THE UNIVERSITY OF MICHIGAN College of Engineering Department of Mechanical Engineering Cavitation and Multiphase Flow Laboratory Report No. UMICH 03371-18-T WET-STEAM TUNNEL FACILITYDESIGN AND PROGRAM OF INVESTIGATIONS by ~$* J. Krzyzaiowski Typing and ReproductionSupported by: National Science Foundations Grant No. GK-730 June 1972 * Dr habil., Institute of Fluid Flow Machinery of the Polish Academy of Sciences, Gdanisk, Poland, on leave in the University of Michigan within the National Academy of Sciences program of exchange.

ABSTRACT A new low-pressure wet-steam tunnel to be constructed and used in the Cavitation and Multiphase Flow Laboratory is described, and the research program for its future use is discussed. TABLE OF CONTENTS Page I. INTRODUCTION....ooo........... *............ 1 II. WET-STEAM TUNNEL FACILITY................ 3 III. PROGRAM OF EXPERIMENTS..................... 4 IV. ACKNOWLEDGEMENTS......................... 7 V. REFERENCES.................................. 8 VI. APPENDICES Appendix A: S o me Selected Parameters of the Test Facility........................... 9 Appendix B: Steam Supply and Location of the Test Facility.................. 16 Appendix C: The Drawings of the Most Important Parts of the Test Facility............. 18

I. INTRODUCTION Recently much attention has been paid to the problems of twophase flow, such as wet steam, fuel droplet-oxidiser, etc. Among others a certain amount of data has already been collected about the interaction between the gaseous phase and a relatively thick liquid layer along channels. Also the dynamics of individual droplets exposed to weak and strong pressure pulses in the gas was investigated. Extensive references may be found, e. g., in (Ref. 1) and (Ref. 2). Also at this university, a series of new experiments has been recently performed on the aerodynamic shattering of individual liquid drops exposed to shock wave impulse (Ref. 3). However, there is still little known about the stability of thin water films on the surface submerged in tranasonic and supersonic flows when surface tension and wettability comes into the picture. Also little has been done on droplet generation in the aerodynamic wake behind a body submerged in such a high speed two-phase flow. In this particular case there is involved: 1. The flow of a thin liquid layer of thickness of the order of -6 100. 10 m or less, with its stability and interaction with the transsonic and supersonic gas flow a question to be studied, and 2. The droplet stream generation, droplet shattering in the aerodynamic wake of a body submerged in transsonic and supersonic two-phase flow, which is proposed for study here, and particularly, the droplet stream structure. These problems are of interest for many different reasons, and their solution could be applied to numerous fields of engineering, i. e., evaporation and dispersion of fluids, different types of heat and mass exchange, cooling and condensation. They are also of interest because 1

2 of the erosion of fluid flow machines, particularly turbines operating on wet steam, metal vapors and vapors of low-boiling liquids. Recently, it has been shown that the structure of the droplet stream has not only significant influence upon the mean value of the droplet impact parameters in the turbine rotor blading (Ref. 4), but also the endurance of the material depends on the droplet stream structure (Ref. 5). Similar observations have been already reported, but the problem deserves further consideration. The investigation of the structure of the droplet stream in the aerodynamic wake have been carried on in IFFM for several years. Some of the results have been published, for instance, in (Ref. 6) and more recently in (Ref. 7). However, the investigations have been carried on using an air stream only. The lack of knowledge of the scaling lows of this complicated phenomena prevents drawing conclusions for other than air-water flows from this data. In particular the extrapolation of the available results for high-speed, low-pressure, wet- steam flow is not possible yet. There are urgently needed: 1. The formulation of the scaling lows for the phenomena under consideration and, 2. Collecting further experimental data as a basis of such generalizations. The purpose of this report is to describe the design of a steamtunnel facility, which would help to achieve these goals, being constructed in the Cavitation and Multiphase Flow Laboratory of the Mechanical Engineering Department of the University of Michigan. The wet-steam flow parameters have been chosen to be as close to the steam-turbine flow parameters in actual wet-steam turbine stage as possible. It will make possible: 1. The repetition of the IFFM experiments for significantly different Reynolds, Mach, and Weber number,:: Institute of Fluid Flow Machines, Polish Academy of Sciences, Gdansk, Poland.

3 2. The collection of the experimental data applicable directly to the theory of droplet impact in the steam turbine blading. Joint use of the experimental data of IFFM and this laboratory will certainly contribute to the achievement of the ultimate generalization of the experiments. It will also contribute to the fruitful future cooperation between both groups. II. WET-STEAM TUNNEL FACILITY It has been assumed that the desired flow parameters in the test section of the wet-steam tunnel facility should match as closely as possible the average steam parameter of the large steam turbines. They are: kG p = I 0.095 2 abs ("1.35 psia) x4 = O. 970 4 t4 = N 44. 50C (' 112F) 3 3 v = 15 (- (^243 ft /lb) 4 kG The rest of the parameters in the test loop depend upon the steam velocity in the test section and its cross-sectional area. Some of these parameters, such as the enthalpy drop h in the nozzle, the pressure, P3, temperature t3, specific volume v3, quality x3, as well as steam- and cooling-water demand, and also a schema of the facility are shown in Appendix A. It has been assumed that the steam expanding in the nozzle from (p3, x3) to (p4, x4) is a wet-steam, generated from saturated steam, suitably throttled and cooled. The steam is The symbols are defined later in the report.

4 cooled by injecting the cooling water into the saturated steam line and if needed, into the silencer. Further details of the steam tunnel depend upon steam supply and the location of the facility. The pertinent argument has been considered in Appendix B. In Appendix C the drawings of the most important elements of the test facility are collected. III. PROGRAM OF EXPERIMENTS A. Minimum Program. This should include the investigation of the size distribution function f of the droplet stream in the aerodynamic wake of a flat n plate as a function of the steam velocity cl. The distance z downstream between the trailing edge of the plate and the area of investigation should be about z = 150 mm. The structure of the droplet stream should be defined by 1 n(r* cl) f(r~) i nc)r=,c1) fn(r* - *Ar N where r - droplet size Ar - the droplet size interval (in order to meet the conditions applied in IFFM take, if possible, As a matter of fact, the structure of the steam in the silencer should be similar to the structure of the steam expanding in the steam turbine. It means that the average droplet size of the liquid phase should be of the order 1. 10-6m or less. This condition may be realized when reducing the steam pressure in a special turbine-expander. Taking into account that the subject of the investigation is the dynamics of the water film and the droplets, it appears that the adopted method here for wet steam generation is adequate. In order to prevent the undesired heat and mass exchange in the test section, attention has been paid to the injection of water droplets: the injected droplets should be as small as possible. One may achieve this by maintaining the injection pressure of water as high as possible.

5 -6 6, Ar = 50. 1 m or 100. 10 m.) n(r*,c,) - the average number of the droplet of the given size r* per unit area per given cl. N - the average number of all the droplets visible per unit area and cl. The average n(r,, Cl), N or n(r*, cl)/N should be established by means of a sufficient number of photographs. The arrangement for taking picture s is shown below: -- ADAPTOR ABOUT 100+150 mm. — l!.c,...d ___ "~''~... DIM SCREEN FLASH The distance between the droplet stream and the lense should be about 100 r 150 mm. The light source: Xenon flash light, the time of exposure: order of magnitude of 10 sec. (a spark-light source may be used). The results should be presented in the form of a curve f = fn(r, cl), where n n I r* and cl is dependent variable and parameter, respectively. For the interpolation of the experimental results, the NukiyamaTanassava function f = NlrNZ eN3 r n 1* should be used. The Weibul function f = 1 - exp(r*N4) may also be used. N1 2 3,3,4 are the sought experimental coefficients.

6 The establishing of the experimental relationship between these coefficients and c1 is the first goal of the experiment. In order to define and maintain chosen flow conditions of the steam expansion at least the following parameters should be measured: GST(t2 P2' APT) - the steam flow, p2 t the pressure and temperature of the steam t2 between the throttle valve, and the orifice, GCW(tC' PCW' Aw) the cooling water flow, this quantity may be measured either by means of a standard orifice or (better) by means of a flowmeter, t - the temperature of the cooling plate, GW P3 = p - AP23 the static pressure of the steam in the silencer, p4 = p3- Ap3j in the test section and in the condenser, G - the flow of water supplied onto the surface of the flat plate,. PL TP - the parameters of the water injected to the condenser, PW P4C - steam velocity distribution in the test section. B. In the next stage of the experiments the variation of the following parameters may be considered. 1. the distance z 2. the static pressure p4 in the test section 3. the water flux over the surface of the plate 4. the shape of the trailing edge of the plate The results obtained may consitute a basis for considering a generalized dimensionless relationship between the parameters of the

7 size distribution function f and a set of Reynolds, Mach, and possibly n Weber numbers. C. Application of the Fast Camera. This may contribute to the understanding of the mechanism of liquid droplet shattering. Attention should be paid to the following que stions: 1. What is the type of the droplet break-up (bag-type, strip-type), and its relationship to the flow parameters? 2. What is the mechanism of the water film break-up in the vicinity of the trailing edge of the submerged body; what is its relationship to the flow parameters and the shape of the body? 3. What is the influence of blowing-out or sucking-in of the water filament formed at the trailing edge of the plate? D. Investigation of the Stability of the Water Film. This should contribute to the understanding of the conditions of the flow, which stimulates the breaking-up of the continuous water layer into rivulets, the onset surface waves pattern etc. Of particular interest might be the influence of the wettability of the plate surface on the film flow pattern and droplet generation. IV. ACKNOWLEDGEMENTS This report has been prepared by the author to describe a large part of the work which he has accomplished during his stay at the Cavitation and Multiphase Flow Laboratory of the University of Michigan, and to help guide the future program resulting therefrom. The stay has been organized within the exchange program between the Polish Academy of Sciences, Warsaw, and the National Academy of Sciences, Washington, D. C. The author is grateful to his host, Professor F. G. Hammitt,

8 whose assistance made this design and the construction of the test facility possible. The author also trusts that the joint efforts of IFFM and this laboratory at UM will contribute to an effective and fast exploration of the problem. The results of the experiments of both laboratories will be complementary and will then be the subject of future periodical joint reviews. V. REFERENCES 1. Brodkey, R. S. "The Phenomena of Fluid Motion, " Addison-Wesley Publ. Comp., Reading, Massachusetts; Menlo Park, California; Don Milos, Ontario; London, 1967. 2. Wallis, G. A. "One-Dimensional Two-Phase Flow," McGraw-Hill, 1971. 3. Ranger, A. A. and Nicholls, T. A. "Aerodynamic Shattering of Liquid Drops," AIAA Journal, Volume 7, No. 2, 1969. 4. Krzyzanowski, J. "The Correlation Between Droplet Stream Structure and Steam Turbine Blading Erosion, " Technical Report No. 147, Department of Mechanical Engineering, University of Delaware, Newark, Delaware, December, 1971. 5. Krzyzanowski, J. "On the Erosion Rate-Droplet Size Relation for an Experimental Stand with a Rotating Sample," Report No. UMICH 03371-17-T, the University of Michigan, 1972. 6. Puzyrewski, R., and Krzeczkowski, S. "Some Experiments on the Droplet Generation and Its Motion in the Aerodynamic Wake, " (in Polish), Trans. of IFFM, No. 29-31, Warsaw, Poland, 1966. 7. Weigle, B., and Severin H. "The Investigation of the Relationship Between the Gas Velocity, the Droplet Stream Structure, and the Erosion Rate-Time Patterns, "(in Polish), Report No. 273/71, IFFM, Gdansk, 1971.

VI. APPENDICES Appendix A: Some Selected Parameters of the Test Facility The Scheme of the Facility --. 34 I I i ip SE MIXING CTONDENSER GPL P iG) G \,t f..W pW Ah j - 4 -. FLOAT VALVE o I ~ ^

10 Some Parameters of the Expansion: ASSUMPTION I 1. Nozzle Cross-Sectional Area: 0.1 x 0.1 m 2. Steam parameters in @: kG 4 = 0.095 cm2 abs (1.35 psia) x4 = 0.970 t = 44.50C (112~F) 3 3 = 15 (243 ft. /lb.) V4 kG5. c4. m 300 200 100 Remarks kcal/kG 10. 70 4. 80 1. 20 h |Btu/lb 19.25 8.64 2.16 h c 380 kcal/kG 0.43 0.19 0. 048 ) zAh Ah =( 1 1) Btu/lb 0.77 0.34 0.086 8380 kG/cm 0.1335 0.1088 0.0983 P3 psia 1. 90 1.55 1.40 X3 - - 0.984 0.976 0.972 ~C 51.10 47.04 45.09 - -~F | 124.0 116.0 113.0 m3/kG 11.30 13.57 14.67 v3 ft3 bs 181 217 235 3 3x3 V3 kG/s 0.20 0.13 0.067 ~lb/h | 1590 1030 523 kG/s 10.0 6.50 3.40 G = 50 G G lb/h 79300 51500 26900 for. surface cond.

11 ASSUMPTION II 1. Nozzle Cross-Sectional Area: 0.08 x 0.08 m 2. Steam Parameters in (: kG p4 = 0. 105 cm2 abs (1. 5 psia) x =0.970 t =46.70C (1160F) 3 3 = 14.1 G (226 ft3/lb) 4 kG kG GP 3.47 k (27500 lb/h) (55 GPM) C4 100 150 200 250 300 350 * 8 kG/s 0.0454 0.0681 0.0909 0.1135 0.136 0.159 GST lb/h 360 540 720 900 1080 1260 ST 1 76 51 38. 2 30.6 25.4 21.8 Gpw PW - --..,,,,,.. 7,... 7 14... kcal/kG 1.19 2.68 4.77 7.45 10.7 14.55 Btu/lb 2.14 4.83 8.60 13.40 19.25 26. 20 kG |- k abs 0.1075 0.1120 0.1205 0.1295 0.1400 0.1605 psia 1. 54 1.60 1. 72 1. 84 2.00 2.30 m3/kG 13.9 13.4 12 5 11.75 10.97 9.60 V3 ft /ib 222.8 214. 2 200.3 188.2 17. 174.1 153.7

12 ASSUMPTION III 1. Nozzle Cross-Sectional Area: 0. 08 x 0. 08 in 2. Steam Parameters in O: kG p4 =0.211 cm abs (3.0 psia) x =0. 970 t4 50 C (142~F) 33 v4 = 7.32 kG (117 ft3/lb) G = 3.47 kG (27500 lb/h) (55 GPM) PW s m c -- 100 150 200 250 300 350 4 s kG/s 0.0875 0.1311 0.1750 0.2190 0.2620 0.3060 GST | lb/h7 695 1040 1390 1735 2080 2430 ST 1 39.5 26.5 19.8 15.8 13.2 11.3 Gp ____________ kcal/kG 1.19 2.68 4.77 7.45 10.7 14.55 Btu/lb 2.14 4.83 8.60 13.40 19.25 26.20 kG abs 0.217 0. 220 0.227 0. 237 0. 248 0.262 c _____ _3_ _7 P3 psia 3.10 3.15 3.25 3.40 3.55 3.75 m3/kG 7.26 7.10 6.86 6.58 6.33 5.98 3 1.3 1.0 1... ft /ib 116. 3 114. 0 110. 0 105. 6 101. 5 95. 8

13 Estimation of the Quantity of Cooling Water GCW CW Assumptions: kG p4 = 0.211 cr 2 abs (3 psia) kG p3 0.246 c2 "abs (3.5 psia) i2 -i = 33.3 kcal/kG ( 60 Btu/lb) 23 r = 556 kcal/kG ("1000 Btu/lb) G = 0.177 kG/s (11400 lb/h) ST i2 - GCW GS. = 0.0106 kG/s CW ST r = 10 GPH Characteristics of a Free Flow Spray Nozzle Injection Pressure psi 30 40 60 100 Water Flow GPH 5. 1 5.9 7.2 9.2 In view of expected small efficiency of mass and heat exchange between the steam and water, 3 nozzles 1/8" (61. 61) have been chosen. McMaster-Carr Catalog, p. 696

14 Estimation of the Pressure in the Mixing Condenser Assumptions: i - c t =n c (t - t.) 4 p4 p o 1 ~ ~/ i4 ='v615 kcal/kG (~1110 Btu/lb) / 4 ti = 15 C (~59~F) to G 3.5 kG/s (27500 lb/h) PW - h - G p 19. 6 G =00 177 kG/s (1400 lb/h) S GST i +nc t. t 4= - 4 44.50C = 112F o c (l+n) P kG p = p4 =P 0.10 c2 = 1.35 psia Estimation of the amount of air sucked into the loop: 1. According to: Dubbel-Technische Taschenbuch, V. 2, 1953, one has to take into account 2-3 liters air/s per 1 kG steam/s. Hence, the amount of air sucked into the loop of the test facility GAR = G (2 r 3) liters/s = (4.4 r 6.5) 104 kG/s AIR ST 2. According to Hoefer K.-Die Kondensation bei Dampfmaschinen, Springer Verl., Berlin, 1925, G G PW ST B AIR 1000 + & 1000 ) 760 where kG G - the quantity of water ( ) AIR G m PW - volumetric flow of cooling water ( h ) 1000 lt = 1.8 + 0.01 L

15 L - the length of the steam line (m) GST the quantity of steam ( ) ST h B - the barometric pressure (mm Hg) As suming: kG Gpw = 3.5 kG/s = 12500 h PW h L = 5m G^ = 0.177kG/s = 637 ST h B = 760 mm Hg One obtains: G = 1.43 kG/h = 4.15 10 kG/s AIR In order to remove this amount of air out of the condenser the water ejector PENBERTHY, size No. 184 A may be used.

16 Appendix B: Steam Supply and Location of the Test Facility The following location for the test facility has been considered: A. The location in the G. G. Brown Laboratory, University of Michigan, North Campus, B. The location in the Automotive Laboratory, University of Michigan, North Campus. The advantages of the variant A are: Abundance of space, Steam supply from the boiler house (the boiler parameters: p = 125 psia, tl =v344 F, saturated steam flow = 10. 042 lb/h; when assuming the return of the condensed steam into the boiler, the steam flow in the test section is limited only by the capacity of the condenser, Existing steam and condenser installation; the condenser capacity: 2, 000 lbs stear/h with 70 F cooling water and vacuum about 2 inches Hg. abs. The disadvantages are: Higher steam parameters demand relatively sophisticated equipment for pressure/temperature reduction (large quantities of water have to be injected into the fresh-steam line), The troublesome moving of a fast camera from the Cell 252 at the Automotive Laboratory to G. G. Brown. Moreover, the fast camera is used in Cell 252 for other experiments, The problems with the efficient organization of the experiments and maintenance of the stand by the personnel of the Cavitation and Multiphase Flow Laboratory. With regard to these disadvantages, it has been decided to locate the facility in the Automotive Laboratory.

17 The advantages of this location are: Convenient operation of the fast camera, Smooth organization of the experiments, Steam supply from the boiler house (the boiler parameters: P = 20 psia, tl =s228 F, saturated steam flow-v8. 750 lb/h; the usable quantity of steam is restricted by the mixing condenser. In view of the discharge of condensed water into the sewer, the quantity of steam has been restricted to about 1, 400 lb/h). The disadvantages are: The need of the design and construction of a mixing condenser, The need of discharge of cooling water and condensed steam into the sewer, The limitation of the time of the operation of the stand to approximately one hour per run because of the waste of water in the boiler, The relatively low reliability of the prototype design of the mixing condenser, In further considerations it has been assumed as realistic to maintain in the mixing condenser the vacuum of the order p = 1.5 3 psia.. The relevant parameters in some selected points of the test facility are presented in the Appendix A. The calculations have been performed for c varying between 100 and 350 m/s. 4

18 Appendix C: The Drawings of the Most Important Parts of the Test Facility 1. The Installation of the Steam-Tunnel Facility in the Cell 252 of the Automotive Laboratory. 2. The Silencer Chamber and Nozzle, 3. The Silencer's Inlet Part 4. The Nozzle 5. The Transparent Test Section and Flat Plate 6. The Diffuser Section 7. The Condenser Originals on file in the Cavitation and Multiphase-Flow Laboratory, University of Michigan.