UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Experimental investigation of subcooled void growth for upflow and downflow at low velocities and low… Bibeau, Eric Louis 1988

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1988_A7 B52.pdf [ 9.42MB ]
Metadata
JSON: 831-1.0080866.json
JSON-LD: 831-1.0080866-ld.json
RDF/XML (Pretty): 831-1.0080866-rdf.xml
RDF/JSON: 831-1.0080866-rdf.json
Turtle: 831-1.0080866-turtle.txt
N-Triples: 831-1.0080866-rdf-ntriples.txt
Original Record: 831-1.0080866-source.json
Full Text
831-1.0080866-fulltext.txt
Citation
831-1.0080866.ris

Full Text

E X P E R I M E N T A L I N V E S T I G A T I O N O F S U B C O O L E D V O I D G R O W T H F O R U P F L O W A N D D O W N F L O W A T L O W V E L O C I T I E S A N D L O W P R E S S U R E by Eric Louis Bibeau B.Sc, McGill University, 1986 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF M A S T E R OF APPLIED SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES Department of Mechanical Engineering We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH C O L U M B I A September 1988 © Eric Bibeau, 1988 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of /MI^CVT. £ ftJ <b , The University of British Columbia Vancouver, Canada Date •Z.C/PS/ 8Q DE-6 (2/88) A b s t r a c t A two-phase experimental loop was designed and built to simulate the op-erating conditions of the SLOWPOKE reactor. Void growth was measured for both upflow and downflow for velocities between 0.07 to 0.46 m/s and at a pressure of 155 kPa. The buoyancy effect causes the Onset of Significant Void, OSV, to occur at higher subcooling for downflow than for upflow. This effect is maximum close to the bubble rise velocity (0.23 m/s) and decreases at higher velocities. The results indicate that bubble detachment is not the only critical parameter affecting OSV. The OSV correlations from the liter-ature did not predict the experimental results well. Investigation of the heat transfer mechanisms indicates that fully developed sub cooled boiling occurs prior to OSV. i i Contents Abstract List of Tables v List of Figures vi Nomenclature viii Acknowledgement xiv 1 Introduction 1 1.1 Research Program Overview 1 1.2 SLOWPOKE reactor 3 1.3 Importance of Void Growth 3 1.4 Void Fraction 4 1.5 Present Research Objectives 6 2 Literature Review 9 2.1 Review of Experimental Work 9 2.2 Void Growth Models 11 3 Experimental Description 17 3.1 Test Loop 17 3.1.1 Open Loop Circuit 21 3.1.2 Other Loop Components 22 3.2 Test Section 22 3.3 Power Supply 27 3.4 Instrumentation 27 3.4.1 Gamma Densitometer 27 3.4.2 Flow-rate 36 3.4.3 Power 36 3.4.4 Temperature 37 3.4.5 Pressure 39 3.4.6 Data Acquisition System 40 iii CONTENTS i v 4 Experimental Investigation 41 4.1 Test Procedure 41 4.2 Parameters Measured and Calculated 43 4.3 Error Analysis 45 4.3.1 Calibration of Gamma Densitometer 46 4.3.2 Dissolution of Air in Water 52 4.3.3 Wall Temperature Measurement 53 5 Sub cooled Boiling Heat Transfer 54 5.1 Introduction to Subcooled Boiling 54 5.2 Computation of Force Convection 56 5.2.1 Governing Differential Equations 56 5.2.2 Computational Details 56 5.2.3 Comparison of Computational Results 58 5.3 Experimental Results 58 6 Void Growth Results 68 6.1 Overview 68 6.1.1 Comparison of Void Profiles 71 6.1.2 Experimental Observations 74 6.2 Discussion of Experimental Results 74 6.2.1 Hydrodynamic Effects on Void Growth 74 6.2.2 Bubble Detachment 78 6.2.3 Flow Direction Effects 82 6.2.4 Heat Transfer Mechanism before OSV 84 6.2.5 Comparison of OSV Results with Models 86 6.3 Heat Transfer Results 91 6.3.1 Heat Transfer Coefficients 91 6.3.2 Flow Direction Effects 91 7 Conclusions and Recommendations 94 7.1 Conclusions 94 7.2 Recommendations 95 7.2.1 Test Apparatus 95 7.2.2 Void Growth 96 Bibliography 97 Appendices 103 A Description of OSV Model 103 B Experimental Void Growth Data 106 List of Tables 2.1 Experimental studies on void growth 11 2.2 Range of applicability of void growth models 12 3.1 Volume of major loop components 22 3.2 Dimension of annulus components 23 3.3 Heat losses through the copper tubes 23 3.4 Flow meter data 36 3.5 Sampling rates 40 4.1 Estimated error for measured quantities 47 4.2 Estimated error for calculated quantities based on Run #7 and Xeq= -7.9% 47 4.3 Void measurements for an annulus completely filled with water 51 5.1 Heat transfer correlations for sub cooled boiling and forced con-vection 65 6.1 Void growth experimental parameters 69 6.2 Range of experimental operating conditions 69 6.3 Void plateau before OSV and zero void check azero 70 6.4 Effects of flow direction on the distribution parameter C0 . . 85 6.5 Experimental OSV results and comparison to detachment model for upflow 87 B.l Nomenclature of symbols from computer output 107 B.2 Tabulated data for Run #15 from computer output 108 v List of Figures 1.1 Schematic diagram of the SLOWPOKE reactor [1] 5 1.2 Void fraction and temperature development in a heated pipe . 7 1.3 Flow patterns in two phase flow [13] 8 3.1 Schematic diagram of two phase loop 18 3.2 Photograph of two phase loop 19 3.3 Schematic diagram of the test section 24 3.4 Photograph of the test section 25 3.5 Spatial positioning and detail diagram of heater assembly . . . 26 3.6 Photograph of the power supply 28 3.7 Instrumentation diagram 29 3.8 Photograph of computer system and amplifiers 30 3.9 Gamma densitometer assembly with counter system 32 3.10 Detail drawings of gamma densitometer 33 3.11 Measurement volume intersected by a horizontal gamma beam and a vertical annulus. The gamma beam is coflimated with a square cross section 34 3.12 Calibration of induction coil 38 4.1 Determination of the OSV point 46 4.2 Error estimate on void fraction and equilibrium quality for Run #7 48 4.3 Gamma densitometer static calibration 49 5.1 Illustration of the subcooled boiling curve 55 5.2 Grid arrangement 59 5.3 Comparison of fully developed velocity profile without heat flux to laminar theoretical profile 60 5.4 Velocity profiles for conditions of Run #101 61 5.5 Temperature profiles for conditions of Run #101 62 5.6 Comparison between experimental and computational results for laminar heat transfer coefficient 63 5.7 Comparison of subcooled boiling experimental results for Run #101 with correlations and computational results 67 vi LIST OF FIGURES ™ 6.1 Comparison of Run #13 void profile with Mcleod: <f>= 30 W/cm2 71 6.2 Comparison of Run #9 void profile with Mcleod: <f>= 60 W/cm2 72 6.3 Comparison of Run #3 void profile with Mcleod: <fi— 90 W/cm2 72 6.4 Repeatability of results for upflow: Runs #3 and 4 73 6.5 Repeatability of results for downflow: Runs #5 and 6 73 6.6 Subcooled boiling photograph: <f>= 60 W/cm2, U= 0.16 m/s and Tin= 25 °C 75 6.7 Hydrodynamic effects on void growth for <fj= 30 W/cm2 and upflow 76 6.8 Hydrodynamic effects on void growth for <f>= 60 W/cm2 and upflow 77 6.9 Hydrodynamic effects on void growth for <f>= 98 W/cm2 and upflow 77 6.10 Hydrodynamic effects on void growth for 98 W/cm2 and downflow 78 6.11 Forces and bubble radius ratios at detachment for upflow . . . 80 6.12 Forces and bubble radius at detachment for upflow and 9— 30 °C 80 6.13 Forces and bubble radius at detachment for upflow and 9= 80 °C 81 6.14 Bubble radius and forces at detachment for downflow 82 6.15 Flow direction effects on void growth for <j>= 98 W/cm2 and U= 0.26 m/s 84 6.16 Flow direction effects on void growth for <f>= 98 W/cm2 and U= 0.37 m/s 85 6.17 Flow direction effects on void growth for <f>= 98 W/cm2 and U= 0.46 m/s 86 6.18 Flow direction effects on void growth for <f>= 60 W/cm2 and U= 0.23 m/s 88 6.19 Flow direction effects on void growth for <f>= 30 W/cm2 and U= 0.22 m/s 89 6.20 Void growth for Run #13: 30 W/cm2, U= 0.07 m/s and upflow 89 6.21 Comparison of Experimental OSV Values to the correlation of Saha and Zuber 90 6.22 Average Heat Transfer Results for Void Growth 93 A. l Contact angle 9 103 B. l Void profile I l l B.2 Wall and water temperature, heat transfer coefficients and Nusselt number 112 B.3 Mass flow rates, velocities, Reynolds number and outlet tem-peratures 113 Nomenclature viii Nomenclature A cross sectional area (m2) A\ parameter for heat transfer correlation ([51]) A2 parameter for heat transfer correlation ([51]) Bo boiling number (<j>/(rh ifg)) d bubble drag coefficient C0 distribution factor Cp specific heat (J/kg°C) Ca empirical factor for surface tension force (equation A.5) C\ empirical factor (equation A.6) C2 empirical factor (equation A.6) C3 empirical factor (equation A.6) Dh hydraulic diameter (m) Di inside diameter (m) DQ outside diameter (m) FB buoyancy force (fJ.N) FD drag force (fiN) Fs surface tension force (fiN) FT empirical factor for surface effects (Equation A.8) / friction factor, based on Fanning definition (TW/(PIV?/2)) g gravitational acceleration (m/s2) G mass flux per unit area (kg/sm2) Grb Grashof number (g(Tw - Tb)D^/u2) Gr* criterion for change in Hcon due to buoyancy (equation 5.6) H heat transfer coefficient (W/°Cm2) Nomenclature ix Hcon convective heat transfer coefficient (W/°Cm2) Hnb boiling heat transfer coefficient (W/°Cm2) heat of vaporization (J/kg) ii enthalpy of liquid (J/kg) if enthalpy of saturated liquid (J/kg) enthalpy of vapor (J/kg) I current (amperes) K conductive heat transfer (W/m°C) I length of heated section (m) tra length at which voids are measured (m) m mass flow (kg/s) N number of counts (counts) Na number of counts for air (counts) Nw number of counts for water (counts) Nud Nusselt number (hKi/Dh) Nu* Nusselt number (defined as </>Dh/(Ki(Taat - Tosv))) P pressure (kPa) Pe Peclet number (GDhCpi/Kt) Ph heated perimeter (m) Pr Prandtl number (fiCpi/Ki) Pw wetted perimeter (m) Q power input (W) R resistance of the heater element (ohms) Real calculated resistance of heater element (ohms) Re Reynolds number (UDH/VI) Ret, bubble Reynolds number at departure (equation A.4) Nomenclature X bubble radius at departure (m) r radius (m) n inside radius (m) To outside radius (m) s sensitivity ratio (equation 4.9) s velocity slip ratio {ug/u) St* Stanton number (defined as <j>/(GCpi(Taat — Toav))) T temperature (°C) Tb bulk temperature {°C) T+ non-dimensional temperature difference (CpipiUrl<f>(Tw T+ at bubble departure (equation A.10) T room temperature compensation (°C) T- inlet temperature (°C) T mean average bulk temperature (°C) Tmb bulk temperature at the onset of nucleate boiling (°C) Tout outlet temperature (°C) Touteal calculated outlet temperature (°C) T wall temperatures (°C) T average wall temperature along the heater (°C) T •Losv temperature at OSV (°C) T bulk temperature at gamma densitometer location (°C) AT, a t Tb-T.at (°C) Ttat - Tb (°C) u velocity of water (m/s) u0 bubble velocity (m/s) U average velocity of water (m/s) Nomenclature xi Ugi vapor drift velocity (m/s) UT shear stress velocity (^J(rw/pi)) U+ non-dimensional velocity (u/uT) V voltage (volts) Va calibration constant for air of void meter (%) Vw calibration constant for water of void meter (%) Vol volume (m3) v cross-stream velocity component (m/s) X true vapor weight quality (%) Xeq equilibrium vapor weight quality (%) Xoav equilibrium vapor weight quality at OSV (%) x distance along heater element (m) xi length of heater element (m) xa length of first copper rod (m) Xb length of first copper rod and heated section (m) y radial distance from the heated wall to inside the annulus (m) yb distance from the wall to the tip of the bubble (m) y+ non-dimensional distance (y/vi\J(rw/pi) ) y~b y+ at bubble departure for y= j/b (equation A. 7) Z position along the heated section (M) Greek Symbols a &OSV C^zero void fraction (%) void fraction at OSV (%) zero void calibration check for void meter (%) Nomenclature xii f3 temperature coefficient of volumetric expansion ("if - 1) H viscosity (kg/ms) v kinematic viscosity (m2/s) <f> heat flux (W/cm2) <f>cai calculated heat flux (W/cm2) ip parameter for heat transfer correlation ([50]) V'o parameter for heat transfer correlation ([50]) p density (kg/m3) a surface tension (N/m) 6 equilibrium contact angle (°) 9a advancing contact angle (°) 9r retreating contact angle (°) TW wall shear stress (N/m2) Subscript 6 bubble cal calculated con convective d departure / liquid g vapor gen generation in inlet to test section nb nucleate boiling out outlet to test section Nomenclature osv onb sat onset of significant void onset of nucleate boiling saturation Acknowledgement xiv Acknowledgement I wish to sincerely thank my advisor, Dr. Martha Salcudean for giving me the opportunity to work on a challenging research project. Her support, guidance, trust and supervision during the research program are greatly ap-preciated. I am thankful for the technical help, friendship and suggestions of Bruce Handson, John Richards, and Ernie Jones during the construction of the experimental loop. I am grateful for the participation and friendship of Dr. Ashok Mal-hotra and Dr. C.S. Yim during their sabbatical year at the University of British Columbia. Discussions and helpful suggestions of Zia Abdullah and Dan Poirier are also appreciated. The collaboration of Zhang Ku-Gang is acknowledged. Finally, I wish to thank Donna Petrovic for her love and happiness she brings me, to my family and friends for their love and comfort and to the meditation group who offer a spiritual dimension in my life. Chapter 1 Introduction Void formation is of considerable importance in the design of nuclear reactors and heat exchangers. This phenomena is proposed to control the power excursions in the SLOWPOKE reactor [1] and is being investigated in the present study. The experimental range of the investigation is relevant to the uprated SLOWPOKE reactor, where low velocities and low pressure prevail. 1.1 Research Program Overview The research program originated at the University of Ottawa, and is now being continued at the University of British Columbia. The overall purpose of the research program is to conduct a study on void growth, critical heat flux, pressure drop and boiling heat transfer for sub cooled conditions at low pressure and low velocities. The application of this research is directed to-wards the understanding of the heat transfer characteristics relevant to the SLOWPOKE reactor. The research may also apply to heat exchangers. The folio wing aspects have previously been investigated at the University of Ot-tawa: 1 CHAPTER 1. INTRODUCTION 2 • Critical Heat Flux: Critical heat flux was found to occur under slug flow conditions, and at lower heat fluxes than those predicted by the appli-cable correlations. Mechanisms for critical heat flux were discussed, and a correlation for its prediction was obtained (Rogers, Salcudean and Tahir [2] and Graham [3]). • Void Growth: Void growth was measured for a wide range of low ve-locities and heat fluxes, and for three different equivalent hydraulic diameters. Considerable void was measured prior to the 'Onset of Sig-nificant Void' (OSV). Experiments at high pressures showed practically negligible void prior to OSV [4,6]. The measurements were compared to the void growth models of Lahey [5], Rouhani [4], Bowring [6], Levy [7], and Saha and Zuber [8] (An overview of these models is presented in Chapter 2). Prediction of void fraction based on these models did not yield satisfactory results (Salcudean et al. [9,10,11], Rogers et al. [12], Mcleod [13] and Abdullah [14]). The void growth models in the literature [6,7] have been primarily devel-oped for pressurized nuclear reactors. The experimental data on which these models are based is for a much higher pressure range than that of interest in this investigation. The poor performance of these models can be attributed to the following factors: • At low pressure both the hydrodynamics and heat transfer mechanisms are different. • Models which predict the OSV reliably at high pressures, based on bubble departure, do not seem to be valid at low pressure. Many in-CHAPTER 1. INTRODUCTION 3 vestigators have recognized that bubbles detach from the heated surface well before void growth occurs (Dix [15] and Serizawa [16]). • These models are mostly empirical and apply only within the range of parameters investigated. • Single phase heat transfer is assumed prior to OSV. This assumption is not valid at low pressure (Section 6.2.4). • Thermal and hydrodynamic conditions as well as surface roughness plays a major role on void generation and growth. These factors have not been simultaneously accounted for in the previous studies. 1.2 S L O W P O K E reactor The SL0WP0KE-1 (prototype) and SL0WP0KE-2 reactors are designed for laboratory irradiation purposes by Atomic Energy of Canada Limited. The principal characteristics of the 2 kW reactor are low pressure, low tem-perature, low critical mass (low ratio of fission power to thermal power), pool type, relatively unattended operation and inherent safety mechanisms (negative temperature coefficient). These features provide an inherent safety mechanism and mechanical devices for cooling are not required. A schematic of the SLOWPOKE reactor is shown in Figure 1.1. Further information is available in [1,17,18]. 1.3 Importance of Void Growth A high negative temperature coefficient controls power excursions in the SL0WP0KE-2 reactor. As the power increases, a subsequent decrease in CHAPTER 1. INTRODUCTION 4 density of the water (moderator) is sufficient to decrease the reactivity of the fuel. The power is thus reduced until an equilibrium is reached. For the up-rated SL0WP0KE-3 reactor (2 MW), the negative void coefficient is also required to moderate the reactivity of the fuel. The negative void coefficient further reduces the neutron scatter due to the subcooled boiling along the fuel rod. The 2 MW reactor is designed for space heating. The negative void coefficient is an essential feature of the SL0WP0KE-3 reactor. A reliable simulation of the thermohydraulics of the reactor must be carried out to provide the designers with enough information to predict the power reduction, following a power surge, to avoid damage to the fuel rods. This simulation requires detailed information on the void fraction growth and critical heat flux over the anticipated range of operation. 1.4 Void Fraction Figure 1.2 shows the temperature, void fraction and bubble layer develop-ment along a uniformly heated channel. As the subcooled liquid enters the heated channel, both the bulk temperature and the wall temperature in-creases linearly. Point A is referred to as the onset of nucleate boiling. This occurs when the wall temperature exceeds the saturation temperature and bubbles start to form on the heated surface. Subcooled boiling occurs in the region between point A and point E. The density of nucleation sites increases in the downstream direction be-yond point A. Vapor bubbles within this region remain attached to the heated wall. Point B corresponds to bubble detachment. The bubbles slide along the wall downstream of point B, and the bubble layer grows until the first bub-ble is ejected into the core of the flow at point C. The OSV occurs at point OUTER CONTROL PLATE MECHANISM GUIDE TUBE G A S PURGING INLET CENTRAL CONTROL ROD MECHANISM CONCRETE SHIELDS SAMPLE STATION REACTOR CONTAINER CONTAINER WATER LEVEL POOL WATER LEVEL BERYLLIUM SHIM PLATES OUTLET ORIFICE BERYLLIUM REFLECTOR SAMPLE TUBE (VARIABLE RADIUS) INLET ORIFICE URANIUM/ALUMINUM FUEL ELEMENTS NEUTRON SOURCE SAMPLE TUBE REACTOR CONTAINER FUEL REGION HEIGHT 10 IN. DIAMETER 9 IN . CROSS-SECTION OF REACTOR POOL CROSS-SECTION OF REACTOR SLOWPOKE REACTOR I 3 § o Figure 1.1: Schematic diagram of the SLOWPOKE reactor [1] CHAPTER 1. INTRODUCTION 6 D where 'significant' void is present. The fluid bulk temperature reaches saturation and bulk boiling begins at point E. Distinct two-phase flow patterns develop as the fluid is heated (Fig-ure 1.3). Bubbly and slug flow are expected in the present investigation. 1.5 Present Research Objectives The objectives of the present investigation are as follows: 1. Design and build an experimental two-phase loop facility for the study of the heat transfer characteristics of the SLOWPOKE reactor. 2. Conduct upflow and downflow void measurements to establish the role of buoyancy on void formation. The buoyancy influence on void growth is uncertain and presently undocumented. To elucidate this topic, the effects of buoyancy on both the void fraction profile and bubble depar-ture are investigated. 3. Investigate subcooled boiling heat transfer. Substantial void is present near the wall prior to OSV, and its effect on the heat transfer mecha-nisms must be established. CHAPTER 1. INTRODUCTION 7 AXIAL POSITION "OCT F L O W rj tf>oV& est nnnggarr°raa o o i — o B D W Cu AXIAL POSITION Figure 1.2: Void fraction and temperature development in a heated pipe CHAPTER 1. INTRODUCTION 8 Bubbly Slug Churn Wispy-annular Annular Figure 1.3: Flow patterns in two phase flow [13] Chapter 2 Literature Review Void growth depends on many different parameters such as mass flow-rate, heat flux, inlet subcooling, pressure, channel geometry and surface roughness. The present investigation focuses mainly on the influence of buoyancy on void growth. Measurements of void growth for downflow are not reported in the open literature to the best knowledge of the author. Void fraction studies for vertical upflow with heat addition are discussed. Literature pertaining to other aspects of the investigation are introduced where applicable. Void growth has been investigated by many researchers due to its impor-tance to the nuclear and chemical industry. Measurements for a large range of parameters have been done and semi-empirical models for void growth have been developed. 2.1 Review of Experimental Work A number of different techniques are available for the measurement of void fraction. Of these techniques the gamma ray attenuation method is more widely used. Transient or time average values of void fraction are measured 9 CHAPTER 2. LITERATURE REVIEW 10 using the differential absorption of gamma rays in water versus vapor [19,20]. Other methods include X-Ray radiography [20], capacitance method [21], hot film method [15,22] and photography [23]. In addition to the void measurements, experimental investigations also provide information on flow patterns and characteristics of the bubble layer. The first experimental investigation to classify void growth into two regions was a photographic study by Griffith, Clark and Rohsenow [23]. These re-gions are described as follows: 1. A highly subcooled region. Liquid temperature near the wall is sufficient to permit nucleation of small bubbles. These bubbles either remain attached to, or slide along the wall and condense very rapidly. Most of the heat within this region is used to raise the liquid bulk temperature (region A to D in Figure 1.2). 2. A low subcooled region. Void fraction increases rapidly as the subcooling decreases. Bubbles survive in the liquid core (region beyond point D in Figure 1.2). The OSV (point D in Figure 1.2) cannot be accurately specified and several explanations have been given, for example, 'vapor clotting' [24] and 'surface covered by several layers of bubbles' [23]. Experimental void growth studies are summarized in Table 2.1. All the experiments are done with water. The experimental velocity, heat flux and pressure range, and geometry used are listed with each study. Most experi-ments were conducted at much higher pressures than in the present investi-gation. The only void measurements at low pressures are those of Edelman and Elias [20] and by Evangelisti and Lupoli [19]. The experiments of Evan-CHAPTER 2. LITERATURE REVIEW 11 Table 2.1: Experimental studies on void growth Source of Data Geometry P <f> G bar W/cm2 kg/m2s This investigation annular 1.5 30-90 70-450 Evangelisti [19] annular 1 44-89 607,1413 Edelman [20] tube 1 2-6 38-114 Ferwell [7] 4-16 24-68 529-1318 Poletavkin [6] tube 7-41 0-290 7-115 Rouhani [4] annular 19-50 60-120 130-1450 Christensen [6] rectangular 27-69 30-69 637-914 Eklund [4] 6-rod cluster 32-51 47-65 1345-1607 Griffith [23] rectangular 34-102 80-860 60-90 Nylund [52] 36-rod cluster 50 22-64 1110-1159 Fogila [6] rectangular 50-88 30-190 20-50 Egen [6] rectangular 138 19-126 9080-1165 Table taken from reference [13] gelisti are for much higher velocities while experiments of Edelman are for much lower heat fluxes. 2.2 Void Growth Models Bowring [6] proposed void growth to be a bulk fluid effect. Griffith, Clark and Rohsenow [23] proposed that void growth could be separated into two distinct regions (separated by the OSV point), and successive models [6,7,12,23,25] utilized this two region concept. Kroeger and Zuber [26] suggest that OSV must be established accurately in order to predict the void growth . Table 2.2 summarizes the models reviewed in this study. Most are semi-empirical, and require many equations and assumptions for their formulation. The models are classified into three categories: CHAPTER 2. LITERATURE REVIEW 12 Table 2.2: Range of applicability of void growth models Models Geometry P G (bar) (W/cm2) (kg/m2s) Levy [7] annular, rect. 19-50 60-120 130-1450 Staub [25] all 1-140 0-200 0-2000 Rogers et al. [12] annular 1.5 30-120 150-500 Unal [29] tube, rect., annulus 1-16 2-20 132-2818 Saha et al. [8] annular, rect. 1-49 2-121 400-1050 Rouhani [4] all 19-138 18-120 130-9080 Bowring [6] all 1-140 0-200 0-2000 Lahey [5] annular 19-50 60-120 130-1450 Griffith et al. [23] rect. 33-100 30-120 600-900 Ahmad [31] all 10-130 30-120 130-1450 Dimmick [33] all 2 30-125 320-1600 Serizawa [16] all 1-150 20-160 100-2000 All = tubes, annular, rectangular and rod clusters A— Transition based models: These models determine the OSV point, which is characterized by bubble departure in models [7,12,25]. After OSV is determined, the vapor weight fraction, Xosv, at the OSV point is calculated. An exponential relationship is assumed between the true vapor weight fraction, X, as function of Xosv and Xeg1. A correlation from the literature, which relate the void fraction, a, to X is chosen. For example, the correlation of Zuber and Findlay [27] relate the void fraction to the vapor weight by determining the slip velocity between each phase. The slip velocity is assumed to be a function of the vapor drift velocity, Ugi, (vapor velocity is influenced by buoyancy) and the distribution parameter, C0 (buoyancy effects axe function of the void location within the 1For a system in non-equilibrium X — mg/(rhg + m/), while for a system in thermodynamic equilibrium X — Xeq = (i/ — if)/ifg CHAPTER 2. LITERATURE REVIEW 13 general velocity profile). The models in this category are summarized as follows: Levy [7]: Levy assumes that OSV corresponds to the point where the tem-perature at the tip of a departing bubble is at the saturation temperature. He determines the bubble radius at departure by balancing the drag and the surface tension force on a single bubble, assuming a single phase turbulent velocity and temperature profile. The buoyancy force is neglected, being small as compared to the drag force. A relationship between the true vapor weight fraction and the corresponding equilibrium value is proposed. The void fraction is determined from the relationship of Zuber and Findlay [27]. Staub [25]: Staub uses a similar approach as Levy to predict OSV. He pro-poses an alternative expression to determine the bubble radius by balancing the buoyancy with the drag force . Rogers et al. [12]: Rogers uses expressions developed by Winderton and Al-Hayes [28] for the forces at bubble departure. He balances both the drag and the buoyancy forces with the surface tension force. The model only predicts OSV. This model is further discussed in Section 6.2.2 and the equations are presented in Appendix A. Unal [29]: Unal derives a simple model for predicting OSV based on dimen-sional analysis. He proposes that single phase heat transfer coefficient, the inlet subcooling and the heat flux are the independent parameters. Saha and Zuber [8]: Saha and Zuber propose that although a bubble may detach, the subcooling may still be too high to initiate growth. In order to predict OSV, they separate the flow into two regions based on the Peclet num-ber, Pe. For Pe less than 70000, OSV is established by thermal conditions (Nu* = 455), while for Pe greater than 70000, OSV is hydrodynamically CHAPTER 2. LITERATURE REVIEW 14 controlled (St* = 0.0065). Their model correlates experimental OSV data to within ±25%. They also assume an expression for the true vapor weight fraction, and using the expression of Kroeger and Zuber [26], they determine the void growth. B - Energy balance Models: Models are developed by assuming different heat transfer mechanisms and bubble condensation expressions within the two regions. An energy balance over an incremental distance along the heater element is performed. Expressions for the true vapor weight quality are assumed or derived from the energy balance. Correlations which relates the true vapor weight quality to the void fraction are used to determine the void growth or a slip velocity derivation is used.2 These models are summarized as follows: Griffith, Clark and Rohsenow [23]: Griffith et al. were the first to develop a model based on two regions. They assume that in the first region heat is removed by single phase heat transfer, and by condensation of the vapor bubbles in the second region. The bubble condensation rate is assumed to remain constant in the second region and the bubbles stay close to the wall. OSV is arbitrarily assumed to occur when 80% of the heat flux is used for steam generation. An expression for the void fraction is derived using a no slip condition. Bowring [6]: Bowring develops expressions for the true vapor weight fraction in each region. The effect of bubble condensation after OSV is shown to be negligible. The void fraction is obtained assuming a high value for the slip ratio. Bowring was the first to suggest that OSV can be determined on the basis of bubble detachment criteria. Rouhani and [4,30]: A steam generation 2 X and a are linked by the slip velocity as follows: cv/(l — a) = (p9/Pi)(X/S{l-X)) CHAPTER 2. LITERATURE REVIEW 15 term is derived by proposing a mechanism of heat removal for subcooled boiling. Bubble condensation is determined by introducing a condensation coefficient. A criterion is established for obtaining both the local subcooling and the void fraction at OSV. True vapor quality is obtained by balancing steam generation with bubble condensation. The void growth is determined using the Zuber and Findlay [27] relationship. Rouhani found poor agreement with experimental results for velocity lower than 0.5 m/s. Ahmad [31]: Ahmad derives an expression for the quenching liquid phase (heat responsible for raising the bulk temperature of the liquid) and for the boiling phase (heat responsible for the formation of steam) based on an energy balance. The temperature distribution is obtained using correlations for the OSV and for the condensation factor [32]. The true vapor weight fraction is expressed in terms of the temperature distribution. Void fraction is determined using a correlation for the slip velocity [32]. Lahey [5]: Lahey evaluates a boiling heat flux term based on an imposed heat flux and an effective single phase contribution. He determines an empirical condensation term. The two terms are then combined to form an expression for the true vapor weight fraction. OSV is determined on the basis of bubble departure. The void growth is determined using the Zuber and Findlay [27] relationship. The performance of the model is good for the high pressure range. C— Other models: Dimmick [33] proposes an empirical correlation for the void growth by fitting the data to an algebraic expression. Serizawa [16] develops an OSV model based on experimental observation of the rapid increase in void down-stream of the first bubble ejection (point C in Figure 1.2). The OSV point CHAPTER 2. LITERATURE REVIEW 16 is determined on the basis of bubble frequency, number of active nucleation cites, bubble diameter at departure and bubble slip velocity Uf,/u. Chapter 3 Experimental Description The test apparatus was built at the University of British Columbia. Simi-lar equipment had been successfully operated over a six year period at the University of Ottawa [13]. Several of the previous original components are incorporated into the present design. These include parts of the test section, power supply, void meter and data acquisition system. A new experimental description is presented in this chapter as several additions and changes have been implemented. The test apparatus can be divided into four parts: test loop, test section, power supply and instrumentation. 3.1 Test Loop Figure 3.1 shows a schematic diagram of the test loop. A photograph is shown in Figure 3.2. The test loop is designed to operate at the pressure, temper-ature and flow-rate range of the upgraded SLOWPOKE reactor. The main function of the loop is to feed water to the test section at pre-determined temperature, flow-rate and pressure. A forced convective flow is generated by the closed loop system. 17 Figure 3.1: Schematic diagram of two phase loop CHAPTER 3. EXPERIMENTAL DESCRIPTION 19 Figure 3.2: Photograph of two phase loop CHAPTER 3. EXPERIMENTAL DESCRIPTION 20 The main pump circulates water through the loop. Water first circulates through two heaters with a combined heating capacity of 12 kW, and is then fed to a 3 kW immersion heater. Each heater can be removed from the loop using a by-pass system. The desired temperature is achieved by setting the controls of the immersion heater. The flow can be diverted back to the pump using a main by-pass fine, possessing a flow control valve and an 'Eagle-Eye' flow meter. The flow-rate at the test section can be varied by adjusting the flow through the main by-pass fine. The water from the immersion heater is filtered, passes through a flow metering system, and is directed to the test section. A high or low flow range turbine meter is used depending on the flow-rate. A flow control valve produces a large pressure drop upstream of the test section, which stabilizes the flow. The instability is caused by the two-phase flow in the test section. All pipes feeding into the test section are made of thick rubber to absorb vibrations and electrically isolate the test section from the loop. The flow can be directed either to the top or the bottom of the vertical test section by the proper selection of valves. This allows upflow or downflow with respect to the heated rod. The two-phase mixture exiting from the heated section is directed to an enlarged cavity to allow the void to collapse. A thermocouple encapsulated in mercury inside the cavity measures the 'average outlet' temperature. The flow exiting the cavity is directed to a vertical cross-flow condenser. Cold water is fed through a copper coil positioned vertically inside the con-denser, and its mass flow-rate and outlet temperature are monitored. The two-phase mixture enters at a midway point in the condenser, and the single phase loop water exits at the bottom. This configuration ensures that no CHAPTER 3. EXPERIMENTAL DESCRIPTION 21 vapor can proceed to the next stage. The condenser simultaneously acts as a degasser by trapping and removing air from the water. The condenser also acts as a heat exchanger and removes heat from the loop water. A 4.5 kW immersion heater inside the condenser removes saturated air from the water when the loop is periodically heated to near boiling point for degassing pur-poses. The water passes through a 20 kW heat exchanger to further control the temperature before returning to the main pump. The flow-rate and the temperature in the heat exchanger are monitored. 3.1.1 Open Loop Circuit The SLOWPOKE reactor operates at constant outlet pressure of 155 kPa. In order to achieve a constant pressure at the outlet of the test section, it is necessary to de-couple static pressure from the influence of loop resistance, mass flow rate and temperature within the closed loop. This is achieved by implementing the method which was used in the previous loop. A filling pump re—circulates water out of and into the storage tank. A flow control valve positioned at the inlet of the storage tank regulates the pressure. This pressurized line is then fed directly to the outlet of the test section, either at the top or at the bottom, depending on the flow direction inside the annulus. Regardless of the conditions prevailing within the test loop, a constant static pressure can be achieved at the outlet of the test section, using the flow control valve to create the desired static pressure. The alternate function of the filling pump is to supply de-mineralized water from the storage tank to the loop when water is removed either via the bleed valves or drains. De-mineralized water is used, as in the SLOWPOKE reactor, to reduce mineral deposits on the heater surface. CHAPTER 3. EXPERIMENTAL DESCRIPTION 22 Table 3.1: Volume of major loop components Unit Diameter (m) Length (m) Volume (m3) Storage tank 0.58 1.0 0.264 Heater 9 kW 0.50 1.4 0.273 Heater 3 kW 0.50 0.58 0.114 Immersion heater 0.32 1.2 0.097 Condenser 0.33 0.92 0.079 1 " pipe 0.03 50 0.025 3/4 " pipe 0.02 25 0.007 r ?otal= 0.92 m 3 3.1.2 Other Loop Components Air bleed valves installed at high points along the loop are used to remove trapped air. The air bleed valves drain back to the storage tank. Temperature probes and pressure gauges are installed at various locations along the loop. A drainage system ensure complete drainage of the loop when the water is changed. Copper pipes, copper or brass fittings, and stainless steel or fiberglass lined tanks are use to prevent rust. Table 3.1 shows the dimensions of all the major loop components. The total volume of de-mineralized water required to operate the loop is approximately 920 liters. 3.2 Test Section The test section (Figures 3.3 and 3.4), is a four legged structure supporting a vertical annulus. Water and static pressure are applied to opposite sides of each of the two plenums. Electrical power is provided through two electrical connectors. The annulus consists of an inner electrically heated tube assem-bly, and an outer Pyrex glass tube, which allows visual observation of the CHAPTER 3. EXPERIMENTAL DESCRIPTION 23 Table 3.2: Dimension of annulus components Annulus Component D0 (m) Di (m) Length (m) Glass tube 0.029 0.0218 0.780 Stainless steel rod 0.0127 0.0085 0.480 Copper top rod 0.0127 0.0069 0.500 Copper bottom rod 0.0127 0.0069 0.540 Table 3.3: Heat losses through the copper tubes Heater Element Components Heat Generation (%) Stainless steel heater tube Upper copper lead tube Lower copper lead tube 97.6 1.2 1.2 void fraction. Table 3.2 shows the dimension of the annulus. The heater ele-ment (Figure 3.5) is fabricated using a 12.7 mm diameter 304 stainless steel tube. The tube is silver-welded at both ends to thick wall copper tubes. A constant wall thickness ensures a constant heat flux at the surface of the stainless steel tube. A small quantity of heat is also generated in the copper tubes. The ratio of the heat generated in the stainless steel tube to that of generated in the copper tubes is listed in Table 3.3. Loop water is fed through either the top or bottom plenum depending on the direction of flow and the static pressure is applied only at the outlet plenum. The plenums rest on rigid adjustable supports which permit the accurate positioning of the heater element with respect to the glass tube. The heater assembly is kept under tension to compensate for the thermal expansion of the metal and to prevent curvature of the long and slender heater element. A horizontal table controlled by a stepper motor coupled to a lead screw, spans the vertical annulus. The table is used for the position-CHAPTER 3. EXPERIMENTAL DESCRIPTION ELECTRICAL CONECTOR (2) STEP MOTOR ADJUSTABLE PLENUM SUPPORT INLET / OUTLET FLOW VOID METER GLASS TUBE GLASS SEAL (2) COPPER ROD INLET / OUTLET FLOW TENSION SPRINGS TOP LUCITE PLATE STATIC PRESSURE 1 FEED LUCITE ADJUSTABLE TABLE LEAD SCREW PLENUM (2) ADJUSTIBLE PLENUM SUPPORT STATIC PRESSURE FEED BOTTOM LUCITE PLATE BASE Figure 3.3: Schematic diagram of the test section CHAPTER 3. EXPERIMENTAL DESCRIPTION Figure 3.4: Photograph of the test section CHAPTER 3. EXPERIMENTAL DESCRIPTION 26 CHAPTER 3. EXPERIMENTAL DESCRIPTION 27 ing of the gamma void meter. The table, the top and bottom plates, and the two plenum supports of the test section are all made of phenolic. This eliminates the high eddy currents that would prevail if a metal plate is used to enclose the A.C. powered heater element. The two electrical connectors are air cooled. The entire test section is enclosed in plexiglass and a panic button to disconnect the power is always easily accessible. 3.3 Power Supply A 64 kVA A.C. power supply steps down 600 volts input line to the 4-32 volt range used. A manually controlled variac provides a coarse setting while a second variac, controlled by a stepper motor, provides a finer adjustment. The power supply is situated four meters from the test section to reduce magnetic interference. Copper bus bars carry the high current to the test section. The power supply is fully encased for safety and an air fan pro-vides the necessary cooling. A photograph of the power supply is shown in Figure 3.6. 3.4 Instrumentation An instrumentation diagram and a photograph of the data acquisition system are presented in Figures 3.7 and 3.8. 3.4.1 G a m m a Densitometer Void fraction is measured using a gamma densitometer. This instrument was designed at the University of Ottawa in collaboration with Atomic Energy of Canada Limited. The gamma densitometer operates on the basic prin-CHAPTER 3. EXPERIMENTAL DESCRIPTION 28 Figure 3.6: Photograph of the power supply CHAPTER 3. EXPERIMENTAL DESCRIPTION 29 Multiplexer Address Bardware A/D Converter Software CHAPTER 3. EXPERIMENTAL DESCRIPTION 30 Figure 3.8: Photograph of computer system and amplifiers CHAPTER 3. EXPERIMENTAL DESCRIPTION 31 ciple of differential absorption of gamma rays in water versus water vapor. Figures 3.9 and 3.10 show the gamma void meter and counter system respec-tively. The instrument rests on the traversing table described in Section 3.2. The void measurements are taken at a position four centimeters from the downstream end of the stainless steel heater (/„ m = 0.44 cm) in this study. A 10 MiCu Co-57 source, which primarily emits gamma rays at 122 keV, is contained in a lead source holder. A metal attenuator controls the intensity of gamma rays which are emitted towards the annulus. The collimator fo-cuses and collimates the gamma rays, intersecting the vertical annulus. The collimator consists of a channel 4.1 cm long, with a square cross section. This cross section is called the window. The dimensions of the square window are equal to the glass tube inner diameter (22 mm by 22 mm). The two-phase flow measurement volume in which the gamma rays are attenuated is there-fore a hollow cylinder with 22 mm outer diameter, 12.7 mm inner diameter and 22 mm height (Figure 3.11). A square window is used instead of the round window in the original design because the latter probably has a bias towards measuring slightly higher void fractions. This can be explained as follows. A round window collimates the gamma rays into a circular beam. The round beam upon intersecting the vertical annulus, produces a measurement volume which is biased to the region close to the heated wall. Since most of the void is located near the heated wall, a higher void fraction is measured using a round window. After the gamma rays have been attenuated by the two-phase mixture inside the measurement volume, they impact on a Nal (Tl) two inch diame-ter scintillator, which produces electron signals proportional to the intensity SCINTILLATOR HOUSING Nal SCINTILLATOR , COLLIMATOR MAGNETIC SHIELD GAMMA ATTENUATION Pre-Amp Power Spectroscopy Amplifier 0-10 Volt Sine Wave Signals Spectrum Stabilizer 0-10 Volt Sine Wave Signals Drift Compensated Single Channel Analyzer Digital^ Output Ratemeter 1000 Volt Power Supply Frequency Counter GO to Figure 3.10: Gamma densitometer assembly with counter system COLLIWTOR Figure 3.11: Detail drawings of gamma densitometer CHAPTER 3. EXPERIMENTAL DESCRIPTION 34 Figure 3.11: Measurement volume intersected by a horizontal gamma beam and a vertical annulus. The gamma beam is collimated with a square cross section of the gamma rays which hit the sodium ionide crystal. A ten tube pho-tomultiplier inside the scintillator amplifies the signal. A photomultiplier base (Ortec 276) further amplifies the signal before being transmitted, via shielded cables, to a counter system. The signal is shaped into sinusodal waves and amplified with a spec-troscopy amplifier (Ortec 452). The entire signal is fed to a stabilizer (Can-berra 2050), which automatically corrects for the system gain shift, by using the energy peak of the spectrum as a reference. The major source of drift can be attributed to the photomultiplier, since it is very sensitive to temper-ature changes and magnetic fields [34]. In order to reduce the gain drift to a minimum, the following procedures are implemented: • The Nal (Tl) scintillator with photomultiplier is shielded with three layers of 'Mumetal' from magnetic fields [14]. CHAPTER 3. EXPERIMENTAL DESCRIPTION 35 • The original collimator and scintillator housings were made of steel and lead. Heat was generated by eddy currents due to the electrical A.C. currents. The collimator was redesigned using aluminium and lead. As a result the heating due to eddy currents is reduced. • The void meter is air cooled to compensate for the heat transfer from the glass tube. The air flow-rate is adjusted so that the temperature inside the void meter does not exceed 30 °C. • The scintillator body and the photomultiplier base are grounded to the instrument. It was found necessary to electrically isolate the scin-tillator from its housing. This also prevents light from reaching the photomultiplier. The signal proceeds to a single channel analyzer (Ortec 406A) which counts only the signals that are produced by the 122 KeV gamma rays. The counter system is able to identify up to 50,000 counts per second. The sig-nal channel analyzer produces a frequency output which is then converted proportionately to a 0 to 10 voltage output by a ratemeter (Ortec 9349). A frequency counter displays the frequency output of the single channel ana-lyzer. The power is supplied to all the instruments by a 'Harshaw NH85' power supply. A 'Bertran 342' high voltage power supply provides the re-quired 1000 volts to the photomultiplier. The void fraction is determined from the following equation: where Vw and Va are the calibration voltage values when the annulus is filled with water and with air respectively. To determine each of these values CHAPTER 3. EXPERIMENTAL DESCRIPTION 36 Table 3.4: Flow meter data Turbine Flow Meters Item Itt Baraton 7285 L fti FLO-4 Linear flow range (kg/s) 0.06 - 0.6 0.006 - 0.06 K factor 1 cts °C (Pulses/m3) 1771657.53 curve K factor 0.75 cts °C (Pulses/m3) 1771657.53 curve Frequency range (Hz) 106.83 - 1068.3 114 - 1273 Linearity (%) ±0.25 nonlinear Repeatability (%) ±0.02 ±0.1 Calibration accuracy (%) ±0.05 ±0.05 24000 samples are averaged prior to a test. The sensitivity of the void meter increases with the increased difference between Vw and Va. 3.4.2 F low - ra te The flow metering system consists of two turbine flow meters with a mea-surement range of 0.006 - 0.6 kg/s. The characteristics of the turbine flow meters are shown in Table 3.4. A relationship between the frequency output of the meter and the flow-rate is supplied by the manufacturer, either as a K factor1 or in graphical form. The frequency signal is converted to a voltage signal using a frequency to voltage converter. 3.4.3 Power The voltage and current measurements are required for the calculation of the power. These measurements are carried out as follows: Voltage: The voltage signal is obtained across the heater element between the 1 Relates mass flow-rate to frequency by m = Kf CHAPTER 3. EXPERIMENTAL DESCRIPTION 37 copper electrical connectors and the plenums. The A.C. voltage signal is conditioned to produce a 0 to 10 D.C. voltage signal. Current: Current is measured using an induction coil (Hammond Class H) which generates an A.C. signal proportional to the current. The output of the coil is conditioned to produce a 0 to 10 V D.C. signal. The coil is calibrated by measuring the current using two different methods, as follows. A resistor is connected in series with a heater element fabricated with in-connel 706. The resistor is pre-calibrated at British Columbia Hydro Power Laboratory (53.1 microhms). The resistance of the inconnel heater is mea-sured within an accuracy of 0.02%. The current is calculated separately using Ohm's law and the voltage signals across the inconnel and the resistor. The conditioned output of the induction coil is simultaneously recorded. The current measurements are plotted in Figure 3.12. The two current measure-ments are within 1%. A polynomial fit between the current and the voltage output of the induction coil is obtained and is also shown in Figure 3.12. 3.4.4 Temperature The following temperatures are measured during each test: i) Inlet temperature: The inlet temperature is measured at one of the two plenums depending on the flow direction. An ungrounded, K type and shielded thermocouple (Thermoelectric K-18-U) is used. ii) Outlet Temperature: The outlet temperature is measured at a location one meter downstream of the outlet plenum to ensure homogenization of the fluid temperature. An ungrounded, K type, shielded thermocouple (Ther-moelectric K-116-U) encapsulated in mercury is used to obtain the average CHAPTER 3. EXPERIMENTAL DESCRIPTION 38 1600 0 200 400 600 800 1000 1200 1400 1600 CURRENT ACROSS INCONNEL Amperes Figure 3.12: Calibration of induction coil CHAPTER 3. EXPERIMENTAL DESCRIPTION 39 temperature of the fluid. iii) Wall Temperature: Six 30 gauge, K type, twisted thermocouples are threaded in the heater element and silver welded at positions specified in Figure 3.5. This technique of welding the thermocouple directly to the elec-trical element is widely used [35,36]. The welds are machined and the heater element is polished to insure a smooth surface finish. An eight channel thermocouple amplifier system, with cold junction com-pensation (Tcomp), is fabricated using: • Two AD2B54 four channel isolation amplifiers • Two AD1215 mounting cards • One AD2B56A cold junction with temperature probe The isolation amplifiers prevents ground loops from forming between the un-grounded wall thermocouples and the amplifier. Tests show that the wall thermocouple signal consists of an A.C. noise superimposed on the temper-ature signal. The A.C. noise is low enough (< 16 mV) to be rejected by the amplifiers for test section currents below 1000 amperes. 3.4.5 Pressure Static pressure is monitored using two Bourdon type pressure gauges ( ± 0.7 kPa). These gauges are placed on movable supports to align them with the top and bottom of the heater. The outlet pressure is maintained at 155 kPa. A differential pressure transducer measures either the inlet, outlet, or dif-ferential pressure across the test section. The output of the pressure trans-ducer is recorded by the data acquisition system. CHAPTER 3. EXPERIMENTAL DESCRIPTION 40 Table 3.5: Sampling rates Signal Channel # Sampling Rate(Hz) Samples/Measurement 1-6 1 50 T-•••in 7 2 100 Tout 8 2 100 T 9 1 50 V 10 1 50 I 11 1 50 p 12 1 50 a 13 40 2000 u 14 40 2000 3.4.6 Data Acquis i t ion System The instruments (thermocouples, pressure, power, flow-rate void meter) are interfaced to an A/D board which resides inside an Intel-II microcomputer. The computer is also connected to an HP plotter and a printer. The data acquisition program is written in Fortran and uses two 8086 Assembler sub-routines to communicate with the A/D board. The sampling details are listed in Table 3.5. The data is stored on floppy disks and processed on an Apollo workstation. Chapter 4 Experimental Investigation 4.1 Test Procedure The loop is heated for two days at 100 °C to de-gas the water prior to a test period. The total time required to do a void growth run is six hours. A run is carried out as follows: • A test begins by pressurizing the bop and bleeding the air trapped in the water. The main pump is activated and the temperature is controlled until the required loop temperature is achieved. • The loop and the instruments are set for either upflow or downflow experiments. • Flow-rate through the test section is adjusted to the desired value and the required heat flux is applied. The data acquisition system is activated. The loop is periodically bled to remove trapped air. • Power is removed and all valves leading to the test section are closed. Water is drained from the annulus and the void meter is calibrated for the air value Va. The system is then repressurized. 41 CHAPTER 4. EXPERIMENTAL INVESTIGATION 42 • Power is applied across the heater element for twenty minutes before calibrating the void meter for Vw. This procedure ensures that the void meter is calibrated as close as possible to the actual operating conditions. • The flow-rate, heat flux, pressure and inlet temperature are adjusted. • The heat flux is readjusted periodically to remain within 1.5% of the desired value. This adjustment is required to account for the fluctua-tions in the unregulated power supply and slight changes in the heater resistance. • The heat balance across the test section is verified by comparing T^t^, (Equation 4.2) and Tmt. If IT^^, - Tmt \ < 1.5° C the heat balance is considered to be acceptable. • The immersion heater and the cooling flow-rate of the condenser are adjusted to obtained a steady rise in inlet temperature (approximately 0.2 °C/min.). Measurements are taken at regular intervals and visual observations axe made. A test is terminated before damage to the test section (as a result of critical heat flux) occurs. • The void fraction reading (ctxero) is measured with water in the annulus to validate the calibration. • The heater element is removed, polished and checked for straightness before re-installation in the test section. The heater element is adjusted for concentricity with respect to the glass tube. CHAPTER 4. EXPERIMENTAL INVESTIGATION 43 4.2 Parameters Measured and Calculated The data acquisition and analysis is carried out by separate computer pro-grams. A description of the experimental parameters are listed below. Fluid properties are determined at Tmtan unless otherwise specified. 1. m : Mass flow-rate measured through the test section for a single measurement, (kg/s) 2. u : Calculated velocity through the test section for a single measure-ment. The velocity during an experiment varies slightly due to changes in the pressure drop across the test section caused by the two-phase flow, (m/s) 3. U : Calculated average velocity for the entire run. The velocity u is averaged over the whole experiment, (m/s) 4. Tin '• Measured bulk temperature of water at the inlet of the test sec-tion. (°C) 5. Tout '• Measured bulk temperature of water at the outlet of the test section. (°C) 6. Toutcal : Outlet temperature calculated from a heat balance across the heater. (°C) Toutcal = ^ + (4.1) 7. Tvmcal : Bulk temperature of water calculated at the gamma densitometer location. (°C) rp _ rp Q(hm/fy n\ lvm«" ~ ± o u i rhCPl ( ' CHAPTER 4. EXPERIMENTAL INVESTIGATION 44 8. a : Void fraction measured by the gamma densitometer from Equa-tion 3.1. (%) 9. Xeq : Equilibrium quality calculated at the densitometer location. Neg-ative for subcooled boiling and positive for bulk boiling. The 'negative quality' has no physical meaning, but it is a generally accepted way [1,17] of expressing subcooling in nuclear engineering. (%) Xeq = C p ' ( r " " " ' ~ r ' a t ) (4.3) 10. azero : Void fraction measured at the end of a run with only water in the test section. This procedure ensures that the void meter is still properly calibrated. (%) 11. V : Voltage measured across the heater, (volts) 12. I : Current measured through the heater, (amperes) 13. (}> : Required heat flux. (W/cm2) 14. 4>cai '• Calculated heat flux (W/cm2). VI +"1 = w w ) ( 4 - 4 ) 15. Q : Power input. (W) Q = $(Phl)10* (4.5) 16. Real : Calculated resistance of the heater element, (ohms) R^i = j (4.6) CHAPTER 4. EXPERIMENTAL INVESTIGATION 45 17. R : Measured cold resistance of heater element and used to perform a resistance balance. R is compared to R^i. At low currents R and R^i are equal and at high currents Rcai is greater than R1. (ohms) 18. Xotv : Equilibrium quality at OSV. (%) Determined from a void growth curve as follows (see Figure 4.1). • Select a point 'j' at which the slope of the void growth curve changes. • determine the intersection between: (a) The line a = constant which is found by averaging the void fraction from experimental points 1 to j (b) The line a = mXeq + 6 where m and b are determined from a least-square-fit for the experimental points j+1 to n. • Xotv is the value of Xeq at the intersection of the two lines. 19. TWl_e : Measured wall temperature at thermocouple locations. (°C) 20. Tw : Average wall temperature along the heater calculated by averaging the wall thermocouples TWl_t. (°C) 21. Toev : Calculated Bulk temperature of the fluid at the void meter loca-tion at OSV. (°C) 4.3 Error Analysis Errors for the directly measurable quantities are obtained either from spec-ifications provided by the manufacturer or estimated from experiments and 1The resistance of stainless steel increases with temperature CHAPTER 4. EXPERIMENTAL INVESTIGATION 46 -8 -7 -6 -5 -4 -3 - 2 - 1 0 1 Xeq Figure 4.1: Determination of the OSV point calibration and are listed in Table 4.1. The error is estimated by the er-ror propagation method [37] for the calculated parameters. The results are shown in Table 4.2 for conditions of Run #7 and for an equiUbrium quality of -7.9 %. A void profile curve with error bars is shown in Figure 4.1 for conditions of Run #7. The error analysis of the void measurement, temperature measurement and effect of air dissolution are discussed in the following. 4.3.1 Cal ibrat ion of G a m m a Densitometer A static calibration [13,19,38] is done using hollow lucite cylinders having a gamma absorption coefficient approximately equal to that of water. Cylin-ders of different diameters are inserted in the annulus to attenuate the gamma rays. A cylinder which completely fills the annulus simulates the Vw reading. The Va reading is obtained from an annulus containing air. Different Void fraction' readings axe simulated using cylinders causing an air gap to occur CHAPTER 4. EXPERIMENTAL INVESTIGATION 47 Table 4.1: Estimated error for measured quantities Measured Quantity Error (%) P 0.88 Tin 0.75 T 3.0 m 0.3 J 1 V 0.1 D0 0.09 Di 0.15 I 0.2 R 0.2 Table 4.2: Estimated error for calculated quantities based on Run #7 and Xeq= -7.9% Calculated Quantity Error (%) Q 1.1 <j>cal 1.8 U 0.9 Real 1.1 Toutcoi 1.2 Xeq 0.16* T 1.2 * values in quality CHAPTER 4. EXPERIMENTAL INVESTIGATION 48 z o u < c u. Q 5 > •7.1 -T •» - I I .« .4.6 -4 -1.1 -» -2.6 EQUILIBRIUM QUALITY % Figure 4.2: Error estimate on void fraction and equilibrium quality for Run #7 either between the stainless steel tube and the lucite (inside air gap), or be-tween the glass tube and the lucite (outside air gap). The volume of the air gap is accurate within 0.6%. The static calibration results are shown in Figure 4.2. The average dif-ference between the measured void fraction and the reference is less than 2% void. Values measured with the inside air gap for all cases over-estimate the true void fraction, while those with the outside air gap under-estimate it. This difference can be partially attributed to the influence of the radial distribution of void fraction [39], but is more likely caused by the calibra-tion of the void meter. The accuracy of the void meter is estimated to be within 2.5% void based on the above analysis. The following errors are not accounted for the static calibration and are discussed in the following: CHAPTER 4. EXPERIMENTAL INVESTIGATION 49 TRUE VOID FRACTION % Figure 4.3: Gamma densitometer static calibration CHAPTER 4. EXPERIMENTAL LNVESTIGATION 50 1. Drift: Photomultiplier drift may cause the calibration constant Va and Vw to change. When the stabilizer is not in operation, the voltage drift for the Vw signal is unacceptable. A voltage drift of 0.14% per hour for Vw is observed when the stabilizer is in operation. This corresponds to an error of less than 0.5% void for a typical two hour operation period. The error induced due to the drift of Va is small at low void fraction (less than 20%) values (see Equation 3.1). 2. Dynamic effects: The error due to dynamic effects of the void are ne-glected. 3. Sampling rate: The sampling rate is six times greater than the time constant of the rate meter, therefore errors in measuring the frequency output of the single channel analyzer are small. 4. Magnetic fields: The error due to magnetic effects is not measurable as the gamma densitometer is well shielded and compensated for drift. 5. Temperature The void meter is air cooled and temperature effects are negligible. Table 4.3 shows typical 'void' measurements for an annulus completely filled with water (a= 0%). The average void fraction measurement is 0.13% void with a standard deviation of 0.69% void. This illustrates that the void mea-surements are repeatable. Statistical error in densitometry is due to the random nature of radio-active decay. Equation 3.1 represents ensemble average values of the count rate. Statistical error in measurement decreases as the sampling size N in-creases, or if the data averaging process is extended [34]. In integral counting, CHAPTER 4. EXPERIMENTAL INVESTIGATION 51 Table 4.3: Void measurements for an annulus completely filled with water Measurement # a (%) 1 0.05 2 0.59 3 -0.68 4 -0.80 5 0.97 6 -0.75 7 0.56 8 1.50 9 0.13 10 -0.02 11 0.48 12 -0.48 Avg= = 0.13% the statistical error is given by [38] : y/N € = (4.7) N sfN Linearizing Equation 3.1 and assuming no statistical error for Va and Vw (averaged over a fifteen minute period) and using Equation 4.7, the statistical error for void measurement CQ can be approximated by [38] : 1 Where the sensitivity s of the system is defined by: N a - N w s = N (4.8) (4.9) In the present work Na « 20,000 and Nw « 12,000 so that the sensitivity is 50%. If Nw is assumed instead of iVa, then the statistical error ta is ±0.2% for measurements taken over a 50 second period. It is concluded that statistical CHAPTER 4. EXPERIMENTAL INVESTIGATION 52 errors are negligible for the present study. A similar analysis and conclusion is reported by Lasschn [34]. The major source of error in the densitometer is due to calibration. Based on the analysis in this section, the measurement error of void fraction is estimated to be ±3-5% void fraction. This is comparable to estimates of Schrok [39] and Edelman [20] who report errors within 5% void fraction. 4.3.2 Dissolution of A i r in Water A small amount of air is dissolved in the water circulating in the loop. Mcleod [13] has shown that the release of this air due to heating of the water does not influence the void measurements appreciably. Water was boiled in the annulus for several hours and air was simultaneously bled from the loop. The void fraction measured during this period was constant. In addition, the 'effective' void fraction due to the release of air when water was heated from 20 °C to Tonb was found to be below the resolution of the gamma void meter. McAdams [40] has shown that the presence of dissolved air causes an earlier appearance of bubbles when water is heated. He reported that the dissolved air can be removed (to 0.3 cc/l) by boiling the water in a de-gassing tank and venting the system. He observed that the onset of nucleate boiling occurs sooner with dissolved gases in the subcooled liquid. Muller-Steinhagen et al. [41] reported that the heat transfer coefficients increase due to the liberation of gases during subcooled boiling. This increase is caused by the greater agitation of the released gases. In the present investigation visual observations revealed small bubbles in the subcooled liquid (near the glass tube) prior to de-gassing. The following CHAPTER 4. EXPERIMENTAL INVESTIGATION 53 procedure is used in the present study: • The loop water is initially circulated for two days at 100 °C. • The storage tank is covered and a rubber disk is placed on the surface of the water to reduce the area of contact between the water and the environment. • The water in the immersion heater is heated to near boiling point every night and air is bled in the morning. • Air is bled from the loop during each run. The volume of dissolved oxygen in the loop is measured to be less than 2.5 cc/l using the membrane electrode method [42]. The solubility of oxygen in water at room temperature is 5.93 cc/l [43], which indicates that the de-gassing procedure removes more than 50% of the dissolved oxygen. 4.3.3 W a l l Temperature Measurement The signal of the wall thermocouples has an A.C. noise which is filtered by the amplifiers. The error induced by the A.C. noise is small for heat fluxes less than 30 W/cm? (1000 A), and grows progressively with current. It is possible that oxidation of the thermocouple surface causes some A.C. rectification at high currents. Results for heat fluxes < 30 °C are presented in this thesis. Chapter 5 Subcooled Boiling Heat Transfer Forced convection and subcooled boiling heat transfer coefficients are mea-sured to enhance the understanding of void growth. Results for a run (#101) with heat fluxes less than 30 W/cm2 are reported. Other results for higher heat fluxes are not presented due to the wall temperature measurement prob-lems discussed in Section 4.3.3. The heat flux is varied from 30 to 1.5 W/cm2 (decreasing heat flux). The velocity is 0.076 m/s and the inlet temperature is 29 °C (83 °C subcooling). Boiling hysteresis (increasing and decreasing heat flux) and effects of dissolved air on the boiling heat transfer coefficient are not examined. 5.1 Introduction to Subcooled Boiling Figure 5.1 illustrates a typical variation of the heat transfer coefficient with heat flux for subcooled boiling [44]. Subcooled boiling can be divided into three regions. In the 'Forced Convective Region' heat is removed by forced convection. The heat transfer coefficient depends on velocity and is relatively 54 CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 55 4> Figure 5.1: Illustration of the subcooled boiling curve independent of the heat flux. The wall temperature increases linearly with the fluid bulk temperature until ATtat is equal to ATonb and boiling is initi-ated. A sharp increase in the heat transfer coefficient at the onset of nucleate boiling causes a decrease of the wall temperature at that point (see Figure 1.2). In the 'Partial Nucleate Boiling Region', convective heat transfer be-tween the nucleation sites and boiling heat transfer are both significant. The wall temperature increases with the heat flux until the 'Developed Subcooled Boiling Region'. In this region the heat transfer coefficient is independent of velocity and subcooling, and dependent on heat flux. The heat transfer mechanisms are similar to those of saturated pool boiling. CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 56 5.2 Computation of Force Convection A finite-difference flow computation is performed to model the force convec-tion region. The computation is only valid for heat fluxes lower than ^onj, (single phase flow). 5.2.1 Governing Differential Equations The flow in the vertical annulus is assumed to be laminar and axi-symmetric. The partial differential equations for the conservation of mass, momentum and energy are written for steady state in Cartesian tensor notation as follows [45]: Mass Conservation Equation duj _ dxj Momentum Conservation Equations 9 , x dp d , ,dui d u j . . Energy Conservation Equation 5.2.2 Computat ional Details The TEACH code [46] is used to solve the governing equations using a finite-volume approach. Figure 5.2 shows the computational mesh and the sym-bols used in this section. A 38x12 staggered grid is used. The entrance (5.1) CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 57 length (0.15 m) and the recovery length (0.15 m) are modelled with the heated length (0.48 m). All properties are calculated based on the tempera-tures at the nodes. The boundary conditions are applied as follows. North Boundary: The North boundary represents the glass tube. Adia-batic wall conditions are used such that at r = ra: u = 0 0 < x < xi v = 0 0 < x < xi dT — = 0 0 < x < x; or South Boundary: The South boundary represents the heater element. Wall conditions with heat flux boundary conditions are used such that at r = r,-: u — 0 0 < x < x/ u = 0 0 < x < xi dT •7— = 0 0 < x < xa and x0 < x < xi Or Q = Qgen xa < x < xb West Boundary: The East boundary represents the inlet to the test section. Slug flow is assumed such that at x = 0: u = Uiniet r, < r < r0 v = 0 TV; < r < r, T = TinUt ri<r<r, o o East Boundary The West boundary represents the outlet. Fully developed flow 'outlet' conditions are used at x = x;: u = v! n < r < ra CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 58 v = v' r, < r < rD T = T Ti<r<r0 where the ' refers to the previous node which is located upstream The energy equation is coupled to the u-momentum equation by the buoyancy effect. The Boussinesq approximation is used: FB=gPVol0(T ) (5-4) Further details on the TEACH program are given in [47]. 5.2.3 Comparison of Computat ional Results Results for the fully developed velocity profile without buoyancy effects com-pare favorably with the theoretical profile for laminar flow given by Lundren [48] (Figure 5.3). The results for the velocity and the temperature field are shown in Figures 5.4 and 5.5 respectively for a heat flux of 2.64 W/cm2 and for conditions of Run #101. The velocity gradient near the heated wall is steeper due to the buoyancy effect. The experimental heat transfer coeffi-cients along the wall are compared with the computed results in Figure 5.6. The computed values agree well with the experimental results. The heat transfer coefficient values axe high at the entrance and become relatively constant downstream when the flow is fully developed. 5.3 Experimental Results Figure 5.7 shows the results for single phase and subcooled boiling heat trans-fer coefficients. The experimental and computational results are presented with correlations from the literature [49,50,51,52]. The experimental heat xa (0.15 m) x\ (0.78 m) xh (0.63 m) L tx I 1 2 3 ... X | Entrance Length _ | Heated Length Recovery Length 1 1 N S Centerline of Annulus Figure 5.2: Grid arrangement Figure 5.3: Comparison of fully developed velocity profile without heat flux to laminar theoretical profile CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 61 VELOCITY PROFILES WITH BUOYANCY EFFECTS FOR RE= 857. PARAMETERS UPFLOW GRAVITY < SCALING INIFT TEMPERATURE ( C (€BT Flirt IK/K..2). INLET VELOCITY IM/S1-l< 29.0 26400. o.oe X SCALING* T SCflLINC-VEL. FACTOR" 0.42 8.37 1000.0 ENTRANCE LENGTH= 0.15 co > 63 •> 7 7 -+ * -» * 37 ^. I 34 • 7i ^ et ^ 110 ^ ice - > B2 49 19 114 ^ 107 k 82 49 19 HERTED LENGTH= 0.48 ZZ 32 Z^ 30 Z> 26 Z> 27 Z> 25 Z> 24 Z> 23 I ^ S4 ^ 60 ^ S7 ^ SS ^ 52 ^ 90 ^ 47 ' > 95 ^. 91 ^ 87 ^. 84 ^ 91 ^ 78 ^ 75 • ^ 119 ^ 110 ^ 108 ^. 106 ^. 104 ^ 103 ^ 101 I ^. 108 ^ 109 ^ 111 ^ 112 ^ 114 ^, 115 ^ 11B ^ 87 ^ 84 ^ 100 ^ 105 ^ 110 ^ 115 ^ ne RECOVERY LENGTH= 0.15 » 22 n 5 . 45 ^ 46 -> 73 > 75 _ ^ 101 ^ 103 US > 122 > 124 ^. 124 103 ^ 13 ^ 5 1 * -40 CDFigure 5.4: Velocity profiles for conditions of Run #101 CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 62 TEMPERATURE PROFILE FDR A CONSTANT HEAT FLL)X= 2B400. PARAMETERS UPFL0V GRfivnr < SCALING INLET TEHPERmuRE= REYNOLDS"- 857. NUMBER 8F ITERATION 29. D 221. X SCRLHffis Y StnLINC-TEtP. SCALE" 0.42 B.37 42.9 ENTRANCE LENGTH= 0 . 1 5 tu 29 29 29 f « 29 29 29 29 25 29 29 29 l a 29 29 J* 29 29 29 29 29 29 29 29 29 29 h 29 29 29 29 HERTED LENGTH= 0 . 4 8 29 29 29 29 29 29 29 29 29 29 2 9 1 1 " 29 29 29 29 23 29 29 29 29 29 1 < 29 29 29 29 29 29 29 29 29 29 29 f 1 2 9 29 29 29 29 29 29 29 29 30 30 1 1 29 29 29 29 30 30 30 31 31 32 32  90 32 33 34 35 35 37 38 39 H 1 37 47 40 42 51 44 53 46 56 SB 46 4 59 9 61 50 62 51 1 63 RECOVERY LENGTH= 0 . 1 5 28 29 29 29 29 29 29 29 30 30 30 31 33 34 35 41 43 54 59 1 51 63 56 52 LU ZD CD Figure 5.5: Temperature profiles for conditions of Run #101 CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 63 1400-1 Z/l Figure 5.6: Comparison between experimental and computational results for laminar heat transfer coefficient transfer coefficients are determined from the temperature readings at ther-mocouple #6 (Z/l= 0.91) to ensure fully developed flow. The boiling heat transfer coefficients are based on (Tw — Tb). The discussion of the results is presented in reference to the three regions described in Section 5.1. For the Forced Convective Region the experimental results (Run #101) are compared to the computational results and the correlation by Sieder [49] for laminar pipes (see Table 5.1). The average Re number in this region is 900. The characteristic length is based on the hydraulic diameter defined as follows: 44 DH = ~ (5.5) At very low heat flux the computational results, the correlation by Sieder [49] and the experimental results compare favorably. As the heat flux increases, CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 64 the experimental heat transfer coefficient increases slightly with heat flux. The increase in the heat transfer coefficient is due to buoyancy effects and is predictable by the criterion proposed by Jackson and Hall [53]. This criterion predicts a minimum of 5% increase in heat transfer if: G r ' = R ^ > 10"5 ( 5 - 6 ) For the present run this corresponds to a wall heat flux of approximately 1 W/cm2. The computational results predict this increase in the heat transfer coefficient. The experimental results for the Developed Subcooled Boiling region are compared to the saturated pool boiling correlation1 of Stephan [51] (see Table 5.2). Good agreement is obtained between the experimental results and the correlation by Stephan. The Onset of Nucleate Boiling occurs in the region of sharp increase of the heat transfer coefficient. The experimental result is compared to a cor-relation proposed by Rohsenow [52]. The correlation slightly under predicts the experimental value. The experimental data axe also compared to the correlation by Shah [50] (see Table 5.1). The convective heat transfer coefficient (for Re = 1600) can be determined using correlations by Sieder [49] and 'modified' Dittus-Boelter [54]. The accuracy of the correlation by Shah in the developed subcooled boiling region depends on the convective heat transfer coefficient used. This is also reported by Muller-Steinhagen et al. [41]. In this investigation, the 'modified' Dittus-Boelter correlation is used, because it gives better agree-ment. Good agreement is obtained with the experimental data. 1The heat transfer coefficient in this correlation is based on (Tw — T,at). In the present work the heat transfer coefficient is calculated based on (Ttt — 7J,). CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 65 Table 5.1: Heat transfer correlations for subcooled boiling and forced con-vection Sieder [49] : Laminar Pipe Flow Nud = 1.86(i?edPr)1/3(<i//)i/3(// /Zu;)o.i4 fttc evaluated at the wall temperature Modified Dittus-Boelter as presented in [54] : Turbulent flow in pipes Nud = 0.023iZe3-8Pra33 Stephan [51] : Saturated pool boiling Hnh = .4l(<^ 2 Al = 2.1A2 = 0.673 taken from diagram ; Enb base on AT,at Shah [50] : Subcooled Boiling Bo > 3 x 10"5 - » Vo = 230-Bo05 Bo < 3 x 10"5 - » xb0 = 1 + 46Bo05 g ^ f > 63000Bo125 - » V = V>o + f j ^ f Otherwise xb = xb„ <t> = HconlpjTy — Tsat)  For the above correlations: Physical properties evaluated at Tmean for forced convection correlations and at T,at for boiling correlations  CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 66 The bulk temperature at point 'A' in Figure 5.7, in the developed sub-cooled boiling region, is 85 °C. This bulk temperature corresponds to an equilibrium quality of -6.0%. This indicates that heat transfer mechanism are those for subcooled nucleate boiling, for a heat flux of 30 W/cm2 and an equilibrium quality -6% . CHAPTER 5. SUBCOOLED BOILING HEAT TRANSFER 67 Legend • RUN 101 SIEDER f 1 (LAMINAR PIPE")  STEPHAN f 1 (POOL BOILING)  SHAH Tm\(SUBCOOL BOILING)_ LAMINAR COMPUTATION _ ROHSENOW (INCIPIENT BOILING) UPFLOW VELOCITY= 0.076 m/s DECREASING HEAT FLUX INLET TEMPERATURE= 29 C 10 35 HEAT FLUX W/cm2 Figure 5.7: Comparison of subcooled boiling experimental results for Run #101 with correlations and computational results Chapter 6 Void Growth Results 6.1 Over vie w Void growth experiments for subcooled boiling are outlined in Table 6.1. The experimental parameters considered are: heat flux, velocity and flow direction. The range of experimental operating conditions are shown in Table 6.2 and the physical dimensions of the test section are presented in Table 3.2. A complete set of results for a typical run (Run #15) are tabulated in Table B.2 and plotted in Figures B.l to B.3 (Appendix B). The outlet temperature, Tout and the calculated Temperature Toutcal in Table B.l differ by an average value of 0.7 °C. This level of agreement in the closure of the heat balance is typical for all of the results. A void growth profile with the characteristic rapid void increase— OSV, is shown in Figure B.l. A constant low void region can be observed prior to OSV. This region was observed previously [9,10,13] (using the experimental set-up at the University of Ottawa). However, it is not discussed in the open literature [4,6]. At moderate pressures, the void profiles of Ferwells [7] (842-1683 kPa) and Evangelisti [19] (101 kPa) show a constant low void region 68 CHAPTER 6. VOID GROWTH RESULTS Table 6.1: Void growth experimental parameters Run # <f> {W/cm2) U (m/s) Flow Direction 1 98 0.26 Upflow 2 98 0.26 Downflow 3 98 0.37 Upflow 4 98 0.37 Upflow 5 98 0.37 Downflow 6 98 0.37 Downflow 7 98 0.46 Upflow 8 98 0.46 Downflow 9 60 0.13 Upflow 10 60 0.16 Upflow 11 60 0.23 Upflow 12 60 0.23 Downflow 13 30 0.07 Upflow 14 30 0.10 Upflow 15 30 0.22 Upflow 16 30 0.22 Downflow Table 6.2: Range of experimental operating conditions Parameter Range Unit <f> 30, 60, 90 W/cm2 P 155 kPa Tin 10-85 °C U 0.07 - 0.46 m/s G 70 - 450 kg/(m2s) m 0.016 - 0.11 kg/m2 CHAPTER 6. VOID GROWTH RESULTS 70 Table 6.3: Void plateau before OSV and zero void check a ; Run <f> U Void Plateau CXrero # [W/cm2) (m/s) % % 1 98 0.26 3.6 -2.32 2 98 0.26 4.6 -2.23 3 98 0.37 2.6 0.11 4 98 0.37 2.1 -1.08 5 98 0.37 5.1 1.91 6 98 0.37 5.4 1.51 7 98 0.46 4.9 2.64 8 98 0.46 5.2 2.02 9 60 0.13 5.9 -1.42 10 60 0.16 3.7 -0.25 11 60 0.23 2.2 -2.52 12 60 0.23 2.5 1.24 13 30 0.07 6.6 -1.50 14 30 0.10 3.6 0.34 15 30 0.22 3.6 1.29 16 30 0.22 3.5 0.27 Czero average= 0.01 % void prior to OSV. The bubble formation and condensation rates are equal in the constant void region [4], and this must be reflected in a phenomenological void growth model. Table 6.3 shows the average void fraction ('void plateau') prior to OSV to be approximately between 3-6% void. These values are lower than those reported by Mcleod [13] but closer to those by Salcudean et al. [9,10]. Lower void fraction measurements prior to OSV are anticipated in this investigation due to various changes on the void meter outlined in Section 3.4.1. The 'zero void check' values for each of the experiments are listed Table 6.3 and a i e r o averages to 0.01 % void for all the runs. CHAPTER 6. VOID GROWTH RESULTS 71 40 35 2 25 O < 20 CC 6 0 Legend • RUN #13, H. Flux= 30 W/em«2, U=0.07 M/s • Mcloed[ ] H. Fkix= 29.7 W/cm»2. U=0.0B M/s • V 6 • • 4 i i i i i i i i • i -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 EQUILIBRIUM QUALITY % Figure 6.1: Comparison of Run #13 void profile with Mcleod: <ff= 30 W/cm2 6.1.1 Comparison of Vo id Profiles Void profiles for Runs # 13, 9 and 3 are compared to similar void growth experiments of Mcleod [13] and the results are presented in Figures 6.1, 6.2 and 6.3 respectively. For Run #13 and 9 (<f>= 30 and 60 W/cm2) the agreement between both profiles is good. For Run #3 (<b= 98 W/cm2) the OSV for the void profile of Mcleod occurs at a much higher subcooling. This can be partially attributed to problems in the previous loop for the heat balance closure and vibration effects at high heat flux. Figures 5.4 and 5.5 show the repeatability of void growth results (Runs #3 and 4 for upflow; Runs #5 and 6 for downflow). The repeatability of results is approximately within the range of experimental error: ±3-5 % void and ±0.2% equilibrium quality. CHAPTER 6. VOID GROWTH RESULTS 40 72 35 30 25 20 15 > 10 5 0 g i-o < cr u. g o Legend • RUN #9 , H. F»ux= 60 W/em»2.U=0.13 M/s • Mek>ed[ ] H. Fkix= 61 W/em»2, U=0.15 M/s • E • S •10 i i i 1 i r— • 8 -7 -6 -6 -4 -3 EQUILIBRIUM QUALITY % -1 Figure 6.2: Comparison of Run #9 void profile with Mcleod: <f>= 60 W/cm2 40-35-30-z o 25-O < 20-EC U. O 15-5 > 10-6-0-Legend • RUN #3, H. Flux= 98 W/em»2,U=0.37 M/s • Mck>*d[ ] H. Flux= 97 W/cm«2, U=0.30 M/s • • • • Q • • • -10 - 9 -2 — T " - 1 -8 -7 -6 -5 -4 -3 EQUILIBRIUM QUALITY % Figure 6.3: Comparison of Run #3 void profile with Mcleod: <f>= 90 W/cm2 CHAPTER 6. VOID GROWTH RESULTS 73 4 0 -3 5 -3 0 -z o 2 5 -t-o < 2 0 -UL 15-Q IOA 10-6 -0 -Legend RUN 0 3 , a Flux= 98 W /cm '2 , U r 0 . 3 7 M/s • RUN #4, H. Flux= 98 W/cm«2, U=0.37 M/s -10 -7 -6 -5 -4 - 3 EQUILIBRIUM QUALITY % Figure 6.4: Repeatability of results for upflow: Runs #3 and 4 40 Z o i-o < cc Q 5 35 30 25 20-15-10 6H 0 Legend RUN #5, H. FLUX= 98 W/cm«2. U=0.37 M /s • RUN #6, K FLUX= 98 W / c m , 2 . U=0.37 M /s -10 - 1 1 1 1 1 r-- 8 -7 -6 -5 -4 -3 EQUILIBRIUM QUALITY % Figure 6.5: Repeatability of results for downflow: Runs #5 and 6 CHAPTER 6. VOID GROWTH RESULTS 74 6.1.2 Exper imental Observations Subcooled boiling is observed on the heated element with the nucleation site density increasing in the upstream direction. A bubbly flow regime with most of the activity centered around the heater element is present throughout the experiment. A churn turbulent flow pattern develops at the end of the heated section which causes vibration at the outlet. The bubbles are produced at the active nucleation cites in a periodic manner through out the test section. Photographic evidence indicates the bubble diameter to be generally less than 2 mm. The periodicity in bubble formation is not related to cyclic AC heating since it is at a much lower fre-quency [55], but can be attributed to the balance between bubble formation and collapse. The bubble trajectory could not be distinguished. Gunther [56] has reported that the bubbles either slide axially (in the direction of the flow) or are ejected radially into the fluid core. Bubble collapse is accompanied by noises, pressure oscillations and vibrations in the test section at high values of subcooling. Bubbles with a diameter larger than 3 mm were noticed for Runs #15 and 16. A subcooled boiling photograph taken at Z/l = 0.9 is shown in Figure 6.6 for conditions of Run #10 and for an inlet temperature of 25 °C. 6.2 Discussion of Experimental Results 6.2.1 Hydrodynamic Effects on Void Growth The influence of velocity on OSV for upflow was investigated by Salcudean et al. [9,10,11]. The subcooling at OSV was found to increase as velocity increased. This result was in contrast with the Saha and Zuber model [8] CHAPTER 6. VOID GROWTH RESULTS 76 40-35-30-z o 25-o < 20-u. 15-D IOA 10-5-o--7 -6 -6 -4 EQUILIBRIUM QUALITY % Figure 6.7: Hydrodynamic effects on void growth for <f>= 30 W/cm2 and upflow since all the experimental conditions were for Pe numbers less than 700001. Figures 6.7, 6.8 and 6.9 (<£= 30, 60 and 98 W/cm2 respectively) show the void growth for upflow, for different velocities and for Pe numbers less than 12000. There is Kttle influence of velocity on OSV at <b= 30 W/cm2 for low velocities. A more noticeable influence of velocity is seen for heat fluxes of 60 and 98 W/cm2 and at higher velocities. The subcooling at OSV increases with velocity as reported previously [9,10,11]. The results indicate that for Pe numbers less than 70000, void growth is influenced by velocity but the effect reduces as the velocity decreases. Results for void growth for downflow are shown in Figure 6.10 for a heat flux of 98 W/cm2 and for velocities— 0.26, 0.37 and 0.46 m/s. Velocity 1For Pe numbers less than 70000, velocity does not influence OSV according to the Saha and Zuber model. CHAPTER 6. VOID GROWTH RESULTS 77 40-35-30-z o 25-1-o 20-< cr U. 15-Q IOA 10-5-0-Legend • RUN #9 , H. Flux= 60 W/cm»2,u =0.13 m/s • RUN #10, H. Flux= 60 W/cm»2, u=0.16 m/s O RUN #11, H. FLUX= 60 W/cm*2, u=0.23 m/s ®PrP , CP ,o o o 0 ) O -10 - 9 -8 -7 -8 -6 -4 -3 -2 EQUILIBRIUM QUALITY % Figure 6.8: Hydrodynamic effects on void growth for <f>= 60 W/cm2 and upflow •1 40-35-30-z o 25-I -o 20-< tr u. 15-Q IOA 10-5-0--7 -8 -5 -4 EQUILIBRIUM QUALITY % Figure 6.9: upflow Hydrodynamic effects on void growth for <f>= 98 W/cm2 and CHAPTER 6. VOID GROWTH RESULTS 78 40-35-30-z o 25-1-o < 20-rr u. 15-Q IOA 10-5-0-• Legend • RUN #2, H. FLUX= 98 W/cm»2,U=0.26 m/s • RUN #5, H. FLUX= 98 W/cm»2,U=0.37 m/s O RUN #8, H.FLUX= 98 W/cm»2, u=0.46 m/s • _Llh „ C •LXfl 1 -10 - 9 •1 -8 -7 -6 -6 -4 -3 -2 EQUILIBRIUM QUALITY % Figure 6.10: Hydrodynamic effects on void growth for <f>= 98 W/cm2 and downflow influences void growth more for downflow than for upflow. The result for velocities of 0.37 and 0.46 m/s show that the subcooling at OSV increases as velocity increases. The results for a velocity of 0.26 m/s show a different trend, namely that for velocities dose to the bubble rise velocity (0.23 m/s [57]), the subcooling at OSV is high. Similar results are reported by Bartolini [58] who observed higher void fraction for downflow for a velocity of 0.26 m/s than for higher velocities (see Section 6.2.3). 6.2.2 Bubble Detachment It has been reported that bubbles first detach and slide along the heated wall and that the OSV occurs further downstream [8,15,16]. Levy [7], Staub [25] and Bowring [6] have developed successful models for the prediction of void growth based on experimental evidence that bubbles depart near the OSV CHAPTER 6. VOID GROWTH RESULTS 79 at high pressures. Recently Rogers et al. [12] developed a method for determining the bub-ble radius at detachment based on a force balance on a bubble. The expres-sions for the surface tension, drag, and buoyancy forces and bubble radius at detachment are presented in Appendix A. The forces and bubble radii are shown in Figures 6.11, 6.12 and 6.13 for velocities between 0.02 to 1 m/s and for equilibrium bubble contact angles of 30° and 80° for the present data. The model [12] was developed for velocities greater than 0.15 m/s but has been applied for lower velocities in this study. The model predicts a rapid change in the relative influence of the buoyancy and drag forces as the ve-locity increases from 0.1 m/s to 0.5 m/s (Figure 6.11). The buoyancy force does not significantly influence bubble detachment (also reported by Levy [7]) for velocities greater than 0.50 m/s. The equilibrium contact angle has an effect on the actual force values as shown in Figures 6.12 and 6.13. These effects are less significant in the ratio of the forces (Figure 6.11). Rogers et al. have shown that the effects of the contact angle on the OSV are not significant. The force balance at bubble departure was reformulated for downflow using the same model as follows: FD = Fs + FB (6.1) The empirical constants for the surface tension force (0a and 0r, defined in Appendix A) are assumed to be the same for downflow. It is not possible to find solutions for velocities lower than 0.35 m/s since the expression for the bubble departure diameter for downflow requires for positive real roots that: CdC2u p > (6.2) CHAPTER 6. VOID GROWTH RESULTS 80 Figure 6.11: Forces and bubble radius ratios at detachment for upflow 0 0.1 0.2 0.3 0.4 0.6 0.6 0.7 0.8 0.6 BUBBLE RADIUS Rb mm Figure 6.12: Forces and bubble radius at detachment for upflow and 0= 30 °C CHAPTER 6. VOID GROWTH RESULTS 81 0 . 3 0 .4 0 . 6 0 . 6 0 . 7 BUBBLE RADIUS Rb mm 0 . 8 0 . 9 Figure 6.13: Forces and bubble radius at detachment for upflow and 8= 80 °C No physical significance is attributed to this velocity limit since the formu-lation for the bubble radius is empirical. Forces and bubble radii at detachment are shown in Figure 6.14 for down-flow. The bubble radius for a given velocity is larger for downflow than for upflow, since the buoyancy force acts in the same direction as the surface ten-sion force and inhibits departure. This implies that models based on bubble detachment would predict2 OSV to occur at lower subcooling for downflow. A thicker thermal layer (lower subcooling) is required for larger bubbles since the tip of the bubbles must be at the saturation temperature. 2Models based on bubble detachment are developed on the hypothesis that bubble detachment is the critical parameter which governs OSV. If this assumption is valid then it must also apply for downflow. CHAPTER 6. VOID GROWTH RESULTS 82 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 BUBBLE RADIUS Rb mm Figure 6.14: Bubble radius and forces at detachment for downflow 6.2.3 F low Direct ion Effects Downflow void measurements are presented to investigate the influence of buoyancy effects on OSV and void growth. A velocity range of 0.20-0.50 m/s is chosen for the following reasons: • The buoyancy force at bubble detachment is significant for velocities less than 0.5 m/s (Section 6.2.2). • Downflow void growth experiments are not feasible for velocities lower than 0.20 m/s since the flow drag is too small to overcome the buoy-ancy of the bubbles. The bubbles rise (in the upstream direction) and cause critical heat flux to occur. This low velocity limit is slightly smaller than the bubble rise velocity in an infinite medium (0.23 m/s Harmathly [57]). CHAPTER 6. VOID GROWTH RESULTS 83 Figure 6.15 to 6.19 show the downflow and upflow void growth results. The OSV occurs at higher subcooling for downflow in all cases. This result is contrary to models based on bubble detachment which would predict a lower subcooling at OSV for downflow (discussed in the previous section). This implies that bubble detachment is not the only critical parameter for OSV for low velocity and low pressure conditions. These results support the concept of a region where bubbles may detach but the subcooling may be too high to initiate growth (Saha and Zuber [8], Pe < 70000). Figures 6.15 to 6.17 are for a heat flux of 98 W/cm2 and for velocities of 0.26, 0.37 and 0.46 m/s respectively. Higher void fraction is measured for downflow than for upflow at a given equilibrium quality near OSV. This effect is more pronounced at 0.26 m/s, and is lower at higher velocities. These results are in agreement with Bartolini [58] and Simoneau [55] who observed higher void at low velocities (0.20 m/s and 0.26 m/s respectively) for downflow than for upflow. It was concluded [58,55] that buoyancy ef-fects are not significant for velocities greater than 0.8 m/s. Simoneau [55] used a transparent square channel and observed that for downflow, bubbles emerging from the heated surface initially flowed upstream before proceeding downstream due to the low velocities near the wall. The experiments were for subcooled boiling but void measurements were not made. Figures 6.18 and 6.19 are for heat fluxes of 60 W/cm7 and 30 W/cm7 for approximately the same velocity. It can be seen that the effects of flow direction are smaller at lower heat fluxes. Table 6.4 illustrates that there is disagreement about the influence of flow direction on C0. Oshinowo [59] reported six different flow patterns for downflow, in which four differed considerably from those for upflow. The CHAPTER 6. VOID GROWTH RESULTS 84 15 10 - Legend • RUN #1, UPf"LOW • RUN 02, DOWNFLOW H. FLUX- 98 W/cm*2, V-0J26 m/s • • • • I I i i i i -10 -7 -6 -5 -4 -3 EQUILIBRIUM QUALITY % -2 -1 Figure 6.15: Flow direction effects on void growth for <b= 98 W/cm2 and U= 0.26 m/s different void profiles obtained for upflow and downflow in the present study imply that the void distribution and flow patterns (therefore the distribution parameter C0) may differ for the two cases. 6.2.4 Heat Transfer Mechanism before O S V Figure 6.20 shows the void growth for Run #13 (<f>= 30 W/cm2, U= 0.07 m/s and upflow). The OSV occurs at Xeqotv= -1.2%. It is shown in Section 5.3 that subcooled nucleate boiling occurs for these conditions. This also applies to the other runs with the same heat flux since subcooled nucleate boiling is independent of velocity and subcooling. Visual observation and constant wall temperature measurements indicate that subcooled nucleate boiling also occurs prior to OSV at higher heat flux. CHAPTER 6. VOID GROWTH RESULTS 85 Legend • RUN #3, UPTLOW • RUN #5. DOWNFLOW H. FLUX- 98 W/CTTT*2, U-0.37 m/i CJ]LXPD • -10 -9 -8 • i -7 —r--4 0 EQUILIBRIUM QUALITY % Figure 6.16: Flow direction effects on void growth for <j>= 98 W/cm2 and U= 0.37 m/s Table 6.4: Effects of flow direction on the distribution parameter C0 Research Flow Pattern Medium Effects* Zuber [27] Bubbly Air-Water No Bhaga,Weber [60] Bubbly Glass-Air-Water Yes aark,Flemmer [63] Bubbly Air-Water No Oshinowo [59] Bubbly, Annular Air-Water Yes Tuorenzi [62] Bubbly Air-water Yes Qark [61] Bubbly Air-Water Yes * Found C0 to vary with flow direction CHAPTER 6. VOID GROWTH RESULTS 86 40-N O 35-NOI 30-l- 25-o < 20-rr 15-Q 10-IOA 6-0 -Legend • RUN #7. UPTLOW • RUN #8. DOWNfLOW H. FLUX- 88 W/cm*2. U-0.46 m/s -10 -7 -6 -5 - 4 -3 EQUILIBRIUM QUALITY % Figure 6.17: Flow direction effects on void growth for <f>= 98 W/cm2 and U= 0.46 m/s 6.2.5 Comparison of O S V Results wi th Models Table 6.5 shows experimental values of Totv, o ,^,, and Xosv. The prediction of Xotv and Tw based on the model of Rogers et al. [12] (for an equilibrium contact angle of 30° and velocities greater than 0.15 m/s), and Tw as pre-dicted by the nucleate boiling correlation3 of Roshenow [64] are also shown. The OSV results predicted by the model of Rogers et al. compare favourably with some of the experimental results. The model does not perform well over the entire range. The explanation for the disagreement is as follows: • Bubble detachment is not a unique criteria for OSV at low velocities 3A nucleate boiling correlation is used instead of the experimental results for T„ since wall temperature measurements are not accurate for heat fluxes greater than 30 W/cm2 CHAPTER 6. VOID GROWTH RESULTS 87 Table 6.5: Experimental OSV results and comparison to detachment model for upflow Parameters Experimental Model Run U Flow T Ctffgv Xotv w X0gV Tw # W/cm2 m/s Dir. °C % % °C % °C 1 98 0.26 Up 100.4 4.50 -2.3 139 -2.7 424 2 98 0.26 Down 91.3 4.80 -4.0 — — — 3 98 0.37 Up 95.5 5.13 -3.2 139 -4.6 334 4 98 0.37 Up 99.8 3.36 -2.4 139 -4.6 334 5 98 0.37 Down 96.8 7.65 -3.0 — — — 6 98 0.37 Down 98.5 7.62 -2.6 — — — 7 98 0.46 Up 95.1 7.51 -3.3 139 -5.29 291 8 98 0.46 Down 87.4 7.26 -4.7 - - — 9 60 0.13 Up 104.1 7.83 -1.6 135 — — 10 60 0.16 Up 102.7 3.95 -1.9 135 0.3 409 11 60 0.23 Up 95.5 4.76 -3.2 135 -1.2 327 12 60 0.23 Down 87.6 3.71 -4.7 - - -13 30 0.07 Up 106.1 10.18 -1.2 131 — — 14 30 0.10 Up 106.8 7.41 -1.1 131 — — 15 30 0.22 Up 103.0 5.39 -1.8 131 -0.5 224 16 30 0.22 Down 98.0 4.65 -2.7 - - -** Based on the corre ation of Roshenow [64] CHAPTER 6. VOID GROWTH RESULTS 88 Legend • RUN #11, UPFLOW • RUN #12. DOWNFLOW K FLUX- 60 W/cm*2, TJ-0.23 m/t -10 -8 -7 -8 -6 -4 -8 3EQUILIBRIUM QUALITY % -1 o Figure 6.18: Flow direction effects on void growth for <f>= 60 W/cm2 and U= 0.23 m/s and low pressure. • The accuracy of the OSV prediction depends on the correct determi-nation of the heat transfer coefficient (see equation A.9). The single phase heat transfer coefficient used in the model does not account for the increased heat transfer due to the agitation by the bubbles prior to OSV. This is reflected in the wall temperatures predicted by the model which are 95-270 °C higher than those predicted by the correlation of Roshenow [64]. Figure 6.21 shows the experimental results and the prediction of the cor-relation of Saha and Zuber for St' versus the Pe number. All values are obtained for the thermally controlled region (Pe less than 70000). Experi-mental values for upflow follow a line of Nu*= 871 , instead of Nu*= 455. CHAPTER 6. VOID GROWTH RESULTS 89 40-35-NOI 30-H 25-O < 20-FR 15-O 10-IOA 5-0-Legend • RUN #15, UPFLOW • RUN #16, DOWNFLOW H. FLUX- 30 W/cnT2, U-0.22 m/» -10 -e -7 -8 -6 -4 -3 EQUILIBRIUM QUALITY •/. Figure 6.19: Flow direction effects on void growth for <f>= 30 W/cm2 and U- 0.22 m/s 30 25- Run #13: <j>= 30 W/cm2, U= 0.07 m/s and upflow 2 20 o < 15-| CC a «• o > 6-i . J —T 1 1 1 1— -7 -6 -6 -4 -3 EQUILIBRIUM QUALITY % -10 -9 -2 Figure 6.20: Void growth for Run #13: <f>= 30 W/cm2, U= 0.07 m/s and upflow CHAPTER 6. VOID GROWTH RESULTS 90 0.1 CO 0.01 0.002 3000 Nu = 871 Nu = 455 i i 10000 Pe f Legend S*R<-ZU8tB • UPTLOW O OOWMTLOW CIFTOMCNTtL 100000 Figure 6.21: Comparison of Experimental OSV Values to the correlation of Saha and Zuber The lower Nusselt number predicted by the correlation Saha and Zuber can be attributed to different surface roughness and to the higher pressure data bank for the correlation. Studies by Salcudean et al. [9,10] also found a higher value for Nu* and experiments at higher pressure (245 kPa) showed the results to agree better with the correlation. Figure 6.21 shows that two experimental points for upflow are very close to Nu* = 455. These two points correspond to Runs #11 and 15 which are for higher inlet temperatures and larger bubbles are observed on the heater surface. This difference in the bubbly flow pattern may modify the OSV mechanism. Further investigation within the high velocity range with low inlet subcooling is required. Photographic experiments by Hinu [54] using an annulus (!?,= 8 mm and D0- 18 mm) with R-113 at 140 kPa, found that CHAPTER 6. VOID GROWTH RESULTS 91 OSV results followed the correlation of Saha and Zuber well. The results were also obtained for low inlet subcooling. This indicates that the degree of subcooling may play a role in determining the OSV. 6.3 Heat Transfer Results 6.3.1 Heat Transfer Coefficients The heat transfer coefficients for the void growth are based on Tw — 21 and averaged over the length of the rod. The results are curve-fitted for the same velocity and shown in Figure 6.22 for <b= 30 W/cm2. The heat trans-fer coefficient is independent of velocity since measurements are in the fully developed subcooled boiling region. The values increase almost linearly due to the linear change of the bulk temperature and the constant wall tempera-ture Tw. The curves are shifted because of the effect of velocity on the bulk temperature. Void fraction does not seem to influence the subcooled boiling heat transfer coefficient since the curves do not change after OSV. 6.3.2 F low Direct ion Effects In Figure 6.22 the heat transfer coefficient is slightly higher for upflow than for downflow (Runs #15 and 16). Larger bubbles were observed on the sur-face of the heater for these runs. Bartolini [58] reports no difference in the heat transfer coefficient between upflow and downflow at 119 kPa except at a 50 °C subcooling with a velocity of 0.2 m/s. These values correspond approximately to the parameters of Runs #15 and 16. Pujol [65] reports sim-ilar results for boiling heat transfer in a serpentine tube geometry where heat transfer coefficients are not markedly different in either direction. Thorsen CHAPTER 6. VOID GROWTH RESULTS 92 [66] reports that for Freon-113 in a heated pipe, the heat transfer coeffi-cients are higher for upflow than for downflow. The reason is attributed to the smaller diameter bubbles for upflow at detachment, which results in a greater micro-convection effect and enhanced heat transfer. CHAPTER 6. VOID GROWTH RESULTS 93 10000 eoooH CM *g 8000 o ^ 7000 6000 5000-4000-< {E 3000 LJ X 2000 1000 o-t Legend O RUN #13 UPFLOW O RUN #14 UPFLOW A RUN #15 UPFLOW V RUN #16 DOWNFLOW U=0.07 m/s Run #13  U=0.10 m/s Run #14 U=0.22 m/s Run #15 & #16 -10 —I 1 1 1 I -7 -6 -6 -4 -3 EQUILIBRIUM QUALITY % Figure 6.22: Average Heat Transfer Results for Void Growth Chapter 7 Conclusions and Recommendations 7.1 Conclusions An experimental two-phase loop facility was designed and built to study the heat transfer characteristics of the SLOWPOKE reactor. Subcooled boiling and void growth for upflow and downflow were investigated for conditions of low velocities and low pressure. The following conclusions are made: • A low void region is observed before OSV which varies approximately between 3-6 % void. This phenomenon can be seen in void growth data at moderate pressures but it is not addressed in the literature. • The heat transfer mechanisms before OSV for <f>= 30 W/cm2 are sim-ilar to those of fully developed subcooled boiling. There is evidence that this is also true for higher heat fluxes. Therefore, the assumption of single phase heat transfer prior to OSV, which is used in models de-veloped at high pressures, does not apply to the present experimental conditions. 94 CHAPTER 7. CONCLUSIONS AND RECOMMENDATIONS 95 • The subcooling at OSV decreases slightly as the velocity increases as previously shown [9,10,11]. • OSV occurs at higher subcooling for downflow than for upflow. • Buoyancy effects on void growth are largest at values near the bubble rise velocity (0.23 m/s) and decrease at higher velocity. • The buoyancy effects on void growth increase with an increase in heat flux. • Heat transfer coefficients are relatively insensitive to flow direction. • Comparison between upflow and downflow void formation showed that for the present range of investigation, bubble detachment is not the only significant criterion for determining OSV. 7.2 Recommendations 7.2.1 Test Apparatus Recommendations for the test apparatus are as follows: 1. A faster computer system with analogue outputs for controlling the power and void meter table location, and digital outputs to shut-off pumps and power before critical heat flux is reached. 2. Visual display of velocity, heat flux, temperatures and pressure values. 3. Dynamic calibration of void meter by using air and water flowing si-multaneously inside the annulus. CHAPTER 7. CONCLUSIONS AND RECOMMENDATIONS 96 4. Install a water distilling apparatus to remove dissolved air and ionic substances. 5. Use Inconnel for the heater element to reduce the current required and spot weld the thermocouples on the inside wall. This may be sufficient to resolve the wall temperature measurement problems. 7.2.2 Vo id Growth 1. Develop a model which is not only based on bubble detachment 2. Use high speed photography to examine the trajectory of the bubbles at detachment. 3. Further investigate the region for low inlet subcooling and low heat flux for which larger bubbles are formed. Bibliography [1] Hilborn J.W., Kay R.E., 'Up-rating the SLOWPOKE Reactor', Atomic Energy of Canada Limited, Chalk River, Ontario, (April 1977). [2] Rogers J.T., Salcudean M., Tahir A.E., 'Flow Boiling Critical Heat Flux for Water in a Vertical Annulus at Low Pressure and Velocities', Seventh International Heat Transfer Conference, Munich, Vol. 4, pp. 339-344, (1982). [3] Graham K.D., 'Investigation of Critical Heat Flux in a Vertical Annulus Under Conditions of Low Pressure and Low Velocities', Ba.Sc. Thesis, University of Ottawa, Mech. Eng. Dept., (April 1982). [4] Rouhani S.Z., 'Calculation of Steam Volume Fraction in Subcooled Boil-ing', J. of Heat Transfer, Vol. 90, pp. 158-163, (Feb 1968). [5] Lahey R.T., 'A Mechanistic Subcooled Boiling Model', Sixth Interna-tional Heat Transfer Conference, Toronto, Canada, Vol. 1, pp. 293-297, (1978). [6] Bowring R.W., 'Physical Model, Based on Bubble Detachment, and Cal-culation of Steam Voidage in the Subcooled Region of a Heated Chan-nel', HPR-29, Institutt for Atomenergi, Halden, Norway. [7] Levy S., 'Forced Convection Subcooled Boiling - Prediction of Vapor Volumetric Fraction', Int. J. of Heat Mass Transfer, Vol. 10, pp. 951-965, (July 1967). [8] Saha p., Zuber N., 'Point of Net Vapor Generation and Vapor Void Frac-tion in Subcooled Boiling', Int. Heat Transfer Conf., 5 t h , Proc, Tokyo, Jpn, Vol. 4, Pap. B4.7, (Sep 1974). 97 BIBLIOGRAPHY 98 [9] Salcudean M., Rogers J.T., Tahir A., Abdullah Z., Graham K., 'Inves-tigation of Two-Phase Flow Characteristics and Heat Transfer Limita-tions in Fuel Elements for the 2 MW SLOWPOKE Reactor', Report to AECL, Dept. of Mech. Eng., University of Ottawa (Feb 1982). [10] Salcudean M., Rogers J.T., Tahir A., Abdullah Z., Graham K., 'Inves-tigation of Two-Phase Flow Characteristics and Heat Transfer Limita-tions in Fuel Elements for the 2 MW SLOWPOKE Reactor', Report to AECL, Dept. of Mech. Eng., University of Ottawa (March 1983). [11] Salcudean M., Rogers J.T., Tahir A., Abdullah Z., 'Void Fraction for Water in a Cosine-Distribution Heated Vertical Annulus at Low Pres-sures and Velocities', Fourth Annual Conference of Canadian Nuclear Society, Montreal, (June 1983). [12] Rogers J.T., Salcudean M., Abdullah Z., Mcleod D., Poirier D., 'The Onset of Significant Void in Upflow Boiling of Water at Low Pressure and Velocities', Int. J. Heat Mass Transfer, Vol. 30, No. 11, pp. 2247-2260, (1987) [13] Mcleod R.D., 'Investigation of Subcooled Void Fraction Growth in Water Under Low Pressure and Low Flow Rate Conditions', Masters Thesis, Carleton University, Ottawa, (1986). [14] Abdullah Z., 'Investigation of Onset of Significant Void and Void Frac-tion Under Conditions of Low Pressures and Low Velocities', Ba.Sc. Thesis, Dept. of Mech. Eng., University of Ottawa, (1882). [15] Dix G.E., 'Vapor Void Fraction for Forced Convection with Subcooled Boiling at Low Flow Rates', Ph.D. Thesis, Univ. of California, Berkeley (1971). Also, General Electric Report Number NEDO-10491. [16] Serizawa A., Kenning D.B.R., 'A Study of Forced Convective Subcooled Flow Boiling', Tech. Rep., Inst. Atomic Energy, Kyoto Univ., (1979). [17] Hilborn J.W., Kay R.E., Stevens-Guille P.D., Jervis R.E., 'SLOW-POKE at the University of Toronto- A Laboratory Reactor for Neutron Irradiation', Atomic Energy of Canada Limited, Chalk River, Ontario, (August 1972). [18] Hilborn J.W., Glen J.S., 'Small Reactors for Low-Temperature Heating', Atomic Energy of Canada Limited, Chalk River, Ontario, (1971). [19] Evangelisti R, Lupoli P., 'The Void Fraction in an Annular Channel at Atmospheric Pressure', Int. J. Heat Mass Transfer, Vol. 12, pp. 699-711, (1968). BIBLIOGRAPHY 99 [20] Edelman Z., Elias E., 'Void Fraction in Low Flow Rate Subcooled Boil-ing', Nuclear Engineering and Design, pp. 375-382, (Feb. 1981). [21] Cimorelli L., Evangelisti R., 'The Application of the Capacitance Method for the Void Fraction Measurement in Bulk Boiling Conditions', Int. J. Heat Mass Transfer, Vol. 10, (1967). [22] Kotohiko S. et al., 'Flow Boiling in Subcooled and Low Quality Regions-Heat Transfer and Local Void Fraction', Int. Heat Transfer Conf., 5"*, Proc, Tokyo, Jpn, Vol. 4, Pap. B4.8, (Sep 1974). [23] Griffith P., Clark J.A., Rohsenow W.W., 'Void Volumes in Subcooled Boiling Systems', ASME paper number 58-HT-19, (1958). [24] Costello CP., ASME-AIChE Heat Transfer Conference, Storrs, ASME paper number 59-HT-18, (1959). [25] Staub F.W., 'The Void Fraction in Subcooled Boiling- Prediction of the Initial Point of Net Vapor Generation', J. of Heat Transfer, Vol. 90, pp. 151-157, (Feb 1968). [26] Kroeger P.G., Zuber N., 'An Analysis of the Effects of Various Param-eters on the Average Void Fractions in Subcooled Boiling', J. of Heat Transfer, Vol. 11, pp. 211-233, (1968). [27] Zuber N., Findlay J.A., 'Average Volumetric Concentration in Two Phase Flow Systems', J. of Heat Transfer, Vol. 87, pp. 453-468, (Nov 1965). [28] Al-Hayes R.A.M., Winderton R.H.S., 'Bubble Diameter on Detachment in Flowing Liquids', Int J. Heat Mass Transfer, Vol. 24, p. 223, (1981). [29] Una! H.C., 'Determination of the Initial Point of Net Vapor Generation in Flow Boiling Systems', Int. J. Heat Mass Transfer, Vol. 18, pp. 1095-1099, (1975). [30] Rouhani S.Z., Axelsson E., 'Calculation of Void Volume Fraction in the Subcooled and Quality Boiling Regions', J. of Heat Transfer, Vol. 13, pp. 383-393, (1970). [31] Ahmad S.Y., 'Axial Distribution of Bulk Temperature and Void Fraction in a Heated Channel With Inlet Subcooling', J. of Heat Transfer, pp. 595-608, (1970). BIBLIOGRAPHY 100 [32] Ahmad S.Y., 'Forced Convection Subcooled Boiling— Prediction of the Onset of Bubble Detachment', Unpublished CRNL Report, (1969). [33] Dimmick G.R., 'Void Measurements for SLOWPOKE', Memorandum to J.W. Hillborn, A.E.C.L., (July 1981). [34] Lasschn G.D., Stephen A.G., Taylor D.J., Wood D.B., 'X-Ray and Gamma Ray Transmission Densitometry', Idaho National Engineering Laboratory, EG&G Idaho, Idaho, (1982). [35] Hawbolt E.B., Chan B., 'Kinetics of Austenite-Pearlite Transformation in Eutectoid Carbon Steel', Vol 14, (Sept. 1983). [36] Schnurr N.M., 'Heat Transfer to Carbon Dioxide in the Immediate Vicin-ity of the Critical Point', ASME 68-HT-32, (1968). [37] Doeblin E.O., 'Meassurement Systems', McGraw-Hill, (1966). [38] Chan A.M.C., 'Transient Two-Phase Flow', P.H.D. Thesis, McMaster University, (1982). [39] Schrock V.E., 'Radiation Attenuation Techniques in Two Phase Flow Measurements', Two Phase Flow Instrumentation, 11th National ASME/AICHE Heat TRansfer Conference, Minneapolis Min., (Aug 1969). [40] McAdams W.H., Kennel W.E., Minden C.S., Carl R., Picornell P.M., Dew J.E., 'Heat Transfer at High Rates to Water with Surface Boiling', Industrial and Engineering Chemistry, (Sept. 1949). [41] Muller-Steinhagen H., Epstein N., Watkinson A.D., 'Effect of dissolved gases on Subcooled Flow Boiling Heat Transfer', Chem. Eng. Process, Vol. 23, pp. 115-124, (1988). [42] McKeown J.J., Brown L.C., Gove G.W., 'Comparative Studies of Dis-solved Oxygen Analysis Methods', J. Water Pollution Control Federal 39:1323, (1967). [43] Messinger A., A Dictionary of Chemical Solubilities, Macmillan Com-pany, N.Y., (1921). [44] Collier J.G., Convective Boiling and Condensation, McGraw-Hill, (1972). [45] Shames I.V., 'Mechanics of Fluids', McGraw-Hill, N.Y., (1982). BIBLIOGRAPHY 101 Benudekax R.W., Gosman A.D., Issa R.I., 'The TEACH-II code for Detailed Analysis of Two-Dimensional Turbulent Recirculating Flow', Dept. Mech. Eng., Imperial College, Rept. FS/83/3, (1983). Djilali N, 'An Investigation of Two-Dimensional Flow Separation with Reattachment', Ph.D. Thesis, Dept. Mech. Eng., University of British Columbia, (1987). Lungren T.S., Sparrow E.M., Trans. ASME, 86:620, (1964). Sieder E.N., Tate C.E., 'Heat Transfer and Pressure Drop of Liquids in Tubes', Ind. Eng. Chem, Vol. 28, p. 1429, (1936). Shah M.M., 'Generalized Prediction of Heat Transfer During Subcooled Boiling in Annuli', Heat Transfer Engineering, Vol. 4, No. 1, pp. 24-31, (1983). Stephan K., Abdelsalam M.,'Heat Transfer Correlations for Natural Convective Boiling', Int. J. of Heat and Mass Transfer, Vol. 23, pp. 73-87, (1980). Roshenow W.M., 'A Method of Correlating Heat Transfer Data for Sur-face Boiling of Liquids', Trans. ASME, Vol. 74, pp. 969-976, (1952). Jackson J.D., Hall W.B., 'Influence of Buoyancy on Heat Transfer in Channels and Bundels, Vol. 2, Mcgraw-Hill, N.Y., (1979). Hino R_, Ueda T., 'Studies on Heat Transfer and Flow Characteristics in Subcooled Flow Boiling- Parti. Boiling Characteristics', Lit J. Mul-tiphase Flow, Vol. 11, No. 3, p. 269, (1985). Simoneau RB. , Frederick F.S., 'A Visual Study and Buoyancy Effects on Boiling Nitrogen', NASA Report. Gunther F.C., 'Photographic Study of Surface-Boiling Heat transfer to Water with Forced Convection', ASME Transactions, Vol. 73, pp. 115-123, (1951). Harmathy T.Z., 'Velocity of Large Drops and Bubbles in Media of Re-stricted or Infinite Extent', AIChE J., Vol. 6, No. 281, (1960). Bartolini R., Gugielmini G., 'Experimental Study on Nucleate Boiling of Water in Vertical Upflow and Downflow', Int J. Multiphase flow, Vol. 9, No. 2, p. 161, (1983). BIBLIOGRAPHY 102 [59] Oshinowo T., Charles M.E., 'Vertical Two Phase Flow', Can. J. Chem. Eng, Vol. 52, (Feb 1974). [60] Bhaga D., Weber M.E., 'Holdup in Vertical Two and Three Phase Flow', Can J. Chem Eng., Vol. 50, (June 1972). [61] Clark N.N., Flemmer R.L., 'Predicting the Holdup in Two Phase Bubble Upflow and Downflow Using the Zuber and Findlay Drift-Flux Model', AIChE Journal, Vol. 31, No. 3, (March 1985). [62] Lorenzi A., Kirkpatrick R.D., 'Comparative Investigation of Some Char-acteristic Qualities of Two Phase Concurrent Upward and Downward Flow', Two Phase Transport and Reactor Safety, Veziroglu and Kakac, eds., Hemispere, Washington, D.C. (1978). [63] Clark N.N., Flemmer R L . C , 'A Technique for Synchronizing Valves and Determining Bubble Rise Velocities in Two Phase Flow', Paper WA-84-FE-11, ASME Winter Meeting, New Orleans (1984). [64] Roshenow W.M., 'Heat Transfer with Evaporation', Heat Transfer— A Symposium held at the University of Michigan during the Summer of 1952. Published by the University of Michigan Press, pp. 101-150, (1953). [65] Pujol 1., Stenning A.H., 'Effect of Flow Direction on the Boiling Heat Transfer Coefficient in Vertical Tubes', in 'Cocurrent Gas-Liquid Flow', Plenum Press, N.Y., (1969). [66] Thorsen R.S., 'A Comparative Study of Vertical Upflow and Downflow in a Uniformly Heated Boiling Fluid', Iht Heat transfer Conf., 5"*, Proc, Tokyo, Jpn, Vol. 4, Pap. B4.3, (Sep 1974). [67] Koumoutsos N., Moissis R, Spyridonos, 'A Study of Bubble Departure in Forced Convection Boiling', J. Heat Transfer, Vol. 90, pp. 223-230, (1968). Appendix A Description of O S V Model The OSV model of Rogers et al. [12], based on bubble departure is described in the following. OSV is postulated to occur when the steam bubbles detach from the wall. Bubble departure occurs when the bubble drag force and the buoyancy force (forces tending to detach the bubbles) overcomes the surface tension force (force tending to hold the bubble). The bubble is assumed to be a truncated sphere with a contact angle at the surface equal to the equilibrium contact angle 6 as shown in Figure A . l . Expressions for the forces are obtained from results of Al-Hayes and Winderton [28] for gas bubble departure from a wall. Figure A.1: Contact angle 6 103 APPENDIX A. DESCRIPTION OF OSV MODEL 104 The expressions axe as follows: 3 FB = plg7^-(2 + 3cos9-cos36) (A.l) O 2 FD = Cdr2b^(-K - 0 + cos0sin0) (A.2) Fs = Ct ^-<T sin 0(cos 9r - cos Ba) (A.3) where • Cd= drag coefficient determined from experimental results of Al-Hayes and Winderton [28] which is function of the bubble Reynolds number defined as: Reb = (A.4) • Ur— velocity at j/b/2. As shown in Figure A. l , yt is equal to the distance from the wall to the tip of the bubble. • Cs is an empirical factor for the surface tension force and given by: ro c ' = « T 5 + a l 4 < A 5 » • 6a and 6r are the advancing and receding contact angles. These angles represent the limiting values for the distortion of the contact angle around the base of the bubble. The distortion is caused by the bubble and drag forces. These angles are approximately equal to 0a = 6 + 10 and 6r = 9 - 10 [67]. The bubble radius can be calculated from a force balance and is given by: 3C2CdU2 8TT2CiC3Ctga 1/2 where C\, Ci and C 3 axe only function of 0, 9a and 6r. The only unknown in equation A.6 is uT and obtained as follows: APPENDIX A. DESCRIPTION OF OSV MODEL 105 1. The friction factor, /, is evaluated from an equation for fully developed single-phase turbulent flow over smooth surfaces [7,12] 2. The wall shear stress, TW, from / = rw/[pi(u2 /2)] 3. The non-dimensional distance at bubble departure, y£, is obtained by: Vt = yb-J{rwlPl) = r6(l+cos e)£LJ(Tw/p,) (A.7) 4. Relate u + to y + by using the universal velocity profile for turbulent single-phase flow [7,12]. The bubble radius at departure is solved by iteration. The subcooling at bubble departure (OSV) is determined by assuming that the temperature at the tip of the bubble is the saturation temperature. The resulting equation is given by: Ttat - Totv = Kj+f- - 7^—) (A.8) where H c o " D h = 0.023Re™ PrfA (A.9) Ki it = 9zj^{Tw_Ttat) (A.10) and • Fr is an empirical factor to allow for surface roughness effects. • T4" is related to y + by the temperature profile given by the Martinelli equations [12]. Appendix B Experimental Void Growth Data A sample run (Run #15) is presented in this appendix and the nomenclature of the computer symbols is shown in Table B.l. The data from the run is tabulated in Table B.2 and various parameters are plotted in Figures B.l to B.3. The tabulated data and the plots are obtained by from the data analysis program. 106 APPENDIX B. EXPERIMENTAL VOID GROWTH DATA 107 Table B.l: Nomenclature of symbols from computer output AVG average value for run Count number of original data points D-B based on Dittus-Boelter heat transfer correlation DIFP differential pressure across test section (Psi) Drod outside diameter of heated section Dglass inside diameter of glass tube Flow flow direction Mux Itslow frequency of void measurements per iteration Itfast number of iteration of data acquisition program Meter selects which flow meter to be used SD standard deviation of void meter reading (volts) SDR standard deviation ratio of void meter readings SD/(Va — Vw)100 VOID a Xeqcon Cpi/ifg APPENDIX B. EXPERIMENTAL VOID GROWTH DATA * * * * * * Table B.2: Tabulated data for Run #15 from computer output EXPRERIKENTAL DATA FOR RUN —> 15 DATE OF EXPERIMENT —> 18-06-88 NUMBER OF DATA POINTS —> 32 P- 22.5 Drod- 0.0127 A- .00025 lvm- 0.4400 * Rod #- 5 Xeqcon- 0.1879 * Tsat- 112.5 Meter- 2 * Q- 5745. Hflux- 30.00 Dglass- 0.0218 Itslow- 40 Va- 5.295 Count- 35. Flow- UPFLOW 1-Itfast-Vw-Dh-0.480 50 3.410 .00910 * * * * * * * * R- .00560 * COMPUTED RESULTS * * * * * * * * * * * MINIMUM VELOCITY m/sec -AVERAGE VELOCITY m/sec -AVG. VOID ON PLATEAU % -# OF THERMOCOUPLES -AVG H.T.C FOR RUN -AVG DIT-BOELT H.T.C. -AVG WALL TEMPERATURE -MASS FLUX (Kg/M**2/S)-.TEMPERATURE AT OSV (C) -STANTON # AT OSV -AVG DENSITY (Kg/M**3) -AVG Cp (J/Kg/C) -AVG K (W/M/C) -0.2113 0.2138 3.6219 6 5239. 2261. 145.6 207.3 103.0 0.0362 969.7 4195.6 0.6733 MAXIMUM VELOCITY m/sec XEQ QUALITY CUTOFF % QUALITY AT OSV % AVERAGE RE # FOR RUN AVG NU# FOR RUN AVG DIT-BOELT NU# ZERO VOID CHECK % VOID AT OSV (%) PECLET # MEAN TEMPERATURE (C) AVG VISCOCITY (Kg/S/M) AVG PR # 0.2170 -1.60 -1.80 5484. 70.8 30.6 1.29 5.39 11756. 83.1 0.000344 2.141 * * * * * * * * * * * * * N VOLTAGE CURRENT Real CALHF Flow rate u Re# 1 6. ,01 966. 5 0.00622 30.31 0. 0518 0. 2151 4665. .4 2 6. .01 966. .7 0.00621 30.32 0. 0519 0. 2153 4692. 6 3 6. .01 965. .9 0.00622 30.29 0. 0520 0. .2158 4761. .7 4 5. .99 964. .1 0.00622 30.17 0. 0520 0. .2158 4827. .2 5 5. .99 962. .5 0.00622 30.11 0. 0520 0. 2164 4914. .1 6 5. .98 961. .7 0.00622 30.05 0. 0519 0. .2163 4981. .4 7 5. .98 961. .7 0.00622 30.05 0. 0521 0. .2168 5111, .4 8 6 .02 967. .8 0.00622 30.44 0. 0519 0. .2170 5273. .4 9 6 .00 964 .4 0.00623 30.24 0. 0514 0. .2150 5293, .1 10 6 .02 967 .0 0.00623 30.41 0. 0513 0. .2145 5337, .9 11 6 .02 966 .4 0.00623 30.37 0. 0513 0. .2144 5400, .5 12 6 .00 963. .6 0.00623 30.20 0. 0512 0. .2139 5451, .9 13 6 .02 966 .3 0.00623 30.36 0. 0512 0. .2147 5542, .5 14 6 .02 966. .6 0.00623 30.39 0. 0511 0. ,2142 5624. .1 15 5 .94 954 .4 0.00623 29.62 0. 0509 0. .2135 5645, .5 16 5 .99 961 .4 0.00623 30.08 0. 0509 0. .2134 5712. .5 17 5 .97 958 .8 0.00623 29.91 0. 0508 0. .2131 5771. .9 18 6 .00 961. .6 0.00623 30.10 0. 0507 0. ,2136 5849. .9 19 5 .98 958 .8 0.00623 29.92 0. 0506 0. .2129 5896. .4 20 6 .01 964, .2 0.00624 30.27 0. 0505 0. ,2127 5924. .3 21 5 .99 961 .4 0.00623 30.08 0. 0505 0. .2125 5966. .2 APPENDIX B. EXPERIMENTAL VOID GROWTH DATA Table B.2: Tabulated data for Run #15 from computer output 109 continued.. 22 5. ,99 960. ,5 0. ,00623 30. ,03 0. ,0504 0. 2123 6008. ,7 23 5. .98 959. ,6 0. .00624 29. .98 0, .0504 0. ,2121 6045. ,0 24 5. .95 953. .8 0. .00623 29, .61 0, .0502 0. ,2130 6142. .2 25 6. 04 968. .1 0. .00624 30. .53 0. .0504 0. .2122 6104. 8 26 5. .95 953. .8 0. .00623 29, ,62 0, .0501 0. .2129 6175. .8 27 5. .96 954. .8 0. .00624 29. .69 0. .0501 0. .2128 6202, .4 28 5. .96 955. .3 0. .00623 29, .71 0 .0500 0. .2124 6227. .9 29 5. .95 954. .8 0. .00623 29. .68 0 .0499 0. .2121 6249, .1 30 5, .95 954. .2 0 .00623 29 .64 0 .0499 0 .2118 6278 .2 31 5. .98 959, .1 0 .00623 29 .94 0 .0498 0, .2116 6310 .7 32 5. .98 959. .7 0 .00623 29 .99 0 .0498 0 .2113 6309, .5 N Tmean Tcomp DIFP SD SDR VOID Xeq 1 69 .3 1 .070 0.40 0. 22 12 .10 2 .09 -5.86 2 69 .7 1, .069 0. ,00 0. 23 12 .50 2 .34 -5.79 3 70 .7 1 .069 0. ,00 0. 23 12 .20 2 .01 -5.60 4 71 .8 1 .069 0. 70 0. 22 12 .00 2 .01 -5.39 5 73 .0 1.069 0. ,20 0. 23 12 .50 2 .16 -5.17 6 74 .1 1 .066 0. .20 0. ,24 12 .80 2 .13 -4.98 7 75 .9 1 .065 0. ,10 0. 23 12 .50 3 .07 -4.65 8 78 .4 1 .065 0. ,20 0. 23 12 .50 2 .48 -4.16 9 79 .5 1 .065 0. ,00 0. 24 12 .90 3 .42 -3.94 10 80 .4 1. .064 0. 50 0. 23 12 .60 2 .77 -3.74 11 81 .4 1. .064 0. 30 0. 24 12 .80 3 .81 -3.57 12 82 .4 1, .065 0. ,20 0. 23 12 .60 3 .81 -3.38 13 83 .7 1. .064 0. ,30 0. 24 13 .00 4. .35 -3.12 14 85 .4 1. .064 0. 40 0. 23 12 .60 4 .67 -2.81 15 86 .1 1. .064 0. 30 0. 24 12 .70 4 .27 -2.73 16 87 .3 1. .064 0. 30 0. 24 12 .90 5 .74 -2.47 17 88 .4 1. .064 0. 40 0. 23 12 .60 3. .34 -2.26 18 89 .4 1. .064 0. 40 0. 23 12. .20 4. .70 -2.08 19 90 .4 1. .064 0. 30 0. 24 13, .00 5. .15 -1.91 20 90 .9 1. .064 0. 00 0. 23 12 .60 5, .64 -1.79 21 91 .6 1. .064 0. 50 0. 25 13. .50 6. .10 -1.67 22 92 .3 1. .064 0. 20 0. 23 12. .50 7. .06 -1.56 23 92 .9 1. .064 0. 10 0. 23 12. .50 5. .70 -1.44 24 93 .9 1. ,064 0. 20 0. 24 12. .70 8. ,37 -1.28 25 93 .7 1.064 0. 20 0. 24 12. .80 9. .03 -1.26 26 94 .4 1. .064 0. 20 0. 25 13. .20 8. .89 -1.19 27 94 .8 1. 064 0. 00 0. 24 12. .70 10. .56 -1.12 28 95 .3 1. .064 0. 50 0. 23 12. .70 9. .34 -1.02 29 95 .7 1. .064 0.40 0. 23 12. .30 10. .70 -0.94 30 96 .2 1. .064 0. 30 0. 22 12. .00 12. ,37 -0.85 31 96 .7 1. .064 0. 10 0. 24 12. .70 13. .26 -0.74 32 96 .8 1. .062 0.40 0. 26 13, .80 14. ,81 -0.71 APPENDIX B. EXPERIMENTAL VOID GROWTH DATA 110 Table B.2: Tabulated data for Run #15 from computer output continued... N Tin Tout Toutcal Tvmcal Tw H Tmean 1 56. ,3 82. .4 82. 7 81. 3 143. 4 4104. 69. 3 2 56. 6 82. .8 83. 1 81. 7 143. 7 4103. 69. 7 3 57. 6 83. .7 84. 0 82. 7 144. 1 4138. 70. 7 4 58. 7 84. .9 85. 1 83. 8 144. 4 4180. 71. 8 5 59. 9 86. 0 86. 3 85. 0 144. 7 4234. 73. 0 6 61. 1 87. .1 87. 5 86. 0 144. 9 4294. 74. 1 7 63 .0 88 .8 89 .3 87. .8 145. .1 4401. 75 .9 8 65 .3 91 .4 91 .7 90, .4 145 .4 4541. 78, .4 9 66 .3 92 .6 93 .0 91, .5 145, .5 4612. 79 .5 10 67, .2 93 .7 93 .9 92. .6 145. .5 4686. 80, .4 11 68 .2 94 .6 94. .9 93 .5. 145, .6 4753. 81 .4 12 69. .1 95 .6 95 .9 94. .5 145, .6 4826. 82, .4 13 70 .4 97 .0 97. .1 95. .9 145, .7 4915. 83 .7 14 72. .1 98 .6 98 .9 97. .5 146. ,0 5040. 85, .4 15 73 .1 99 .0 100. .0 98. .0 145, .4 5180. 86, .1 16 74. .2 100 .5 101. .1 99 .4 145 .9 5239. 87 .3 17 75. .3 101 .5 102. .3 100. .5 145, .8 5356. 88 .4 18 76 .4 102 .5 103 .3 101, .4 146 .0 5443. 89 .4 19 77. .3 103 .4 104 .4 102, .4 146, .0 5543. 90 .4 20 77 .8 104 .0 104, .9 103, .0 146 .2 5579. 90 .9 21 78. .5 104 .7 105 .6 103, .6 146, .2 5657. 91 .6 22 79. .3 105 .3 106 .4 104, .2 146 .1 5756. 92 .3 23 79. .9 105 .9 107, .0 104, .8 146, .2 5818. 92 .9 24 81. .0 106 .8 108 .1 105 .7 145 .9 5971. 93 .9 25 80. .5 106 .9 107, .6 105 .8 146 .5 5853. 93 .7 26 81. .5 107 .2 108 .7 106 .1 145 .9 6034. 94 .4 27 81 .9 107 .6 109 .1 106 .5 146 .0 6079. 94 .8 28 82 .4 108 .2 109 .7 107, .1 145 .9 6157. 95 .3 29 82 .9 108 .6 110 .1 107 .5 145 .9 6219. 95 .7 30 83. .4 109 .1 110.7 108. .0 145 .9 6293. 96 .2 31 83 .7 109 .7 111 .0 108 .6 146 .1 6319. 96 .7 32 83. .8 109 .8 111, .2 108. .7 146 .0 6339. 96 .8 — > COMMENTS: - VERY BIG BUBBLES AT SURFACE OF ROD - NO SLUG FLOW APPENDIX B. EXPERIMENTAL VOID GROWTH DATA 111 RUN * 15HERT FLUX= 30.0 W / C M ~ 2 RVG VEL= 0.214 M/S UPFLOW VQID GRQWTH PROFILE -z. o <_J c r o r Q CD 60-j 55-50-45-40-35-30-25-20-15-10-5-0 XEQ DSV X = -1.80 VOID M PQS (M) = 0.44 DIR ROD (M) = 0.013 Dlfl GLASS (M) = 0.022 ROD LENGTH (M) = 0.480 ROD NUMBER = 5 VOID PLPTEAU X = 3.62 -8 — r - 7 i 1 1 1 1 1 r -6 -5 -4 -3 -2 -1 0 1 2 EQUILIBRIUM QUALITY % Figure B.l: Void profile APPENDLX B. EXPERIMENTAL VOID GROWTH DATA 112 RUN * 15 HEAT FLUX= 30.0 W / C M . . 2 AVG VEL= 0.214 M/S UPFLBW 200-190_ 180 _ 170 _ S In 160 _ _ J LU ^ ISO-% 140-E 130-U J ¥— 120-1)0. 100. AVG. WALL TEMPERATURE ^ A A A A A AAAAA T WALL flVC= 145.6 -8 120-no. 100. 90 S I .oj UJ 4 0 . 30-20. —i 1 1 1 1 1 r -7 -6 -5 - 4 -3 -2 -1 E Q U I L I B R I U M O U R L I T Y X WATER TEMPERATURES CP 0 O TOUT A TMEAN • T IN —i 1 1 i 1 1 r -8 -7 -6 -5 - 4 -3 -2 -1 20 -IB . (000 )6_ • 14 _ u 12-• • x: 10. » e_ (J 6 . X 4_ 2_ 0 AVG. HEAT TRANSFER COEF. H.T.C. flVG.- 5239 H.T.C. D-B.= 2251 AAA A A A A * ' -8 200 180. ISO _ 140. £ 120. 100 _ £ 8 0 . 60 4 0 . 2 0 . 0 -8 —i 1 1 1 1 1 r -7 -6 -5 -A -3 -2 -1 E Q U I L I B R I U M O U R L I T Y X AVERAGE NU » NU • RVG.= 71 NU « D-B.= 31 AAA A & & A A ' 1 1 1 1 1 1 r -7 -6 -5 - 4 -3 -2 -1 E Q U I L I B R I U M O U R L I T Y X E Q U I L I B R I U M O U R L I T Y * Figure B.2: Wall and water temperature, heat transfer coefficients and Nus-selt number APPENDIX B. EXPERIMENTAL VOID GROWTH DATA 113 RUN * 15HERT FLUX= 30.0 W / C M « 2 RVG VEL= 0.214 M/S UPFLOW Li to ir> ,15 14 _ , 13 _ 12 _ 11 _ 10 _ 09 _ 06 _ ,07 _ .06 _ ,05 _ ,04 _ ,03_ ,02_ .01 _ .00 _. - 8 20 _ 18 _ 16 _ 5 14 J O 2 12 _ * 10 _| e 6 _ 4 _ 2 _ 0 0£ M A S S FLOW R A T E M A A A A A A AAV)ftxY\ — i 1 1 1 1 1 r - 7 - 6 - 5 - 4 - 3 - 2 - 1 E Q U I L I B R I U M Q U A L I T Y X RE « RE « AVG= 5484 M A A A A A A AA6 tf £ A 1.50 1.40 _ 1.30 _ 1.20 _ 1.10 _ to 1-00 _ * 0.90 _ E o.eo _ o 0.70 _ S? 0.60 _ 0.50 _ 0.40 _ 0.30 _ 0.20 _ 0.10 _ 0.00 _ 0 — i 1 1 1 1 1 1 1 - 8 - 7 - 6 - 5 - 4 - 3 -2 - 1 0 E Q U I L I B R I U M D U A L I T Y X 120 110 _ 100 _ 90 _ <g <r> 80 _ _i UJ UJ h 50 _ UJ t-40 _ 3D _ 20 _ V E L O C I T Y VEL RVG M/5= 0.214 A A A A AA AA AAWtfA - 8 -T 1 1 1 1 1 r 7 - 6 - 5 - 4 - 3 - 2 -1 E Q U I L I B R I U M Q U A L I T Y X O U T L E T T E M P E R A T U R E ES B1 • T C A L 0 A T O U T — i 1 1 1 1 1 r - 8 - 7 - 6 - 5 - 4 - 3 - 2 -1 E Q U I L I B R I U M Q U A L I T Y X Figure B.3: Mass flow rates, velocities, Reynolds number and outlet temper-atures 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0080866/manifest

Comment

Related Items