Sculpher, P., "The Formation of Water Drops Which Cause Turbine Blade Erosion," Inst. of Mech. Engr, Thermodynamics and Fluids Mechanics Convention, Liverpool, England, April, 1966. D D1. Deich, M. E., "Investigation of the Velocity of Sound and Decrement of Attenuation of Sound-Waves in TwoPhase Media," Proc, Inst. Fluid Flow Mach., nos. 29-31, pp. 141-159, 1966.. D2. Denny, V. E., Millsf A, F. and Jusionis, V. J., "Laminar Film Condensation from a SteamnAir Mixture Undergoing Forced Flow Down a Vertical Surface, J, Heat Trans., Trans. ASME, pp. 297-304, 1971. D3. Duckler, A. E., "Fluid Mechanics and Heat Transfer in Vertical Falling Film Systems," Chem. Engr. Prog. Symp., ser, 65, 1-10, 196Q. D4. Depooter, K., Brundrett, E. and Strong, A. B., "The Calibration of Preston Tubes in Transpired Turbulent

244 Boundary Layers, t J, Fluid Engr., Trans. ASME, vol. lQ0., pp 16. 1- 978, D5. Dodd, K. N., "On the Disintegration of Water Drops in an Air Stream," J. Fluid Mech, vol. 9, pp. 175, 1960, D6. deNevers, N., "R'ate Data and Derivatives," A.I.Ch.E.J., vol, 12, no. 6, pp. 1110-1115, Nov. 1966, E El. Engel, 0. G., "Fragmentation of Waterdrops in the Zone Behind an Air Shock,"1 J. Res, NBS, vol. 60, no. 3, Mar. 1958. F Fl. Fenyes, Peter, "Liquid Film Thickness Measurement, Design and Use of a Conductance Probe," DRDA Rept. No. UMICH 12449-6-I, Dec. 1974, F2. Fillipov, G. A., Povarov, 0. A. and Pryakhin, V. V., "Investigation and Calculation of Wet Steam Turbines," (Russian), Moscow, 1973, G Gl. Gyarmathy, G. and Meyer, H., "Spontane Kondensation," Forsch V.D.I., no. 508, 1965. G2. Gyarmathy, G., Burkhard, H-P, Lesch, F. and Siegenthaler, A., "Spontaneous Condensation of Steam at High Pressures: First Experimental Results," Instn. Mech. Engrs., London, Conf. on Heat and Fluid Flow in Steam and Gas Turbine Plant, Univ. Warwick, April 3-5, 1973, paper C 66/73, pp. 182-186. G3. Gyarmathy, G.,. "Zur Wachstumsgeschwindigkeit kleiner Flussigkeit in einer ubersattigten Atmosphare," Zeitschrift Fur Angewandt Mathematik und Physik (ZAMP), vol. 14, no. 3, 1963, pp, 280-293, G4. Gardner, G. C., "Events Leading to Turbine Blade Erosion," Proc. Inst. Mech. Eng,, vol. 178, pto 1, no, 23, pp. 593-624, 1964. G5 Ganic, E. N, and Rohsenow, W. M., "On the Mechanism of Liquid Drop Deposition in Two-Phase Flow," ASME paper no, 76.6WA/HT-t18, 19766, G6, Gregor, W. and Rumpf, Ht "Velocity of Sound in TwoPhase Media," Int. J. Multiphase Flow, vol. 1, pp. 75376&9, 1975. G7. Gater, R. A. and L'Ecuyer, M. R., "A Fundamental

245 Investigation of.the Phenomena that Characterize Liauid-Filim Cooling," Inst. J. Heat Mass Trans., vol, 13, pp. 1925-1939 1970.: G8. Gill, L, E. and Hewitt, G, F, "Sampling Probe Studies of Gas Core in Annular Two-Phase Flow III," UKAEA Rept. AERE-M 1202, 1966, H H1. Hill, P. G,, "Condensation of Water Vapor During Supersonic Expansion in Nozzle," J. Fluid Mech., vol. 25, 1966, pp, 593-620, H2. Hartley, D. E. and Murgatroyd, W., "Criteria for the Break-up of Thin Liquids Layers Flowing Isothermally over a Solid Surface," Int. J, Heat Mass Transfer 7, 1003, 1964. H3. Hobler, T., "Minimal Surface Wetting in a Centrifugal Force Field," Chemia Stosow 2B, 265, 1968. H4. Hewitt, G. F. and Lacey, P. M.C., "The Breakdown of the Liquid Film in Annular Two-Phase Flow," Int. Heat Mass Transfer, vol. 8, pp. 781-791, 1965. H5. Hewitt, G. F., Kearsey, H. A., Lacey, P.M.C. and Pulling, D. J., "Burn Out and Nucleation in Climbing Film Flow," Int. J. Heat Mass Transfer, vol. 8, pp. 793814, 1965. H6. Hughmark, G. A., "Film Thickness, Entrainment and Pressure Drop in Upward Annular and Dispersed Flow," A.I. Ch.E. J., vol. 19, no. 5, pp. 1062-1065, 1973. H7. Henstock, W. H. and Hanratty, T. J., "The Interfacial Drag and the Height of the Wall Layer in Annular Flows," A.I.Ch.E. J., vol. 22, no. 6, pp. 990-1000, 1976. H8. Hinze, J. O., "Fundamentals of the Hydrodynamic Mechanism of Splitting in Dispersion Process," A.I.Ch.E. J., 1, 289, 1955, E9. Hammitt, F. G., Keller, A., Ernst, G., "Steam Tunnel Initial Operation and Res.ultq," DRDA Report no. UMICH Q12449-l1-T, Jan. 1974. H10. HamrIitt, F. G,, Hwang, J. B., Mancuso, A.^ Krause, D., and Blore, S., "Liquid Fil!n Thickness Tests - Wet Steam Tunnel," DRDA Report no, UMICH. 012449-9-T, June 1975. Hll. Hammitt, F. G., Hwang, J-B and Kim, W., "Liquid Film Thickness Measurements in University of Michigan Wet Steam Tunnel," 1976 ASME Cavitation and Multiphase Flow --.......

246 Forum, May 1976., H12. Head, M. R. and Rechenberg, I., "The Preston Tube as a Means of Measuring ki n Friction," J. Fluid Mech., vol. 14, pp. 1-17, 1962, H13. Hewitt, G. F, and Hall-Taylor, N. S., "Annular TwoPhase Flow," pp, 136-.148, Pergamon Press, N,Y,, 1970. H14. Hartley, D. E. and Roberts, D. C., "A Correlation of Pressure Drop Data Two-Phase Annular Flow in Vertical Channels," Nuclear Res, Memo Q6, Queen Mary College, Univ. of London, 1961. H15. Hinze, J. O., "Critical Speeds and Sizes of Liquid Globules," Appl. Sci. Res., vol. Al, pp. 273, 1949. H16. Hassler, G., "Breakup of Large Water Drops Under the Influence of Aerodynamic Forces in a Steady Stream of Steam at Subsonic Velocities," Proc. of 3rd Intl' Conf. on Rain Erosion and Related Phenomena, Hartley Whitney, Hampshire,England, Aug. 1970. H17. Hanson, A., Domich, E., and Adams, H., "Shock Tube Investigation of the Break up of Drops by Air Blasts," The Physics of Fluid, vol. 6, no. 8, pp. 1070-1080, Aug. 1963. H18. Hammitt, F. G., Lihn, D. N., Tully, J. E., Ernst, G. C., Hwang, J-B, Krause, D. A. and Bhatt, N. R., "Film Thickness Electrical Conductivity Gages - Detailed Utilization Procedure for U-M Wet Steam Tunnel," DRDA Report no. UMICH 012449416-I, April 1975. H19. Hewi'tt, G. F., King, R. D. and Lovegrove, P. C. "Techniques for Liquid Film and Pressure Drop Studies in Annular Two-Phase Flow," Atomic Energy Res. Est., R3931, 1962. H20. Hammitt, F. G., Kim, W. and Bou-Maroun, A., "Thin Shear Driven Water Film Disintegration and Wavelet Celerity," DRDA Report no, UMICH. 014571-13-I, Jan. 1978, I I1. Ihnatowicz, E,; Gunkowski, S. and Mikielewica, J,, "Experimental Study of Evaporation and Breakdown of Thin Liquid Films Driven by Shear Stress," ASME Winter Annual Meeting, Atlanta, Ga., Novy 27-Dec, 2, 1977. I2, Ishii and Grolines, "Inception Criteria for Droplet Entrainment in Two-Phase Concurrent Film Flow," A.,I.Ch.E. J, vol. 21, no. 2, pp, 308-318, 1975. _._.,.....

247 J J1. Jenkins, D. C,, and Booker, J. D, "Time Required for Righ-speed Air Streams to Disintegrate Water Drops," Aero. Res., Council Current Papers, C.P. No. 827. J2. Jeffreys, H., "Resonant Phenomena in Gravity Waves," Proc- Royal Soc., A1Q7, 189, 1925. K K1. Khoshaim, B. H. and Ryley, D. J., "Rivulet Flow Over Steam Turbine Blade," CERI meeting, Ft. Lauderdale, Fla., Oct. 18-20, 19.76, K2. Krantz, W. B., and Zollars, R,, "The Linear Hydrodynamic Stability of Film Flow Down a Vertical Cylinder,t" A..Ch.E. J, vol. 22, no. 5, pp. 930-934, 1976. K3. Kutateladze, S.S., "Elements of the Hydrodynamics of Gas Liquid System," Fluid Mech. - Soviet Res. 1 (4), 29, 1972. K4. Kim, W. and Hammitt, F, G., "Observation of Liquid Fingers and Drop Formation from Trailing Edge of Simulated Turbine Blade in Wet Steam," DRDA Report no. UMICH 014571-14-I, April 1978. K5. Krzyzanowski, J., "Wet-Steam Tunnel Facility-Design and Program of Investigations," DRDR Report no. UMICH 03371-18-T, June 1972. K6. Kim, W., Hammitt, F. G. and Krzecskowski, S., "Investigation of the Behavior of Thin Liquid Film with CoCurrent Steam Flow," Two-Phase Transport and Reactor Safety, Proc. Two-Phase Flow and Heat Transfer Sym. Workshop, Oct. 18-20, 1976, Ft. Lauderdale, Fla, vol. IV, pp. 1213-1230, Hemisphere Publ. Co., 1978. K7. Kim, W., Hammitt, F. G., Blome, S., and Hamed, H., "Thin Shear Driven Water Film Wavelet Characteristics," 1977 ASME Cayitation and Multiphase Flow Forum, June 1977. K8. Krzeczkowski,,, Kim, W. and Hanriitt, F. G., Investigation of Secondary Liquid Phase Structure in Steam Wake," submitted to ASME, J. Fluid Mech. K9. KQrdyban, E.^ "Interfacial Shear in Two-Phase Wavy Flow in Closed Horizontal Channels " J. Fluid Engr., Trans. ASME, pp, 9.7-1iQ02 June, 1974. K10. Kosky, P, G., "Thin Liquid Films Under Simultaneous Shear and Gravity Forces," Inst. J. Heat Mass Transf.,

248 vol. 14, pp. 122Q-1224, 1971: Kil. Krzeczkowski, S. and Hanmi tt, F. G., "Experimental Investigation of Liquid Droplet Break-Up Duration," Cavitation and Multiphase Flow Forum, 1977, K12. Krause, W. E,, "Water Drop Deformation and Fragmentation Due to Shock Wave Impact," Ph.D, thesis, Univ, of Florida, 1970, K13, Kim, W. and Hammitt, F, G,, "Observations of Liquid Finger and Drop Formation from the Trailing Edge of Simulated Turbine Blade in Wet Steam," DRDA Report no. UMICH 014571-14I, April 1978. L L1. Lin, S. P., "Instability of a Liquid Film Flowing Down an Inclined Plane," The Physics of Fluids, vol. 10, no. 2, pp. 308-313, 1967, L2. L-n, S. P., "Finite Amplitude Stability of a Parallel Flow with a Free Surface," J. Fluid Mech., vol. 36, pp, 113-126, 1969, L3, Lin, S, P., "Profile and Speed of Finite Amplitude Waves in a Falling Liquid Layer," The Physics of Fluids, vol. 14, no. 2, 1971. L4. Levich, V. G., "Physiochemical Hydrodynamics," Prentice-Hall, Englewood Cliffs, N.J., 1962. L5. Levy, S., "Prediction of Two-Phase Angular Flow With Liquid Entrainment," Int. J. Heat Mass Trans., vol. 9, pp. 101-188, 1966, L6. Lamb, H., "Hydrodynamics," Dover Publications, 1932. L7. Langmuin, I. and Blodgett, K. B., "Mathematical Investigation of Water Droplet Trajectories," Tech. Rept. No. 5418, Air Materials Command, AAF, Feb. 1946. L8. Lane, W..R., "Shatter of Drops in Streams of Air," Ind. and' Chgr, Chem,, vol, 43, pp. 1312, 1951. M1. Miller, E, H,, Schofield, P.^ "The Performance of Large Steam Turbine Generators with water Reactors," ASME Winter Annual Meeting, N.Y,, Nov. 26-3Q, 1972. M2. Moses, C, A. and Stein, G. D., "On the Growth of Steam Droplets Formed in a Laval Nozzle Using Both Static Pressure and Light Scattering Measurements," Condensation in High-Speed Flows, ASME Jt, Appl. Mech. Fluid

249 Fluid Eng. and BiQeng. CQnf Yale Uniy., New Haven, Conn., June 15-17, 1977. M3, Mikielewicz, J. and Moszynski, J. R., "Minimum Thicks ness of a Liquid Film Flowing Vertically Down a Solid Surface," Inst. J. Heat and Mass Transfer, vol. 19, no. 7, pp. 771-776, 1976. M4. Miles, J. W, r "The Hydrodynamic Stability of a Thin Film of Liquid in Uniform Shearing Motion," J. Fluid Mech., 13, p, 433N448, 1962. M5. Miles, J. W.,'"On the Generation of Surface Waves by Shear Flows," J, Fluid Mech., 13, p. 433-448, 1962. M6. Miles, J. W,, "On the Generation of Surface Waves by Shear Flows, Part I, II and III," J. Fluid Mech., vol. 3, pp. 185-204, 1957, vol. 6, pp. 568-582, 1959, vol. 6, pp. 583-598, 1959. M7. Mills, A. F. and Chung, D. K., "Heat Transfer Across Turbulent Falling Films," Int. J. Heat Mass Transfer, vol. 16, pp. 694-696, 1973. M8. Mikielewicz, J. and Moszynski, J. R., "Breakdown of a Shear Driven Liquid Film," Trans. Inst. Fluid Flow Machinery, Polish Acad, of Sci., vol. 66, pp. 3-10, 1975. M9. Mancuso, A., Hammitt, F. G., Blome, S. and Hwang, J-B, "Data Reduction Procedures - Steam Tunnel Film Thickness Tests," DRDA Report no. UMICH 012449-20-Ia. M10. Mikielewicz, J. and Hammitt, F. G., "Generalized Characteristics of Electrical Conductance Film Thickness Gages, " Prace IMP, pp. 87-95, 1977. Mll. Miya, M., Woodmansee, D. E. and Hanratty, T. J., "A Model for Roll Waves in Gas-Liquid Flow," Chem. Eng. Sci., vol. 26, pp. 1915-1931, 1971. M12. Murgatroyd, W., "The Role of Shear and Form Forces in the Stability of a Dry Patch in Two-Phase Film Flow," Int. J. Heat Mass Transfer, v. 8, pp. 297-301, 1965. M13. Milne-Thomson, L, M., "Theoretical Hydrodynamics," The MacMillan Co,, 196Q, M14. Margle J. T,, "Investigation of Secondary Liquid Phase Droplet Accelerations," DRDA Reort no. UMICH 01457117I, April, 1978. M15. Moszynski, J. R,, Personal Communication, 1978.

250 N N1. Norman^ W,. $S and McIntyre, y:Y, "Heat Transper to a Li quid Filrn on a Vertical Surface," Trans. Inst. Chem and Chem. Engrs. v. 38, 301, 19.6.0 N2, Nusselt, W., ZVD 1 6G, 541 and 569, 1916. N3. Nayfeh, A. H, and Saric, W~. Ct, "Stability of a Liquid Film,"' AIA J., vol, 9, no, 4, pp. 750-752, 1971. N4. Nayfeh, A, H. and Saric, W. C., "Non-Linear KelvinHelmholtz Instability,1 J, Fluid Mech., vol. 46, pp. 209-232, 1971, N5. Nayfeh, A. H. and Saric, W. C,, "Nonlinear Stability of a Liquid Film Adjacent to a Supersonic Stream," J. Fluid Mech., vol, 58, pp. 39-51, 1973. N6. Nachtsheim, P. R., "Stability of Crosshatched Wave Patterns in Thin Liquid Films Adjacent to Supersonic Streams," The Physics of Fluids, vol. 13, pp. 24322447, 1970. N7. Newitt, D. M., Dombrowski, N. and Knelman, F. H., "Liquid Entrainment: I, The Mechanism of Drop Formation from Gas or Vapor Bubbles," Trans, Inst. Chem. Eng., vol. 32, p. 244, 1954. N8. Nicholson, J. E., "Drop Breakup by Airstream Impact," published in Rain Erosion and Associated Phenomena," RAE Report N68-19401-427, Farnborough, England, 1967. N9. Niedzielski, R. J., "Computer Analysis Programs, Computer Graphic," DRDA Report no. UMICH 014571-15-1, Ap. 1978. 0 01. Orell, A. and Bankoff, S. G., "Formation of A Dry Spot in a Horizontal Liquid Film Heated from Below," Int. J. Heat Mass Transfer 14, 1834, 1971. P P1, Petr, V.,. "Measurement of an Average Size and Number of Droplets during Spontaneous Condensation of Supersaturated Steam," lInstn Mech. Engr., London, Thermodynamics and Fluid Mechanics Group Convention, Univ. Glasgow, Mar. 23-25, 1970, paper 10, pp. 73-79. P2. Petr, V., "The Propagation of Small Disturbances in Wet-Steam," Fifth Nat. Conf. Czech. Sii-Eng. Soc., Mech. Eng. Grp,, Pilsen, Oct, 1972,...:-,-..-,

251 P3. Preston, t H., "The Determination of Turbulent Skin Friction by Means of Pitot Tubes," J. Royal Aero Soc., vol, 58, pp, 109-121, 1954. P4. Patel, V, C. "Calibration of the Preston Tube and Limitations on its Use in Pressure Gradients," J. Fluid Mech,, vol. 23, ppt 185-208, 1956, P5, Phillips, O. Mi, "Resonant Phenomena in Gravity Waves," J. Fluid Mech,, 2, p, 417, 1957, P6. Pouchot, W, D., Kothman, R. E., 2nd ed., "Basic Investigation of Turbine Erosion Phenomena," NASA CR-1830, Nov. 1971. R R1. Ryley, D. J. and Tubman, K, A., "Spontaneous Condensation in High-Pressure Expanding Steam," Proc. Inst. for Turbomachinery Res., Gdansk, Polish Aca. Sci., vol. 7072, 1976, pp. 273-284. R2. Reynolds, W. C. and Potter, M. C,, "J. Fluid Mech., 27, 465, 1967, R3. Roshko, A., "On the Development of Turbulent Wakes from Vortex Streets," NASA Rept 1191, 1954 R4. Ranger, A. and Nicholls, J., "Aerodynamic Shattering of Liquid Drops," AIAA J., vol. 7, no. 2, Feb. 1969. R5. Reinecke, W. G., Waldman, G. D., McKay, W. L. and Ziering, M. B., "Shock Layer Shattering of Water Drops and Ice Crystals in Reentry Flight," AFNL-TR-75-71. S S1. Smith, T. N. and Tait, R. W. F., "Interfacial Shear Stress and Momentum Transfer in Horizontal Gas-Liquid Flow," Chem. Eng. Sci, vol. 21, pp. 63-75, 1966. S2. Saric, W, S. and Marshall, B. W., "An Experimental Investigation of the Stability of a Thin Liquid Layer Adjacent to a Supersonic Stream, AIAA J, vol. 9, no. 8, pp. 1546-19553, 1971. S3. Saric, W. S, and Nayfeh, A, H. and Lekoudis, S, G,, "Experiments on the Stability of Liquid Films Adjacent to Supersonic Boundary Layers," J. Fluid Mech., ol. 77, pp. 6:33-80Q 1976, S4. Steen, D. A. and Wallis, G. B. ", The Transition From Annular to Annular-Mist Cocurrent Two-Phase Down Flow," AEC Rept, NYO-3114-2, 1964.

252 S5. Sherman,. P., Klein, J,, p.nd Trihus, M,, "DetermTination of Drop Trajectorpes by Me.ans of an ExtensiQn of Stokes. Law" Eng ResInst., Air Res. and Dey, Command, USAF't Uni:y( Mich, APpr. 1952, Contract No, AF 18 C60I-51. S6. Serafini, J. S,, "ITpingement of Water Droplets on Wedges and Doublewedge Airfoil at Supersonic Speeds," NACA Report no. 1159, 1954, S7. Simpson, R. L. and Whitten, D. G., "Preston Tubes in the Transpired Turbulent Boundary Layer," AIAA J, vol. 6, pp, 1776-1777, 1968. S8. Schlichting, H,, "Boundary Layer Theory," NY., McGraw-Hill, 1960. S9. Smith, A., Kent, R. P. and Armstrong, R. L., "Erosion of Steam Turbine Blade Shield Materials," ASTM Special Tech. Publ. no. 408, Mar. 1967. S10. Spink, L. K., "Principles of Flow Meter Engineering," Foxborough Co., Foxborough, Mass., 1958. T T1. Traupel, W., "Zur Theorie der Nassdamfturbine," Schweiz. Bauzg, vol. 77, pp. 324, 1959. T2, Tam, C. K. W., "The Drag on a Cloud of Spherical Particles in Low Reynolds Number Flow," J. Fluid Mech., vol. 38, part 3, pp. 537-546, 1969. T3. Tailby, S. R. and Portalski, S., "The Hydrodynamics of Liquid Films Flowing on a Vertical Surface," Trans. Inst. Chem. Engr. (London), 38, pp, 324-330, 1960. T4. Tyler, A. L. and Salt, D. L., "Periodic Discontinuities in the Acceleration of Spheres in Free Flight," ASME. paper no, 77-WA/FE-3, T5. Tseng, R,, "Investigation of the Acceleration of Droplets in a Steam Wake Behind a Stationary Turbine Blade," DRDA Rept, no. UMICH 014571-19 I, Apr, 1978. T6, Trier, E. and Blomer S,, "Wet Steam Tunnel Trailing Edge Observations," DRDA Rept. no, UMICH 014571-7-I, June 1977,

253 Y Yl. Yellot, J, I., "Supersaturated Steam, " Trans. ASME, Vol. 56, 1934, pp. 411-430. Y2, Yellot, J, I. and Holland, C, K, "'The Condensation of Flowing Steam,"- Trans., ASME, vol, 59, 1937, pp. 171-183, Y3, Yih, C.S., Proc, 2nd. U.S. Congress, Appl. Mech., ASME, p, 623, 1954. Y4. Yih, C. S,, "Stability of Liquid Flow Down an Inclined Plane," Physics of Fluids, vol. 6, no. 3, pp. 321-334, 1963. Y5. Yablonik, R. and Khaimov, V. A., "Determination of the Velocity of Inception of Droplet Entrainment in TwoPhase Flow," Fluid Mech. Soc. Res. 1 (1), p. 130, 1972. U U1. Ueda, T. and Tanaka, T., "Studies of Liquid Film Flow in Two-Phase Annular and Annular-Mist Flow Regions; Pt 1 and 2", Trans. JSME, vol. 39, no. 325, p. 2842, 1973. V V1. Valha, J. and Ryley, D. J., "Optical Studies of Nucleation in High-Pressure Expanding Steam," Condensation in High-Speed Flows, ASME Jt. App. Mech. Fluids Eng. and Bioeng. Conf., Yale Univ,, June 15-17, 1977, pp. 27-42. V2. Van Rossum, J. J,, "Experimental Investigation of Horizontal Liquid Films," Chem. Eng. Sci, vol. 11, pp. 3552, 1959. W W1. Woodmansee, D. E. and Hanratty, T. J,, "Base Film Over Which Roll Waves Propagate," A,I.Ch.E J, pp. 712-716, 1969. W2. Woodmansee, D. E. and Hanratty, Tt J., " Mechanisms for the Removal of Droplets from a Liquid Surface by a Parallel Air Flow,' Chen Eng, Sci, vol 24, pp. 299307, 1969., W3, Woodmansee, D. E. and Hanratty, T. J., "A Model for Roll Waves in Gas-Liquid Flow," Chem. Eng. Sci, vol. 26, pp. 1915-1931, 19.71,

254 W4, Wurz, D.^'Flow. Behavior of Thin Water Films Under the Effect of a Cocurrent Air Flow of Moderate to High Subsonic Ve.locities," Proc. of 3rd Int'l Conf. on Rain Erosion and Associat'ed Phenonena,. pp. 727'745, England, 1970 W5. Wicks, M. and Dukler, A. E,, "Entrainment and Pressure Drop in Concurrent Gas-Liquid Flow: 1. Air-Water in Horizontal Flow," ATIChE' J, 6, 463! 1960. W6. Wilson, C" T, R., "Condensation of Water Vapor in the Presence of DustrFree Air and Other Gases," Phil. Trans. Roy, Soc., London, A189, 1897, pp. 265-307, W7. Wallis, G. B,., "The Onset of Droplet Entrainment in Annular Gas-Liquid Flow," GE. Rept. 62 GL 127, 1962. W8. Wallis, G. B, "One-Dimensional Two-Phase Flow," pp. 320, 345-351, 376-391, McGraw-Hill, N.Y., 1969. W9. Wolfe, H. E,, "Photographic Study of the Break-up of Liquid Drops," J. SMPTE, vol. 75, pp. 738, 1966. W10. Waldman, G, and Reinecke, W., "Particle Trajectories, Heating and Breakup in Hypersonic Shock Layers," AIAA J, vol. 9, no. 6, pp. 1040-1048, June, 1971. Wll. Waldman, G., Reinecke, W. and Glenn, D., "Raindrop Breakup in the Schock Layer of a High-Speed Vehicle," AIAA J, vol. 10, no. 9, pp. 1200-1204, Sept. 1972. W12. Wegenka, M., "Multiphase Flow and Cavitation Computer Analysis Program," DRDA Rept. no. UMICH 014571-11-I, Dec. 1977. W13. Whitaker, Sir Edmund and Robinson, G., "Interpolation with Unequal Intervals of the Argument," The Calculus of Observations, 4th ed., Blackie & Sons, Ltd., London, 1944, pp. 20-34. W14, Wurz, D. E,. "Subsonic and Supersonic Gas-Liquid Film Flow," Proc. AIAA Annual Meeting, Seattle, Wash., July 1978, z Zl. Zuber, N, and Staub. F. W, "Stability of Dry Patches ForRing in Liquid Film Flowing Over Heated Surfaces," Int, J, Heat Mass Transfer, 9, 897, 1966, Z2, Zhivaikan, L, Y.,, "Liquid Film Thickness in Film-Type Units,"'Intern. Chem, Eng, 2 C3., 337, 1962.

255 Z3. Zuber, N.? "On the AtorQization and 3ntrainment of Liquid Films in Shear Flow, " G. E Rept. 62 CL 153, 196 2 Z4. Zanelli, S, and Hanratty, T. J., "Relationship of Entrainment to Wave Structure on a Liquid Film," Inter. Symp. Two-Phase Flow, paper 2-1, Haifa, Israel, Aug 1971.

APPENDIX A DESIGN, CALIBRATION AND GENERALIZED CHARACTERISTICS OF'ELECTRICAL CONDUCTANCE FILM THICKNESS GAGES The behavior of thin film is of considerable interest because the gas-liquid interface is often the controlling feature for heat-mass transfer, The most important parameter for the investigation in this line of interest is the thickness of film and variation of its values with respect to time. Various techniques to measure film thickness have been (T3) developed before i,e., hold-up measurement, radioactive absorption and emission, conductance probe, needle contact method. Light absorption and Fluorescence-Spectrometer method and so on. An extensive review on film thickness (C13) measurement technique was provided by Collier and Hewitt. Among many possible techniques, an electrical conductance gage was employed at the University of Michigan Steam Tunnel facility. The unique feature of the gage is that no part of it protrudes into either the liquid or the gas. The gage has been successfully utilized to study the interface wave characteristics and the thickness of a liquid film flowing over a horizontal plate with a high-speed vapor flow over the liquid face. The gage consists of 3 concentric parts; ie., a center conducting electrode, of radius Xl, an insulating ring of outside radius r2, and finally a conducting infinite plate, (Figs. Al and A2). Electrical connections are made to the center electrode and the plate, 256

257 Fig. Al: Schematic Representation of Conductance Probe 100KCI 100KC. 0 Audio oscillator Fig. A2: Design Drawing of Gauge Fig. A3: Wheatstone Bridge and Related Equipment Liquid — Electrodes Fig. A4: Schematic of Calibration Method

258.t Q I Mikielewicz and Hamnitt, soyling Laplace equation for electrical potential distribution in the filmn concluded that, i = C h 6 Au (A.1) where, i =, ar1 j - electric current 6- specific conductivity of liquid j - density of electrical current flow Cg - geometric constant of the gage h - film thickness Au- electrical potential difference between r and r2 Based on the experimental results of Collier and Hewitt, (C13) C = 2.918 6 h (A.2) where C is the film conductivity between the probe electrodes, they assumed the following relation for this range of thickness. C = C 6C fCh.) (A.3) where f(h) is unknown function of film thickness, which can be easily established experimentally by calibrating the gages. The design of our gage is shown in Fig..A2. Collier and Hewitt C19641 pointed out the importance of gage size. As the gage is reduced in size, the local sensitivity is increased while the maximum measureable thickness is decreased. They suggested that electrode spacing should be about 10 times the maximum expected film thickness. Thus the diameter of the center electrode was chosen to be 2.3 mm, and the outside diameter of insulating ring 4.8 mm.

259 Measurements were conducted using an out-of-hbalance Wheatstone bridge (Fig. A31^ excited by an A-C audio oscillator, Frequency of 1 KHa was used, producing a sinusoidal wave amplitude of 1Q volts. The bridge output voltage was fed to an oscilloscope. The output voltage was a function of the film resistance. Known film thicknesses were obtained by placing a pre-measured circular shim around the gage (Fig. A.4), putting a few drops of water on the gage, and finally setting a circular non-conducting plate over fixed shims, trapping the water over the gage, and squeezing to the desired shim thickness. Measurements were made with distilled water and 0.05% salt Cby mass) addition. This solution strength was necessary to avoid variations in water conductivity, at the different temperatures required, ranging from 70~ to 150~F. (C 21 to 66~C). Typical calibration curves are shown in Fig. A.5 The function, f(h), depending only on film thickness for a given gage, was found using values from calibration curves. The resulting function f(h), along with the dependence upon the factor Cg versus temperature, are presented in Figs. A.6 and,A.7. The small scatter of data points shown on Figs. A 6 and A. 7 indicates the general precision level of calibration, and the realism of the calculating procedure used. Function f (.l is general for all gages, but the factor C is different for various gages: One can obtain the thickness of liquid film, using the generalized characteristic curves of Fig..A.6 and A.7.

Ci);RO 4J) 4-).:j ~ra.r P4 N 0 1 3 4 5 6 7 Film Thickness - (Mils ) Fig. A5 Typical Calibration Curves (Gauge No. 1)

f(h)= -C. x I~ X I I I I I I X- 1 gauge 0- 2 gauge A- 3 gauge - 4 gauge o' 1-" I I I I 1 2 3 4 5 6 7 8 9 10 h(mil) Fig. A6 Dependence of f(h) on Film Thickness, h~ for Gauges No. 1,2,3,4

6 g 800 700 600 snn X- 1 gauge 0- 2 gauge A- 3 gauge 0- 4 gauge 0 0 0 m4 2 o o _ _ - r --, V V 401 300 200 O PMA-~ ~ ~ ~ ~ ~~ C' I ~". A di 00 0, cN I I I 70 80 90 100 110 120 Fig. A7 Fig A7 Dependence of Product f Geometric actor and ater Conductane on Temperature for Gauge No. 1,2,3, and uctance 70 ~ ~ ~ ~ ~ ~ 123,ad4 t(F)

263 For example, if the Qutput of the.Wheatstone bridge is 1200 mV and the temperature of the gilr; is 14-5F (63QC). one can find directly from the calibration curve CFio. A.5) the thickness of the film, 4,75 mils C 121 l)., The same result can be obtained using the generalized characteristic. For film temperature of 145~F, the corresponding factor Cg CFig. A,7) is 320. Dividing output of bridge (1200 mV) by value of factor Cg (320), one can find the value of the function f (h)., and the measured thickness of the film equals about 4,75 mils (Fig. A.6). The procedure is shown by the dotted lines on Figs. A.6 and A.7. Using this generalized characteristic, one can easily see the influence of temperature on the thickness measurement. Error of correlating experimental data of film thickness is given in Table A.1,

264 Thickness Predicted Minimum Maximum Range of measured measur mea seaured error value of value of I. value of f - fm / ^m ft (h) | M) frdh) |% It - fr _ 3937 39. 37 38.1 40.64 + 3.1 50.80 50.80 48.26 53.34 +5.0 73.66 69.85 69.85 78.74 0, +12.6 99.06 86.36 76.20 88.90 -11.8, +14.6 140.97 107.95 96.52 116.84 -10.6, +8.3 191.77 132.08 121.92 137.16 -7.7, +3.8 240.03 147.32 142.24 154.94 -3.4, +5.2 * —------------— j —-— I —-— I Table Al. Error of Thickness Gage Calibration

265 APPENDIX B CALCULATION OF NORMAL AND TANGENTIAL PERTURBATION STRESSES UPON WAVY FILM Estimates of normal and tangential perturbation stresses upon the wavy films from the parallel gas phase flow is extremely important for author's present data on wavy film characteristics (B41 (C10) Based on the Benjamin's analysis4, Craik simplified the expression for perturbation stresses. Author's wavy (C10) film data has been compared with Craik's theoretical analysis. Introducing normal II, and tangential Z, perturbation stress parameters, he solved the "Orr-Sommerfeld" equation for perturbation velocity stream function. 11 =/ P [I - S + i(fi - s) (B.) f i (4SI) ei 7/3 a3 ( Re)4/3 = 3fe i (mR) (B.2) where, I (U 2 Y d( YZoH/2h 2yh21 e 2/d(ay) = (-) e- d(ay) =2m 2e) d( ) o Jo (B.3) s = 0.644 I (B.4) 21 k is 2/3 e2 = 2_1 2/3Ca Re 3 2 (B.5) i V f.08 [/ 2/3 P_ 1/3 4 = 2,744 (-Ti yl (B.6) /3 f f 2r h kh=2h a -7 = Nondimensional wave number V fh 2q Vfh = 2q Fijlm Reynolds number

266 f. Friction coefficient estimated in 1 Section B of Chapter IV I Integral representing the effect of turbulence gas velocity profile using 1i/ rule h film thickness H Height of the channel (- 4 cm) Siubscript s and f represent steam and liquid properties used, Finally he obtained wave growth rate and wave propagation velocity as follows: (dRef [II + 32 - T - G] aCiRef = r 1 (B.7) 7 3 o(dCiRef) -'RefII + + CRe r) Cr - 1 = - (B.8) r 3 -+ 0( + ( C.oCiRef) Twh ere, T= TfVf2h G= gh Vf and subscript r and i indicate real and imaginary part respectively. Critical (or cut-off) wave number can be obtained by setting oCiRef = Q. List of computer program and output are as follows,

267 MICHIGAN TERMINAL SYSTEM FORTBAN G(21.8) MAIN 06-22-78 15:51 C "*** PRESSURE-SHEAR PERTUABATION PROGRAM **m.* C THIS PROGRkA SOLVES AN INTEGRAL TO DETEBRINE VAVE STABILITY. C TWO DIFFERENT FUNCTIONS ARE INTEGRATED BY THE SUBROUTINE DCAL 0001 REAL x(180),Y(180),Z(180),INPS,ITPS,H,Q,INTENS 0002 REAL*8 BP(4) 0003 COMPLEX*16 IP(4) 0004 INTEGER NPS/120/ 0005 COMMON ALPHA, RINTENS,SHT,G C 0*** PRINT INITIAL HEADING *~ 0006 (A)I (A/C) * (2./7.) *EIP (-A) 0007 WITE (6,1000) 3008 1000 FORMAT('1',T15,'SOLUTION TO PRESSURE-SHEAR PERru&BATION AT A LIQUI 1D -GAS INTERFACE',///) 0009 YRITE (6,2000) 0010 2000 FORMAT('ALPHA',6X, I'9,X, INTENS',7X,'RNPS',7,'INPS',7 I.RTPS,7X,'ITPS',7X,'GBORATE,6ZX, CR-,2,' 3*ITPS/ALPA',2X, WB'Z,'REAL * SOLN',//) 0011 WRITE(6,201) 0012 201 FORMAT (Vz21. 4',4X,'QO0.063) C ** READ COMMON PARABETERS sse 0013 M 1 0014 1 READ(5,2004) SH,V,R,ALPHA, B,G,H 0015 2004 FORMAT(4F10.5,2F10.8,F10.5) 0016 SR0. 00254*SH 0017 T=1./WB 0018 Ca (ALPHA*H) /(2. SH) 0019 IF ( C GE. 180.) C=180. C ** SOLVE EQUATION ONE BY INTEGRATION *** 0020 DO 10 I=1,NPS 0021 (I) (C/NPS)*(I-1) 0022 10 Y(I) =(X(I)) 0023 CALL QTFG(X,T,Z,NPS) 0024 INTENS=2*Z(120) 0025 CALL STRESS(RNPS,IMPS,RTPS,ITPS,DD, LPAR,CRMSI,Tr) 0026 =B+1 0027 IF(M.EQ. 6) WRITE(6,202) 0028 202 FORRAT(//'V-21..4',4X,,' 0.104') 0029 IF (H.EQ. 9) ERITE(6,203) 0030 203 FORMAT(//,'V=36.6',41,' 0.021 ) 0031 IF ( M.EQ. 14) WRITE(6,204) 0032 204 FORMAT(//,'Y36.60,4X,' t0.063 ) 0033. IF (M.EQ. 18) WRITE(6,205) 0034 205 FORMAT(//,'V=36.6'S,4X,'Q0. 104') 0035 IF(.-EQ. 19) WRITE(6,206) 0036 206 TORRAT(//,'V=61.0,4X,'Q=2.4') 0037 IF(B.EQ. 22) BRITE(6,207) 0038 207 FORMAT (//,'T61.0,4., Q' 0. 063 ) 0039 IF (N.,EQ. 29) IBITE(6,208) 0040 208 FORMAT(//, V-6 1.0', 4X, * 0.104 ) 0041 IF(m.EQ. 32) WRITE(6,209) 0042 209 FORMAT(//,V' 93.3',41X,Q0.021') 0043 IF (B.EQ. 37) WRITE(6,210) 0044 210 FORMAT(//, V=93.3, 4 Q=0.042t) 0045 IF (f.EQ. 42) YRITE(6,211)

268 BICRIGAl TERBIWAL SSTTE FORTRAN G(21.8) RHIB 06-22-78 15:51 0046 211?ORAT(/,'V=93.3',X,.'OQ=0.063') 0047 I? (a.,Q. 46) YBITE(6,212) 0048 212 FORRAT (//,'V 1 59',,' 0. 01 0') 0049 I1 (K.EQ. 50) WRITE(6,213) 0050 213 FORSAT(//,'V=159',4X,'QSO.021') 0051 IF (B.EQ. 55) BRITE(6,214) 0052 214 ORMAT(//,' V=159',4X,' Q0.042') 0053 IF(f.EQ. 65) WRITE(6,215) 0354 215 FOBlAT(//,'V=252',4X,'Q-0.010I) 0055 IF(B.ZQo 68) VRITE(6,216) 0056 216 FORBAT(//,'V=252',4X,'Q=0.021') 0057 IF (f.EQ. 71) BVITE(6,217) 0058 217 FORMAT(//,'V=252'S4X,'QO0.042o) 0059 IF(B.EO. 75) WRITE(6,218) 0060 218 FORMAT(//,IY=252t,41,9Q=0.063') 0061 IF((.EQ. 78) WRITE(6,219) 0062 219 FORRAT (//,' V=298 # 4, Q=0. 021') 0063 IF(R.EQ. 83) WRITE(6,220) 0064 220 FO7RAT(//,'V=2989,4X,'Q=0.042') 0065 IF(B.EQ. 85) YRITE(6,221) 0066 221 FORBAT(//,'V=298",41X,Q=0.063t) 0067 IF ( 5.GE. 89) GO TO 300 0068 RP(1) =3. tITPS 0069 RP(2)=RNPS-G 0070 RP(3) -0. 0071 RP(4) -T 0072 CALL ZPOLY(3,RP,IPIR) 0073 R= DREL (IP (1) ) 0074 RITE (6.2001) ALPHA,,INTeZS,RNPSINPS, RTPS,ITPS, LPAR,CRRNS1,TT, *BRR 0075 2001 FORBAT(12(E10.3,1Z)) 0076 GO TO 1 0077 300 STOP 0078 END *OPrrONS IN EFFECT* ID,EBCDIC,SOUaCENOLIST,NODECK,LOAD,NOHAP *OPTIONS IN EFFECT* NAME = AIN, LIHECNT = 57 *STATISTICS* SOURCE STATEMENTS s 78,PROGRAB SIZE = 5160 *STATISTICS* NO DIAGNOSTICS GENERATED NO ERRORS IN BAIN

269 HICHIGAN TERBINAL SYSTEB FORTRAN G(21.8) STRESS 06-22-78 15:52 0301 SUBROUTINE STRESS(RNPS,IXPS,RTPS,ITPSDD, ALPAR,CRB'S1,TT) 0002 REAL INPS,ITPS,INTENS,FRC 0003 COBBON V,R,IITBES,SHT,G C**THIS SUBROUTINE CALCULATES THE NORBAL AND SHEAR SrRESS 0004 FD01.0 0005 SVT0.93 0006 rT0. 0074 0007 SD-0.00011 0008 FRCO0.005* (1SH R/5.33) 0009 DA= (2*INTERS) /F8C 0010 DBS (SV/Pf) ** (2./3.) 0011 DC= ((U*R) ** (-2./3.) ) (W**2) 1 0012 DD DA*DB*DC 1 0013 S-0. 644*DD*INTENS 1 0014 RNPSi (2. *W) / (*FC) 1 0315 RNPSB=INTENS-S*SQRT (3.) 1 0016 RBNPSRNPSA*BRPSB 1 0017 IN PSt ( (2. W) / (R*FRC) * (FRC-S) 1 0018 BA (SV/FV)** (2./3.) 1 0019 BB= (SD/FD) **(1./3.) 1 0020 U2.744*BA*BB 1 0021 RTPSz (*0. 5/FRC) * (**3.) * ((*R)* ) (-./3.)) 0022 ITPS=RTPSS QRT (3.) 1 0023 TT=3. *ITPS/W 1 0024 ALPAR (WR ) **2.* (RNPS+3. *ITPS/W-T*W**2. -G)/3. 0325 CAA& (-7./8.) ALPA&R-*R*ZIPS+ (3./2.) *R*RPS 1 0326 CAB=3. (27./5.) ***2.+(12./5.)*ALPAB 1 0027 CRaNSluCAA/CAB 1 0028 201 PORMAT(E12.5) 1 0029 RETURN 1 0030 ED 1 cOPTLOHS IN EFFECT* ID,EBCDIC,SOURCE,NOLIST,NODECK,LOAD,NOMAP *OPTIONS IN EFFECT* NAHE STRESS, LINECNT = 57 *STATISTICS* SOURCE STATEMENTS = 30,PROGBAB SIZE = 1408 *STATISTICS* RO DIAGNOSTICS GENERATED NO ERRORS IN STRESS SRUN -LOAD+NAAS:SSP+NAAS:OLDLIB EXECUTION BEGINS 15:52:11

SOLUTION TO PIESSURR-SHIAR PERTUABATION AT A LIQUID -GIS INTERFACE ALPHA B V=21.4 Q=0.063 0.112E*00 0.169E*02 0.840B-01 3.169Eo02 0.710E-01 0.169E*02 0.135E+00 0.169Eo02 V=21.4 0Q0.104 3.483E-)1 3.169E*02 0.231E+00 0.282E.02 3.246E*30 0.282E 02 T136.6 Q-0.021 ).7691e03 3.282BE02 0.153ge00 0.560E*01 3.256E,33 0.560E+01 0.1531E00 0.560e+01 0.256E100 0.560E,01 =36.6 Q00.063 0.192E*30 0.560E091 0.221E*00 0.169E*02 3.265*+30 0.169E+02 0.265e300 0.169E+02 V=36.6 0=0.104 3.221E+00 0.169E+02 V=61.0 Q=2.4 3.352E*30 0.2821*32 0.420E-01 0.560.E01 3.990-)31 0.560E+01 V=61.0 0=0.063 0.990-01 3. 560E+01 0.1191e00 0.169E*02 0.166E30 0. 169E+02 3.119E+)30. 169E+02 0.2071+00 0.169E102. 111E*130 0.169E*02 0.119E100 0.169E*02 T=61.0 00. 104 0.1191+00 0.169E~02 3.228E*00. 169E*02 0.190EO00 0.169EL02 INTENS RNPS 0.837e,00 -0.1671e03 0.914E,03 -0. 102E103 0.962E100 -0.7561eE02 0.789E,00 -0.230E*03 0.107E0 1 -0.374E1 32 0.734E.00 -0.298E,03 3.7198,00 -0.331E03 0.460e*0) -0.194E,04 0.625E*00 -0.123E*04 0.509E100 -0.272Eo04 0.6251e00 -0.123804 0.509E100 -0.272E*04 0.574E100 -0.176* 04 0.667E+00 -0.522E*03 0.6251+00 -0.701E*03 0.625B+00 -0.701E 03 0.667E*00 -0.5221E03 0.6251E03 -0.578E103 0.833E100 -0.105E03 0.626+000 -0.4451E03 3.626E109 -0.415E*03 0.702E100 -0.1361E03 0.624E#03 -0.234e.03 0.702E*03 -0. 136E103 0.575E100 -0.333E+03 0.718E*00 -0.1211EJ3 0.702E*00 -0. 136E133 0.702Er00 -0.13bEJ03 0.6251e03 -0.493E*33 0.667E *00 -0. 366E 03 IMPS -0.978E102 -0.596E132 -0. 446q 02 -0.13 4E*03 -0.2231E02 -0.173 103 -0.192E+03 -0.112E*04 -0.714E*03 -0.158E04 -0.7 14B.03 -0. 158BE04 -0.102B*04 -0.30 3 03 -0. 407e*03 -0.4 07 03 -0.303e103 -0. 335E03 -0.622 E02 -0. 2591+03 -30.259e03 -0.794E,02 -0.137E*03 -0.794B*02 -0.194 B03 -3.707E*32 -0.794 e02 -0.7941*02 -0. 286193 -0.21 3e33 RTPS 0.197B*00 0. 122BE00 0.9231-01 0.269E*00 0. q481E-01 0. 333 1 00 0. 370*00 0. 2471E01 0. 145E*01 0. 34f2101 0. 1451+01 0.342L1*01 0. 212E01 0.6 13e100 0.829E*00 0.829E,00 0.613E+00 0.672 B00 0. 1681900 0.702E+00 0. 02E100 0.219E+00 0.381e*00 0. 219E00 0.550E100 0.195E103 0.219E100 0.219E100 0.646 R 03 0.476EO00 ITPS 0. 342e00 0.212E 00 O. 160B100 0.4671E00 0.833E-01 0.5761B00 0.640e*00 3. 4281E01 0.251E+01 0.592E+01 0.2511*01 0. 592e+01 0,366E+01 0. 106E*01 0.144E101 O. 1441*01 0.106e*01 3.1161*01 0.2911e30 0.122E*01 0. 122E*01 0.3791t900 0.659E+00 0.379E+00 0.9521*00 0.3371E00 0. 39E*,00 0. 379E*00 3.112E101 0.8251s00 GROVRATE CR-1 3*ITPS/ALPHA MD RBAL SOLH -0.190E*03 -0.635E*02 -3. 332e802 -0.384e+03 -0.711* 01 -0. 413e+04 -0.5221 04 -0. 305B 06 -0. 29 1E*03 -0.183e*04 -0.291E,33 -0. 183e1 04 -0. 663B+03 -0. 238E 04 -0.461E*04 -3.461E 04 -0.239E*04 -0, 1881E05 -0. 156e*01 -0.4210*02 -3. 421E*02 -0. 1711+03 -0.5871E03 -0. 171E+03 -0. 131E104 -0. 31E*03 -0. 711e*03 -0. 171E03 -3. 237E+34 -0. 122E*04 0.8591+01 0.1631B02 0.2391E02 0.566e*01 9.6582E02 0.79 1e00 0.646E*00 -0.286E*00 0.1361e02 0.4321E01 0. 136E*02 0.432 e01 0.8 27E01 0. 176E01 0.1081O01 0. 108E* 1 0.176e101 0.1038e00 0.116E104 0.3638102 0.36 3*02 0.769E01 0.359e+01 0.769E+01 0.211B01 0. 899E01 03.769E*01 0.769E901 0.164E*01 0.258B 01 0.9 15E01 0. 756101 0.676 *01 3.104E+02 0.520+*01 0.749e*01 0.781E+01 3. 167E*02 3.492E*02 0.6941*02 0.492E*02 0.6941*02 0.5731*02 0. 1441B02 0.1631E02 0. 16 3102 0. 144l*02 0.9921Z01. 2081E02 0.3681*02 9.3680*02 0.954e101 0.119E*02 3.9542*01 0.138 E02 0.911 *01 0.9542+01 0.954 +01 0.147E+02 0. 130E*02 0. 1331-01 0. 130e-01 0. 1301-01 0. 130e-01 0. 1301-01 0. 2641-01 0. 2641-01 0. 264e-01 0. 254u-02 0. 254E-02 0. 25u4-02 0. 2541-02 3. 2541-02 0. 133E-01 0. 133-01 0. 133e-01 0. 1331-01 0. 277E-01 0. 388E-02 0. 3881-02 0. 388e-02 0. 212E-01 O. 212E-01 0. 2121-01 0. 2 12e-0 1 0. 2121-01 O. 212E-01 0. 212E-01 0. 427-01 0. 427E-)1 3. 153]*00 0. 156E*00 0.159*,00. 1521+00 0. 167*00 0. 145e+00 0. 145100 3.165E+00 0. 153*00 0.1631*00 3.153 *00 0. 1631+00 O. 1561*00 0. 153B00 0. 154*00 0. 1542Q00 3. 153B+03 0. 1511*00. 2081+00 0.2051+00 0.2051*00 0.210 *00 0.2111*00 0.2101*00 0. 2151E00 0.210B*00 9.2131 )03 0.210E*00 0. 170E* 00 3. 169E+90

V93. 3 0. 228E.00 0. 893 -31 0. 104E0OO 0. 890E-1. 780E-31 V=93. 3 0. 890E- 01 3. 1741E00 0. 109e+00 0.145Eo00 O. 109E+30 V=93. 3 0. 1 74 00 3. 129E130 0. 135E100 0. 155E.00 Q=0.021 0.169E102 3.560E101 0.560E01 0.5601E01 3. 560E31 O=0.042 0.560E+01 0. 1301E01 0. 130Et01 0.130E101 3. 130E101 0=0.063 0.1 30E101 0. 169E102 0. 169E102 0.169E+02 VZ159 Q50.010 0. 135E*00 0.169E132 0.6008-02 0.280E+01 ). 2930g-.31 0. 280E+01 0.290E-31 0.280E101 V=159 Q=3.021 0. 290- 01 0.280E+01 3. 5931-01 0.560 E31 0. 830E-01 0.5 60E101 0.830e-01 0.560Et01 3.830e-3 0. 560+0Ot V= 159 QO0.042 0.830E-01 0.560E+01 3.145E+30 3.113E102 0.145E+00 0.113E.02 0. 121E00 0.113E.02 3.121E+30 0.113E102 0.1611E00 0.113E+02 3. 161E*30 3.113E102. 236E*33 0.169E+32 3.2211e00 0.169E.02 3. 141E3)) 3. 169E+32 0.625E+00 -0.493E103 -0.286E103 0.544E+00 -0.261E,33 -J.153,J33 0.509E100 -3.330E,03 -0.1931e03 0.544EB00 -0.2b1E103 -0.153E+03 0.573E+00 -0.213E133 -J.125E103 0. 544E00 -0.261E+03 -0.153E,03 0.469E03 -0.107E*35 -0.619E104 0. 5741E03 -0.537E+04 -0. 31 P.04 0.510E+00 -0.827E104 -0.4 78+04 0.574B+00 -0.5371+04 -0.311E,34 0.4691+00 -0.107E+05 -0.619.O04 0.584E+00 -0. 113e03 -0.664EB02 0.574E100 -0.1221+03 -0.7131E02 0.543E100 -0.1511e03 -0.881E,02 0.574E100 -0.122E+03 -0.713E+02 0.1031e01 -0.472E101 -0.323Et01 3.625E100 -0.7866+02 -0.4681+02 0.625E100 -0.786E102 -0.468E102 0.625E+00 -0.716E+02 -0.4686E02 0.545E00 -0.997e+02 -0.5891E02 0.469E100 -0.164E*33 -0.965E102 0.469E+00 -0.164E+03 -0.965E102 0.469e100 -0.164E,03 -0.965E+02 3.469E100 -0.164E.03 -0,965E102 0.418E100 -0.1511E03 -0.88U2102 3.4 18E00 -3.151E103 -0.082E102 0.4591E00 -0.11 9E03 -0.698E+32 0.459E103 -30119E103 -0.698E102 0.394E+00 -0.111E+03 -0.100E103 0.394E+00 -0.17E1+3 -0. 100E03 0.351E+03 -0.168E 33 -0.982E.32 0.366E+00 -0.157E103 -0.9171E02 0.469E+00 -0.898E302 -3.5271E02 0.646E300 0.588E100 0.762E103 O.588OEO0 J.472E*00 0.5881E00 3. 126E+02 0.578E+01 0.9301:01 3.5 79E01 0.126E+02 0.250E+00 0.213E100 0.340E.00 3.270 E03 0.166E-01 0.229E100 0.229 E00 0.229E100 0. 296 B300 0.524 E 30 0.524 E00 0.524E 00 O. 524E 00 0.523E00 0.52)E+00 0. 38E 00 3.3817E00 0.622E100 0.622 B00 3.685E100 0.614 o00 0.293 +03 0.1121E01 -0.2317E04 0.1)21E01 -0.189E102 3. 1321+01 -3.332E102 0.1021E01 -0.189E102 0.818E100 -. 116E+02 0.102E 01 -0. 189902 3.218E+02 -0.176E 03 0.100102 -0. 341B*02 0.1611,02 -0.9401E02 0.130OE02 -3.34E*02 0.218BE02 -0.176E103 3.433E,00 -0.172E103 0.467E100 -3.203E+03 0.5881E00 -0.335Et03 3.4617E00 -3.203E+03 3.287E1-01 3.904E-03 0.3961e00 -0.832E-01 0.396E100 -0.8321-01 0.J96E100 -0.832E-01 0.513e100 -0.269B101 0.907E100 -0.953B01 0.907E100 -0.953E+01 0.9071t00 -0.953e101 0.907E100 -0.953E101 0.905E+00 -0.118E103 0.905E+00 -0.118E103 0.670E100 -0.636E102 0.670E100 -3.6361E02 0.108 l +0 -0.167E103 0.108E101 -0.167E+03 0.119E301 -0.822g,03 0.106E+01 -0.672E103 3.533E100 -0.151E,03 0. 164E101 0.497E.02 0.343E*02 0.497E102 3.694 B32 0.497E+02 0.460e,02. 127E103 0.680E*02 0.1271803 0.4601B02 0.6301E01 0.567E101 0.411EO01 0.561E*01 -0.6581E01 -0.116E103 -3.116E+03 -0.1161.03 0.2371e03 0.669E102 0.669E102 0.669E102 0.669E102 0.821.01 0.8211E01 0.124E102 0.124E+02 0.646E+01 0.646B101 0. 16)3B01 0.190E101 0.561E101 0. 1417E02 0. 343E 02 3. 381E02 0.343E102 0. 314E102 0.343E+02 0.376E*03 0.2751+03 0. 333E+03 0. 275E03 3. 376 E03 0.1011+02 0.104E102 0.1141e02 0.104 E02 0.143E.02 0.410Eo02 0. 410E102 0.4 10 o.02 0.2611E02 0.328ZE02 0. 3281002 3.328b102 0. 328E102 0. 1817E02 0. 1871E02 0. 166E*02 0.166B,02 0.201E.02 0.201E 02 3.1t51S02 0. 144E~02 0.1 07E102 0.427E-31 0.624E-02 0.6248-02 0.624e-02 0.624E-02 0. 624E-32 0. 181e-01 0. 181E-01 0.181E-01 0. 1811-01 O. 1811-01 0. 325E-02 0. 3251-02 0. 325E-02 0. 3251-02 0. 335E-02 O. 335B-02 0. 335E-02 0. 3351-02 0. 941E-02 0. 94 1I-02 0. 941E-02 0.94 11-02 0. 94 1E-02 0. 268E-01 0. 26 8-0 I 0. 268E-01 0. 2682-01 0. 268e-01 0. 268E-01 0.268E-0 0. 268E-01 0. 268E-01 0. IZ3E130 0.292E,00 3. 30E*00 0.292F'00 0. 288 00 0.292E100 0. 153E00 3. 143JE33 O. 146100 0. 140* 00 0. 1531300 0.287E100 0. 2t8E*00 0.2931*00 0.288B*00 0. 4471 00 0.37 8E00 0.378E100 0.37181E33) 0.386E100 0.4 14 e00 0. 14E+00 0.4141E00 0.414el00 0.4501.00 O.qSOEe 00 0.450B100 0.422E100 0. 4221 00 0.472E+00 0.472E100 3.5291E03 0.508E100 0.420B100 V=252 Q=3.010 0.126E100 0.169E.02 ). 523E-)1 J.280E+01 0.620B-01 0.2801E01 V-252 Q03.321 3. 443E- 31 0.280EO01 0.494e00 -3.766E.02 -0.450E102 0. 241 E100 O.417E100 -0.102E103 0.72901 0.9901 0.2681-31 0.408E0OO 0.298EL00 -0.731E+32 -0.4171.E02 3.605E*00 0.105E*01 -0.7721-01 -0.102E103 0.605E,02 0.197B-02 0.111E*01 0.2561030 -3.781F.02 -0.464Et02 3.812E,.3 O. 14eO01 -0.121E+00 -. 117E103 O.68010B2 0.197E-02 0.133f131 3.3'.8E00 -0.607E1J2 -O.3b2B,02 0.458P.00 0.1941.00 -0.380e-01 -0.857E102 0.5411E02 0. 197R-02 0.969E+03

0.693E-01 0. 5603 01. 830 E-01 0.560E 01 V=252 Q00.042 0. 930E-01 0.560EO01 O. 150E 03 0. 113t02 0.1351e00 0.113Eo02 0.104E 00 0.1133202 V=252 Q=0.063 3. 113 00 0.11 3102 0.120E*00 0.169E002 0.130E300 0.169 E02 V=298 Q=0.021 J.130E 30 0.169Eo02 0.150eB00 0.560E.01 0. 143o00 0.560E,01 0.9308-01 0.5603E01 0.830E-01 0.560Eo01 VT298 Q=0.0q2 3. 15000 0.560E+01 0.166e*00 0.113E+02 VT298 Q=0.063 0.139E00 0.113E302 0.3123800 0.169tE02 0.257E100 0.169Eo02 3.2078100 0.1693.02 EXECUTION TE R HATED SSIGNOFr 0.299Be00 -0.429E.02 -0.2563B02 0. 2553o00 -0.48 13 02 -0.2863E02 0.232e300 -0.520E*02 -0.309E302 0.232E+00 -0.4971302 -0.294e802 0.255EO00 -0.469B)02 -0.270B302 0.317E.00 -0.393E302 -0.234eo02 O.298E00 -0.421E*02 -0.2493.02 0.317Eo00 -0.278E*02 -0.166E*02 0.298E00 -0.2971E02 -0.1773102 0.298Eo00 -0.297B102 -0.177Z,02 0.232.800 -0.161e.f03 -0.942BC02 0.232e300 -0.144E303 -0.843E302 0.325eo00 -0.103BE03 -0.60Sto02 0.352e,00 -0.920E102 -0.543r.02 0.232E300 -0.161E803 -0.9423B02 0.255E+00 -0.766E+02 -0.4 50E+02 0.298E+00 -0.6871E02 -0.4053O 02 0.232E300 -0.14813E03 -0.825.o02 0.232eo00 -0.8953+02 -0.525e 02 0.2563e00 -0.655B+02 -0.3585302 15:52:29 T=.91 BC=0 S.29 0. 385 EO00 0.524 eo00 0.6331o00 0.553B.:00 0.464EO00 0.301E 00 O. 34S5E00 0.222Et00 0.254 +00 0.254EO00 0.140E+01 0.1301t01 0.633EB00 0.524 E 00 0.143E 01 0.655E 00 O. 487E00 0.109 E01 0.790oe00 0.551E.00 0.667e*00 -0.7333E00 -0.2763e03 0.907E000 -3.1193o01 -0.2133*04 0.110eo01 -0.1633*01 0.479eB03 0.958 300 -0.3053:02 0.1093.02 0.8048300 -3.233E002 0.138E+02 0.5218,00 -0.115E002 0.257E+02 3.598B+00 -0.1463Z02 0.208E+02 0. 384e00 -0.2571e02 0.111Be02 0.4393E00 -0.325e+02 0.9173+01 0.4 39E00 -0.3253102 0.917 101 0.243e101 -0.267E+02 0.2053B02 0.2253E01 -0.209e302 0.2383902 0. 113E 01 -0.613E+01 0.7071002 0.9071+00 -0.430E.01 0. 102e* 03 0.2431*001 -0.2673102 0.2051+02 0.113, 01 -3.660B 02 0.74q2301 0.844E.o00 -0.416EB02 0.1101*02 0. 1893001 -0.115E304 0.7753B00 0.137.o01 -0.468E*03 0.146e*01 0.9548+00 -0.2133E03 0.264ZB01 0. 2903E02 0.328B 02 0.3548302 0.1923E02 0. 1793*02 0. 1503102 0.1593.02 0.96 l1EOI 0.101oE02 0. 101e02 0.486Bt02 0. 471E02 3.35483*02 0.328e*02 3.486*B02 0.205Eo02 0.1823*02 0.182 202 0. 1603102 0.138 E02 0. 588e-02 0. 588B-02 0. 588e-02 0. 1458-01 0. 1485-01 O. 145E-01 0. 145-01 0. 282B-01 0. 282E-01 0. 282 -01 0. 1571-01 0. 1571-01 0. 1571-01 0. 157e-01 0. 1571-01 0. 471e-0 1 0. 471B-01 0. 847B-01 0. 847 -01 0. 847e-01 0. 116G01 0. 1408031 0. 156E301 O. 144e801 0. 128E01 0.990e.00 0. 1063.01 0.103BO01 0.111 E301 0.111 e01 0. 113E01 0. 117E01 0.801Eo00 0.739Eo00 0.113eo01 0.111 1*01 0.921B*00 0. 100B01 0. 1153,01 O. 109E01 ti — j

273 APPENDIX C COIPUTER PROGRAM FOR DATA REDUCTION ON ACCELERIATION AND BREAK-UP TIME OF DROPLETS WATERDROP is a computer program developed to analyze liquid droplet behavior as observed in steam tunnel. The size, orientation and kinematics of a droplet are determined by analyzing four basic parameters in a film frame of movie film. The equations used in the program are developed and presented in section B.1, Chapter V. The experimental results physically define the particle in each observation frame. The experimental and analytical accelerations calculated are compared in the program to determine if the theoretical relations can be used to predict the actual acceleration for a specific particle size and steam velocity. A basic understanding of the program requires a knowledge of how it categorizes experimental data. Each data point belongs to three interrelated groups or sets as shown in Fig. C1. In general, each point is classified to be in a set. A set includes all of the data taken at a specific liquid flow rate and steam velocity. The next major SET SUBSETS. C R D pi CtOBSERVATIONS Fig. C1 WATERDROP Data point categories

274 classification is a subset, A Subset includes all of the data points describing a specific particle motion from the time it was first observed until it either disintegrates or vanishes from view. The last group is the observation. The instantaneous size and position of a particle in a reference frame is recorded. SUBSET RESULTS A sample listing of these results are given. Each subset has experimental results printed under the twelve headings at the top of the computer printout. The metric CGS units used in the program are listed below each heading. The first six columns define the particle postion and size, while the last six columns define its motion. The second set of headings lists the calculation results and the comparison between experimental and analytical accelerations according to Serafini correlations (S6) The last four columns allow a direct comparison of the experimental and analytical accelerations in three ways. First, the experimental acceleration is reprinted directly after the theoretical to allow a comparison at each observation. Second, the difference between these accelerations is printed. Third, a ratio between them is printed to allow an order of magnitude comparisQn. Statistics of the subset experimental accelerations are printed next. The arithmetic mean, the mean deviation, and the standard deviation of all accelerations in a subset show an average value and how it varies, The arithmetic mean of

275 the acceration ratio is also printed to show the. ayeage value Qfr the. ubset. Information in the last four lines of the subset results define the operating conditions for the subset. Steam flow, rate and velocity are given in both metric and English units for clarity. Particle density, steam density, and the kinematic viscosity of the steam are printed to further define the flow conditions. Movie film framing rate and a data conversion factor are given to show the equivalence factors necessary to convert the raw laboratory data into usable results. The last line describes the particle stability, and the elapsed time and position of the observation. SET RESULTS The program determines an average value for experimental acceleration and acceleration ratio for each flow and velocity condition. A sample listing from these printouts are also listed. The arithmetic mean, mean deviation, and standard deviation are calculated for the average experimental accelerations of each subset, In addition, the arithmetic mean of all average acceleration ratios from each subset is printed. The average experimental accelerations from all subsets are printed to keep all the average values on a single page. Xi'X-i t [ X2*iMX +?*.-.+ XnXI *Mean deviation - [x,. -.... —. —... m0

276 Variable Niarle ABAR (II' I ACCEL CJ) ARATIO (J) AREA (J) ARR (II) ARSUM ASUM ATOT ATOTB CD(J) CON CONFAC DELA (J) DELV (J) DELX (J) DIA CI DTAVG WATERDROP VARIABLE AND SUBSCRIPTS LIST Description of Variable Average acceleration for each subset EXperimental accelerations from three successive film observations Ratio of experximental to analytic acceleration for each observation Water particle area at each observation Average ratio of experimental acceleration ratios of a subset Sum of all experimental acceleration ratios of a subset Sum of all experimental accelerations in a subset Total sum of experimental acceleration for a subset Average experimental acceleration of a set Analytic drag coefficient at each observation Equivalence factor for converting feet to centimeters Equivalence factor for converting film measured quantities to centimeters Change in acceleration in four consecutive observations Change in velocity in three consecutive observations Change in displacement between two consecutive observations Diameter of the particle at each observation Average change in time in experimental acceleration calculation

277 Variable Name DVM FILMV FN (I) Kl MASS (J) MEANDV N NPL NP2 P14 Q QCHECK RBAR RE(J) RHOS RHOW RSUM SD SFLOW S I GMA SMD Description of Variable Mean deviation of experimental acceleration calculation Velocity of the film Number of film frames: between observations A test point to determine whether a particle was observed to split Mass of the particle at each observation Mean deviation of experimental acceleration for each subset Total number of data points in a subset One less than the total number of data points Two less than the total number of data points One-fourth of the constant Steam flow rate Test point to determine the equality of two flow rates Average acceleration ratio for a set Reynolds number for each observation Steam density Water density Sum of acceleration ratios for a set Absolute value of the difference between average experimental accelerations of a set and of a subset Steam flow rate Standard deviation of experimental acceleration for each subset Sum of the differences between experimental and average acceleration for each set

278 Variable Name SMD2 STDV SUMDIF SUMD2 SUDMD SVEL T (I) TD (J) TOTHRD TTOL UDOTW (J) V VCHECK VEL(J) VISNU VOL (J) Subscript Character I II J JJ K Description of Variable Square of the sums of the experimental and average acceleration difference for each subset Standard deviation of the experimental acceleration for each set Absolute value of the difference between the average and experimental acceleration in each subset Square of the sum of the experimental and average acceleration difference for each subset Sum of the experimental and average acceleration difference for each subset Steam velocity Water particle thickness Elapsed time between observations A constant, 0.6666... Total particle observation time Analytic Acceleration Steam velocity A test point to determine the equality of two steam velocities Velocity between two consecutive observations Kinematic viscosity of the steam Volume of an observed particle Description of Subscript Input reading iteration Set calculation iteration Subset calculation iteration Average set calculation iteration Output printing iteration

279 WATERDROP FLOW CHART PI4, TOTHRD CON, VISNU HOS, II I l I I I SUMMD, SUMDZ, SIGMA, ARSUM I I 1 I 1 DTAVG DELVj-2 ACELj-2 DELAj-2 ARATIOj_2 FORCEj_ ASUM ARSUM I I _ J I _J

280 \ NP2 I Rk \ /I UDOTI, DEI. IL__ ACCEL.,: SUMF=.- I SUMMD= ARATIO SUMD2 = I - E 0 __ ABARii,.. SFLOW,.. MEA NDV= L T — " -. -.- T K1 - 1? NP2 > 1? ) SIGMA= —--- I Of F F TSo.- -' TTOL DIA1, T1, X1 Xn-1 AREA1. VOL1 MA SS1, L/TTOL k=2, 3,. r K1 0? -— I NPL 1. 1.F b - TTOL TD, DELX, Xn VEL, T, DIA, AREA, AREA —-. VOL, MASS, XCHARATE DELV,ACEL,HRCTE FORCE~ L- STRING

281 WATDROOP PPROGRAM LISTING C V AT_2.D OP C **' ^-ATS?-?AP.TICLZ DATAX..B DUCTIOH PSOGRA3 *- ** C * I* TES PFOSPA. IS TO r.TER2IN2 TH2 VLDOCITY,AC=ZLZEATIOB, C " AND FOCE ON A WATE2 PASTICLE AS IT TPAVELS TEED A C t** ST2.EB TIUNNL. IT ALSC CALCULATES AN ANALYTIC C m* ACC.LE.RATION BY TBH PFOLOUING ILATION: C C 2 C DU 2a 2/3 3 DENSITY VELDCIT= C -- - ( 1.0 + 0.17*.! - ---- C DT E! 4 RATIO DIA S Tw C IEPLICIT R.tAL8 (A-H,S,O-Z) DIT.NSION ACEL (18),APEA(20),DELV(18),DZL(20), DIA (20),ASS (20), 1 T(20),TD (2), EL(20), VOL(20),X (2), FORCZ (18),ABA (100), 2 FN(2_), 5R(2), CD(20), UDOTW(20), DELA(18), ARATIO(1 8).,A RE(99) PU DAR SIN(1.CDO)/2. CON T 30.48 TO TTR D 2. / 3. VISNU =.93469634 PROS =.0001135215 StAD(5,540) FIL!V, RHOW, CONTAC, VCHECK, OCRECK tF a= 0.0 10C TTsO=0. I=IO. ASUT = 0.0 xSOan ~. = 0.0 SlTD2 = 2.0 SI GA = 0.0 C RA** READ ON! SUBSET OF DATA A*=* DO 119 T1,20 P TDr5, 5, ) FR O(), DIA (I), T(I), X (I) C **** CO.NVP.T TH2 DATA TO ER TIC CIZENSIO.NS **= DI A (I) = CONAC*DIA (I) T () = CONF.ACT (I) X (I) C COPACIX(I) C =**'TESTE TO SEZ IF THE PAPTICLE SPLIT ***= IF( T(I).LT. 0.0 ) GO TO 120 C "'= TEST TO SE 1f THIS IS TEE LAST DATA POINT OF TH2 SET ** I ( X(I).LT. 0.0 ) GO TO 130 11 CON.TINn1 120 K1=1 T T) =DA3 (S(T ()) 1?0 N = I X(tN) =DABS(X T))?7AD(5,,54) V, Q TF ( V.E. V:-TTC?..OP. Q.NE. QC.2CX ) GO TO 20C iuC I?( N.L. 2 ) GO TO 1 ^3 TT = II * - C 2"22 C..A CTLATN' 0?.'F.PATxOTS ** SV_?L V= YON SLOW = O*60. / CO't** c 222* i:-:ATV_ SUBEST CALCLIAT!ONS ==22

282 DO 15 J=*,R AP!A(J) = PIf*DIA(J) **2 VOL(J) AEZ A (J) *T (J) 16. / 3. nASS () a VOL (J) *PBRO P_ (J) = DIX(J) t SVI / VISNU CD (J) = 24. / RE (J) * ( 1. + 0.17*32 (J) *TOTBRD) UDOTW(J)= CD(J) * RHOS/SHOI *3./4. /DIA(J) *SVEL**2 I?( J.LZ. 1 ) GO TO 150 D (J-1) mPN (3)J / FXLMV TTOL L * TD(.-1) D2LX(J-1)= X1() - X(J-1) VTL (3-1) DELX (J- 1)/TD (J-1) IF( J.LE. 2 ) GO TO 150 rTAG - (TD(J-1) + TD(J-2)) / 2. D2LV(J-2) VL (3J-1) - VEL (J-2) C2 (tJ-2) = D L LV (J- 2 /2./UTA VG DPLA (J-2) =ODOT (J-1) -ACE (3-2) ARATO (J-2) AC'L (J-2) / UDOT(J- 1)?ORCZ (J-2) = XCEL (3-21.ASS (3-1) AS U O ACEL (-2) + ASVE A? SU AA TIO (J-2) * ARSDU 150 CONTIHNUi C "*** SUBST S- AT!STICS CXICULATIONS **** NP2 2 N - 2 ABA (II)' A Sm / NP2 AP! (11) - ARSU! / WP2 C *** STOP.E SUBSZT P.SULTS FOR PLOTTING **** TT (1,660) BAR (I) B (,AS (II), VCEHCK, QCH2EC DO 160 J=1,NP2 SUEDIF = DABS( ACEL(J) - ABAR(I) ) S-UitD = SUEDIP + SU 3D 160 S!5D2 = SUDD2?I 2 + SUMD2 MTANDV = SI. DD / NP2 17 ( NP2.GT. 1 ) SIG2A =DSORT( S'ID2 / (RP2 - 1. )) C **** PRINT OUIT P!SULTS OF TEE SET ***L S IT? (6,550) DIA (1), T(1),APSA (1) VOL (1), ASS (1), X (1) NP Lu N- 1 DO 190 K=2,NPL 180?PIT2: 6, 560) TD (-K-1),DELX (-1),VrL (K-1), DIA (K), r (K), AP.EA K), I VOL(K),!ASS(K), XI), D!IV(K-1), ACTL (K-1),FORC2(K-1) uPT~(6S6,5) TD T(?L),DELX (PL), VEL(NPL), DIA(N), T(), AREA(), 1 VO (W),.IASS(N) X (N) C =*** PZNT ANALYTTC RSESUITS ANC COEPARISONS **=t WP.TT(6,1SI)) 52(1), CD(1), UDOTW(1) DO 190 K=2,NPL 190?.IT!2(6,70) EZ(X), CD (), TDOTW(K), ACEL(K-1), DLAX(X-1), ARATIO(K-1) w. T t(6, 570) E Z(), CD(N)), nWCOT t (N) PZTE(6,r650) ABA?. (Z I), aATDV, SIGMA, A.(IrI) w IT (6,62 0) SFLOW,Q,V,SVEL,PHOf,?HOSVISNUFiL V,CONFAC t? ( K1.O. I ) W:IT2(6,580) TTOL,1(N-1) TT"OL.'0t + TD (J) ITf t. 0. ) ZTI-(6,590) TOL, X(.N) fITS ( 6, 5^1C) C *** GO pAD A.OTFZS DATA ST **~' GO T:) 1On' C *** P=._OSt EMT STATISTICS CA ICLATIONS ** 2'CO ATO = 0.^ SED = o." STr2 = ^.n TT _O.C

I]O o~ U * ".4 N,..-.I - Ni *L N,Uo 0.4., Ch IG O 0'.- U) * # It-, ~. H Ka lc 6,.. ~ P u %'1,- * B I C cc c) n > 4 ~n r 10 I) LM 01 lI; r< o O, o %P 13 *r eq oG In I/ \H %, n.4P f~ I in 0 b MN IJ,-C SK in H N % I-,4 * "-,,'-,* l. l TU0 -U II 0 VM Wi N ol lob *4 c* * r4 * U r 0 N OD oo, F tn %. 4..;) CO Yp P1 - o M 0 <, P...- H. cn.. sF1. etU t jS H (,C co #' — C *'O''OU r- 4 U Ml 1C H S e~)U V<o P~ 1 O q 0'WB3 PH100 P -,eL.- U.,w o" 0 be S I0,1,u,-,C. c0 0 Cto 11o. n P. P4.. -' H. a.r - in, I 4 3 1- - 0,%vY) (il rIw (4,.4O- t,-1N5I - g.- a IVt vHe *41 01 H',o 3 b r 1; t14 c- s N - lb- P. 1 Il IHI - V I P- II VI 1 r- InZ Iv r to f.M K u 0. rir piPAl I I - I 3'4 l 4 I."; II0 4 I/ E.4 bl.. U Cin r 4i 4 -V) P * H P2 ~ v- UP p. P1pi< PA Sr. V- N.- PO W |S re 1M:_ - S i~O-P~5A "IOU.- NOIPMIb- H0(-4rC1- coN W Mr PS 0 U IU ini v- PiA -I-MM I r-. A S " i-N N - II i nM. WN U e- < uC?~WMUII%* N. O4 V |..- n H- A" VSHU P.-, Ui nO N Co " - ( 0 IL9 I/ sp 0 Ch- r', OBPItM,., P, - l N VI, P, r"CP 4W co Sq II W CO (q Pi r Kt in U -~-z D w- c-IM MW I-o.I(VVU boVI V OPiJ Usr4n Ufl C'~) N~~ "~ t H t IHa-Uv- P * PLnrI x i PaBsr - USH P- SW PcWH'.N j N PI C) (4 P1 H - H *., l-' v) cPv ) p-oa. U Pi in - D.. HO)a - 4 N U SIV l b-4 1' S t I j u-. - b - - I.IMU i H'M1U 4 Hi i- H P o P 0 M 0*n o 0 Q.-14'-..l-5 N P1 -H- UPOA 1 * 11- b 01.'ONZIt-I' S V1 -I I' 4-..-,-I ~' E-^-Ik O (I-4. Pl h. Sl'l(A, ltF/I -I 4eO Ulou~ C <:. fpS 9c U c _ H pi r 1 en el I'- 1v, 0:o oP.r.~'IoPH'S P -.. -.- - 0 < 1`2 VI P H H 0 10 F-I V) ~ r- U Pol V ^ r P *N L' O- * S,,H IH- -~r-,.-,MH-...r-.14: IIMMtP ~-loHP, lE.I-.I tdD S" t I e IP 54 I C < -5k.'~- bh-.OUl- r4Vl * H,-1HMU#, — I C'SCUSE IE-s I r. SI'1, W UUrl~U O3 S01-1o. U.1 — S VJ/l1, U lN'UP~-pi' e.%,,1(". I PVO Na 0' MIU eI. PP((.- H.t M1 F-N. JPl.-r U-LHP- VI-I -r - 4 I/;- /-I SCI /U PUHH OH 0P P; V) fI OU-0 U If 4 C~ V U I FH - H *O N Pe\. 0 Kt ( S I.) 4Ql USL- I XV/ I H AI. MH I1 %k- P\i C t -11. 4 N 0 0 -1 K * nOC c H I-nt. PS ^' OH KVMr-fe PM Hen.-/UP S l,,'- e* -.- I4/-'p 1, l I NN tZO N11.M1i t -1~-1 o44 H in to' IHHH U1 5t MOP) UUr - OU ~ BA 0a O IL' S 4-W'W. 5' t I5' HH < S S o III OIv'P. 3 i1~ 1 l-H -lU K. AU 1.-~P14 P- 14. - C A-I a I9) p Pi CHUiCtB.4 P. SC -. -. Ui f'-' rUHCi tt- 0- I- / NH %- CHh bN 0- 5- OH Pfj' in S -g4o VHPHvS Oi eH t~ V VI r-.1- v P1 5'- VI O H I U tiInHf 0 V W O I rK-. t-A M F N 0 - S 11SP1 PS SCU o lu I V)eIL4IP -1. t- % % [- 4 i SCc.- g4 P iP-is.- f-t Hi P 01 14)3U us P.\P tOI-opf-IO r o <o (a Cc O~~) 44.u SC O. Iuu Pr. r-cI4 *y P S r -.-.- r-.- 3 > - pI. *P o 04- " oOvH P. o< cv v- C3 +i Oin- ~ a w- BA.trPCrV m. rO - % n I v CC 0 t - tp II ti A iM p -i01 b I.. to 14 C 3 # 1 fs * HIPt.OVS-n MU).v-C 1 ||l- 0 CI/ 1-, 1 ql1- t5-r-. o'P. in,1 v -.o-xa0-e H.- I4 ICFrPaOH O'S P3.1SIPI-..'CL A: 3l.-.,-.I/1.-. PLY K C''. ~,5!I h.PU4. 1 1M - f'E-I.- H*5 S 5U II P,5-I o 1S4nS ^~<0& ~^K ^(n^ U a. ~K:O~-^ ^K *^ ^~nic OMO. ~ (^*t^-Htl^~ rri u v*/- I Po P, H r k 4r aC,oiVC0 0 0 I >0 D u-...r-',(ir b0O-.- Pt!(irIPnl l OI'U 5. piP SIcL -.' N -- toW U P1S'4 W 10 be * P4.-1fl.,as NJ u-toUV1 p f-4'- cW P N S, Ic1-I4.-5-4'L Is C. Pa C) u C) ( n %.O a V 4 0 1 ILI: in t-I 0`4 t ~' r W pi - C V3 $1 U h< ~P o ) I>~I,., Hto1. W,,O" o * o,...7 -.rb.,, a'""O H "" "'"ll "", dKI) P:l. X,,h e d t::li *)E-406 I. 5eb:r, M1P1- (I- llrSlUI-IOr l'SVPS'I II U OOOSCH HR 11- H# H H*,< K'U ""' %, r 0"' W:,K <U < Ps: -< eC f #W p. I. oC) o 0 V o o,CU-cr CC)v i C - 0 I43( 0 t- 0 Ct- 0' O.' W- I ('4 ('4 Alln Iii i S I C. %L|L pSp.5: U U U U

284 640 FO?.PeAT f 6 r2).8//) 650 FOP5!x (//TUt,' STVTSTICS 0F EXPlPIM ETAL ACCSLERArTON'// 1TI l,'AR IZ-'ATC LAPI ='.1 2. 3,T4 5,'AN DEVIATION a oF12.3, 2T8O,'STAHDAP.D DEVIA.TION a,,?12.3/T35, AVEBASZ A-CzLE2ATIO0?N, 3' PAO',?20.1 3) 6605'O (E 1 T 5.5, t ) ERD

nl&MFTCQ rH TIfK^' S ( cM ( f Cm APEA VnLUME MASS PnSITION OELTA I[ME DELTA X (rm**2?l CM**31 (G.AMS) (CM FROM FIN1 (SECONDS) ( CM ) VELOCITY DEL VEL ACCELERATION FORCE (CM/SECI ICM/SEC) (CM/SEC**21 (DYNESI 0. I Soo 0. I 9'00 0. 13530 0. I nOO0 0.15 000 0.! 3500 0.13500 0. 15900 0. IOO O.00000 0. I 3IO o. Ivoo DEYNrLLO N I lu I r p 1342. Sq q 1 34 7.'OR I 70o. 2 5? O 14'7. 0 6q 1342. 5o6Pq 34 2. 6 Q6 P 9 1 4?7,SQt PP 1 34 Z.5 W P R a95. 1li I5 1 4?.26'SPn 134?7,.AoQ 0. oo000 o.0 00on O. 1i500 o. oo000 0.0~000 0.0o.01) 0. 0000q o. 0o000 o. oQ000 0. 0' 500 0. 0n5o00 0.01L 3130 0.006B7066 0.01o 3139 0.00687066 0.0OL 3130 0.0103C590 0.0294469 0.01221451 r.0l3139 0.00687066 0.0143139 0.00343533 0.01l3139 0.00687066 0.01'3139 0.00687066 0.006o36t17 0.00305363 0.01 3139 0.00343533 O.01 3139 0.00343533 0.00697066 0.00687066 0.01030599 0.01221451 0.00697066 0.00343533 0.00687066 0.00697066 000 030536 0.00343533 0.003^3533 0.45000 0.49500 0. 58500 0.67500 0. 31000 0.94500 1.17000 1.39500 1.71000?.07000 2.3 500 O.001O 0.0010 0.0010 0.0010 0.0310 0.0010 0.0010 0.0010 O.OJ10 0.0010 0.045000 0.090000 0.090000 0.135000 0.135000 0.225000 0.225000 0.315000 0.363000 0.315000 45.000 90.000 90.000 135.000 135.000 225.000 225.000 315.000 360.000 315. 000 45.0000 0.0 45.0000 -0.0000 90.0000 0.0000 90. 0000 45.0000 22500.0 0.0 22500.0 -0.0 45000.0 0.0 45000.0 22500.0 154.590 0.0 274.827 -O.000 154.590 0.000 309.180 68.07 -77.295 -45.0000 -22500.0 nORtr, ANALYTIC EXPERIMENTAL ACCELERATI3N C.CFFPCIFNT ACCFLEATION ACCFLERATION CIFFERENCE (fM/SFC**?) (CM/SEC*? ) (CM/SEC**21 O. -r'031 21131.7 0. 397"031? 21131.7 22500.00 -1368.300.3 P770317 21131.7 0.0 21131.700 0.,4n 17? 142A3.7 22500.00 -9216.290 O.? P!317 21131.7 -0.00 21131.700 O. 37'0317 2111.7 45000.00 -23868.300 0.3P-70317 21131.7 0.00 21131.700 O.?877J3l7 21131.7 450:)0.00 -3868.300 O,.'t 50. 36003.P 22500.00 14303.841 0.31q7J317 21131.7 -?7500.00 43631.700 0. A'Ol 3 I 1 31. ACCEL RATION RA 0O 1.0648 0.0 1.5752 -0.0000 2.1295 0. 0000 2.1295 0.6113 -1. 0648 00 U1 A I. TJ.Al^T ^r Tr:: ST.TI TICS nF FX0OERIMFNTAL ACCELEQAFION 15000.000 MFAN DEVIATIN 19333.333 STANDARO DEVIATION AVFoQr,E PUCELEPITION RATIOn 0.7151749322211 22500.000 r- I r Q A': ol. ni (1.; q c A n 1.0)0000 CM**3/SC; STEAM VELOCITY = 305.00000 FT/SEC OR 9296.4 CM/SEC

0IAMTF r - ra(l CK'sF 5. I CM ( r'4 I APFA VOLUME MASS POSITION OELTA TI[E DELTA x ICM**21 (CM**3) (GQAMS (CM CROM FINI (SECONDS) ( CM I VELOCITY DEL VEL ACCELERATION FORCE (CM/SEC) (CM/SECI (CM/SEC**21 (DYNESI O,.o0n0 0.0:-500 3.0' 503 0. 04 5 ) 0* o-500 0.01n00 0. 0 500 OO.L n1 0. 0L 500 0. 0000o 0. 0 500 0. 0 500 0. 01000 0.006361 7 0.0015934 0.0015o04 0.0063617 0.0015904 0.0015004 0.0n03617 0. 001526PI 0.00034170 0.00038170 0.00305363 0.00038170 0.00038170 0. 0030363 0.0015 26i 0.00031170 0.00038170 0.00305363 0.00038170 0.00038170. C0305363 0.49500 0.67500 0. 90000 1,26000 1.62000 2.02500 2.61000 0.0010 0.0010 0.0010 0.0010 0.0010 0.0010 0.183000 0.225000 0.360000 0.360000 0.405000 0.585000 180.000 225.000 360.000 360.000 405.000 585. 000 45.0000 135.0000 -0.0000 45.0000 180.0000 22500.0 67500.0 -0.0 22500.0 90000.0 8.588 25.765 -0.000 8.588 34.353 OaEYN1Lf) DQA, JANALYTIC EXPERIMENTAL ACCELERATION Nfjtc.q rnFFFICIF^NT CCELCQATION ACCELERATION DIFFFRENCE (Cr/SEC**2 I(CM/SFC**2) (CM/SEC**2) Qqc,11 2? O.t 50nqqlq 36PO3,. 6&7.55563 0.r'0087 959984.5 22500.00 73484.454 LL7.5<^653 O.'F'7n77 95aQ4.% 67500.00 28484.454 9sC.1 1? 0./501 iqqo 16903.R -0.00 36803.894 4&7.i.3 0.P"700R'2? 958q4.5 22500.00 73484.454'4-', l O. PEOOR'2 95CR4.5 90000.00 5Q84.454 R9;.l12 0.-5,:Ol90 36803.8 ACCELERATION RATIO 0.2344 0.7032 -0.0000 0.2344 0.9377 03 0' STATISTICS OF EXPERIMENTAL ACCELERATION tP ITt4MATr MC:N = 40500.000 MEAN )EVIATION = 30600.000 STANDARD DEVIATION m 36971.273 AVEPRAGE ACCEIERATION PArTIrn 0.421433271900 ct L ),.'c = O.Olnr) rcm ro 5.00GOO CM*'3S/S:C STEAM VELOCITY a 305.00300 FT/SEC OR 9296.4 CM/SEC o&Qlr(I. nc jclry = 1.30000 ST AM DENSITY =.0001135215 GM/CM**3 KINEMATIC VISCOSITY - 0.93469632 CM**2/SEC CILM V'Lr ITY c' 10O, FO&aFS/SEC DATA CONVFRSION FACTOQR 0.450 TH: Do^o rvrL Wic 0'P- 1pcrcVrn TO PQREF UP IN 0.00700 SECONDS PR 2.61000 CM. FROM THE END OF THE FIN. *** * *a ft. *i *t +*** ** * **** * 4 * * * *a* * * * Jt * *** ***o

D^4RirClr')EISITY l.CnOO() STEAM DFISITY a.0001135215 GM/CM**3 KINEMATIC VISCOSITY * 0.93469632 CM*~2/SEC CTLM VCL'TInrY = cOO. FIoAUr/cFC. DATA CONVESIONr FACTOR - O.LIU THE OAQETICAI weC. trnt ncnovrn Tn PREAK Un IN 0.01o10 S-CnNnS.R 2.38500 C., FROM THE END) OF THE FIN. * e * * * * * * * t *** * * e * * * * * * * * * * * * * * * * + * * * * * * * + * * * * * * * * * * * * * * * 0 * * * * * 0 * * tNJ 00 "-.

* *** * * **** r* **** ** ** ** ** ** *** *** ***4 ** **** ** ** ** *** *** * * ******** &VFOAhrc FYf)P!IMF% iL rA.rFltPA' ION ST&TISTICS FnR A FLnW &ATE PF 5. CMH*3/MIN AND STEAH VELOCITY 305. FT/SEC STITISTICS O1F EXPERIMENTAL aCCELEQATION AP I TMTIC M:AN 26928.01339286 AC'CLFRATIrN ATl O, 0.78721570486 MEAN OEVIATION J 14922.36328125 ~TANnARO nlcVlATION = 17R34.16571408 AVFPAGE ACCELEqATION OF l6 OATA SETS I ^7;. OO000 OC. OOO 52 500. 00000000 3OC 0000,00000000 37500.00000000 22500.00000000 1to0, 0oooooo 000 000.00000000 t0O00.00000000 22500.00000000 6000.00000000 18000.00000000 oon000.o000ono 0.0 5035.71428571 196R7.50000000 A**o * * iA A ** ****** **4******* **4* 4* * * * *** ***. * * * * * **** *** * ***

nIA tETcQ TH ICW'FSS I CM 1 ( CM 1 ARFA VOLUME MASS POSITION DELTA TIME DELTA X (fC**?7 (CM**3) (GRAMS) (CM FROM FIN) (SECON3S) I CM I VELOC ITY (CM/SEC) DEl. VEL ACCELERATION FORCE ICM/SEC) (CM/SEC**21 (DYNESI o. Oo000 0. 00300 0.00000 o. 0o000 0. 0O000 0.0o000 0.o00000 0.00n371' 0.0030'16 0.00305361 0.0-500 0.003617 0.001526P1 0.00152691 0.00000 0.0063617 0.00305363 0.00305363 0.01000 0.0063617 0.00305363 0.00305363 0.00000 0.0063617 0.00305'63 0.00305363 O.0QOO 0.0063617 0.00305163 0.00305363 0.36000 0. 54000 0.76500 1.03500 1.39500 1,93500 0.0010 0.0010 0.0010 O.0010 0,0010 0.183000 0.225000 0.270000 0.360000 0.540000 180. 000 225. 000 270.000 360.000 540.000 45.0000 45.0000 90.0000 1 80. 0000 22500.0 22500,0 45000.0 90000.0 34.353 68.707 137.413 274.827 ocyVNLnC NI lUMR C 152,|. 1?7'l V?6. 1lfl 1526. 17t' 1 1526. l 7 lc76.17cl1 1;26.l2.c l nPA hr ANALYTIC EXPEPIMENTAL ACCELERATION rnFF F!CIFNT 6CCELEERATION ACCELERATION DIFFERENCE ICM/SFC**2 I(CM/SEC**21 (CM/SEC**2) 0.37010119 87953.6 0.1701011o 87953.6 22500.00 65453.645 0.37010110 87953.6 22500.00 65453.645. 0. 7010ll9 87953.6 45000.00 47q53.645 0.37010110 87953.t 90000.00 -2046.355 O.?701011n 87953.6 ACCELERATION RAT IO 3.2558 0.2558 0.5116 1.0233 00'.0 STATISTICS OF EXPERIMENTAL ACCEI ERArION X IT4HATTr MFFAN 45000.000 MEAN DEVIATION * 22500.000 STANDARD DEVIATION ~ 31819.805 AVEPAGE ACCELERATION RATIO = 0.5115331458401 cLOW qarc 0.011'8 CFm np 15.00000 CM**3/SEC STEAM VELOCITY - 520.00000 FT/SEC OR 15849.6 CM/SEC oAarTCLr IFNSITY = 1.00000 STAM DEtNSITY =.0001135215 GM/C4**3 KINEMATIC VISCOSITY - 0.93460632 CM**2/SEC tlLuw lCLrr ry' rOO;). FPPAMq/SEC PATA CONVERSION FACTOQ = 0.450 THC oaD41lCL WAr ORS=QVFen TO RREAK IP AT 0.00500 SECONDS ANO 1.93500 CM. FROM THE END OF FIN. *,*,*, * ~ e * * b ~ ~* ~ * * -w * * * * ~ * * * * * * * * 4 * { I 4 * * * W 4 t I { I + { A ~ ~ I $ + a ( + * * ~ $ e

*4*3* * * * *^ 3* 3* 3 ** ** ***** ** **** * * * ****** *** ***** *** **** **** VFQAE YP;O lMcDIrtAI ACCFLFPAT ION STATISTICC FnR A FLOW QATE OF 15. CM**3/MIN AND STEAM VELOCITY 520. FT/SEC STATISTICS CIP r-'( PR IFNTAL AC.CF LEAr I ON APITTMATIC MFAN 8 92940.21739130 ACCeLSP ATION RATIO = 1.02963352162 MFAN'FVI AT iN = 49834.59357278 STANOnA nDEVIATION = 71 33.81289208 A4FPAE ACCEI.ERATION OF 23 DATA SErs 6 7; 00. )OOOO()O0 57500,100o0000 -00D00.00000000 40cO.O00000000?2 0000.00000300 ooo0000. oocoo0 tr 75 00. oooooooo 73125, 00000000 aS++it>4.~l<~ 135000.00000000 101250.00000000 146 750. 00000000 41250.00000000 *3*4l2.4O**4Q * + * * *+ * * l 202500.OJJU0000 90000.00J00000?70000.00OO0000 45000.00000000 * *3*3* * * 54000.00000000 45000.00000000 -11250.0000000 45000.00000000 44 *3*3*3* $~ 90000.00000000 58500. 00000000 75000.00000000 roj 0

nlAMFTTn TLlCK NrSC ( CM ) ( C. I AqFA VOLUJME MASS POSITInN DELTA TIME DtLTA X ICu**2 ) (Cm**3I~ (rAMSI ( CM FROM FIN) (SECnN)S) (' CM ) VELOCITY DEL VEL ACCELERATION FORCE ICM/SEC) iCM/SEC) ICM/SEC**2) IDYNES O. 0'~ 5 N 0. 0&900 0.04500 0.0 530 O. 04500 0. 0 o500 0. 0 500 0. 0t 500 0. 04 5nO 0.04500 0. 0 500 0)o0063617 0.0015904 0.0015004 0.0015004 0.001 004 0.001 5064 0. 001268Pl 0.00038170 0.00009170. 00038170 0.00039170 0.00038170 0.00152681 0.00038170 0.00038170 0.00038170 O.00038170 0, 0003 170 0. 18000 0.36000 0. 4530 0.8 1000 1.35000 2.25000 0.0010 0.0010 0.0010 0.00 10 0.0010 0.180000 0.135000 0.315000 0.540000 0.900000 180.000 135.000 315.000 540. 000 900.000 -45.0000 180.0000 225.0000 360.0000 -22500.0 90000 0 112500.0 180000.0 -8.588 34.353 42.942 68.707 OEYn'l VS ORAr. ANALYTIC EXPERIMENTAL ACCELERATION NIFllFo CPEFCICI-,NT ACFFLERATION ACCELERATION DIFFERENCE (CM/SFC**2? (CM/SEC**2) (CM/SEC**21 22.1.25-&6 0.31375316 187682.0 1210.,?q33 0. 02.64 004 481705.4 -22500.00 504205.413 1210.t313 O.t024)400 4 491705.4 90000.00 391705.413 1210.62P33 0.-02?600'A8l705.' 112500.00 369205.413 1210. 22833 0. 0? 400 0 9 1705.4 180000.00 30l70'5.413 1210. t2At3 0. ^076400t' 4R1705.4 ACCELERATION RATIO -0.0467 0.1868 0.2335 0.3737 rD v<D STATISTICS OF EXPERIMFNTAL ACCELERATION AcITHMAT1i uyA = 900U0.000 MEAN DEVIATION 56250.000 STANDARD DEVIATION - 84187.291 AVERAGE ACCELERATION RATIO - 0.1.68361815152 FLOW RArc - 0.10594 CFM OR 50.00000 CM**3/SEC STEAM VELOCITY. 925.00000 FT/SEC OR 25146.0 CM/SEC DRrTICL:?FNSITY - 1.00000 STEAM DENSITY =.0001135215 GM/CM**3 KINEMATIC VISCOSITY - 0.93469632 CM**2/SEC FILu VCL"CIrY - C000. CRAMFS/SEC DATA CONVERSION FACTOR - 0.450 THc oDAo0T L WAS r)9S-RVo TrO BREAK I1o AT 0.00500 SECONDS AND 2.25000 CM. FROM THE END OF FIN. * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *

* * r + * I I t a * * L * $ e + O I I * f*** * a I * * * * * + 4 * * ** *4 4 * 4 * * * * ****** * &\IFPDA.E )YDCC TcJ^AL ACCELFp^ATICN STATISTICS FnR A FLOW R4TE nF 50. CM**3/MIN ANn STEAM VELOCITY A?2. FT/SEC STATISTICS lF FYPEQIMENTAL ACCELERATION AOITHMATIC =EAN t 147197.50000000 ACC LERATI ON PATIn I 1.02707793883 PFAN OEVIATI'^J' 66796.8750J000 STANrAQO r)OVIATION 9 78995.61690948 AVFPAGF ACCELERATIOlN OF 0 DATA SETS 20.00.00000o 0 no 2500.00000000 75000.00000000 90000.00000000 112500.00000000 90000.00000000 1012o0.000000o n 213750.0000000 * * i * * * * * * * * * * * * * ******* *** I * * * *** * ** I * *** * * 4+. * I I * * * *** 4 ~ * I ** i s * L *I +

293 APPENDIX D STATISTICAL LINEAR REGRESSION ANALYSIS OF DROPLET ACCELERATION Due to the nature of the physical phenomenon under consideration and the large amount of experimental data collected, certain statistical tools were used to better understand the experimental results. Reduction of the initial raw data (Appendix C) resulted in 855 droplet accelerations for the flow conditions outlined in Table 5.3. These 855 cases were broken into 327 subsets for which average droplet accelerations were obtained. When performing a first-order regression of E.A, (Experimental Acceleration) on Q the 327 data set was arranged into the following cases corresponding to constant steam velocity, Us, and stored in data file "AVGACCEL"; Cases Us (ft/sec) 1-45 305 46-129 520 130-185 825 186-246 975 247-327 1100 In a similar manner the "327" data set was rearranged in order to perform a first-order regression of E.A. on U with liquid flow rate, Q, being held constant. These are stored on data file "QACCEL",

294 Cases. Q (_cm 3/min. 1!3 Q 4-58 5 59-130 1Q 131-198 15 19g-250 2Q 251-294 30 295-302 40 303-327 50 MIDAS "REGRESSION" command was used to perform the regression analyses mentioned throughout the test, From the MIDAS results, it is found that E.A. is not affected by changing Q. However, it is seen that E.A. is strong function of Us. For example, "goodness of fit," R-SQR shows that for Q = 15, there is a variation of 76% in E.A., which can be explained by the variation in U. Full s details of this analysis is available from on of our previous reports. (M14) Regression analysis output is followed.

295 REGRESSION OF E.A. ON @ at CONST. Us VARIABLE LABELS DATA FILE: AVGACCEL AA- Ave. "327" DATA SET Acceleration, Ft/Sec R- ACCEL, RATIO (Measured/Senafini) U- Steam Velocity, Ft/Sect Q- Liquid Flow rate, cm3/min

296 RES COMMA,tO?HIST V=AA CASES=1-45 OP=HISTT INTERVAL EXPRESSION -- #INT:(MIN.MAX) (MIN,MAX)/WIDTH PER/(MIN, MAX) HISTOGRAM MIDPOINT 0. 676.67 1353.3 2030.0 2706.7 3383.4 4060.0 TOTAL CASES=CASE#: 1 -45 HISTZ COUNT FOR 1.AA (EACH X= 1) 20.0 9 +XXXXXXXXX 46.7 21 +XXXXXXXXXXXXXXXXXXXXX 22.2 10 +XXXXXXXXXX 8.9 4 +XXXX 0. 0+ 0. + 2.2 1 +X 45 (INTERVAL WIDTH: 676.67) COMMAND?HIST V=AA CASES=1-45 OP=HISTZ INTERVAL EXPRESSION -- #INT: (MIN, MAX) (MIN,MAX)/WIDTH #PER/(MIN,MAX) HISTOGRAM CASES CAS E: 1-45 MIDPOINT HISTZ COUNT FOR 1.AA (EACH X: 1) 0. 20 0 9 +XXXXXXXXX 676.67 46.7 21 + XXXXXXXXXXXXXXXXXXXXX 1353.3 22.2 10 +XXXXXXXXXX 2030.0 8.9 4 +XXXX 2706.7 0. 0 + 3383.4 0. 0 + 4060.0 2.2 1 +X TOTAL 45 (INTERVAL WIDTH= 676.67) COMMAND?3ESCRIBE V=AA CASES=1-45 DESCRIPTIVE MEASURES CASES =CASE#: 1 -45 VARIABLE N MI NI MUM MAXI MU MEAN STD DEV

297 1.AA 45 0. 4060.0 851.28 C0MMAND 7DiST V=AA CASES=1-45 PROB=.5 DISTRIBUTIONAL ANALYSIS CASES CASE#: 1-45 CUMULATIVE SAMPLE DISTRIBUTION OF 1.AA N= 45 OUT OF 45 1.00000 + *.90000 + 3 * 2.80000 + 3.70000 + ** * 3.60000 + 2 2.50000 + * 2.40000 + 3.30000 + 5.20000 + - 2 *.10000 + **.73 5.72 * 0. * ** + 0. 1624.0 3248.0 AA 812.01 2436.0 4060.0 PROB QUANTILE.5000 632.73 COMMAND?REGRESS V-AA;Q CASES=1-45 LEAST SQUARES REGRESSION CASES=CASE#:1-45 A ALYSIS OF VARIANCE OF I.AA N= 45 OUT 0F 45 SOURCE DF SUM SQRS MEAN SQR F-STAT SIGNIF REGRESSION 1.25527 +6.25527 +6.46589.4985 ERROR 43.23561 +8.54793 +6 TOTAL 44.23816 +8

298.ULT R=.10353 R-SQR=.01072 SE= 740.22 VARIABLE PARTIAL COEFF STD ER0R T-STAT SIGNIF CONSTANT 944.69 175.80 5.3736.0000 4.Q -.10353 -5.9627 8.7357 -.68256..4985 C0MMAND 7

299.60000 + 3 5.50000 + 6 5.40000 + 7.30000 + 3 * 4.20000 + 2 9.10000 + ** * * * 0. + —-— t-.... —~ — -~ —-- -. —-. ( --— t. -— (.- — (. —-. ~....+ -7381.8 1476.4 10335. AA -2952.7 5905.5 14764. PR0B QUANTILE.5000 2509.8 C iMnMAND 7REGRESS V=AA;Q CASES46-.-=46-129 LEAST SQUARES REGRESSION CASES=CASE#:46-129 ANALYSIS 0F VARIANCE 0F 1.AA N= 84 OUT 0F 84 S0URCE DF. SUM SQRS MEAN SQR F-STAT SIGNIF REGRESSI0N 1.11675 +8.11675 +8 2.0719.1538 ERR0R 82.46205 +9.56347 +7 T0TAL 83.47372 +9 MULT R=.15699 R-SQR=.02464 SE= 2373.8 VARIABLE PARTIAL C0EFF STD ERROR T-STAT SIGNIF C0NSTANT 3402.7 446.23 7.6255.0Q00 4.Q -.15699 -29.388 20.417 -1.4394.1538 COMMAND? --

300?HIST V:AA CASES=130-185 OP=HIST% INTERVAL EXPRESSION -- INT:(MIN, MAX) (MlN, MAX)/WIDTH #PER/(MIN,MAX) HISTOGRAM CASES CASE: 130-1 85 MIDPOINT HIST7 COUNT FOR 1 AA (EACH X= 1) 738.19 1.8 1 +X 2952.8 8.9 5 +XXXXX 5167.3 19.6 11 +XXXXXXXXXXX 7381.9 28.6 16 +XXXXXXXXXXXXXXXX 9596.5 23.2 13 +XXXXXXXXXXXXX 11811. 12.5 7 +XXXXXXX 14026. 3.6 2 +XX 16240. 1.8 1 +X T0TAL 56 (INTERVAL WIDTH: 2214.6) COMMAND 7?ESV-SCRIBE V=AA CASES=130-185 ERROR -- INVALID COMMAND: DESSCRIBE' COMMAND CANCELLED COMMAND?DESCRIBE V=AA CASES=130-185 DESCRIPTIVE MEASURES CASES =CASE#: 130-1 85 VARIABLE N MINIMUM iAXIMUM MEAN STD DEV I.AA 56 738.19 16240. 7871.8 3195.2 COMMAND?DIST V=AA CASES=130-185 PROB=.5 DISTRIBUTIONAL ANALYSIS CASES:CASE#: 130-185 CUMULATIVE SAMPLE DISTRIBUTION OF 1.AA N= 56 OUT OF 56 1.00000 + * 3:0i0:,,5.90000 + 3

301.80000 + 5.70000 + 4*.60000 + 4 5.50000 + 2.40000 + 5.30000 + 4.20000 + ** 4.10000 + * 2 * * 0. + +-.+- _- _ —. --- - -— +._ ----— +. --- -+. +- - _ — 738.19 6939.0 13140. AA 3838.6 10039. 16240. PROB QUANTILE.5000 7381.9 COMMAND?REGRESS V=AA;Q CASES=130-185 LEAST SQUARES REGRESSION CASES=CASE#: 130-1 85 ANALYSIS OF VARIANCE OF 1.AA N= 56 OUT OF 56 SOURCE DF SU. SQRS MEAN SQR F-STAT SIGNIF REGRESSION 1.12325 +9.12325 +9 15.185.0003 ERROR 54.43828 +9.81163 +7 TOTAL 55.56153 +9 MULT R=.46849 R-SQR=.21949 SE= 2848.9 VARIABLE PARTIAL C0EFF STD ER0RR T- STAT SIGNIF C0NSTANT 9991.2 663.89 15.050.0000 4.Q -.46849 -104.57 26.835 -3.8968.0003 C0MMA N?

302 HIST V=AA CASES:186-246 I-OP=HISTZ INTERV@L EXPRESSI0N - #INT:(MIN,MAX) (MIN,MAX)/WIDTH #PER/(MIN,MAX) HISTOGRAM CASES=CASE#:186-246 MIDPOINT HIST2 COUNT FOR 1.AA (EACH X= 1) 2214.6 1.6 1 +X 4640.0 26.2 16 +XXXXXXXXXXXXXXX 7065.5 31.1 19 +XXXXXXXXXXXXXXXXXXX 9491.0 13.1 8 +XXXXXXXX 11916. 14.8 9 +XXXXXXXXX 14,42. 8.2 5 +XXXXX 16767. 3.3 2 +XX 19193. 1.6 1 +X TOTAL 61 (INTERVAL WIDTH: 2425.5) COMMAND?DESCRIBE V= 86-246#?DESCRIBE V:AAB. CASES=186-246 DESCRIPTIVE MEASURES CASES =CASE#: 186-246 VARIABLE N MINIMUM MAXIMUM MEAN STD DEV 1.AA 61 2214.6 19193. 8525.5 3661.7 COMMAND?DIST V=AA CASES=186-246 PROB=.5 DISTRIBUTIONAL ANALYSIS CASES=CASE#:186-246 CUMULATIVE SAMPLE DISTRIBUTION OF 1.AA N= 61 OUT OF 61 1.00000 + 2 3*.90000 + 4*.80000 + 4.70000 + 2 5

303. UUUU -t.50000 + 2.40000 + * 2 4.30000 + 2 6.20000 + 3.1 0000 + 2 3 0. +* 4.- --- ---— + -- - — + - ----- -+ —---- - - -+-....+ —— +- - -—'42214.6 9005.9 15797. AA 5610.2 12402. 19193. PR0B QUANTILE.5000 812 0.1 C0MMAND?REGRESS V=AA;Q CASES=186-246 LEAST SQUARES REGRESSI0N CASES=CASE#:186-246 ANALYSIS 0F VARIANCE 0F 1 AA N: 61 OUT 0F 61 S0URCE DF SUM SQRS MEAN SQR F-STAT SIGNIF REGRESSI0N 1.26170 +9.26170 +9 28.446.0000 ERR0R 59.54280 +9.92000 +7 TOTAL 60.80450 +9 MULT R=.57035 R-SQR=.32530 SE= 3033.1 VARIABLE PARTIAL COEFF STD ERR0R T- SAT SIGNIF C0NSTANT 13526. 1014.8 13.328.0000 4.Q -.57035 -265.25 49.733 -5.3335.0000 COMMAND 7 - — ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~-l

304 HIST#?HIST V=AA CASES=247-327 OP=HISTZ INTERVAL EXPRESSION -- #INT: (MIN,MAX) e. ( MI N,MAX)/WIDTH #PER/( I N, MAX) HISTOGRAM MIDPOINT 2952.8 5085.3 7217.8 9350.4 11483. 13615. 15748. 17881 20013. 22146. TOTAL CASES=CASE#:247-327 HIST% COUNT FOR 1.AA (EACH X= 1) 1.2 1 +X 12.3 10 + XXXXXXXXX 22.2 18 + XXXXXXXXXXXXXXXXXX 25.9 21 + XXXXXXXXXXXXXXXXXXXX 19.8 16 +XXXXXXXXXXXXXXX 2.5 2 +XX 9.9 8 +XXXXXXXX 3.7 3 +XXX 1.2 1 +X 1.2 1 +X 81 (INTERVAL WIDTH= 2132.5) C0MMAND?DESCRIBE V=AA CASES=247-327 DESCRIPTIVE MEASURES VARIABLE I.AA CASES =CASE#:247-327 N MINI MUM MA XI MUM 81 2952.8 22146. MEAN 1 0201. STD DEV 3819.4 COMMAND 7REGRESS V=AA;Q CASES=247-2-327 LEAST SQUARES REGRESSION CASES=CASE^:247-327 ANALYSIS F07 VARIA, SOURCE REGRESSION ERROR TOTAL MULT R=.0624 VARIABLE CO NS TA NT 4.Q 0F 1.AA N= 81 OUT 0F 81 DF SUM SQRS MEAN SQR 1.45454 +7.45454 +7 79.11 625+ 10.1471 5 +8 80.11 670+1 0 R-SQR=.00389 SE= 3836,0 F-STAT 3 0890 T-STAT 14.934 -.55579 SIGNIF.5799 SIGNIF.0000.5799 PARTIAL COEFF 10512. -.06241 -17.273 STD ERROR 703.91 31.079

305 C0MMAND?DIST V=AA CASES:247-327 P0.-ROB=.5 DISTRIBUTIONAL ANALYSIS CASES CASE#:247-327 CUMULATIVE SAMPLE DISTRIBUTION OF 1.AA N= 81 OUT OF 81 1.00000 + * 5.90000 + 3 * *.B0000 + 5 4.70000 + 6 9*.60000 +.50000 + 5 52.40000 + 7.30000 +.20000 + 3* 2*.10000 + 3* * 0. PR0B.5000 COMMAND 3.- - -t —2952.8 10630. 18307?. AA 6791.3 14469. 22146. QUANTILE 9596.5

306 REGRESS V=AA;Q CASES T0 SELECT =ALL LEAST SQUARES REGRESSION ANALYSIS OF VARIAN S0 URCE REG R ESSI N ERR0R TOTAL MULT R=.100O VARIABLE CBNSTA NT 4.Q OF I.AA N= 327 OUT 0F 327 DF SUM SQRS MEAN SQR 1.71320 +8.71320 +8 325.69509+10.21387 +8 326.70222+10 R-SR:.01016 SE= 4624.6 F-STAT 3.3347 T-STAT 15.546 -1.8261 SIGNIF.0688 SIGNIF.0000.0688 PARTIAL C0EFF 6998.8 -.10078 -37.216 STD ERR0R 450.21 20.380 C0MMAND 7TRANS TRANSrFRMATI0N EXPRESSION (E.G., V.=V2+V5 0R =V1IO=13GIO(AA) L=LA ERROR -- SYNTAX: "VIO=LOG1O(AA) L=LA' ENTER NEW VALUE FOR TRANSFORMATION EXPRESSI0N V10 = SQRT(V8) ) V1 O=SRT(V8) ) (E.G., V9 V2+V5 OR ERROR -- ILLEGAL VALUE THE EXPRESSI0N STARTS WITH tVK:,. WHERE 0 < K < 10000; THE RIGHT-SIDE MAY BE. A", A OP B" F, FCN CN(A)' FCN(AB), WHERE A AND B ARE VARIABLES AND/OR CONSTANTS WITH DECIMAL POINTS, FCN IS FUNCTION NAME. ENTER NEW VALUE FOR TRANSFORMATION EXPRESSION (E.G., V9=V2+V5 OR VI O0 SQRT(V8) ) =V20: LOGI O(AA) LABEL FOR THE RESULT VARIABLE(S) =LAA LG1 O TRANSFORMATION VARIABLE TOTAL 20.LAA 327 VALID 323 MISS 4 C0MMAND?REGRESS V=LAA;Q CASESE.-ALL

307 LEAST SQUARES REGRESSIBN ANALYSIS OF VARIANCE 0F 20.LAA N= 323 OUT OF 327 S0URCE DF SUM SQRS tIEAN SQR F-STAT SIGNIF REGRESSION 1.39460.39460 1.9284.1659 ERR0R 321 65.685.20463 T0TAL 522 66.080 MULT R=.07728 R-SQR:.00597 SE=.45236 VARIABLE PARTIAL COEFF STD ERR0R T-STAT SIGNIF CONSTANT 3.6946.44400 -1 83.211 0. 4.Q -.07728 -.27783 -2.20007 -2 -1.3887.1659 C 0 MMA ND?DIST V=LAA CASES=ALL OP==.-HIST% ERROR -- INVALID KEYWORD: OP=HIST%" DISTRIBUTIONAL ANALYSIS CUMULATIVE SAMPLE DISTRIBUTION OF 20.LAA N= 323 OUT OF 327 1.00000 + 52* X 9.901 $.07 USED. CONTINUECYES/NO) %0200 DATA CHECK. 7%0200 DATA CHECK. r??ES Y OR N "?N COMMAND?HIST= V?HIST V=LAA CASES=ALL OP=HIST% INTERVAL EXPRESSIN -- #INT: (MIN, MAX) (MI N,MAX)/WIDTH PER/(MIN, MAX) HISTOGRAM MIDPOINT HIST% COUNT FOR 20.LAA (EACH X= 2).8682.6 2 +X 2.0139 0. 0 + 2.1596 0. 0 + 2.3053.6 2 +X 2.4510 1.2 4 +XX 2.5967 1.9 6 +XXX 2.7424 3.1 10 +XXXXX 2.8882 1.5 5 +XXX 3.0339 2.5 8 +XXXX 3.1796 6.8 22 +XXXXXXXXXXX 3.3253 7.7 25 XXXXXXXXXXXXX 3.4710 7.7 25 +XXXXXXXXXXXXX

308 3.61b 9.b 31 + AAAAAAAAAAAAAAAA 3.7 624 15.0 42 +XXXXXXXXXXXXXXXXXXXXX 3.9081 17.6 57 +XXXXXXXXXXXXXXXXXXXXXXXXXXXXX 4.0539 19.2 62 XXXXXXXXXXXXXXXXXX XXXXXXXX XXXXXXXXX 4.1996 5.9 19 +XXXXXXXXXX 4.3453.9 3 +XX MISSING 4 T0TAL 327 (INTERVAL WIDTH:.14571) C0MMA ND?READ N0TE, THIS DATASET CURRENTLY HAS 5 VARIABLES AND 327 CASES FILE'CONTAINING DATA OR * TO ENTER DATA HERE =WGABAR FORMAT SPECIFICATION OR * TO ENTER DATA SEPARATED BY COMMAS VARIABLES TO READ (E.G., 1-10) =1-4 LABELS C0RRESP0NDING TO THESE VARIABLES * FOR STANDARD LABELS =AAC,RC, U, Q CASES TO ASSIGN TO THE DATA BEING READ (E.G., 1-98) =1-327 READ OBSERVATIONS 1-327 VARIABLES BY CASE 327 CASES READ FOR 4 VARIABLES CASES CHANGED FOR 4 EXISTING VARIABLES COMMAND?REGRESS V=AAC;Q CASES=ALL LEAST SQUARES REGRESSION ANALYSIS OF VARIANCE OF I.AAC N= 327 OUT OF 327 SOURCE DF SUM SQRS MEAN SQR F-STAT SIGNIF REGRESSION 1.66258+11.66258+11 3.3347.0688 ERROR 325.64576+13.19869+11 TOTAL 326.65238+13 MULT R=.10078 R-SQR=.01016 SE=.14096 +6 VARIABLE PARTIAL COEFF STD ERROR T-STAT SIGNIF CONSTANT.21332 +6 13722. 15.546.0000 4.Q -.10078 -1134.4 621.19 -1.8261.0688 C0MMA N D?TRANS TRANSFORMATION EXPRESSION (E.G., 9 =V2+V5 OR VIO=SQRT(V) ) =V30= LG O0(AAC) LABEL FOR THE RESULT VARIABLE(S) =LAAC

309 L0GI O TRANSFORMATION VARIABLE TOTAL VALID MISS 30.LAAC 327 323 4 COMMAND?REGRESS V=LAAC;Q CASES=ALL LEAST SQUARES REGRESSION ANALYSIS OF VARIANCE OF 30.LAAC N= 323 OUT OF 327 SOURCE DF SUM SQRS MEAN SQR F-STAT SIGNIF REGRESSION 1.39460.39460 1.9284.1659 ERROR 321 65.685.2 0465 TOTAL 322 66.080 MULT R=.07728 R-SQR=.00597 SE=.45236 VARIABLE PARTIAL COEFF 5TD ERROR T-STAT SIGNIF CONSTANT 5.1786.44400 -1 116.64 0. 4.Q -.07728 -.27783 -2.20007 -2 -1.3887.1659 C0MMAND S I G NOFF ERROR -- INVALID COMMAND: SIGNOFF' C0MMAND CANCELLED C0MMAND?STOP USE $RES TO RE-ENTER MIDAS #SIGNOFF #ODLU 15:39:40-17:22:21 FRI MAR 31/78 T ERM, NR MAL,UNI V ELAPSED TIME 102.685 MIN. $2.40 #CPU TIME USED 3.021 SEC. $.49 #CPU STOR VMI 4.116 PAGE-MIN. $.24 #WAIT STOR VMI 141.476 PAGE-HR. #DRUM READS 5806 #APPROX. COST OF THIS RUN IS $3.23 0ISK STORAGE 128 PAGE-HR. $.01 #APPROX. REMAINING BALANCE. $73.29 (@

310 REGRESSION OF /El ON Us AT CONST @ VARIABLE LABELS DATA FILE: QACCEL V1.AA - A V2.R - V3.U - V4.Q - V20.SRAA AVG ACCEL. Ft/sec2 ACCEL RATIO STEAM VELOCITY, Ft/sec LIQUID FLOW RATE, cm3/min - SQUARE ROOT AVG. ACCEL, /Ft/sec2

311 SIGNON ODLU #ENTER USER PASSWORD. 7STEAM TERM, NORMAL, UNIV #**LAST SIGNON WAS: 00:19:25 TUE APR 04/7F f USER'ODLU' SIGNED ON AT 17:27:51 ON WED APR 05/78,F AVGACCEL CUMACCEL LF LINEFIT MDPRDATA PFILE PLOT2 QACCEL TDATA TEMP WATERDROP WD WDGRAF WG WGABAR WGAVG #LIST QACCEL > 1 19192.91406 0.75F39 1100.00000 > 2 15501.96875 0.61254 J100.00000 3> 7381.89063 0.29169 1100.00000 > 2030.01978 2.24796 305.00000 > 5 132F.74023 1.10043 305.0000 6 1722.44116 1.426AF 305.00000 ~ I >ATTN!'RUN STAT;:.MIDAS EXECUTION BEGINS PLOTI WDATA 0.0 0.0 0.0 5.0 5.0 5.0 MIDAS STATISTICAL RESEARCH LABORATORY UNIVERSITY OF MICHIGAN 17:30:46 APR 5, 1978 COMMAND?READ FILE CONTAININ3 DATA OR * TO ENTER DATA HEPE =*QA CCEL FORMAT SPECIFICATION OR * TO ENTER DATA SEPARATED BY CnMMAS =4(F15.5, IX) VARIABLES TO READ (E.G., 1-10) =1-4 LABELS CORRESPONDING TO THESE VARIABLES OP * FO STANDARD LABELS =AA,R,U, CASES TO ASSIGN TO THE DATA BEING READ (E.G.. 1-9P) =1-327 READ OBSERVATIONS 1-327 VARIABLES BY CASE 327 CASES READ FOR 4 VARIABLES COMMAND 7WRITE V=ALL CASES=1-6

312 FILE TO RECEIVE DATA OR * TO WRITE DATA HERE FORMAT SPECIFICATION OR * TO LIST DATA WITH HEADINGS WRITE OBSERVATIONS JARIABLES BY CASE 1. 2. AA R 19193..75839 15502..61254 7381.9.29169 2030.0 2.2480 1328.7 1.1004 1722.4 1.4265 6 CASES WRITTEN FOR CASES=CASE: 1-6 3. 4..U Q 1100.0 0. 1100.0 0. 1100.0 0. 305.00 5.0000 305.00 5.0000 305.00 5.0000 4 VARIABLES COMMA ND?DESCRIBE V:AA CASES=ALL'ESCRIPTIVE MEASURES VARIABLE N MINIMUM I.AA 327 -73P1.R MAXIMUM 22146. MEAN 6322.2 STD DEV 4641.2 COMMAND 7DIST V=AA CASES=ALL PROB=NOGRAPH:.25,.5.,75 DISTRIBUTIONAL ANALYSIS CUMULATIVE DISTRIBUTION OF I.AA N= 327 OUT OF 327 PROB QUANTILE.2500 2399.1.5000 5659.4.7500 9596.5 COMMA ND

313 HIST V:AA CASES:ALL OP=HISTZ INTERVAL EXPRESSION -- #INT:(MIN,MAX) (MIN,MAX)/WIDTH rPEP/(MIN,MAX) HISTOGRAM MIDPOINT HIST7. COUNT FOR 1.AA (EACH X: 2) -7381.8.3 1 +X -5741.4 0. 0 + -4101.0 0. 0 + -2460.6.3 1 +X -820.14.6 2 +X 820.28 15.9 52 +XXXXXXXXXXXXXXXXXXXXXXxXX 2460.7 16.2 53 +XXXXXXXXXXXXXXXXXXXX XXXXXY 4101.1 10.7 35 +XXXXXXXXXXXXXXXXXX 5741.5 9.5 31 +XXXXXXXXXXXXXXXX 7381,9 15.0 49 +XXXXXXXXXXXXXXXXXXXXXXXXX 9022.4 9.5 31 +XXXXXXXXXXXXXXXX 10663. 7.6 25 +XXXXXXXXXXXXX 12303, 6.1 20 +XXXXXXXXXX 13944. 2.8 9 +XXXXX 15584. 3.1 10 +XXXXX 17224 1.5 5 +XXX IFF65..6 2 +X 20505. 0. 0 + 22146..3 1 +X TOTAL 327 (INTERVAL WIDTH= 1640.4) COMMAND 7REGRESS V=AA;U CASES=ALL OPTION=MEANZERO LEAST SQUARES REGRESSION ANALYSIS OF VARIANCE OF i.AA N= 327 OUT OF 3^7 SOURCE DF SUM SaRS MEAN SQR F-STAT SI^NIF REGRESSION 1.16705+11.16705+11 1607.R C. ERROP 326.33571+10.10390 +P TOTAL 327.20092+11 OPT: MEANZERO R-SQR=.57491 SE= 3223.4 VARIABLE PARTIAL COEFF STD EPRO? T-STPT SIGNIF 3.U.91182 8.6755.21636 40.09F 0. tOMMAND

314 REGRESS V=AA;U CASES=ALL LEAST SQUARES REGRESSION ANALYSIS OF VARIANCE OF l.AA N= 327 OUT OF 327 SOURCE DF SUM SQRS MEAN SMn F-STAT SINIF RERRESSInN 1.40371+10.40371+10 A39.55.0000 ERROR 325.29851+10.91P4E +7 TOTAL 326.70222+10 MULT R=.75R23 R-SQR=.57491 SE= 3030.6 VARIABLE PARTIAL COEFF STD ERROR T-STAT SI^NIF CONSTANT -3204.5 4P4.32 -6.6164.0000 3. U,75P23 12.325.5F787 20.965.0000 COMMAND 7

315 7SCATTER V=AA U CASES=ALL SCATTER PLOT N= AA 22146. + 327 OUT OF 327 1.AA VS. 3.U + 16240. + 10335. + 4429.2 +* 2 + -1476.5% + 4 * 2 * 6 5 X X X * 3(k 2 3 4 2 6 P 7 5 6 5 2 3 2, 5 3 * X 3 * * C PX 7 7 3 7 2 4 X 6 5 4 3 5 z~ t -7351.F + * 305.00 623. 00 a!. 0 U 464.00 72.00 1100.0 COMMAND 7

316 TRANS TRANSFORMATION EXPRESSION (E.G., V9=V2+V5 OR =V20:SQRT(AA) LABELS=SRAA ERROP -- SYNTAX: "V20=SQRT(AA) LARELS:SRPAA ENTER NEW VALUE FOR TRANSFORMATION EXPRESSION VI OSQRT(VF) ):V20=SQRT(AA) LABEL FOR THE RESULT VARIABLE(S) =SRAA VIO:SQRT(VF) ) (E.".,. V9V2-+V 0' SQRT TRANSFORMATION VARIABLE TOTAL VALID 20.SRAA 327 324 COMMAND?WRITE V=SALL CASES=1-6 FILE TO RECEIVE DATA OR * ORMA SPEC CA OR FORMAT SPECIFICATION OR * MISS 3 TO WRITE DATA HERE TO LIST DATA WITH HEADI'NS WRITE OBSERVATIONS VARIABLES BY CASE io. o2. AA R 19193..75R39 15502..61254 7381.9.2q169 2030.0 2.240O 132P.7 1.1004 1722.4 1.4265 6 CASES WRITTEN FOR CASES=CASE: 1-6 3. 4. U Q 1100.0 0. 1100,0 0. 1100.0 0. 305.00 5.0000 305.00 5.0000 305.00 5.0000 5 VARIABLES 20. SRM. 13P.54?P4.5 I 05 1* l 45.05e 36 o45 41.502 COMMAND?DESCRIBE V:SRAA CASES=ALL DESCRIPTIVE MEASURES VARIABLE N MINIMUM 20.SRAA 324 0, MAX IMUM 14P.P 1 M7EA 1 73.991 STr DEV 32. < U COMMAND

317 HIST V=SRAA CASES=ALL OP=HIST. INTERVAL EXPRESSION -- #INT:(MIN,MAX) (MIN,MA X)/W!DTH f PER/ (M!, MAX) HISTOGRAM MIDPOINT HIST% COUNT FOR 20.SRAA (EACH X= 1) 0..3 I +X 8.2675.6 2 +XX 16.535 3.4 11 +XXXXXXXXXXX 24.g02 4.9 16 +XXXXXXXXXXXXXXXX 33.070 3.4 11 +XXXXXXXXXXX 41.337 7.7 25 +XXXXXXXXXXXXXXXXXXXXXyXXY 49.605 9.0 29 +XXXXXXXXXXXXXXXXXXXxXXXXXX XX 57.872 7.7 25 +XXXXXXXXXXXXXXXXXXXXXXXXX 66.140 6.5 21 +XXXXXXXXXXXXXXXXXXXXX 74.407 9.3 30 +XXXXXXXXXXXXXXXXXXXXXXXXXXYXXX 82.675 8.3 27 +XXXXXXXXXXXXXXXXXXXXXXXXX'XX 90.942 12.0 39 +XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXYXX 99.209 F.6 28 +XXXXXXXXXXXXXXXXXXXXXXXXXX 137.48 7.1 23 +XXXXXXXXXXXXXXXXXXXXXXX 115 74 5.2 17 +XXXXXXXXXXXXXXXXX 124.01 3.A 11 +XXXXXXXXXXX 132.28 1.5 5 +XXXXX 140.55.6 2 +XX 148.81 3 1 +X MISSING 3 TOTAL 327 (INTERVAL WIDTH: F.2675) COMMAND 7DIST V:=RSRAA CASES=ALL PROB=NOGRAPH;.25,.5,.75 DISTRIBUTIONAL ANALYSIS CUMULATIVE DISTRIBUTION OF 20.S!AA N= 324 OUT OF 3?7 PROB QUANTILE.2500 49.605.5000 76.847.7500 97.962 COMMAND?

318?REGRESS V:SRAA;ll CASES=ALL OP:MEANZERO LEAST SPUARES REGRESSION ANALYSIS OF VARIANCE OF 20.SRAA N= 324 OUT OF 327 SOURCE DF SUM SORS MEAN SAP F-STaT REGRESSION 1.19PFO +7.lQO01 +7.527.6 ERROR 323 97979. 303.34 TOTAL 324.20781 +7 OPT: MEANZERO R-SQR=.SF 150 SE= 17.417 VARIABLE PARTIAL COEFF STD ERROR T-STAT 3.U.97614.94625 -1.1171I -2 P0.794 COMMAND? S!,IIF 0. SI GIF C.

319?SCATTER V=SRAA;U SCATTER PLOT N:,RAA 148.81 + 324 OUT OF 327 20.SRAA VS. 3.U l r 4 7 + 29.763 +* X +3 x 9 + 2 * 2 2 5 5 9 x X X X 2 * 2 A 3 7 5 4 7 5 5 2 4 2 2 * 2 * * 4 A 5 x.5 X * 6 9 2 3 3 A x 4 7 7 7 4 4 3 *k * 0. +* +-........4- -.-.. -.-4 — -+.-.-+-.4 — +- --- -4-...-*+- -— * 305.00 623.00 4 1.00 U 464.00 72.00 1 100.0 COMMAND 7

320 HIST V=RSRAA CASES 1-3 I=* HISTOGRAM CASES:CASE#: 1-3 MIDPOI NT COUNT FOR 20.SRAA (EACH X= 1) P5.918 1 +X 138.54 2 +XX TOTAL 3 (INTERVAL WIDTH: 52.621) COMMA ND?DIST V:SRAA CASES:!-3 PROB=NOG;APH;.25,.5,.75 DISTRIBUTIONAL ANALYSIS CASES=CASEA:1-3 CUMULATIVE DISTRIBUTION OF 20.SRAA N: 3 OUT OF 3 PROB QUANTILE.2500 85.918.5000 124.51.7500 138.54 COMMAND 7DESCRIBE V=SRAA CASES=1-3 DESCRIPTIVE MEASURES CASES:CASE;:I-3 VARIABLE N MINIMUM MAXIMUM MEAN STD DEV 20.SRAA 3 PF591F 13.o54 116.3.5 r7.24, COMMA ND?

321 REGRESS V:SRAA U CASES=1-3 OP=MEANZERO LEAST SQUARES REGRESSION CASES=CASE-:1-53 ANALYSIS OF VARIANCE OF 20.SRAA N= 3 OUT OF 3 SOURCE DF SUM SORS MEAN SPR F-STAT REGRESSION I 40592. 40592. 54.670 ERROR 2 1485.0 742,45 TOTAL 3 42077. OPT: MEANZERO R-SQR=.00000 SE: 27.249 VARIABLE PARTIAL COEFF STD ERROR T-STAT 3.U.98220.10575.14302 -1 7.3939 COMMAND?SCATTER V=SRAA;U CASES=1-5 SCATTER PLOT CASES:CASE#:1-3 INVALID INTERVAL FOR 3.U: 1100.0, 1100.0 N= 0 OUT OF 3 20.SRAA VS. 3.U COMMA ND? ST!NIF.01 7.01 7F -

322 SCATTER V=SRAA;U CASES=4-58 SCATTER SRAA 130.30 PLOT CASES=CASE#:4-5P N= 54 OUT OF 55 20.SRAA VS. 3.U + 3a ok + 104.24 + 78.181 + * 2 3 * 2 + 52.120 + 2 A 4 2 3 * 2 * 26.060 +4 2 0. +* 4..-.+.-.- 4.-.-+I,-.-..+-... —,+ -. —+ — 4 —-- 305.00 623.00 ~4 1.00 U 464.00 7 2.00 1100.0, COM1MAND

323 HIST V=SRAA CASES=4-58 OP=HISTZ I:= HISTOGRAM CASES=CASE,: 4-5P MIDPOINT HISTZ COUNT FOR 20.SRAA (EACH X= 1) 0. 1.9 1 +X Il.61A4 14, P +XXXXXXX 37.229 1, 5 10 +XXXXXXXXXX 55.43 14 +XXXXXXXX 74.45e 11.1 6 +XXXXXX 93.072 25.9 14 +XXXXXXXXXXXXX 111.69 7.4 4 +XXXX 130.30 5.6 3 +XXX MISSI NG I TOTAL 55 (INTERVAL WIDTH= 19.614) COMMAND 7DIST V:SRAA CASES=4-54 PROB=NOGRAPH;.25,.5,.75 DISTRIBUTIONAL ANALYSIS CASES=CASE:4-5? CUMULATIVE DISTRIBUTION OF 20.SRAA N= 54 OUT OF 55 PROB QUANTILE.2500 36.452.5000 64.677.7500 94. 11 COMMA ND

324 DESCRIBE V-SRAA =S.. RAAA_ CASES=4-58 DESCRIPTIVE MEASURES CASES=CASE-#:4-5F VARIABLE N MINIMUM MAXIMUM MEA N STD DEt 20.SRAA 54 0. 130.30 5,p50 33.,73 COMMA ND 7REGRESS V=SRAA:U CASES:AL_4-5F OP:MEANZERO LEAST SQUARES REGRESSION CASES=CASEa.4-5P ANALYSIS OF VARIANCE OF 20.SRAA N: 54 OUT OF 55 SOURCE DF SUM SQRS MEAN SQO F-STAT S!T;IF REGRESSION 1.27783 +6.277B3 +6 1029.?.OGOrf ERROR 53 14299. 269.79 TOTAL, 54.29213 +6 OPT: MEANZERO R-SQR=.760gO SE= 16.425 VARIABLE PARTIAL COEFF STD ERROR T-STAT SIGNIF 3.U.97522.95737 -I.FpA34 -2 32.090.0000 COMMAND 7

325 SCATTER V=SRAA;U CASES=59-130 SCATTER SRAA 124.51 PLOT CASES-CASE:59-130 N= 72 OUT OF 72 20.SRAA VS. 3.U 4. * 101.32 + /.,,' +. 78.141 + / 2 2 2 -/ * * 2 // 2t* m4 3 2 2 2 3 2 3 3 2 *c *,I 31.775 + +3 8.591 0 + - 305.00 623.00 041.00 U 464.00 782.00 1100.0 COMMA D 7DESCRIBE V=SRAA CASES=59-130 DESCRIPTIVE MEASURES CASES=CASE:59-130 VARIABLE N MINIMUM MAXIMUM 20.SRAA 72 8.591F 124.51 MEAN 77.396 STD DEV 30.637 COMMAND?

326 HIST V=SRAA CASES=59-130 OP=HIST2 I=* HISTOGRAM CASES=CASE#:59-103 MIDPOINT HIST: COUNT FOR 20.SRAA (EACH X= 1) 8.591P 2.R 2 +XX 23.0 1 6.9 5 +XXXxr 37.571 4.2 3 +XXX 52.00O 1.1 13 +XXXXXXXXXXXXX 66.549 6.9 5 +XXXXX 1.039 12.5 9 +XXXXXXXXX 95.52E 23.6 17 +XXXXXXXXXXXXXXXXX 110.02 22.2 16 +XXXXXXXXXXXXXXXX 124.51 2. 2 +XX TOTAL 72 (INTERVAL WIDTH= 14.4P9) COMMAND?DIST V=SRAA CASES=59-130 PROR=NOGRAPH;.25,.5,.75 DISTRIBUTIONAL ANALYSIS CASES=CASE#:59-130 CUMULATIVE DISTRIBUTION OF 20.SRAA N= 72 OUT OF 72 PROB QUANTILE.2500 52.026.5000 P5.91F.7500 101.66 COMMA ND

327 REGRESS V=SRAA;U CASES:59-130 OP=MEANZERO LEAST SQUARES REGRESSION CASES=CASE#:59-130 )NALYSIS OF VARIANCE OF 20.SRAA N: 72 OUT OF 72 SOURCE DF SUM SQRS MEAN SOR F-STAT SIG IF REGRESSION 1.4P093 +5.4093 5+ 2007.4.0000 ERROR 71 17010. 239.5s TOTAL 72.49794 +6 OPT: MEANZERO R-SQR=.75104 SE= 15.478 VARIABLE PARTIAL COEFF STD ERROR T-STAT STGNIF 3.U.98277.10040.2240P -2 44.F04.0000 COMMAND

328 SCATTER V=SSR_SRAA;U CASES:131-198 SCATTER PLOT CASES=CASE#: 131- 1F N= 66 OUT OF 68 20.SRAA VS. 3.U jRAA 13P.54 + 2 + * * 118.19 + * * * 3 4- * * 2 * 97.58 +* * * 2 * * * * + 3 * * * 57.13P +* 520~00 752.00 s=4. 00 U 3 2 2 2 36.7PP 4 520o00 752-.00 4. 00 U 636.00 P68.00 1100.O COMMAND 7? 7DESCRIBE V=SRAA CASES=131-198 DESCRIPTIVE MEASURES CASES:CASE:- 131-19F VARIABLE N MINIMUM MAXIMUM MEAN ST1r DEV 20.SRAA 66 36,78P 138.54 P4.547 27.p25 COMMAND

329 RE#?HIST V:SRAA. CASES:131-19 OP:HISTZ I=* HISTOGRAM CASES=CASE: 131-19? MIDPOINT HIST. COUNT FOR 20.SRAA (EACH X= 1) 36.78p 10.6 7 +XXXXXXX 49.507 10.6 7 +XXXXXXX 62.226 13.6 9 +XXXXXXXXX 74.944 6.1 4 +XXXX 87.663 12.1 F +XXXXXXXX 100.38 22.7 15 +XXXXXXXXXXXXXX 113.10 15.2 10 +XXXXXXXXXX 125. 2 7.6 5 +XXXXX 158.54 1.5 1 +X MISSING 2 TOTAL 68 (INTERVAL WIDTH: 12.719) COMMAND?DIST V:SRAA CASES:131-198 PROB=NOGRAPH;.25,.5,.75 DISTRIBUTIONAL ANALYSIS CASES:CASE#: 13 1- 1 9F CUMULATIVE DISTRIBUTION OF 20.SRAA N= 66 OUT OF 6F PROB QUANTILE.2500 66.552.5000 F5.918.7500 105.23 COMMA ND?REGRESS V=SRAA;U CASES:131-198 OP=MEPNZERO LEAST SQUARES REGRESSION CASES=CASE 11 - 19, ANALYSIS OF VARIANCE OF 20.SRAA N: 66 OUT OF 68 SOURCE DF SUM SQRS MEAN SQR F-STAT SIVNIF REGRESSION 1.50215 +6.50215 +6 1i35.1.0000 ERROR 65 19962. 307.10 TOTAL 66.52211 +6 OPT: MEANZERO R-SQR:.60911 SE= 17.524 VARIABLE PARTIAL COEFF STD ERROR T-STAT SIGNIF 3.U.98070.10115.25016 -2 40.437.0000 COMMAND

330 7?SCATTER V=SRAA CASES=199-250 ERROR -- WRONG I OF VARS:'SRAA' ENTER NEW VALUE FOR VARIABLES -- VERTICAL; HORIZONTAL SRAA;U SCATTER PLOT CASES=CASE# 199-250 N= 52 OUT OF 52 20.SRAA VS. 3.U SRAA 130.30 + 2 + 108.45 + 86.599 + 64.748 + * * * 2 3 2 2 3 2 2 * 42.897 + * 2 2 3 * * +2 * 4 21.046 +2 +4. — 4 — - _+ —-_+... —_ —+_ + —_ —--------..-_ 305 00 623.00 4 1. 00 U 464.00 7F2.00 1100,0 COMMAND DESCRIBE VSRAA CASES19925?D~SCRIBE V=SRAA CASES=199-250 DESCRIPTIVE MEASURES CASES=CASE: 199-250 VARIABLE N MINIMUM MAXIMUM 20.SRAA 52 21.046 130.30 MEAN 71.219 STD DEV 31.705 COMMAND

331?i?HIST V:SRAA CASES=199-250 OP=HISTZ HISTOGRAM MIDPOINT 21.046 36.653 52.261 67 869 3. A477 99.0F5 114.69 130.30 TOTAL CASES=CASE#: 1 99-250 HIST% COUNT FOR 20.SRAA 13 5 7 +XXXXXXX 9.6 5 +XXXXX 15 4 +XXXXXXXX 13.5 7 +XXXXXXX 17.3 9 +XXXXXXXXX 17.3 9 +XXXXXXXX 5 53 +XXX 7.7 4 +XXXX 52 (INTERVAL W (EACH X= 1) IDTH= 15.60) COMMAND?DIST V=SRAA CASES=199-250 PROB=NOGRAPH;.25,.5,.75 DISTRIBUTIONAL ANALYSIS CASES=CASE#:199-250 CUMULATIVE DISTRIBUTION OF 20.SRAA N= 52 OUT OF 52 PROB QUANTILE.2500 47.059.5000 71.8F4.7500 94. 11 COMMAND 7

332?#?REGRESS V=SRAA;U CASES_199-250 OP:MEANZERO LEAST SQUARES REGRESSION CASES=CASE'-199-250 ANALYSIS OF VARIANCE OF 20,SRAA N= 52 OUT OF 52 SOURCE DF SUM SQRS MEAN SQR F-STAT STIrI* REGRESSION 1.301 +G.030OIt +6 1173.1.0000 ERROR 51 13124. 257.34 TOTAL 52.31502 +6 OPT: MEANZERO R-SQR=.74400 SE= 16.042 VARIABLE PARTIAL COEFF STD ERROR T-STAT SI' IF 3.U.97895.91948 -1.26;45 -2 34.251.CGOO COMMAND

333 7??SCATTER V=SRAA CAS t;U CASES=251-294 SCATTER PLOT CASES=CASE#:251-294 N= 44 OUT OF 44 20.SRAA VS. 3. U iRAA 127.44 + + 107.38 + + * 3 * 27.330 + 3 *DESCRIBE V:SRAA CASES:251-294 + * * * * * 6 2 67.277 + * * * +* * * 2 47.223 + +2 27.170 + * 520.00 752.00 9p4.0no U 6356.00 P68.00 OC1100.0 COMMAND?DESCRIBE V=SPAA CASES-251-294 DESCRIPTIVE MEASURES CASES=CASE#:251-294 VARIABLE N MINIMUM MAXIMUM MEAN STD DE" 20.SRAA 44 27.170 127.44 73.952!1 COMMAND

334 7??HIST V=SRAA CASES=251-294 OP=HISTZ HISTOGRAM MI DPOI NT 27.170 453.81 60.592 77.303 94.014 110.73 127.44 TOTAL CASES: CASE:25 1-294 HISTZ COUNT FOR 20.SRAA (EACH X= 1) 4.5 2 +XX 11.4 5 +XXXXX 13.6 6 +XXXXXX 3E.6 17 +XXXXXXXXXXXXXXXXX 25.0 11 +XXXXXXXXXXX 4.5 2 +XX 2.3 1 +X 44 (INTERVAL WIDTH= 16.711) COMMAND 7DIST V:SRAA CASES:* PROB=NOGRAPH;: * ERROR -- INVALID CONSTANT: "NOGRAPH:*' ENTER NEW VALUE FOR PROBABILITY POINTS _NOGRAPH;.5,.25,.75 DISTRIBUTIONAL ANALYSIS CASES=CASE#:251-294 CUMULATIVE DISTRIBUTION OF 20.SRAA N= 44 OUT OF 44 PROB QUANTILE.5000 71.884.2500 60.753.7500 FS.040 COMMAND 7

335?REGRESS V=SRAA;U CASES=251-294 OP:MEANZERO LEAST SQUARES REGRESSION CASES=CASE#Z25 1-294 ANALYSIS OF VARIANCE OF 20.SRAA N= 44 OUT OF A4 SOURCE DF SUM SQPS MEAN SnR r-STAT SIGNIF REGRESSION I,.4F16 +6.24816 +G o00.49.o000 ERROR 43 11733. 272.06 TOTAL 44.259F9 +6 OPT: MEANZERO R-SQR=.42784 SE= 16.51F VARIABLE PARTIAL COEFF STD ERROR T-STAT SIGOIF 3.U.97717.83657 -1.27740 -2 30.15P.000CC COMMAND?

336?7?SCATTER V=SRAA CASES=295-302 ERROR - WRONG # OF VARS: "SRA A ENTER NEW VALUE FOR VARIABLES -- VERTICAL; HORIZONTAL SRAA,_; U SCATTER PLOT CASES:CASE:295-302 INVALID INTERVAL FOR 3.U: 305.00, 305.00 N= 0 OUT OF P 20.SRAA VS. 3.U COMMAND?DESCRIBE V=SRAA CASES:* DESCRIPTIVE MEASURES CASES:CASE:295-302 VARIABLE N MINIMUM MAXIMUM MEAN STD DEV 20.SRAA 8 9,0565 42.959 24.f30 11.166 COM MAND?DIST V=SRAA CASES:* PROB:NOGRAPH.25,.5,,.75 DISTRIBUTIONAL ANALYSIS CASES:CASE: 295-302 CUMULATIVE DISTRIBUTION OF 20.SRAA N= F OUT OF F PROB QUANTILE.2500 16.384.5000 22.184.7500 25.616 COMMAND?

337?HIST V:SRAA CASES=* OP=* I=* HISTOGRAM CASES=CASE:295-302 MIDPOINT COUNT FOR 20.SRAA (EACH X= 1) 9.0565 2 +XX 26.00F 4 +WXXX 42.!59 2 +XX TOTAL g (INTERVAL WIDTH= 16.951) COMMAND?REGRESS V=SRAA;U CASES=* OP=MEANZERO LEAST SQUARES REGRESSION CASES=CASE#:295-302 ANALYSIS OF VARIANCE OF 20.SRAA N= B OUT OF F SOURCE DF SUM SORS MEAN SQR F-STAT SIGMIF REGRESSION 1 4932.1 4932.1 39.557.0004 ERROR 7 S72.77 124.68 TOTAL S 504.B OPT: MEANZERO R-SQR=.00000 SE= 11.166 VARIABLE PARTIAL COEFF STD ERROR T-STAT SI'NIF 3.U.92176.A1409 -1.12944 -1 6.2R95.0004 COMMAND

338 SCATTER V=SRAA CASES=303-327 ERROR -- WRONG 4 OF VARS: * SRAA' ENTER NEW VALUE FOR VARIABLES -- VERTICAL HORIZONTAL JRAA;U SCATTER SRAA 148.l 1 PLOT CASES CPSE#:303-327 N: 25 OUT OF 25 20.SRAA VS. 3.U 4 * 126.07 + 103.32 + + * * 80.571 + * + *c 57.2 24 +* 2 2* 2 * * * 2 * 35.076 +* 4-....+....-I-.- - - -+...+,....~-4-....+ —.... -.....+,....q..... 4- 520.00 752.00 aP4.00 U 636.00 868.00 1100.0 COMMA ND?DESCRIBE V=SRAA CASES=* DESCRIPTIVE MEASURES CASES=CASE#:303-327 VARIABLE N MINIMUM MAXIMUM 20.SRAA 25 35.076 148.RI COMMAND 7 MEA N 70.386 STD DEV 2.3 15

339 DI_HIST V:SRAA CASES-* OP=HISTZI=__ I=* HISTOGRAM CASES=CASE,:3033-27 MIDPOINT HISTZ COUNT FOR 20.SRAA (EACH X= 1) 35.076 16.0 4 +XXXX 57,824 44.0 11 +XXXXXXXXXXX 80.571 20.0 5 +XXXXX 103.32 12.0 3 +XXX 126.07 4.0 1 +X 148.81 4.0 1 +X TOTAL 25 (INTERVAL WIDTH= 22.748) COMMAND?DIST V=SRAA CASES=* PROB=NOGRAPH;.25,.5,.75 DISTRIBUTIONAL ANALYSIS CASES=CASE:3053-327 CUMULATIVE DISTRIBUTION OF 20.SRAA N= 25 OUT OF 25 PROB QUANTILE.2500 48.981.5000 60.753.7500 F3.742 COMMAND?

340?7#?REGPESS V=SRAA;U CASES=* OP:MEANZERO LEAST SQUARES REGRESSION CASES=CASES:303-327 ANALYSIS OF VARIANCE OF 20.SRAA N= 25 OUT OF 25 SOURCE DF SUM SMRS MEAP SOP F-STAT SIrNIF REGRESSION I.13409 +6.1340o +6 357.18.0000 ERROR 24 9009.8 375.41 TOTAL 25.14310 +; OPT: MEANZERO R-SQR=.53178 SE= 19.375 VARIABLE PARTIAL COEFF STD ERROR T-STAT SIr.NIF 3.U.96801.85139 -1.45049 -2 F1.P99.0000 COMMAND 7

341 APPENDIX E COMPARISON OF EXPERIMENTAL DROPLET ACCELERATIONS WITH THOSE USING SERAFINIf ABRAHAM AND CHURCHILL'S EMPIRICAL CORRELATIONS FOR THEE DRAG COEFFICIENT, C VARIABLE NAME AA (M) AABAR AASM ABAR (M) AC (M) ACBAR ACCSM ACEL (J) ACSM AS (M) DESCRIPTION OF VARIABLE Analytical acceleration as calculated by Abraham Average of all Abraham accelerations for a particular steam velocity and flow rate Sum of all Abraham accelerations for a particular steam velocity and flow rate Average of the measured accelerations for one droplet Analytical acceleration as calculated by Churchill Average of all Churchill accelerations for a particular steam velocity and flow rate Sum of all the measured accelerations for a particular steam velocity and flow rate Measured acceleration of one droplet for three successive film readings Sum of all the Churchill accelerations for a particular steam velocity and flow rate Analytical acceleration as calculated by Serafini ASBAR ASUM CDA (M CDC C(T) CDS CM) CON Average of all measured accelerations for a particular steam velocity and flow rate Sum of all the measured accelerations for one droplet Abrahaml's expression for the drag coefficient Churchill's, expression for the drag coefficient Serafini's expression for the drag coefficient Constant conversion factor, 30.48, the number of cm/sec in one ft/sec.

342 VARIABLE NAME CONFAC DELV (J DELX (J DIA(J) DPBAR DPSM DPSUM DRPU (M) DTAVG FILMV FIVNIN FN(I) IM2 M Q QCHECK RA CM RABAR DESCRTPTION OF VARIABLE Constant conversion factor,.45, the number of centimeters in one measured screen unit The difference between two consecutively measured velocities The difference between two consecutively measured distances The diameter of the water droplet for a particular observation The average of all the diameters for a particular steam velocity and flow rate The sum of all the diameters for a particular steam velocity and flow rate The sum of all the diameters observed for a particular droplet The average of all diameters observed for a particular droplet The average time between each successive observation The film speed at which the pictures were taken Constant, which equals 5/9. Number of frames between each successive observation The number of observations in a set minus two A counter which checks how many different cases were observed for each steam velocity and flow rate Flow rate of water entering the test section A control that is used to see if the flow rate has changed The ratio of measured acceleration to Abraham acceleration, for a particular water droplet The average of all measured/Abraham acceleration ratios for a particular steam velocity and flow

343 VARIABLE NAME DESCRIPTION OF VARIABLE rate RASM The sum of all measured/Abraham acceleration ratios for a particular steam velocity and flow rate RBAR The average of all measured/Serafini acceleration ratios for a particular water droplet RC(M) The ratio of measured acceleration to Churchill acceleration for a particular water droplet RCBAR The average of all measured/Churchill acceleraration ratios for a particular steam velocity and flow rate RCSM The sum of all measured/Churchill acceleration ratios for a particular steam velocity and flow rate RE(M) Reynolds number for a particular water droplet REBAR The average of all Reynolds numbers for a particular steam velocity and flow rate RESUM The sum of all Reynolds numbers for a particular steam velocity and flow rate RHOW Density of water RS(M) The ratio of measured acceleration to Serafini acceleration for a particular water droplet RSUM The sum of all measured/Serafini acceleration ratios for a particular steam velocity and flow rate SEBAR Average of all Serafini accelerations for a particular steam velocity and flow rate SVEL Steam velocity in cr/sec. T (_I Thickness of the water droplet for a particular observation TD CJ_ Elapsed time between observations TOTHRD Constant, which equals 2/3 TTOL Total time that observations on a particular

344 VARIABLE NAME DESCRIPTIJON OF VARIABLE drop were made V Steam Velocity VEECK A control which checks if the steam velocity has changed VEL(J) Measured velocity of one droplet for two successive film readings X(I) Distance an individual particle was away from the edge of the blade

345 SSIG ODLU T=3 P=20 SB *FTN 5=YWATA C **~*~~SiI *~- ACCELERATICN BATIC EFCGRAM,**,iISS ~~l*f*ls C THIS PROGEBA CALCULATES THE M.EASURED ACCELERATION AND THHEE C ANALYTIC ACCELERATIONS:SERAFINI, CEURCHILL, AND ABRAHAf. C IT ALSO FINDS THE ACCELERATION RATICS IND DETERMINES AVERAGE C VALUES FOB EACH STEAt VELOCITY AND FLCW BATE. C **~as**t**as ***~***~g***se s**** *a*ss****~ ******s**s DIMENSION ACEL(20),FN(20),DIA (20),T (20),X 20),4ID(20),DELX (20), 1 VEL(20),DELV(20), ABAE (25), RPU (25),RE(25),CDS(25),.S(25),aS (25), 2 CDA(25),AA(25),RA(25),CDC(25),AC(25),C(25) WRITE(6,199) TOTHRD=2./3. ANIN=9./5. FIVNIN=5./9. CON=30. 48 READ(5,101) FILa~V,EHCW,CONFAC,VCHECK,CCHECK M=0.0 1 ASU.=O.0 TTOL=O. DPSU=O0.0 C'**"* READ THE DATA IOB EACH CBSERVAIICN ***t, DO 110 I=1,20 READ (5,101) FN (I),DIA(I),I), ()(I) DPSUM=DPSUM+DIA (I) IF(T(I).LT. 0.0) GO TO 120 IF(X(I).L7. 0.0) GC TO 130 110 CONTINUE 120 T(I)=ABS (T (I)) 130 X(I)=AES(X (I)).EAD(5,101) V,Q IF(I.EQ. 2) GO TO' C *~*** CHECK TO SEE IF THERE ARE NE CC:NDITIGOS **"* IF(V.NZ. VCHECK.OB.Q.NE.QCHECK) GO TO 160 100 o =a+1 DEPU (N)=DPSDi/I*. 45 C **,** CALCULATE THE MEASURED ACCELERCATICN.*.DO 150 J=2,I TD (J-1) =0.001 TTOL=T'OL+TD (J-1) DELX (J-1)= (X(J)-X(J- 1) )*0.45 VEL (J-1)=DELX (J-1)/TD (J-1) IF(J.EQ.2) GO TO 150 DTAVG=(TD (J-1) +TD (3-2) ) /2.0 DELV (J-2)=VEL (-1) -VEL (J-2) ACEL (J-2) =DELV (J-2)/2./UTAVG ASUM=ACEL (J-2) +ASUM 150 COKTI;UE IS2=I-2 ABAR (a)=ASUO/I52 SVEL=V*CON C ~**,, CALCULATE THE ANALYTIC ACCELERATICNS "'"-* R E()=V*DRPU ()) *32.61 CDS (N) =24./BE (") * (1. +0. 17.*R (N) **TCTHFD) ASs) () =(CDS (M) *0.75) *.0001135/DE U ( ) *SVEL**2 RS () = A A ()/ S ( S) CDA (1) =.292 ( (1.+ (9.06/ (2'RE (" ))**Z.5)) *'2) AA () = (CDA (M) 0O.75).001135/DRFD0 (3) *SVEL**2 RA (:) =ABAR (4) /nA (w) CDC () =24. / E E ( ) * ( (1.+ (. (I() /60) **IYVIN) *A'. I;)

346 AC ()= (CDC (,) *.75).C001135/DBP O(M) *SVEL**2 RC () =ABBL (") /AC (1) WBITE (6,200) V,Q, DBPU (B),RE (B),ABAR (), AS (t), BS (),AA (),RA (), 1 AC (),Rc(5) GO TO 1 C ***"* DETERMINE AVTERGIS'*"'* 160 DPSB=0.0 RESUO=O.O aCC$S-O. O SESUH=O.O AASB=0.0 ACSK=O. 0 HCSM=0.0 DO 170 N=1,B DPSM =DPS +DEPU (N) RESUE=BESUMt+RB (R) ACCSB=ACCSM+ABAR(N) SESUO=SESZS+AS (N) RSut S=R Su. + BS (N) AASB=AASM+A A (N) BASM=RASl+BA (N) ACSl=ACSM+AC (N) RCSB=BCS3+BC (N) 170 CONTINUE DPBAR=DPSB/3 BEBAB= RES Uof/M ASBAE=ACCS. /M SEBARB=SES-UM/ HBAR=RSUM/M AABAR= AASa1/a RAB A= RAS a/. ACBAB= ACS?I/ RCBAB=RCS/C WRITE (6,201) QCBECK,DPBAR,REREA, ASB.A,SIEiR,REAR,AAEAR, 1 kBEAR, ACBABR,CEA B=0. 0 VCHECK=V QCECK= Q IP(I.EQ.1) CALL EXIT GO TO 100 101 FORMAT (6 (10.5)) 199 FORlAT ('1,'STEA~',4X,'FLO',3X,'DRCP',4,'REYNOLDS',3X, 1'SEASURED',5X,'SERAFINI SERAFINI ABEAHAN ABRAHA~M',3X, 2'CHURCHILL',2X,'CBURCHLL'/' VELOCIT', 2X,'RATE DIAMETER', 3 3X,'NoUBER',2X,'ACCELERATION ACCELERATIC' 2X,'RATIO',3X, 4'ACCELERATION RATIC ACCELERATICN',3X,'BATIO'/' (FEET/S)', 5 I1,' (CC/M) (CI)', 1 X' (Ca/SIC**2) (Cl/SEC*-2)', 6 11X,' (C3/SEC*22)' 11X,' (C'/S!C**2)') 200 FORSAT ('O',1X,F6.1,3X,F4.1,2X,f6.4,3X,F8.3 lX,Ft11.3,3X, 1 F10.3,3X,6.3,3X,F10.3,3X,F6.3,2X,FO. 2, X,F6.3) 201 FORSAT ('0',', AVG FCB C^',F3o0t1XF6. 4,3X X F 1x,I.'11e.3,3., 1 F10.3,3XF6.3,3X,F10.3,3X,F6.3,2X,F10.2,4X,6.3) END SENDFILE $3 -LOAr 5=WDATA 3 EN F ILE SSIG

STEAl PLOW DROP iEYNOLDS VELOCITY RATE DIAlETER NUfMBER (FEET/S) (CC/M) (Cr.) 305.0 305.0 305.0 305.0 305.0 305,0 305.0 305.0 305.0 305.0 305.0 305.0 305.0 30 5. 0 30 5.0 305.0 305.0 **AVG FOR 305.0 305.0 305.0 305.0 305.0 305.0 305.0 305.0 305.0 305.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 0= 5. 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 0.0975 0.0900 0.0900 0. 0900 0.0956 0.0956 0. 1260 0. 1350 0. 1350 0.0643 0.0450 0. 1350 0.0900 0.0900 0.0900 0.0787 0.0900 0.0963 0.1237 0.0900 0. 09'.0 0.0900 0.0990 0. 1087 0. 1215 0. 1150 0. 0844 0. 110. 969.738 895. 143 895. 143 895. 143I 951.089 951.089 1253. 199 1342. 113 1342.713 639. 388 447.572 1342. 713 895. 143 895. 1143 895. 143 783. 250 895. 143 958. 202 1230.822 895. 143 904.657 893. 14J 984.657 100 1. 630 1208. 443 1143. 79J 839. 197 1 1)1. 71J MEASUPED ACCELERATION (C./SEC**2) 61874.953 40499.949 52499.957 30000.008 37499.977 37499.977 22499.969 11249.969 14999.965 40500.012 22499.984 6000.000 17999.973 44999.969 0.001 8035.691 19687.492 27549.848 123750.063 67499.938 37500.082 11249.973 11249.977 2249.981 0437.492 6428.551 11249.977 01881.13 SERAFINI ACCELERATION (Cn/SEC**2) 32969.531 36796.660 36796.621 36796.660 33859.242 33859.242 23213.598 21127.598 21127.598 58492.781 95965.688 21127.617 36796.660 36796.660 36796.660 44211.266 36796.660 37854.715 23791.879 36796.621 32286.652 36796.660 32286.652 28388.996 24395.758 26299.715 40208.293 2768J.238 RATIO 1.877 1. 101 1.427 0.815 1. 108 1. 100 0.969 0.532 0.710 0.692 0.234 0.284 0.489 1.223 0.000 0.182 0.535 0.782 5.201 1.834 1. 161 0.306 0. 348 0.079 0. 346 0.244 0.280 0.296 ABRA iAR ACCELERATION (C /S EC* *2) 32030. 258 35184.652 35184.633 35184.652 32767.145 32767. 145 23777.996 21962.758 21962.758 52493. 148 81025.875 21962.758 35184.652 35184.652 35184.652 41196. 277 35184.652 35778.723 24277.504 35184.633 31463. 141 35184.652 31463.164 28198.461 24797.535 26426.887 37965.684 27601.067 CHURCHILL RATIO ACCELERATION (CR/SEC* 2) 1.932 42732.91 1.151 46961.88 1.492 46961.81 0.853 46961.88 1.144 43720.39 1.144 43720. 39 0.946 31691.73 0.512 29267.38 0.683 29267.38 0.772 70244.75 0.278 108885.81 0.273 29267.40 0.512 46961.88 1.279 46961.88 0.000 46961.88 0.195 55033.64 0.560 46961.88 0.807 47797.92 5.097 32359.11 1.918 46961.81 1.192 41973.14 0.320 46961.88 0.358 41973.11 0.080 37602.10 0.340 33054.10 0.243 35232.34 0.296 50693.89 0.296 36803.90 RATIO 1. 448 0.862 1.118 0.639 0.858 0.958 0.710 0.384 0.513 0.577 0. o07 0.205 0.383 0.958 0.000 0.146 0.419 0.605 3.824 1.437 0.893 0.240 0.268 0.060 0.255 0.182 0.222 0.222 WL

**AVG FOR 305.0 105.0 305.0 305.0 305.0 305.0 305.0 305.0 305.0 305.0 **ATG FOR 305.0 305.0 305.0 305.0 305.0 305.0 305.0 305.0 **AVG FOB 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 Q=10. 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Q0=20. 40.0 40.0 40.0 40.0 40.0 40.0 40.0 40.0 Q=40. 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 0. 1042 0. 0900 0. 1600 0. 1000 0.1125 0. 1200 0. 0450 0.1181 0. 0900 0. 0950 0.0450 0.0976 0. 0900 0.0675 0. 1309 0.09175 0. 1294 0.0787 0. 1800 0. 1558 0. 1162 0. 0787 0. 0630 0. 0900 0. 0900 0. 0900 0. 0900 0.04150 0. 0900 0.0720 103b. 519 095. 1143 1591. J365 9914. 603 1118.928 1 193.523 447. 572 1174.816 895. 143 944. 873 447. 572 970. 358 895. 143 67 1. 357 1302.024 969.738 1 286. 767 783. 250 1790. 285 154 9. 285 1155. 981 1335. 378 1068. 302 1526. 145 1526. 146 152u. 14o 1526.1146 763. 073 1526. 146 1220. 917 28779. 754 17999.97 3 19285. 703 32 14 2. 848 44999.969 33750.000 19285.676 22499.969 13499.973 16071.383 14999.965 23453.523 56249.984 15000.008 199999-969 45000.016 18749. 988 11250.000 2500.012 8181.813 22116. 453 135000.063 52499.977 -22499 7.000 67499. 938 168749.875 112500.000 127500.000 112499. 938 112500. (000 30893.4 18 36796.660 1676 1.746 31844.957 27102.238 24813.371 95965.563 25353.141 36796.660 34165. 176 95965.563 42556. 480 36796.660 54679.961 2203J. 504 32969.531 22390.852 44211.227 14280.930 17384.426 30593.352 105488.875 143097.000 87936.500 87936.4 38 87936.438 87936.438 227116.563 87936.4 38 119215.313 1.010 0. 489 1. 151 1.009 1.660 1.360 0.201 0.887 0. 367 0.470 0. 156 0.775 1.529 0. 27 4 0.908 1.365 0.837 0. 254 0.175 0.471 0.727 1.280 0.367 -2.559 0.768 1.919 1.279 0.561 1. 279 0.9411 30256.324 35184. 652 18084.6 17 3 1095. 617 27 109.320 25156.207 81025. 813 25618.730 35184.652 33019.957 81025.813 39250.523 35184.652 49507.641 22753.770 32030.258 23064.613 411 196. 258 15821.797 18645.242 29775.500 109529.500 14 775.375 94000. 813 94000. 750 94000.750 94000. 750 210583.938 94000.750 1211441. 2iO 1.014 0.512 1.066 1.034 1.660 1.342 0.238 0.078 0.384 0.487 0. 185 0.779 1.599 0. 303 0.879 1.405 0.813 0.273 0.158 0.4 39 0.7 34 1.233 0.370 -2. 394 0.718 1.795 1. 197 0.605 1.197 0.926 40361.51 46961.88 24093.04 411480. 80 36145.03 33533.53 108885.88 34151. 73 46961.88 44059.30 108885.88 52515. 87 46961. 88 66219.50 30323.62 42732.91 30738.83 55033.70 21077. 14 24840.59 3974 1.00 145959.94 189064.31 125237.25 125237.06 125237.06 125237.06 281370.75 125237.06 161871.19 0. 760 0.383 0.800 0.775 1.2415 1.006 0. 177 0.659 0.287 0.365 0. 138 0.584 1.198 0.227 0.660 1.053 0.610 0.204 0.119 0.329 0.550 0.925 0.278 -1.797 0.539 1.3417 0.098 0. 453 0.898 0.695 00 O0

520.0 520.0 520.0 520.0 **AVG FOR 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 520.0 **ATG FOR 520.0 520.0 520.0 5.0 5.0 5.0 5.0 0Q 5. 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 0=10. 15.0 15.0 15.0 0. 0900 0. 0900 0.0900 0.0675 0. 0805 0.0900 0.0675 0.0525 0. 0900 0.0900 0. 0900 0. 0900 0.0900 0. 0900 0.0450 0. 0675 0.0750 0. 0900 0.0450 0. 0900 0.0900 0.0750 0.0750 0.0900 0.0600 0.0776 0. 0975 0.0810 0.0900 1526. 146 1526. 145 1526. 14o 1 144. 609 1364.725 1526. 146 114 4.60? 890. 252 1526. 146 1526. 146 1526. 146 1526. 14o 1526. 146 1526. 145 763. 073 1144. 609 127 1. 780 1526. 146 763. 073 1526. 145 1526. 146 1271. 788 1271. 788 1526. 145 1017. 431 1316.299 1653. 324 1373.532 1526. 14b 82499.1175 115000. 000 10 1249.9 38 84 114.938 75144.375 97499. 938 157500. i25 84374.938 172499.875 82500.000 157499.813 101249.938 90000.000 73124.938 82499.938 135000.000 270000.000 67500. 188 74999.938 90000.000 74999.938 50625.000 73124.938 50625.016 112499.938 104906.000 67500.000 119999.875 135000.000 87936. 438 87936.500 87936.4)8 130208.188 109893. 500 87936.438 130208. 188 183732.438 87936.438 87936.438 87936.438 87936.438 87936.438 87936.500 227116.563 130208. 188 112751. 688 87936.438 227116.563 87936.500 87936.438 112751.688 112751.688 87936.500 152984.563 117845.813 78860. 563 101512.500 87936.438 0.938 0.512 1. 151 0.64118 0.699 1. 109 1.210 0.459 1.962 0.938 1.791 1.151 1.023 0.832 0.363 1.037 2. 395 0.768 0.330 1.023 0.853 0. q449 0.649 0.576 0.735 0.983 0.856 1. 182 1.535 94000.750 94000.813 94000.750 130857.250 112783.813 94000.750 130857. 250 175494. 438 94000.750 94000.750 94000.750 94000.750 94000. 750 94000.813 210583.938 130857.250 115855.000 94000.750 210583.938 94000.813 94000.750 115855.000 115855.000 94000.813 150071.625 119500.750 85813.375 106043. 188 94000.750 0.878 0.479 1.077 0.645 0.671 1.037 1.204 0.481 1.835 0.878 1.676 1.077 0.957 0.778 0. 392 1.032 2.330 0.718 0.356 0.957 0.798 0.437 0.631 0.539 0.750 0.943 0.787 1.132 1.436 125237.06 125237.25 125237.06 17411458.31 150355.38 125237.06 1744518.31 234244. 19 125237.06 125237.06 125237.06 125237.06 125237.06 125237.25 281370.75 174458.31 154407.63 125237.06 281370.75 125237.25 125237.06 15411407.63 154407.63 125237.25 200170.06 159344.88 114320.63 141305.63 125237.06 0.659 0.359 0.808 0.484 0. 504 0.779 0.903 0.360 1.377 0.659 1.258 0.008 0.119 0.5811 0.293 0.771 1.749 0.539 0.267 0.719 0.599 0.328 0. 174 0. 404 0.562 0.708 0.590 0.849 1.078 520.0 15.0 0.0750 1271.788 202500.125 112751.688 1.796 115855.000 1.7U8 154407.63 1.311

520.0 15.0 0.0900 520.0 15.0 0.0900 520.0 15.0 0.1012 520.0 15.0 0.0900 520.0 15.0 0.0900 520.0 15.0 Q.0900 520.0 15.0 0.0900 520.0 15.0 0.0900 520.0 15.0 0.0750 520.0 15.0 0.0787 520.0 15.0 0.1350 520.0 15.0 0.0900 520.0 15.0 0.1012 520.0 15.0 0.0900 520.0 15.0 0.0900 520.0 15.0 0.0750 520.0 15.0 0.1012 520.0 15.0 0.0900 520.0 15.0 0.0900 **AVG FOB 0=15. 0.0909 520.0 20.0 0.0900 520.0 20.0 0.0900 520.0 20.0 0.0900 520.0 20.0 0.0900 520.0 20.0 0.0900 520.0 20.0 0.0900 520.0 20.0 0.0900 520.0 20.0 0.0900 520.0 20.0 0.0900 *'AVG FOR Q=20. 0.0900 1526. 145 54000.012 1526. 145 90000.000 1716.913 52499.977 1526.146 89999.938 1526.146 101250.063 1526.146 90000.000 1526.145 44999.938 1526.145 58499.961 1271.7d8 -90000.125 1335.378 157499.938 2289.219 146249.938 1526. 146 270000.000 1716.914 -11250.000 1526.146 74999.938 1526.145 49500.000 1271.788 73124.938 1716.913 41249.969 152b.145 44999.957 1526.145 44999.969 1541. 627 82939.813 1526. 146 82499.938 1526.145 71249.938 1526.145 84374.938 1526.145 78749.938 1526.145 44999.973 1526.145 53999.973 1526. 145 67499.938 1526. 146 67500.000 1526. 14b 74999.938 1526.145 69541.563 87936. 500 87936.500 74915. 188 87936.438 87936.438 87936.438 87936.500 87936. 500 112751.688 105488.875 50697.652 87936. 438 74915.125 87936. 438 87936. 500 112751.688 74915. 188 87936.500 87936. 500 88814.250 87936. 438 87936. 500 87936. 500 87936.500 87936.500 87936.500 87936.500 87936.438 87936.418 87936.438 0.614 1.023 0.701 1.023 1. 151 1.023 0.5 12 0.665 -0.798 1.493 2.885 3.070 -0. 150 0.853 0.563 0.649 0.551 0.512 0.512 0.966 0.938 0.8 10 0.959 0.896 0.512 0.614 0. 768 0.768 0.853 0.791 9ql000.813 94000.813 82215. 125 94000.750 94000.750 94000.750 94000.813 94000.813 115855.000 109529.500 59468.723 94000.750 82215.063 94000.750 94000.813 115855.000 82215.125 94000.813 94000.813 94655.125 94000.750 94000.813 94000.813 94000.813 94000.813 94000.813 94000.813 94000.750 94000. 750 94000.750 0.574 125237.25 0.957 125237.25 0.639 109524.88 0.957 125237.06 1.077 125237.06 0.957 125237.06 0.479 125237.25 0.622 125237.25 -0.777 154407.63 1.438 145959.94 2.459 79233.44 2.872 125237.06 -0.137 1095211.69 0.796 125237.06 0.527 125237.25 0.631 154407.63 0.502 109524.88 0.479 125237.25 0.479 125237.25 0.897 126117.13 0.878 125237.06 0.758 125237.25 0.898 125237.25 0.838 125237.25 0.179 125237.25 0.574 125237.25 0.718 125237.25 0.718 125237.06 0.798 125237.06 0.740 125237.06 0.431 0.719 0.479 0.719 0.808 0.719 0.359 0.467 -0.583 1.079 1.846 2. 156 -0. 103 0.599 0.395 0.474 0.377 0.359 0.359 0.673 0.659 0.569 0.674 0.629 0.359 0.43 1 0.539 0.539 0.599 0.555 LJ 0

520.0 30.0 0.0900 1526. 145 76499.938 520.0 30.0 0.0450 76J.13 112499.875 520.0 30.0 0.0990 167U. 760 90000.000 520.0 30.0 0.0900 1526.146 97500.000 520.0 30.0 0.0900 1526.146 82500.000 520.0 30.0 0.0900 1526. 145 29999.945 520.0 30.0 0.0771 IJ08.125 44999.913 520.0 30.0 0.0900 152b6.145 44999.977 520.0 30.0 0.0900 1526. 145 56249.969 520.0 30.0 0.0900 1526.146 157500.000 **IVG FOB 0=30. 0.0851 1443.297 79274.938 520.0 50.0 0.0900 1526. 146 59999.977 520.0 50.0 0.0900 1526.145 73124.938 520.0 50.0 0.0975 1653.324 106874.938 520.0 50.0 0.0900 152b.145 44999.984 520.0 50.0 0.0900 1526. 145 67499.875 520.0 50.0 0.0900 1526.145 58499.973 520.0 50.0 0.0900 1526.14qb 73124.938 520.0 50.0 0.0731 1239.993 37499.988 **IVG FOR Q=50. 0.0888 1506.273 65203.055 825.0 5.0 0.1050 2824.838 450000.000 825.0 5.0 0.0900 2421.290 337499.750 825.0 5.0 0.0562 1513.306 247499.813 825.0 5.0 0.0450 1210.646 269999.750 825.0 5.0 0.1050 2824.838 247500.250 825.0 5.0 0.0600 1614.194 292500.250 825.0 5.0 0.1050 2824.0d3 225000.000 825.0 5.0 0.0750 S 017.741 270000.000 825.0 5.0 0.1050 2H24.8Jd 247499.625 #*AVG FOR Q= 5. 0.0829 2230.723 287499.750 87936.500 0.870 227116.563 0.495 77239.813 1.165 87936.438 1.109 879 36. 4 38 87936.500 108498.625 87936.500 87936.500 87936.438 1020841.000 87936.438 87936.500 78860.563 87936.500 87936.500 87936.500 87936.500 116717.438 90399.563 152311.813 187645.250 355236. 188 481610.813 152311.813 325359.813 152311.813 240260.813 152311.813 0.938 0.34 1 0.4 15 0.512 0.640 1.791 0.828 0.682 0.832 1.355 0.512 0.768 0.665 0.832 0.321 0.746 2.954 1. 799 0.697 0.561 1.625 0.899 1. 4177 1.124 1.625 94000. 813 210583.938 84338. 313 94 000.750 94000.750 94000.813 112157.375 94000.813 94000.813 94000.750 106508.500 94000.750 94000.813 85813.375 94000.813 94000.813 94000. 813 94000.813 119287.125 96138. 125 187948.688 223069.188 379027.438 489730.063 187948.688 352153.813 187948.688 27J602.000 187948.688 0.814 0.534 1.067 1.037 0.878 0.319 0.401 0.479 0.598 1.676 0.780 0.638 0.778 1. 245 0.479 0.71 8 0.622 0.778 0. 314 0.697 2.394 1.513 0.653 0.551 1.317 0.831 1. 197 0.987 1.317 125237.25 281370.75 112354.56 125237.06 125237.06 125237.25 149469.06 125237.25 125237.25 125237.06 141985.38 125237.06 125237.25 114320.63 125237.25 125237.25 125237.25 125237.25 158992.63 128092.06 250509.06 297231.13 504983.19 652785.75 250509.06 469149.06 250509.06 364491.56 250509.06 0.611 0.400 0.801 0.779 0.659 0.240 0.301 0.359 0.449 1.258 0.586 0. 479 0.584 0.935 0.359 0.539 0.467 0.584 0.236 0.523 1.796 1.135 0. 490 0.414 0.988 0.623 0.898 0.74 1 0.988 244372.875 1.418 274374.875. 1196 365630.50 0.897

825.0 10.0 0.1200 32,28.J07 224999.500 127160.250 825.0 10.0 0.0750 2017.741 427500.000 240260.813 825.0 10.0 0.1500 4035.485 314999.750 94093.750 825.0 10.0 0.0900 2421.290 180000.000 187645.250 825.0 10.0 0.1350 3631.935 292500.250 108464.500 825.0 10.0 0.1050 2824.838 202499.875 152311.813 825.0 10.0 0. 1050 224.838 382500.000 152311.813 825.0 10.0 0.0750 2017.741 179999.150 240260.813 825.0 10.0 0.0900 2421.290 360000.250 187645.250 825.0 10.0 0.1350 3631.935 382499.750 108464.500 **AVG FOR Q=10. 0.1080 2905.546 294749.563 159861.688 825.0 15.0 0.0900 2421.290 224999.625 187645.250 825.0 15.0 0.1350 3631.935 281249.938 108464.500 825.0 15.0 0.0600 1614.194 494999.500 325359.813 825.0 15.0 0.0900 2421.290 269999.750 187645.250 825.0 15.0 0.0600 1614.194 404999.500 325359.813 825.0 15.0 0.0750 2017.742 359999.500 240260.375 825.0 15.0 0.0900 2421.290 202500.000 187645.250 825.0 15.0 0.0900 2421.290 135000.063 187645.250 825.0 15.0 0.1012 2723.951 168750.063 159991.500 825.0 15.0 0.0900 2421.290 135000.000 187645.250 825.0 15.0 0.1200 3228.387 225000.000 127160.250 825.0 15.0 0.0900 2421.290 135000.375 187645.250 825.0 15.0 0.0600 1614.194 225000.125 325359.813 825.0 15.0 0.0562 1513.306 134999.875 355236. 188 825.0 15.0 0.1200 3228.387 292500.000 127160.250 *$AVG 0OR Q=15. 0.0885 2380.934 245999.625 214681.313 825.0 20.0 0.1500 403J.4d5 179999.750 94093.750 825.0 20.0 0.1050 2824.83Jd 202499.500 152311.813 825.0 20.0 0.0900 24121. 290 292499.750 187645.250 1.769 162178.063 1.779 273602.000 3.348 126981.438 0.959 223069. 188 2.697 142492.250 1.330 187948.688 2.511 187948.688 0.749 273602.000 1.919 223069.188 3.526 142492.250 2.059 194338.188 1.199 223069.188 2.593 142492.250 1.521 352153.813 1.439 223069. 188 1.245 352153.813 1.498 273601.813 1.079 223069.188 0. 7 19 223069.188 1.055 195679.875 0.719 223069.188 1.769 162178.063 0.719 223069.188 0.692 352153.813 0.380 379027.438 2.300 162178.063 1.262 247335.313 1.913 126981.438 1.330 187948.688 1.559 223069.188 1.387 216234.50 1.562 364491.56 2.481 169428.56 0.807 297231.13 2.053 190055.50 1.077 250509.06 2.035 250509.06 0.658 364491.56 1.614 297231.13 2.684 190055.50 1.636 259023.56 1.009 297231.13 1.974 190055.50 1.406 469149.06 1.210 297231.13 1.150 469119.06 1.316 364490.81 0.908 297231.13 0.605 297231.13 0.862 260792.94 0.605 297231.13 1.387 216234.50 0.605 297231.13 0.639 469149.06 0.356 504983.19 1.804 216234.50 1.056 329574.69 1.418 169428.56 1.071 250509.06 1.311 297231.13 1.041 1. 173 1.859 0.606 1.539 0.808 1.527 0. 494 1.211 2.013 1.227 0.757 1.41.0 1.055 0.908 0.663 0.988 0.681 0.454 0.647 0.454 1.041 0. 454 0. 480 0.267 1.353 0.792 1.062 0.808 0.984 L( NJ

825.0 825.0 825.0 *$AVG FOR 825.0 825.0 825.0 825.0 825.0 825.0 825. 0 825.0 *$AVG FOR 825.0 825.0 825.0 825.0 825.0 825.0 825.0 825.0 *.*AVG FOR 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 20.0 20.0 20.0 Q=20. 30.0.30.0 30.0 30.0 30.0 30.0 30.0 30.0 0=30. 50.0 50.0 50.0 50.0 50.0 50. 0 50.0 50.0 Q0=50. 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 0. 1350 0.0750 0. 0900 0. 1075 0.0750 0. 1125 0.0675 0. 1050 0.0990 0.0750 0. 0900 0. 0900 0.0892 0. 1200 0. 1200 0.0900 0.05 25 0. 1012 0. 0900 0. 1275 0. 0900 0.0989 0.0600 0.0900 0.0900 0. 1050 0.00450 0. 1350 0.0900 0. 0600 3b61. 9J5 20 17. 741 242 1.290 2892.095 2017.7141 3026. 613 1815.967 2824.838 2663. 418 20 17. 74 1 21421.290 21421. 290 21401. 111 3228. 387 3228.387 21421.290 1412. 420 2723. 951 21421.290 3430. 159 2142 1. 290 2660. 896 1907. b684 286 1.525 2861. 525 3JJ8. 446 11430. 763 4292.285 da6 1.525 1907. 6H4 360000.000 269999.500 315000.250 269999.625 382499.500 179999.938 213749.938 22500.121 142499.938 247500.000 202499.875 157499.875 19J593.375 202499.875 292499.750 75000.000 89999.938 112500.000 90000.000 101250.000 2137119.813 1147187.375 315000.000 405000.000 359999. 750 247499.750 179999.750 3371499.750,69999.750 270000. 250) 108464. 500 240260.8 13 1U645. 250 161736.875 240260.813 138747.063 277213.250 152311.813 164930.875 240260.813 187645.250 187645.250 198626.750 127160.250 127160.250 187645.250 390248.063 159991.500 187645.250 1171611. 250 187645.250 185582.375 427957.4 38 2417062.688 24 7062.688 200604.375 632929.000 142919.500 247l 062 688 4127951. 438 3. 319 1. 124 1.679 1.820 1.592 1. 297 0.771 0.1148 0.864 1.030 1.079 0.839 0.953 1.592 2. 300 0. 400 0.23 1 0.703 0. 480 0.864 1. 139 0.964 0.736 1.639 1. 1457 1.2314 0.2811 2.361 1.09 3 0.631'12492. 250 273602.000 223069. 188 196193.625 273602.000 174141.313 308121.500 187948.688 200621.875 273602.000 223069.188 223069. 188 233021.875 162178.063 162178.063 223069.188 4 10140.625 195679.875 223069. 188 151716.063 223069.188 218887.375 481070.313 305834.438 305834.438 258000.375 667063.813 195965.063 305834.1438 481070.313 2.526 0.987 1.4 12 1.455 1.398 1.034 0.694 0.120 0.710 0.905 0.908 0. 706 0.809 1.2149 1.8014 0.336 0.219 0.575 0.1403 0.667 0.958 0.776 0.655 1.324 1.177 0.959 0.270 1.722 0.883 0.561 190055.50 364q491.56 297231. 13 261491.00 3641491.56 2321145.144 410466.00 250509.06 267367.06 364491.56 297231.13 297231.13 310491.50 216234.50 216234.50 297231.13 5146495.25 260792.941 297231.13 202321.144 297231.13 291721.25 640864.75 407645.88 407645.88 344029.75 888813.69 261531.38 407645.88 6408614.75 1.8914 0. 741 1.060 1.092 1.049 0.775 0.521 0.090 0.533 0.679 0.681 0.530 0.607 0.936 1.353 0.252 0. 165 0. 431 0.303 0.500 0.719 0.582 0.1492 0.994 0.883 0.719 0. 203 1.290 0.662 0.1421 Un LC

975.0 975.0 915.0 975.0 975.0 975.0 975.0 975.0 * *AV POR 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 *AVYG pPll 975.0 975.0 975.0 975.0 975.0 975. 0 975.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 0=10. 15.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 15.0 Q= 15. 20.0 20.0 20.0 20.0 20.0 20.0 20.0 0. 0750 0.0750 0. 1050 0. 0900 0. 0'150 0.0900 0. 0750 0.0600 0. 0825 0.0750 0.0600 0. 1050 0.0600 0.0600 0. 1200 0.0750 0.0750 0.0600 0. 0900 0. 1350 0.0750 0.0825 0. 1050 0. 0600 0.0600 0.0750 0. 0600 0. 0900 0. 1012 2J814.603 238114.604 3338.446 2861. 525 23814.603 2861. 525 2 384. 603 1907. 6814 262J3.063 238J4. 60J 1907. 684 3338. 446 1907.684 1907.684 3815.366 2384.603 2384.603 1907.684 286 1. 52 5 4292.285 23814.603 2623.063 3J38. 4416 1907. 684 1907. 6811 23 J8411. 603 1907. 684 28b 1.525 3219. 214 J14999.500 4011999.750 247500. 125 247500. 125 404999.500 359999. 750 33711499.750 292499.500 312186.813 337499.500 179999.750 517500.000 427499.750 292500.000 359999.500 24 7500.000 337499.750 270000. 250 135000.000 382499.250 517500.000 3337149. 625 472500-250 247499.750 2417500.000 269999.750 112499. 750 157500.063 157500.063 316206.81 3 316206.438 200604. 375 247062.688 316206.813 247062.688 316206.813 427957.1438 310066.11438 316206.813 427957.1438 200604.375 427957. 438 427957.1438 167520. 188 316206..8 13 316206.813 427957.1438 247062.688 142919.500 316206.813 311229.875 2006014.375 41127957.41138 427957.1438 3 16206.813 427957.438 24 7062.688 210703.938 0.996 1.281 1. 234 1.002 1. 281 1. 457 1.067 0.683 1. 152 1.067 0.421 2.580 0.999 0.683 2.1419 0.78 3 1. 0671 0.631 0. 546 2.676 1.637 1.270 2. 355 0.578 0.578 0.8511 0. 263 0.637 0.7147 37411532.813 374532.375 258000. 375 305834. 438 374532. 813 305834.438 374532.813 481070.313 365596.063 374532.813 4181070.313 258000.375 481070.313 481070.313 222849.125 3714532.813 374532.813 481070.313 30583411.438 195965.063 37411532.813 367088.063 258000.375 41181070.313 481070.313 374532.813 481070.313 305834.438 268537.250 0.84111 1.081 0.959 0.809 1.081 1.177 0.901 0.608 0.938 0.90 1 0. 37 4 2.006 0.889 0.608 1.6 15 0.661 0.901 0.561 0. 1 4 1 1.952 1.382 1.0214 1.831 0.514 0.514 0.721 0.234 0.515 0.587 499038.56 499037.75 344029.75 407645.88 499030.56 407645.88 499038.56 640864.75 487210.75 499038.56 640864.75 344029.75 6408641.75 640864.75 297284.69 499038.56 499038.56 640864.75 407645.88 261531.38 499038.56 489174.81 344029.75 6408614.75 640864.75 499038.56 640864.75 407645.88 358042.75 0.631 0.8 12 0.719 0.607 0.8 12 0.883 0.676 0. 456 0.704 0.676 0.281 1.504 0.667 0. 456 1.211 0. 496 0.676 0. 421 0.331 1.463 1.037 0.768 1.373 0.386 0.386 0.511 0. 176 0.386 0.440 0.54 1 975.0 20.0 0.0750 2384.601 269999.750 316206.813 0.851 374532.813 0.721 499038.56

975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 **AVG FOR 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 975.0 9715.0 **AVG FOR 1100.0 1100.0 1100.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Q=20. 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 30.0 0=30. 0.0 0.0 0.0 0. 1050 0. 1050 0. 0600 0. 0900 0.0900 3.0787 0.0900 0. 0900 0.04 50 0.0812 0.0750 0. 0900 0.0562 0. 07 50 0.0600 0.0675 0. 1050 0. 0675 0.0600 0.0900 0. 0900 0. 0900 0. 0750 0. 0900 0.0900 O. 01150 0. 0766 0.0600 0. 0600 0.0562 3J38. 446 3338.446 1907. b84 2861.52 5 2861.525 2503.034 28b1. 525 286 1. 525 1 430. 763 2580.982 2384. 603 2861.525 1788.453 2384.604 1907. 68411 2 146. 142 3338.441b 2 146. 142 1907.6811 2861. 525 2861.52 5 2861. 525 2384.603 286 1.525 2861. 525 1430. 1763 2436.766 2 152. 258 2 152. 250 2017.742 359999.500 247499.875 2021199.875 1 46250.000 112499.875 168750.000 67500.000 134999.938 179999.875 209117.063 247500.000 180000.000 146250.000 247499.875 225000. 000 236249.688 157499.875 168749.875 157500.41138 20241199.813 146250.000 157499.875 247500.125 157499.750!, 23624119.813 112'99.875 189140.125'584999.750 472499. 750 225000.063 200604.375 200604. 375 41127957.41138 24 7062.688 2417062.608 295988.375 247062.688 247062.688 632929.000 312999.000 316206.813 247 1062.688 467170.375 316206.438 427957.41138 36474114.375 200604.375 36411744. 375 427957.438 247062.688 247062. 688 247062.688 316206.813 247062.688 247062.688 632929.000 332318.563 521766.813 521766.813 569506.000 1.795 1.234 0.473 0.592 0.455 0.570 0.273 0.546 0. 2811 0.770 0.783 0.729 0.313 0.783 0. 526 0. 648 0. 785 0. 463 0. 368 0.820 0.592 0.637 0.783 0.637 0.956 0. 1 78 0.625 1.121 0.906 0. 395 258000.375 258000. 375 481070.313 305834.438 305834.438 354705.875 305834.438 305834. 438 667063.813 368636.41138 374532.813 305834.438 517455.938 374532.375 481070.313 421385.000 258000. 315 41121385.000 41181070.313 305834.438 305834. 438 305834.438 37114532.813 305834.438 305834.438 667063.813 387876.875 603194.11438 603194.11438 648537.500 1.395 0.959 0.421 0.478 0.368 0. 476 0.221 0.441 0.270 0.627 0.661 0.589 0.283 0.66 1 0.468 0.561 0.6 10 0.400 0.327 0.662 0. 478 0.515 0.661 0.515 0.772 0. 169 0.521 0.970 0.783 0.3417 344029.75 344029.75 640864.75 407645.88 4076415.88 472658.13 407645.88 40764115.88 888813.69 491256.38 499038.56 4110764115.88 689334.19 499037.75 6110864.75 561393.81 344029.75 561393.81 640864.75 407645.88 407645.88 407645.88 499038.56 40764115.88 407645.88 888813.69 516854.63 803612.63 803612.63 863978.31 1.046 0.719 0.316 0.359 0.276 0.357 0. 166 0.331 0.203 0.471 0.4196 0.1442 0.2 12 0.496 0.351 0. 41121 0.458 0.301 0.246 0.4197 0.359 0.386 0.41196 0.386 0.580 0.127 0.391 0.728 0.583 0.260

**AVG FOB Q= 0. 0.0587 2107.419 1427499.625 537679.625 1100.0 5.0 0.0450 1614.194 157500.000 771222.625 1100.0 5.0 0.0600 2152.258 382499.500 521766.813 1100.0 5.0 0.0600 2152.258 404999.500 521766.813 1100.0 5.0 0.0900 3228.J87 202500.000 301417.000 1100.0 5.0 0.1050 3766. 452 494999.250 244788.813 1100.0 5.0 0.1050 3766.452 247500.000 244788.813 1100.0 5.0 0.1050 3766.452 224999.875 244788.813 1100.0 5.0 0.0900 3228-387 224999.075 301417.000 1100.0 5.0 0.0900 3228.387 292500.000 301417.000 1100.0 5.0 0.0900 3228.387 517499.500 301417.000 1100.0 5.0 0.0900 3228.381 337499.500 301411.000 1100.0 5.0 0.0900 3228.387 315000.000 301417.000 1100.0 5.0 0.1050 3766.452 270000.000 244788.813 1100.0 5.0 0.0600 2152.258 134999.875 521766.813 1100.0 5.0 0.0750 2690.321 270000.000 385666.938 1100.0 5.0 0.0900 3228.387 168750.000 301417.000 *AYVG FOR Q0 5. 0.0844 3026.612 290390.188 363203.625 1100.0 10.0 0.0750 2690.321 247499 750 385666.930 1100.0 10.0 0.0750 2690.321 337499.750 385666.938 1100.0 10.0 0.0562 2017.742 326250.000 569506.000 1100.0 10.0 0.0600 2152.250 179999.875 521766.813 1100.0 10.0 0.0450 1614.194 247500.125 771222.625 1100.0 10.0 0.0600 2152.258 315000.000 521766.813 1100.0 10.0 0.0675 2421.289 314999.875 444789.313 1100.0 10.0 0.0600 2152.258 J359999.750 521766.813 1100.0 10.0 0.0900 3228.387 247499.750 301417.000 1100.0 10.0 0.0600 2152.258 225000.125 521766.813 1100.0 10.0 0.0750 2690.321 337499.750 385666.938 1100.0 10.0 0.0600 215Z. 25d 314999.150 521766.813 0.807 618308. 313 0.204 834734. 813 0. 733 603194.438 0.776 603 194. 438 0.672 384421.875 2.022 324569.188 1. 011 324569.188 0.919 324569.188 0.746 384421.875 0.970 384421.875 1.717 384421.875 1.120 384421.875 1.045 384421.875 1.103 324569.188 0.259 603194.438 0.700 470272,250 0.560 384421.875 0.910 443988.250 0.642 470272.250 0.875 470272.250 0.573 648537.500 0.345 603194.438 0.321 834734.813 0.604 603194.438 0.708 528757.875 0.690 603194.438 0.821 J84421.875 0.431 603194.438 0.875 470272.250 0.604 603194.438 0.700 823734.31 0.189 1112056.00 0.634 803612.63 0.671 803612.63 0.527 512555.94 1.525 432961.69 0.763 432961.69 0.693 432961.69 0.585 512555.94 0.761 512555.94 1.346 512555.94 0.878 512555.94 0.819 512555.94 0.832 432961.69 0.224 803612.63 0.574 626740.31 0.1439 512555.94 0.716 591835.06 0.526 626740.31 0.718 626740.31 0.503 863978.31 0.298 803612.63 0.297 1112056.00 0.522 803612.63 0.596 704548.75 0.597 803612.63 0.644 512555.94 0.373 803612.63 0.718 626740.31 0.522 803612.63 0.525 0.142 0. 476 0. 504 0.395 1. 143 0.572 0.520 0.439 0.571 1.010 0.658 0.615 0.624 0. 168 0.431 0.329 0.537 0. 395 0.539 0.378 0.224 0.223 0.392 0.447 0.448 0.483 0.280 0.539 0.392 L4J 0N

1100.0 10.0 0.0600 2152.25d 225000.000 521766.813 1100.0 10.0 0.0750 2690.321 360000.000 385666.938 1100.0 10.0 0.0750 2690. 21 472499.750 385666.938 1100.0 10.0 0.0900 3228.387 337499.500 301417.000 **AVG FOR Q=10. 0.0677 2429.696 303046.375 465454.875 1100.0 15.0 0.0600 2152.258 494999.250 521766.813 1100.0 15.0 0.0600 2152.258 584999.750 521766.813 1100.0 15.0 0.0600 2152.25a J37499.750 521766.813 1100.0 15.0 0.0600 2152.258 315000.000 521766.013 1100.0 15.0 0.0787 2824.838 292499.750 361035.125 1100.0 15.0 0.0600 2152.250 270000.000 521766.813 1100.0 15.0 0.0900 3228.387 314999.750 301417.000 1100.0 15.0 0.0600 2152.258 292500.000 521766.813 1100.0 15.0 0.0750 2690.321 472499.750 385666.938 1100.0 15.0 0.0750 2690.321 247500.125 385666.938 1100.0 15.0 0.0600 2152.25$ 359999.750 521766.813 1100.0 15.0 0.0600 2152.258 134999.875 521766.813 1100.0 15.0 0.0900 3228.387 224999.938 301417.000 1100.0 15.0 0.0600 2152.258 292500.000 521766.813 1100.0 15.0 0.0450 1614.194 382499.500 771222.625 1100.0 15.0 0.0450 1614. 194 427499.500 771222.625 1100.0 15.0 0.0450 1614l.194 382499.500 771222.625 1100.0 15.0 0.0450 1614.194 382499.500 771222.625 **AVG FOR Q=15. 0.0627 2249.407 344999.313 528665.813 1100.0 20.0 0.0600 2152.258 517499.500 521766.813 1100.0 20.0 0.0600 2152.258 202500.250 521766.813 1100.0 20.0 0.0750 2690.321 292499.500 385666.938 1100.0 20.0 0.0600 2152.258 382499.750 521766.813 1100.0 20.0 0.0562 2017.742 281249.750 569506.000 1100.0 20.0 0.0600 2152.25t1 337500.000 521766.813 0.431 603194.438 0.933 470272.250 1.225 470272.250 1.120 384421.875 0.700 546962.188 0.949 603194.438 1. 121 603194.438 0.647 603194.438 0.604 603194.438 0.810 445508.063 0.517 603194.438 1.045 384421.875 0.561 603194.438 1.225 470272.250 0.642 470272.250 0.690 603194.438 0.259 603194.438 0.746 384421.875 0.561 603194.438 0.496 834734.813 0.554 834734.813 0.496 834734.813 0.496 834734.813 0.690 606809.750 0.992 603 194.4 38 0.388 603 194.438 0. 373 803612.63 0.766 6267140.31 1.005 626740.31 0.878 512555.94 0.583 728816.50 0.821 803612.63 0.970 803612.63 0.560 803612.63 0.522 803612.63 0.657 593799.38 0.448 803612.63 0.819 512555.94 0.485 803612.63 1.005 626740.31 0.526 626740.31 0.597 803612.63 0.224 803612.63 0.585 512555.94 0.4185 803612.63 0.458 1112056.00 0.512 1112056.00 0.458 1112056.00 0.458 1112056.00 0.588 808506.75 0.858 803612.63 0.336 803612.63 0. 280 0.574 0.754 0.658 0.438 0.616 0.728 0. 420 0.392 0.493 0.336 0.615 0.364 0.754 0.395 0.448 0. 168 0.439 0.364 0.344 0.384 0.344 0.344 0. 442 0.644 0.252 0.467 0.476 0. 326 0.420 tU1 Ln — sj 0.758 470272.250 0.622 626740.31 0.733 603194.438 0.634 803612.63 0.494 648537.500 0.434 863978.31 0.647 603194.436 0.560 803612.63

1100.0 20.0 0.0600 2152,.,50 494999.750 521766.813 1100.0 20.0 0.0600 2152.25d 315000.000 521766.813 1100.0 20.0 0.0600 2152.258 247500.125 521766.813 1100.0 20.0 0.0600 2152.258 1495000.000 521766.813 **AVG FOR Q=20. 0.0611 2192.612 356624.688 512930.108 1100.0 30.0 0.0900 3228.3187 495000.000 301417.000 1100.0 30.0 0.0750 2b90.321 J59999.750 385666.938 1100.0 30.0 0.0750 2690.321 2417499.625 385666.938 1100.0 30.0 0.0600 2152.258 315000.000 521766.813 1100.0 30.0 0.0450 1614.194 270000.000 771222.625 1100.0 30.0 0.0562 2017.742 157500.000 569506.000 1100.0 30.0 0.0562 2017.71142 191249.938 569506.000 1100.0 30.0 0.0900 3228.387 269999.750 301417.000 1100.0 30.0 0.0450 1614.194 157500.000 771222.625 1100.0 30.0 0.0600 2152.258 90000.000 521766.813 **&VG FOR Q=30. 0.0652 2340.5a0 255374.608 509915.500 1100.0 50.0 0.0900 3228.387 517500.000 301417.000 1100.0 50.0 0.0600 2152.25U 202500.125 521766.813 1100.0 50.0 0.0600 2152.258 674999.750 521766.813 1100.0 50.0 0.0562 201 7.742 213750.000 569506.000.1100.0 50.0 0.0600 2152.25d 134999.750 521766.813 1100.0 50.0 0.0600 2152.258 314999.750 521766.013 1100.0 50.0 0.0787 2824.83d 281249.013 361035.125 1100.0 50.0 0.0540 1931.0J3 142499.875 601952.563 1100.0 50.0 0.0750 2690.321 180000.000 385666.938 **AVG FOi Q=50. 0.0660 2367.4d3 295832.750 478515.625 EXECUTION TERMINATED 15:16:01 T=1.893 RC=O $.59 0.9149 603194. 438 0.604 603194.1438 0. 41174 603194.1438 0. 949 603194.1438 0.699 594436.063 1.6142 384421.875 0.933 1470272.250 0.642 470272.250 0.604 603194. 438 0.350 834734.813 0.277 6148537.500 0.336 648537.500 0.896 3814421.175 0.204 834734.813 0.172 603194.438 0.606 588231.688 1.717 384421.875 0.388 603194.4138 1. 294 603194.1438 0.375 648537.500 0.259 603194.1438 0.604 603194.1438 0.779 445508.063 0.237 679042.563 0. 467 470272.250 0.680 560061.875 0.821 803612.63 0.522 803612.63 0.410 803612.63 0.821 803612.63 0.602 791961.50 1.288 512555.94 0.766 626740.31 0.526 6267140.31 0.522 803612.63 0.323 1112056.00 0.241 863978.31 0.295 863978.31 0.702 512555.94 0.189 1112056.00 0.149 803612.63 0.500 783788.25 1.346 512555.94 0.336 803612.63 1.119 803612.63 0.330 863978.31 0.224 803612.63 0.522 803612.63 0.631 593799.38 0.210 904600.50 0.383 626740.31 0.567 71146235.63 0.616 0. 392 0.308 0.616 0. 452 0.966 0.574 0.395 0.392 0.2143 0. 182 0.221 0.527 0.1142 0.112 0.375 1.010 0.252 0.840 0.2147 0. 168 0.392 0. 1474 0. 158 0.287 0.1425 Un 00

359 APPENDIX F COMPUTER PROGRAM AND OUTPUT, OF DROPLET BREA~K-UP TIME DATA REDUCTION LIST OF VARIABLES VARIABLES NAME DIA CI) DIST DRPF DRPI FN (I) Q QM QMTB RE SMACH T(I) TB THKF THKI TIME V VEL WE X (I) DESCRIPTION OF VARIABLE Diameter of an individual droplet at a particular observation The distance that the particle travelled since the last observation Final observed diameter of a particular droplet Initial observed diameter of a particular droplet Number of frames between each successive observation Flow rate of water in the test section Ratio of dynamic pressure Square root of the dynamic pressure ratio times dimensionless break-up time Reynolds number for a particular particle Mach number of the steam through the test section Thickness of an individual droplet at a particular observation Dimensionless break-up time Final observed thickness of a particular droplet Initial observed thickness of a particular droplet Real time measured between observations Steam velocity in the system Droplet velocity across the test section Weber number for a particular particle Distance of an individual droplet from the edge of the blade at a particular observation.

360 BREAK-UP TIME ANALYSIS PROGRAM r —--- zA) —- I I 30 1 - - —..i _ - l V, Q, DRPI, THKI, DRPF, THKF, VEL, TIME, SMA CH, WE, RE, TB, QM, QMTB TIME DRPF DIST THKI VEL THKF DRPI

361 SSIG ODLO T=3 P=20 SB *FTN 5=WEATA(2,1873) C *'"'"" *****" BREAK-UP TIME ANALTSIS FPOGRAM ""*=*'"~***" C THIS PROGRAa CALCULAIES THE INITIAL ARD I]NAL C DIMENSIONS OF A DROP AND ITS VELOCITY! OR SPLIT CASES ONLY. C IT ALSO FINDS WEBER NUMBER, REYNCLDS UEERfi, DIMENSIONLESS C BREAK-UP TIDE, AND TB E YNASIC PRESSURE RITIO. DIMENSION FN(20), DIA (20), T(20), 2(20) WRITE(6, 19 5) C ***** BEAD THE DATA TOE EACH CBSBRVATICN ***$* 1 DO 110 1=1,20 READ (5,101) FNCI), DIA(I), T (I), 1(1) I (T(I).LT.0.0) GO TO 130 I? (1(I).LT.0.0) GC TC 140 110 CONTINUE 130 I (I.EQ. 2) GO TO 140 TI.5E= (C- 1-0) *0.001 DIST= (X (I) -X (1)) *0. 45 VEL=DI ST/TIME C ***** FIND THE INITIAL AND FINAL DROP DIMENSIONS *.** DRPI=DIA (1) *0.45 DSPF=DIA (I) *0. 45 THKI=T (1) *0.45 THKF=ABS (T (I) ) *0. 45 BEAD(5,101) V,C C ***** DETERMINE THI STEAM 3ACH NUMBE2 ***D IF (V.EQ.305.0) S3ACE=0.23 17 (V.EQ. 520.0 ) SMACH=0.38 IF (V.EQ.825.0) SMACB=0.61 IF (V.EQ.975.0) SMACH=0.72 IF(V.EQ.1100.0) SMACB=0.81 E — (V**2) *DRPI* 1.435E-3 BE=V*DBPI*32. 61 TB= (TIME*V)/DRPI*0. 323 Q=0O.78+1.47/(1.+2. 1*SMACH**3.4) QMTB=SCRT (CM) TE WRITE (6,200) V,Q,DRPI,THKI,DRPF,THKF,VZL,'TIME,SMACH,WE,EE,TB, 1 QnQMTE GO TO 1 140 READ(5,101) V,Q GO TO 1 101 FORMAT (6(F10.5)) 199 FORMAT ('1',2X,'STEAM',4X,'FLCV',2X,'INITIAL' 3X, 1'INITIAL',4X,'FINAL',6X,'FINAL',61,'DROP',5X, 2'TOTAL MACH',5X,'E'6X,'RE',6X,'B','C',4X,'QM.STB'/ 3 21,'VELOCITY RATE DIA3EIEF TBICKNESS rlA3ETER THICKNESS', 4 2X,'VELOCITY TIME NUMRER (RS*V**2 (Y*DO/', 5 1,' (T*V/DC PRESS'/23,' (FEET/S) (CC/") (CM)',6X, 6' (CM)',61,' (CM)'6X,' (CM)',5X,' (C/S)',41,' (SEC)',8X, 7'"DO/SIG) KVIS) *RBS*.5) RATIO') 200 FORSAT ('0',F8.3,21,F4.1,2X,F8.5,IX,F9.5,2X, 1 FS.5,2X,F9.5,2X,F8.2,3XF4. 3,2X,v4.231,fX6.2,2XF6.1, 2 2X,F6.3,2X,F5.3,2X,F5. 2) END SENCFILE S3 -LOAD 5=WDATA(2,1873) 3SNEFILE $3IG

.STSAM Ft.nW t[T[AL IPI"I&L VELOCITY nRTE 0IVETTRI TPICK^'SS (r EET/S) (CC/MI (c.i) (cI) 305. 000 5. 305.o00010,0 105.000 20.0 520.000 5.0 520.000 5.0 520.000 10.0 520.000 10.0 520.0900 15.0 520. 00 15.0 520.000 15.0 520.000 15.0 520.0n0 20.0 520.000 20.0 825. 000 5.0 R25.000 5.0 825.000 5.0 825.000 5.0 825.000 5.0 825.000 5.0 825.000 5.0 825.000 5.0 825.000 5.0 825.000 10.0 825.000 10.0 825.n000 10.0 825.000 10.0 825.000 10.0 925. 00n 10.0 0. oqf)0o O.'9000 0. 1500) 0. 0900, 0. 09000 0. 09000 0. 09000 0.09000 0.09000 0. 09000 0.09000 0. 09000 0. 09000 0.09000 0. 01500 0. 09000 0. 08500 0. 09000 0.04500 0. 09000 0.11500 ). 004500 0.13500 0.0900 O. 0900) n. so(00) 0.0 9000 0. 09000 0). 04500 0. 09000 o. oqoo 0. 08 5)00 0. 09000 0. 09000 0. 09000 0. 0900o 0. O900q 0.09000 0. 09000 0. 09000 0.09000 0. 0'1500 0.00500 0. 09000 0.04u500 0. 09000 0. 04500 0. 090800 0. 09000 0. 04500 0.1 500 0. It 500 0. 090o0 0. 09000 FI tlAI. nDIO ^MFrrP (CIM 0.09030 0. 1.3500 0. 22500 0. 09010 0. 09000 0.09000 0.09000 0. 09000 0. 09000 0. 1 500 0.09000 0. 09000 0. 09000 0. 1 1500 0. 09000 0.01000 0. n08 500 0. I 1500 0. 09000 0. 1.1500 0. 09000 0. 1 3510 0. 1 )500 0. 09000 0. 1 3500 0.Qo000 0. 1 300 0. 11500 TH ICK P F. S (C1) 0.09000 0.09000 0. 09000oo 0.09000 0.09000 0.04500 0.09000 o.oqoo0 0. 09000 0. 04500 0. 09000 0.09000 0.09000 0.09000 0.04500 0. 08 500 0. n s500 0. 04 500 0.09000 0. o000 O. qoo00 0.11500 O.OqOO0 0. J9000 0.13500 0.09000 0.09000 O. 04 sn5 F't" AL DROP TOTL IACHI 1wr PR TB 0 Q01 T8 TTMR H198Rn (R9S*V**2 p7*nO/ (T*V/DO PRESS (SEC) *00/SiC) KVISq) *n,.S5 RATIO VRLOCITY (CM/Sl 210.000 163.616 121.4 29 525.000 405.000 6415.000 106.000 49q.000 277. 500 295.71 4 315.000 414.000 262. 500 0945.000 877.500 660.000 585.000 607.500 697.500 675.000 765.000 877.500 1090.000 022.500 1015.000 4 50. 000 697.500 967.500.006 0.21.011 0. 23.007 0.21.00 3 0. f3.004 0.q3.003 0.38q.005 0.381.003 0.389.006 0.39.007 0.38.005 0. 38.005 0.38.006 0. 38.002 0.61.002 0.61.003 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61 12.01 8q5. 1 12.01 895.1 18.02 11142.7 34.92 1526. 1 14.92 1526.1 17.46 763.1 3 4.92 1526.1 3 4.92 1526.1 3 4.92 1526. 1 34.92 1526.1 348.92 1526.1 34.92 1526.1 34.92 1526.1 87.90 2421.3 87.90 2q21.3 41.95 1210.6 483.95 1210.6 87.90 2421.3 41.95 1210.6 87.90 2412 1. 1 411.9; 1210.6 87.90 2421. 1 131.85 3631.9 41.05 12191.6 131.85 3631.9 87.90 2421.1 87.90 21181.3 87.90 247 1. 6. 568 2. 229 12.041 2.229 5. 108 2. 229 5.509 2.11143 7. 4 65 2. 14 3 11.197 2. 1'13 9. 31 2. 143 5.599 2. 14 3 11. 197 2. 143 13. 064 2. 141 9.331 2. 141 9.331 2.14 3 11.197 2.14.1 5.922 1. 817 5.922 1.837 17.765 1.837 11.843.1 1. R17 5.922 1.837 11.84.3 1.837 5.922 1.837 11.843 1.837 5.922 1.817 3. 948 1. 837 11.841l 10837 3.948 1. 837 5.922 1.837 5.922 1.837 5.922 1.837 9.81 17. 08 7. 61 8. 20 10.93 16..39 13.66 8.20 16.39 19. 11 13.66 13.66 16. 39 8.01 8.03 24. 08 16. 05 8.01 16.05 8.0] 16.05 8.01 5.35 16. 05 5. 35 8.03 8.03 8.01 0s

F125 001 10.0 0.oqooo I) 0 000.09o n.0o01000 5960.000.002 0.61 87.90 2121. 1 5. 922 1. 817 8.01 825. 000 1.0 825. 000 15.0 825.000 15.0 825.000 15.0 825.000 15.0 825.000 15.0 825. 000 15. 0 825.000 15.0 825.000 15,0 825.000 15.0 25. 000 15.0 825.000 20.0 825.000 20.0 825.000 20.0 825.000 10.0 825.000 30.0 825.000 10.0 925.000 10.0 625.000 30.0 825.000 50.0 825.000 50.0 825.000 50.0 825.000 50.0 825.000 50.0 825.000 50.0 975.000 10.0 975.000 10.0 975.000 10.0 975. 000 10.0 0. 13501 0. 03000 0. 04500 0. o0l50s 0. 04500 0. 39000 0. 09000 0. 09000 O. 04500 0. 014500 0.09000 0. (9000 0. 09000 0. 09000 0. 09000 0. 04500 0. 09000 0. 04500 0. 09000 0.13500 0.09000 o. oqooo o. oqooo 0. I0ooo 0. 09000 o. 09000 0.011500 0.09000 0. 09000 n. 09000 0. 180')0 0.09000 0. 1800n 0.04500 0.09000 0. 04500 0.09000 0.04500 0.04500 0.04500 0.09000 0. 04O500 0. 04500 0. 09000 0. 04500 0. 04 500 0.09300 0. 0 qoo0 0. 04500 0. 09000 0.04500 0.04500 0. 0500 0.09000 0.04500 O. Oqooo 0. 09000 0. 09000 0. 04500 0. 04500 0. 09q00 0. 13500 0.09000 0. 13500 0.09000 0.04 500 0.09000 0. 1MOO0 0.09000 0.09000 0. 1o000 0.09000 0.13500 0.09000 0.09000 0.09000 0.09000 0. 04 s500 0. 18000 0. 09000 0. 1ROOn 0.09000 0.090130 0.09000 0. 04500 0. 09000 0.09000 0.09000 0.09000 0. 09000 0. 04 500 0.09000 0. 04 500 O. 04 500 0.13500 0. 09000 0.09000 0.09000 0. 13500 0.04500 0.09000 0.09000 0.09000 0.09000 0. 09000 0. 04500 0.09000 0.09000 0. I800O 0.09000 0.09000 0.09000 105 7.500 825.000 1170.000 810.000 765.000 450.000 765. 000 610.000 585.000 100.000 607.500 697.500 855. 0 00 855.000 600. 00 570.000 506.250 4 72. 500 562.500 652.500 4 16.250 4 14.000 6 15.000 472.500 405.000 1034. 9 99 765.000 855. 00.002 0.61.003 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.002 0.61.003 0.61.002 0. 61.002 0.61.002 0.61.002 0.61.003 0.61.003 0.61.004 0.61.002 0.61.002 0.61.002 0.61.004 0.61.005 0.61.003 0.61.004 0.61.005 0.61.002 0.72.002 0.72.002 0.72 131.R5 3631.9 87.90 2421.3 43.95 1210.6 4 3. 95 1210.6 43.95 1210.6 87.90 2421. 3 87.90 24 21.3 87.90 2421.3 43.95 1210.6 43.95 1210.6 87.90 2421.3 87.90 2421. 1 87.90 2421.3 87.90 2421.3 87.90 2421. 43.95 1210.6 87.90 2421.3 43.95 1210.6 87.90 2421.1 131.85 3631.9 87.90 2421.1 87.90 2421.1 87.90 2421.3 87.90 2421.3 87.90 24121.3 61.19 1410.9 122.77 2861.5 122.77 2861.5 3.948 1.837 R.802 1.837 11.843 1.831 11.P43 1.837 11.843 1.837 5.922 1.817 5.922 1. 37 5.922 1.817 11.843 1.817 17.765 1. 817 5.922 1.8317 5.922 1.837 5.922 1.837 5.922 1.837 8.882 1.837 17.765 1.837 11.843 1.837 11.843 1.837 5.922 1.837 3. 948 1. 837 11.841 1.837 14.804 1. 837 8.682 1.837 11.843 1.837 14.804 1.837 13.997 1.651 6. 998 1.651 6.998 1.651 5.35 12.04 16.05 16. 05 16.05 8.0.1 8.03 8.03 16.05 24.08 A.03 8.03 8.01 6.01 12.04 24.06 16. 05 16.05 8.03 5. 15 16.05 70. 06 12.o0 16.05 20. 06 17.99 8.99 8.99 0' W 0. 13500 0.0ol500 0.09000 0.09000 427.500.002 0.72 184.16 64292.3 4.666 i1.651 6.00

975.000 10.0 975.000 10.0 975.000 10.0 975.000 10.0 975.000 10.0 975.000 10.0 75. 000 10.0 975.000 10.0 975.000 10.0 975.003 10.0 175. 000 10.0 975.030 10.0 975.000 15.0 975.000 1.5.0 975.000 15.0 97,.000 15.0 975. 000 15.0 975.000 15.0 975.000 19.0 975.000 15.0 975.000 15.0 975. 000 15o0 9-17;. 000 15.0 975. 000'. 0 975.000 20.0 975.000 20.0 975.000?0.0 975.000 20.0 975.000 20.0 975. 1100 2n.0 0.0 r) 4) 0.045000 0. 0500 0.0450)0 415.000.0)2 0.72 61.19 1410.9 13.997 1.f51 17.99 0. 1 15f) 0. 09003 0. 04 50 0. 09000 0. 04500 0. 0900:). 411500 0. OSn 00 0. 0O 500 0. 0450 ) 0. 04S00 0.:)500 0. n4500 0.09000 0. 04150n 0. 01500 0.09000 0. 09000 0. 09000 0. 94Ot500 0. 0900( 0. 04l501'. 09000 0. 018500 0. 0450'0 0. 090001 0. 09000 0. 09)0o 0. 0)00) 0.109000 0. 09000 0.04 500 0. 04500 0. 0'500 0. 09000 0.01500 0. 04500 0.0, 500 0. 01500 0.041500 n. 0' 500 0. 018500 0. 041500 0. 011500 0. 04500 0. 041500 0. 04 500 0. 04500 0. 01 500 0. (9000 0.011 00 0. 0 1500 0. 04500 0. 01530 0. o09000 0 090000 0. oQ900 0.09000 0. 11 50 0. 09000 0.09000 0.09900) 0. 11500 0. 13500 0. 1 3500 0. q09000 0. 13500 0.09000 0.09000 0.09000 0.09000 0.1 3500 0.09000 0.09000 0.1.1510 0. 09000 0. 09000 0.09000 0. 09000o 0.09000 0. o1300 0.09000 0.09000 0. o045)0 0. )9000 0. 13900 0.09000 0.09001 0. 014 500 0. 09000 0.0114500 0.09000 0.09000 0. 04500 0.13500 0.09000 0.04 500 0.09000 0.09000 0.09000 0.09000 0,09000 0. 09003 0.09000 0. 09000 0. 011500 O. 0900o 0.011400 0.09000 0.09000 0. 04500 0. 09000 0. 04 50 0. 09000 0. 11500 922. 500 720. 000 720.000 765.000 1014.999 652.500 652.500 Q00. 000 900.000 652.500 8 12.500 787.500 160.000 1057. 500 70 7. 4 99 562. 500 765.000 517.5 00 967. 500'585.000'150.000 1057. 500 1012. 99 472.500 697.500 540. 000 182.500 510.000 111). 0 00. 002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.002 0. 72.002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.00 2 0. 72.002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.002 0.72.002 0. 72.002 0.72.002 0. 72.002 0.72. 001 0. 72. 003 O0. 72 184.16 114292.1 4.666 1.651 6.00 122.77 2961.5 6.998 1.651 8.99 61.19 1430.8 1i.997 1.651 17.99 122.77 2861.5 6.998 1.651 A.99 61.19 11130.8 13.997 1.651 17.99 122.77 29161.5 6.99R 1.651 8.99 61.19 141110. 13.997 1.651 17.99 61.39 14110.8 11.997 1.651 17.99 61.39 1430.81.1.997 1.651 17.99 61.39 1010.8 13.997 1.651 17.99 61.39 14110.8 13.997 1.651 17.99 61.39 141110.8 13.997 1.651 17.99 61.19 1410.8 13.997 1.651 17.99 122.77 2861.5 6.998 1.65 1 8.99 61.39 11110.8 13.997 1.651 17.99 61.39 1430.5 13.997 1.651 17.99 122.77 2861.5 6.998 1.651 8.99 122.77 28f61.5 6.0998 1.651 8.99 122.77 2861.5 6.998 1.651 8.99 61.39 1430.8 13.997 1.651 17.99 122.77 2961.5 6.998 1.651 8.99 61.39 1430.q 13.997 1.651 17.99 122.77 2861.5 6. 991 1.651 8.99 61.39 14310. 13.997 1.651 17.99 61.19 14130.9 1.1.997 1.651 17.99 122.77 2q61.5f 6.998 1.f651 8.99 122.77 2861.5 6.99q 1.651 8.99 122.77 2861.5 10.497 1.651 13.49 122.77 28161.5 10.4197 1.691 11.149 0J 0s cr

975.000 20.0 975.000 20.0 975.000 20.0 975.000 20.0 975.000 20.0 975.000 20.0 975.000'0.0 975.000 20.0 975.000 20.0 975.000 20.0 975.000 10.0 975.000 ]0.0 975.000 30.0 975.000 30.0 975.000 30.0 975.000 30.0 975.000 10.0 975.000 30.0 975.000 30.0 975. 000 10.0 975.000 30.0 975.000 ]0.0 975.000 10.0 975.000 30.0 975.000 30.0 975.000 10.0 1100.000 0.0 1 100.000 0.0 1100.0 0 0.0 1100.000 5.0. 09000 0. )9000 0.09000. 0114500 0. 09000 0.09000 0.09000 0.09000 0. 09000 0. 0s500 0. 0U500 0. 09000 0. 0411500 0.011500 0. 04500 ). 04500 0. 09000 0.04500 0. 011450 a. 09000 0. 09000 0.09000 0. 041500 0. 09000 0. 09000 0. 04500 0.0114500 0. 01 500) 0..)4 500 0. 01'500 0. 01u500 0. 0qO00 0. 09000 0. 011500 0. 09000 0. 09000 0. 01500 0. 09000 0.09000 0. 04500 0.0411500 0. 09010 0. 04500 0. 04 500 0. 011500 0.0411500 0. 09000 0. 04500. 0411500 0. 09000 0.09000 0.011500 0. 014500 0. 09000 0.0114500 0. 041500 0.0411500 0.04500 0. 04500 0.-O4500 O. oo000 0. 1 35 ) 0. 13500 0.09000 0. 09000 0. 09000 0. 09000 0.09000 0. 09000 0.01104500 0.09000 0.13500 0. 09000 0. 11500 0.011500 0.09000 0. 11500 0.09000 0.09000 0.09090 0.09030 0.09000 0.09000 0. o0900oo0 0.09000 0.04510 0.09000 0.01000 0.09000 0.011 500 0.09000 0. 09000 0. 13500 0. 041500 0.09000 0. 0oo000 0.011500 0.09000 0. oqoo0o 0. 04500 0.09000 0. 04500 0.04500 0.13500 0.04500 0.09000 0.13500 0. 04 500 0. 09000 0. 04500 0. 09000 0. 04500 0.011500 0. 09000 0.09000 0.004500 0.041500 0.09000 0.09000 0.041500 6 75.000 765. 000 607.500 652.5 00 585.000 405. 00 610.000 2110. 000 585.000 525.000 517.41199 600.000 o 00.000 697.500 585.000 555. 000 382.500 4 50.000 697.500 585. 00 1165.000 607.500 562.500 427.500 705.000 41120.000 855.000 922.500 510.000 2417.500.002 0.72.002 0.72.002 0.72.002 0.72.003 0.72.003 0.72.001 0.72.003 0. 72.003 0.72.00] 0. 72.002 0.72.003 0.72.003 0.72.002 0.72.002 0.72.001 0.72.002 0.72.003 0.72.002 0.72.003 0.72.003 0.72.002 0.72.002 0.72.002 0.72.003 0.72.003 0. 12.002 0.81.002 0.81.003 0.81.002 0.81 122.77 2861.5 6.998 1.651 8.99 122.77 2861.5 6.998 1.651 8.99 122.77 2861.5 6.998 1.651 8.99 61.39 1430.8 13.997 1.651 17.99 122.77 2861.5 10.497 1.651 13.4q 122.77 2861.5 10.497 1.651 13.49 122.77 2861.5 10.n97 1.651 13. 49 122.77 2861.5 10.497 1.651 13.49 122.77 2861.5 10.41197 1.651 13.149 61.39 1430.8 20.995 1.651 26.98 61.39 1410.8 13.997 1.651 17.99 122.77 2861.5 10.497 1.651 13.49 61.19 141110.8 20.995 1.651 26.98 61.39 1430.8 13.997 1.651 17.99 61.39 141110.8 13.997 1.651 17.99 61.39 1430.R 20.995 1.651 26.98 122.77 2861.5 6.998 1.651 8.99 61.39 1430.8 20.995 1.651 26.99 61.39 1430.8 13.997 1.651 17.99 122.77 2861.5 10.497 1.651 13.49 122.77 2861.5 10.497 1.651 13.49 122.77 2861.5 6.998 1.651 8.99 61.39 1410.8 13.997 1.651 17.99 122.77 2861.5 6.998 1.651 8.99 122.77 2861.5 10.497 i1.651 13.49 61.39 1430.8 20.995 1.651 26.98 78.14 1614.2 15.791 1.506 19.38 78.14 1614.2 15.791 1.506 19.38 78.14 1614.2 23.687 1.506 29.06 79.1 161114.2 15.791 1.506 19.38 U,

1100. 000 1100.000 1100.000 1 on100. 000 1100. 000 1100. 00n 1 100. 000 1100. 000 1100.000 1100. 000 1100.000 1100. 000 1100.000 1100. 000 1100. 000 1 100. 000 1100.000 1 100.000 1 100. n000 1100. 000 1100. 000 1100.000 1100.000 1 100. 000 1 100.00') 1100o 000 1100. 000 1100.000 I 100. 000 1 100. 000 5.0 5.0 5.0 5.0 5.0 5.0 %.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 0.049500 0. 04500 o. 09000 0. 09000 0. 09000 0. 09000 0. 09000 0. 09000 0. 09000 0. 09000 0. 09000 0. 09000 0. 011500 0.0oq50o 0. 09000 0.0114500 0.09000 0. 0411500 0. 041500 0. 0114500 0. Un500 0. 014500 0.04 500 0. 09000 0. 09001 0. oqooo 0. o0' 500 0.0OU500 0.041500 0.00 0O) 0. 04500 0.0 4500 0. 0 9000 0. 045s0 0. 09000 0. 04500 0.004500 0.0'1500 0. 09000 0. 04500 0. 04500 0.09000 0.0114500 0.004500 0.0114500 0. 04500 0.114500 0. 041 500 0. 04500 0.04 500 0.011500. 0411500 0.04 500 0. 014500 0.011500 0. 04 50 0. 04500 0.04 500 (0.0 1100 O. OU 100 0.09000 0.09030 0. 09000 0. 13500 0. 1 35100 0. 1 3500 0.O0OO0 0.09000 0. 13500 0.09000 0.09000 0. 1 3500 0.09000 0.09000 0. 13500 0.09000 0.09000 0.09000 0.09000 0.04500 0.090o00 0. OQOO0 0.09000 0.09000 0.0114500 0.09000 0. 09090 0. 09000 0. l910 0. 000nn 0. 09000) 0.09000 0.04 500 0.13500 0. 01500 0. 01 500 0.09000 0.09000 0.04 500 0.09000 0.09000 0..1500 0.09000 0. 09000 0. 09000 0.09000 0. 04500 0.09000 0.09000 0. 01 500 0. 04500 0.09000 0.0411500 0. 09000 0.011500 0. o 500 0.041500 0. 09000 0. 09000 0. o 500 652.500 675.000 382.500 675.000 517.500 450.000 54 0.000 607.500 71142. 41 99 787.1199 720.000 6 75.000 315.000 4110.000 495.000 41172.500 R32. 500 795.000 495.000 562.500 765.000 870.000 900.000 562.500 675.000 517. 4199 4 50.0 00 160.000 610.000 n12.500.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81,002 0.81.002 0.81.002 0.81.002 0.81.003 0.81.002 0.81.002 0. 81.003 0.81.002 0.81.002 0.81.002 0. 81.003 0.81.002 0.81.002 0.81.002 0.81.002 0.81.902 0. 81.002 0.81.002 0.81.002 O.R1 78. 11 1614.2 78.14 161'f.2 156.27 3228.4 156.27 3228.4 156.27 3228.411 156.27 3228. 4 156.27 3220. 4 156.27 3228. 4 156.27 3228. 4 156.27 3228. 4 156.27 3228.411 156.27 3228.4 78.14 1614.2 78.14 1614.2 156.27 3220.4 78.111 1614.2 156.27 322q. 4 78. 111 1614.2 78.14 1614.2 78.14 1614.2 78. 141 16411.2 78.1 4 16113.2 78.14 1614.2 156.27 3228.4 156.27 1229. q 156.27 3228.4 78.14 16114.2 78.14 161411.2 78.14 16111.2 156.27 1221. 1 15. 791 15.791 7.896 7.896 7.896 7. 896 7.896 7.896 7.896 7.896 7.896 7.896 15.791 15. 791 11. q1 15.791 7.896 23.687 15. 791 15.791 15. 791 23. 687 15.791 7.896 7. R96 7. 896 15. 791 15.791 15. 791 7.896 1. 506 19.38 1. 506 19.38 1.506 9.69 1.506 9.69 1.506 09.69 1.506 9.69 1.506 q.69 1.50 6 9.69 1.506 9.69 1.506 9.69 1.506 9.69 1.506 9.69 1.506 19.38 1. 506 19. 38 1.506 111.5. 1. 506 19.38 1.506 9.69 1.506 29.06 1. 506 19.38 1. 506 19.38 1.506 19.38 1.506 29.06 1. 506 19. 38 1. 506 9.69 1.506 9.69 1.506 9.69 1.506 19.18 1.506 19.18 1. 506 19.38 1.506 9.69 uO ON

1100. 00 10.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.000 15.0 1100.030 15.0 1100.000 15.0 1100.000 15.0 1100.000 20.0 1100.000 20.0 1100.000 20.0 1100.000 20.0 1100.000 20.0 o100.000 20.0 1100.000 20.0 1100.000 20.0 1100.000 20.0 1 1(10.000 20.0 0. 09009 0. n 000 0.0900() 0.09000 797.500.002 0.81 156. 27 1228. 4 7.896 1.506 9.69 0. n01O00 0. 01so00 O. 011504 0. 4 500 0. 04 500 0. 04500 0. qo900 0. 04500 0. 34500 0.09000 0.04 500 0. 04500 0. 09000 0.014500 0. 04500 0. 014500 0. 04500 0. 34500 0. o0 500 0. 09000 O. O500 0. 04500 0. O'85')0 0. 014500 0. 04500 0.')4500 0. 041500 0. 04500 0. 011500 0.04 500 0. 04O00 0.04 500 0. O0l500 0. 04500 0. 0o 5 0S 0.04500 0.04 500 0.041500 0.04500 0. 041500 0.011500 0. 011500 0.014500 0.014500 0.041500 0.04500 0. 048500 0. 04500 0. 081500 0. 04500 0. 04 500 0. 0485(0 0.048500 0. 04 500 0.04500 0. 04500 0. 09o00 0. 09000 0. 09000 0.09000 0.09000 0.09000 0. 09000 0.09000. 09000 0.09000 0.09000 0.09000 0. 09000 0.09000 0.041500 0. 01500 0. 04500 0.004500 0. 09000 0.04500 0.09000 0. 09000 0. o000ooo O. nqnno 0. 09000 0.09000 O, oqnoO 0.09000 0.09000 0. 011500 0. 09000 0.09000 0.09000 0. 09000 0. 04500 0. 09000 0.09000 0. 09000 0. 09000 0.09000 0.04 500 0. 084500 0. 04500 0. 048500 0.04 500 0.04500 0.04 500 0. 09000 0.04 500 0. 09000 0. 04500 o. oqono 0. 99000 O. 0o000 0. 01500 675.000 765.000 472.500 495.000 615.000 900.000 585.090 652.500 1057.500 697.500 585.000 450.000 525. 000 652. 500 7482.4 99 832.500 787.1199 6 97. 500 742.4 99 697.500 877.500 1012.500 600.000 607. 500 765.000 585.000 562.500 10184.999.002 0.81.002 0.81.002 0.81.002 0.81.001 0.81.002 0.81.002 0.81.002 0.81.002 0.81.002 0.R1.002 0.81.002 0.81.003 0.81.002 0.81.002 0.81.002 0. 8.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81.003 0.81.002 0.81.002 0.81.002 0.81.002 0.81.002 0.81 79. 14 16 148.2 15. 791 1.506 19. 18 78. 14 1614.2 15.791 1.506 19.18 78. 14 1614.2 15.791 1.506 19.31 78. 14 1614.2 15.791 1.506 19.38 78. 14 16114.2 23.687 1.506 29.06 78.14 1614.2 15.791 1.506 19.38 156.27 13229.4 7.896 1.506 9.69 79. 14 16114.2 15.791 1.506 19. 18 78. 18 1614.2 15.791 1.506 19.381 156.27 3228. 11 7.896 1.506 9.69 78.14 1614.2 15.791 1.506 19.38 78.14 1614.2 15.791 1.506 *19. 38 156.27 3228.4 11.843 1.506 14.53 78.14 1614.2 15.791 1.506 19.18 78.14 1614.2 15.791 1.506 19.38 78.14 16i14.2 15.791 1.506 19.38 78.14 1614.2 15.791 1.506 19.38 78.14 16114.2 15.791 1.506 19.38 78. 14 1614.2 15.791 1.506 19.38 156.27 3228.84 7.896 1.506 9.69 78.14 1614.2 15.791 1.506 19.38 78.14 1614.2 15.791 1.506 19.38 78.14 1614.2 23.687 1.506 29.06 78. 14 1614.2 15.791 1.506 19.38 79.14 1614.2 15.791 1.506 19.38 78l. 11 1614.2 15.791 1.506 19.38 78.14 1614.2 15.791 1. 06 19.38 70. 14 1614.2 15.791 1.506 19.38 Du I 0.(n f( 0 (f(( (.0)a,) n gO 0.04S500 1125.000.002 0.I1 156.21 322q." 1.q89 1.506 1.69

1100.000 10.0 100oo00o 10o. 1100.000 30.0 1100.000o 0.0 1 100.000 10.0 1100.000 10.0 1100. 0)0 30.0 1100.001 10.0 1100.000 30.0 1100.000 50.0 1100.000 50.0 1100.000 50.0 1100.000 50.0 1100.000 50.0 1100.000 50.0 1100.000 50.0 1100.000 50.0 1100.000 50.0 0. 09000 0. 04500 0. 0s500 O. 011500 0. 01500 0.014500 0.09000 0. 01500 0. 04500 0. 09000 0. 04 500 0. 01500 0. 041500 0. 04500 0.04500 0. 09000 0.0114500 0. 09000 0.041500 0. 04 500 0.0411500 0. 04500 0. 014500 0. 04500 0.09000 0.0 o500 0.04500 0. 04500 0.0411500 0. 01500 0. 011500 0.04500 0. 011500 0. 09000 0.04 500 0.04500 0. 090001 0. 0000 0. 09000 0. 01500 0. oqooo 0.09000 0.09000 0.01 500 0.0 9000 0.09000 0. 0900 0.09000 0.09000 0. 090o0 o. oqoo0 0.09000 0.09000 0.09000 0. 011500 0. 014500 0.01500 0. 01 500 0.04500 0. 04500 0. 04500 0.0 500 0. 09000 0. O01 500 0. 01500 0.04500 0. 01 500 0.09000 0. 0114500 0. 04 500 0.04 500 0. 09000 5R5.000 877.500 855.000 630.000 51 0. 000 525. 000 855.000 697.500 450.000 1012.500 517.500 1080.000 705.000 160.000 585.000 705.000 596.250 505.000.002 o. 81.002 0.81.002 0.1l.002 0. 1l.001 0.8f.003 O. n1.002 0.81.007. A I.002 0.81.002 0.81.002 0.81.002 0.81.003 0.81.002 0.81.002 0.81.003 0.81.004 0.81.002 0.81 156.27 322.11 4 7.896 1.506 9.69 78.14 161'.2 15.791 1.506 19.38 78.111 1611.2 15.791 1.506 19.38 79. 14 161'.2 15.791 1.506 19. l8 78.14 16111.2 23.687 1.506 29.06 78. 114 1614.2 23.687 1.506 29.06 156.27 1228.411 7.896.. 1.506 9.69 7. 14 61114.2 15.791 1.506 19.18 78.14 1614.2 15.791 1.506 19.38 156.27 322q. 4 7.896 1.506 9.69 78.14 1614.2 15.791 1.506 19.38 78.141 1614.2 15.791 1.506 19.38 78.14 16111.2 23.687 1.506 29-06 78.114 1614.2 15.791 1.506 19. 38 78.11B 1614.2 15.791 1.506 19.38 156.27 1228.11 11.843 1.506 14.53 798.14 1614.2 11.582 1.506 38.75 156.27 1228.4 7.896 1. 506 9.69 00 00 END OF FTLE ON WDATA(.18173) CAUSES A PrETTIP' TO MTS. EECIITrO9 TER!INATED 00:20:29 T=1.509 RC=0 S.J,

369 APPENDIX G WAVE GROWTH THEORY ON LIQUID DROPLET SURFACE TQ understand the droplet break!-up mechanism, the use of unstable wave growth theory has been suggested. (B,' R5) A summary of these theories is presented here. Considering the moving liquid droplet as the frame of reference, the following schematic describes droplet motion. Uf. u~ I Acc -d f - -- a a The gas-liquid interface is assumed smooth and potential flow to exist in the gas. Then wave displacement can be expressed, =Y o + = yo + oei(kX-C5t) From the solution of Lamb(L6) or Chandrasekhr(C17) Pu*g.f ^V 2, t~f1/2 _ PfUfl+ g | (U[Uf) 2 p (f-) P + KS \ k J g \f gg )?f 9 (Pf + Lg (f +g) (G.1) (G.2) Replacing and where, PfUf + pgUg 9f + 99 9 = - A DO3 2 C2) UC2 -I1 4 o ) (G.3) (G.4) 3 U 2 A=' C I 2 (-?1f DO (G. 5)

370 Equation (G. 2. becomes, 1/2 =v | + ^CT^12 \ pf LA. kS CG.6) k V r T 2+ _ 2 5 p+?'Y, Cl) C2) C3) If the sum of terms within the square root is negative, the wave will become unstable and grow. Each term in Eq. (G.6) can be interpreted as follows: (1) Relative velocity term between gas and liquid on the drop (Kelvin-Helmholtz term) (2) Acceleration term due to drag (Rayleigh-Taylor term) (3) Surface tension term Here (1) and (2) are destabilizing factors, while (3) is stabilizing. Since 9f >> pg (U >> Uf), (2) 9 f A 1 Pf 3 u2 1 (1) Eg k (UgU f2 Pg CD 9f D= 2i U g (2) _ 3 CD (G.7) * (1) 8 QD (G7 D) where, 2 - I_ Dynamic pressure of gas at drop surface -Q 2. Dynamic pressure of gas ahead of drop?gUCO (G.8) (Cll C9x + g) - Ul2 A pgUg2 rg? o g 2U L C32IT.2 0 2r o C) -2 A. WebQ

371 where?g U2 Do Web = - S-,-. CG 10) For unstable case, from Eq, CG, 61 i 6 - kV = iU 9 9f 2_T 2''f [L + T4 Do 217 21 D 1/2 A. Webt j (G ll) Rearranging Eq, (G.ll1 the dimensionless time, T6 is T = i2l <y C-kV Uo Jf Do ~ /A l1 (/2J3 D3 2DQ Q(4 Do 2IT 21 Do -1/2 Web s D(Q +3 CD 4 2H DO 21r Web D -1/2 A. (G.12) The fastest growing wave length, %fi, is obtained by setting, " T = 0 Then Then Af D o 8T9 9C D 1/2 CDQ ( 4 Q2Web -1] (G. 13) Also the critical (or cut-off) wavelength, c is obtained by setting, PT, 00 then c - 0 4 Q1 3"tD'l 3 C 1/2 + D1 II-1] Q We Q eb (G.14) For large vaues of Web! two limiting cases can be obtained as, il Q= 0 Cl I Ac D 2* 2 QWeb

372 A;- 3'lr D0 QWeh ^Ta. = X rCL 3/2WebDi ii) Q2 < Web Xc 4 T Do | 3CDWeb o - - -OD Xf 4 T Do ~ CDWeh Tf = 4 t'f CD3/4w 1/4 CD Web

UNIVERSITY OF MICHIGAN 3 9015 03023 0018