• TABLE OF CONTENTS
HIDE
 Title Page
 Dedication
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Nomenclature
 Abstract
 Introduction
 Experimental facilities
 Heat transfer and pressure drop...
 Incipience and hysteresis
 Nucleation site density
 A unified model for vapor bubble...
 Probability density functions of...
 Vapor bubble growth rate
 Conclusions and recommendations...
 Appendix
 Reference
 Biographical sketch
 Copyright














Title: ebullition process in forced convection boiling
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00090917/00001
 Material Information
Title: ebullition process in forced convection boiling
Series Title: ebullition process in forced convection boiling
Physical Description: Book
Creator: Zeng, Ling-Zhong,
 Record Information
Bibliographic ID: UF00090917
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 001938273
oclc - 30947296

Table of Contents
    Title Page
        Page i
    Dedication
        Page ii
    Acknowledgement
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
        Page vii
    List of Tables
        Page viii
    List of Figures
        Page ix
        Page x
        Page xi
        Page xii
    Nomenclature
        Page xiii
        Page xiv
        Page xv
        Page xvi
    Abstract
        Page xvii
        Page xviii
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
    Experimental facilities
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
    Heat transfer and pressure drop in saturated flow boiling
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
    Incipience and hysteresis
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
    Nucleation site density
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
    A unified model for vapor bubble detachment
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
    Probability density functions of vapor bubble detachment diameter
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
    Vapor bubble growth rate
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
    Conclusions and recommendations for future research
        Page 157
        Page 158
        Page 159
        Page 160
    Appendix
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
    Reference
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
    Biographical sketch
        Page 177
        Page 178
        Page 179
    Copyright
        Copyright
Full Text











THE EBULLITION PROCESS
IN FORCED CONVECTION BOILING


















By

LING-ZHONG ZENG


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1993


UhsvfYSrrY OF FLORIDA LIERIMItS
























TO MY WIFE TANG YONG














ACKNOWLEDGEMENTS

My greatest appreciation goes to Professor James

Klausner, chairman of the supervisory committee, for all the

support and encouragement during the course of this research.

Dr. Klausner, a mentor and role model, has spent a great deal

of time and effort in helping me with the fabrication of the

experimental facility, analysis of the experimental data, as

well as editing this dissertation. I also want to extend my

appreciation to Professor Renwei Mei, member of the

supervisory committee. Dr. Mei has always been accessible and

has provided many useful suggestions throughout the course of

this research. I would also like to thank professors C.K.

Hsieh, D.Y. Goswami, and S. Anghaie for serving on the

supervisory committee. Their useful suggestions have improved

this dissertation.

I also want to thank Dave Bernhard and Boby Warren,

fellow graduate students and friends. Dave and Bob have

provided substantial support in calibrating instrumentation

and planning experiments.

I can not express enough love and appreciation to my

wife, Tang Yong. Without her support, understanding, and

sacrifice, I would not have finished my Ph.D. program.

Finally, I want to thank my parents, brothers, and sisters for


iii








their support throughout the entire course of my education.

It is my parents who inspired me to pursue the education I

have obtained.















TABLE OF CONTENTS

ACKNOWLEDGEMENTS ........................................iii

LIST OF TABLES ........................................ iii

LIST OF FIGURES ........................................ ix

NOMENCLATURE ...........................................xiii

ABSTRACT ...............................................xvii

CHAPTERS

1 INTRODUCTION ....................................1

2 EXPERIMENTAL FACILITIES .........................5

2.1 Flow Boiling Test Loop .....................5
2.2 Construction of Transparent Test Section ..10
2.3 Development of Capacitance Based
Film Thickness Sensors ....................11
2.3.1 Introduction .......................11
2.3.2 Design and Fabrication of
Film Thickness Sensor .............. 13
2.3.3 Instrumentation and Calibration ....15
2.4 Data Acquisition System ...................24

3 HEAT TRANSFER AND PRESSURE DROP
IN SATURATED FLOW BOILING ......................28

3.1 Introduction ............................. 28
3.2 Experimental Results and Discussions ......30
3.3 Conclusions ...............................37

4 INCIPIENCE AND HYSTERESIS .................... 38

4.1 Introduction and Literature Survey ........38
4.2 Experimental Results ......................42
4.3 Theoretical Analysis for Boiling
Incipience ......................... ......44
4.3.1 Boiling Initiation ................. 44
4.3.2 Sustaining Incipience Superheat ....54
4.3.3 Hysteresis of Boiling Incipience ...56
4.4 Conclusions ..................... ........ 57









5 NUCLEATION SITE DENSITY .....................59

5.1 Literature Survey ......................59
5.2 Optical Facility and Measuring
Technique .................................64
5.3 Experimental Results ......................66
5.4 Discussion of Results ..................... 78
5.5 Conclusions ...............................83

6 A UNIFIED MODEL FOR VAPOR BUBBLE DETACHMENT ....84

6.1 Introduction ..............................84
6.2 Literature Survey ......................... 86
6.2.1 Pool Boiling Departure Diameter
Correlations ....................86
6.2.2 Flow Boiling Detachment Diameter
Correlations ....................... 92
6.3 Development of Departure and Lift-off
Model .....................................94
6.3.1 Formulation ........................ 94
6.3.2 Expressions for Bubble Departure
and Lift-off Diameter ............. 105
6.4 Comparison with Experimental Data ........107
6.4.1 Pool Boiling Data ................. 108
6.4.2 Flow Boiling Data ................. 119
6.5 Conclusions ...........................127

7 PROBABILITY DENSITY FUNCTIONS OF VAPOR
BUBBLE DETACHMENT DIAMETER .................... 130

7.1 Introduction ............................. 130
7.2 Formulation ..............................132
7.3 Comparison with Experimental Data ........134
7.4 Conclusions ...........................138

8 VAPOR BUBBLE GROWTH RATE ...................... 140

8.1 Introduction .............................140
8.2 Facility and Methodology ................. 141
8.3 Results and Discussions .................. 144
8.4 Conclusions ........................... 156

9 CONCLUSIONS AND RECOMMENDATIONS
FOR FUTURE RESEARCH ...........................157

9.1 Accomplishments and Findings ............. 157
9.2 Suggestions for Future Research ..........159

APPENDICES
A HEAT TRANSFER COEFFICIENT, PRESSURE DROP,
AND LIQUID FILM THICKNESS IN STRATIFIED
TWO-PHASE FLOW ................................161









B NUCLEATION SITE DENSITY IN FORCED
CONVECTION BOILING ............................164

REFERENCES .................................. .......... ..... 167

BIOGRAPHICAL SKETCH ...................................... 177


vii














LIST OF TABLES


Table

6-1 Summary of forces appearing in momentum
equations ..........................................106

6-2 Mean deviation tabulated for present bubble model
as well as other correlations reported in
literature .................. ...... ....... .....109

6-3 Comparison of measured and predicted vapor bubble
departure diameter for elevated pressure data
using present model ................................117

6-4 Comparison of measured and predicted vapor bubble
departure diameter for reduced gravity data
using present model ............................... 118

6-5 Measured and predicted departure diameters
based on high speed cinematography data ............ 123

8-1 A summary of parameters controlling vapor bubble
growth rate in flow boiling ....................... 155


viii














LIST OF FIGURES


Figure

2-1 Schematic diagram of flow boiling facility ...........6

2-2 Calibration curve for flowmeter ......................8

2-3 Calibration curve of heat loss for preheaters ........9

2-4 Isometric view of transparent test section ..........11

2-5 Cut-away view of liquid film thickness sensor .......14

2-6 Prediction of relative film thickness vs capacitance
using model of Chun and Sung (1986) .................18

2-7 Calibration curve for film thickness sensor .........19

2-8 Temperature calibration for film thickness sensor
filled with pure liquid ............................. 21

2-9 Temperature calibration for film thickness sensor
filled with pure vapor ............................. 22

2-10 Close-up view of stratified two-phase flow using
CCD camera (flow direction is from left to right) ...23

2-11 Comparison of liquid film thickness measured with
CCD camera and capacitance sensor ...................24

2-12 A schematic diagram of data acquisition system ......26

3-1 Microconvection heat transfer for saturated forced
convection nucleate boiling ......................... 32

3-2 Macroconvection heat transfer coefficient in
saturated forced convection boiling ................. 33

3-3 Pressure drop in horizontal two-phase flow ..........34

3-4 Zuber and Findlay's (1965) correlation for void
fraction in horizontal stratified two-phase flow ....36

4-1 Nucleate pool boiling hysteresis constructed from
the data of Kim and Burgles (1988) ..................39

ix








4-2 A typical saturated flow boiling plot of q, vs
AT, for G=180 kg/m2-s, X=0.156, and 6=4.8 mm ......43

4-3 Measured saturated flow boiling incipience
wall superheat ......................................45

4-4 An idealized sketch of a vapor embryo in a
conical cavity ......................................47

4-5 Variation of vapor temperature with vapor embryo
volume during expansion inside a conical cavity .....49

4-6 An idealized sketch of a vapor embryo in a
reentry cavity ......................................52

4-7 Variation of minimum vapor temperature and
initiation superheat with cavity reservoir
mouth radius .......................................53

5-1 Pool boiling nucleation site density data from
Griffith and Wallis (1960) ..........................61

5-2 A diagram of optical facility for measurement
of nucleation site density ..........................65

5-3 A typical photograph of nucleation sites on
a boiling surface ...................................67

5-4 Nucleation site density as a function of wall
superheat for constant heat flux and saturation
temperature ........................................68

5-5 Pool boiling nucleation site density as functions
of wall superheat and heat flux .....................69

5-6 Nucleation site density as a function of vapor
Velocity for constant heat flux and liquid film
thickness ........................................ 71

5-7 Nucleation site density as a function of mass
flux ................................................ 72

5-8 Nucleation site density as a function of liquid
velocity ........................................ 73

5-9 Nucleation site density and liquid film thickness
as functions of vapor velocity ......................74

5-10 Nucleation site density as a function of liquid
film thickness ......................................75








5-11 Nucleation site density as a function of heat
flux ................... ........................... 77

5-12 Nucleation site density as a function of wall
superheat ........................................ .78

5-13 Nucleation site density as functions of saturation
temperature and wall superheat ......................79

5-14 Nucleation site density as a function of critical
radius for constant heat flux and vapor velocity ....80

5-15 Nucleation site density as a function of critical
radius for all flow boiling data ....................81

6-1 A typical picture of vapor bubble departure and
lift-off in flow boiling ............................85

6-2 A schematic sketch of vapor bubble detachment
process in flow boiling .............................95

6-3 Comparison of predicted and measured vapor bubble
departure diameter for subatmospheric pressure
data using present model ...........................112

6-4 Comparison of predicted and measured vapor bubble
departure diameter for subatmospheric pressure
data using Cole and Shulman 2 correlation ..........113

6-5 Comparison of predicted and measured vapor bubble
departure diameter for atmospheric pressure data
using present model ............................... 114

6-6 Comparison of predicted and measured vapor bubble
departure diameter for atmospheric pressure data
using Cole and Shulman 2 correlation ............... 115

6-7 Departure diameter variation with mean liquid
velocity at constant AT, .........................121

6-8 Comparison between predicted and measured
departure diameters ............................... 122

6-9 Departure diameter variation with mean liquid
velocity and AT ...................................... 125

6-10 Predicted inclination angle variation with
predicted departure diameter .......................126

6-11 Predicted inclination angle variation with
mean liquid velocity and ATt .....................127








6-12 Comparison between predicted and measured
lift-off diameter ............................ ..... 128

7-1 Statistical distribution of bubble lift-off
diameter in flow boiling ...........................135

7-2 Statistical distribution of bubble departure
diameter in flow boiling at constant AT, .........136

7-3 Statistical distribution of bubble departure
diameter in flow boiling at constant u, ............137

8-1 A schematic diagram of the high speed facility
for filming vapor bubble growth rate ..............142

8-2 Time history of bubble growth ......................145

8-3 Time history of bubble growth .............. ........146

8-4 Time history of bubble growth .....................147

8-5 Time history of bubble growth .......................148

8-6 Time history of bubble growth ......................149

8-7 Time history of bubble growth ......................150

8-8 Time history of bubble growth .......................151

8-9 Time history of bubble growth .....................152

8-10 Time history of bubble growth .......................153

8-11 Time history of bubble growth ....................154


xii














NOMENCLATURE

a, a(t) Radius of a growing vapor bubble

C Capacitance

CD Drag coefficient for a freely rising vapor bubble
in an infinite liquid

Cp Liquid specific heat

C, Empirical constant, equals 20/3

d Vapor bubble diameter

d, Diameter of contact area

D Inside dimension of the test section or diameter

F Force

g Earth gravity

G Mass flux

h Heat transfer coefficient

hfg Vaporization latent heat

Ja Jakob number

k Thermal conductivity

K Power law bubble growth constant as in a(t)=Ktn

m Mass

M Molecular weight

n Power law bubble growth index as in a(t)=Kt"

n/A Nucleation site density

P Absolute pressure or Polarization factor


xiii








p(x) Probability density function

q Heat flux

r Radius of the liquid/vapor interface

r, Mouth radius of the cavity

r2 Mouth radius of the cavity reservoir

R Engineering gas constant

Re Reynolds number

t Time

T Temperature

u Mean velocity

U(y) Velocity profile

V Volume

X Vapor quality

Greek Symbols

a Void fraction

6 Liquid film thickness

AP Pressure drop

AT Superheat

1 Liquid thermal diffusivity

0 Contact angle

0i Bubble inclination angle

A Dynamic viscosity

p Density

a Surface tension coefficient or standard deviation

Half cone angle of the cavity

e Relative permittivity


xiv








Subscripts

b Bulk or buoyancy

cp Contact pressure

d Bubble departure

da Dynamic advanced

dF Departure diameter predicted from Fritz's model

dr Dynamic receded

du Force due to bubble growth

g Non-condensible gas

G Garolite material

h Hydraulic

inc,i Incipience, initiation

inc,s Incipience, sustaining

L Bubble lift-off or lift force created by bubble
wake

e Liquid phase

m Mixture of vapor and liquid

mac Macroconvection

max Maximum

mic Microconvection

min Minimum

s Surface tension

sa Static advanced

sat Saturation

sL Shear lift

sr Static receded

24 Two-phase








v Vapor phase

w Wall

x X-direction, i.e., horizontal direction

y Y-direction, i.e., vertical direction


xvi














Abstract of Dissertation Presented to the Graduate School of
the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


THE EBULLITION PROCESS
IN FORCED CONVECTION BOILING

By

Ling-zhong Zeng

August, 1993

Chairman: Professor James F. Klausner
Major Department: Mechanical Engineering

A forced convection boiling facility with Refrigerant

R113 was designed and fabricated in order to experimentally

study the ebullition process in horizontal flow boiling.

Capacitance sensors were developed for measuring the liquid

film thickness for stratified and annular two-phase flow.

Measurements of heat transfer coefficient, pressure drop, and

liquid film thickness in stratified two-phase flow with and

without boiling have been obtained. The experimental data

have conclusively demonstrated that microconvection, which is

the heat transfer due to the ebullition process, is

significant in almost all phases of saturated flow boiling.

The initiation and sustaining incipience superheats of

saturated flow boiling with R113 were found to be insensitive

to the fluid convection but they strongly depend on the system

pressure as well as the cooling history of the heating surface

xvii








prior to boiling. Nucleation site density of saturated

forced convection boiling was measured using a CCD camera.

The mean vapor velocity, heat flux, and system pressure appear

to exert a dominant parametric influence on the nucleation

site density. The critical cavity radius is an important

parameter in characterizing the nucleation process but by

itself it is not sufficient to correlate nucleation site

density data for saturated flow boiling. Based on

experimental observations and theoretical reasoning, an

analytical model has been developed for the prediction of

vapor bubble detachment diameters in saturated pool and flow

boiling. The vapor bubble growth rate is a necessary input to

the model. It is demonstrated that over the wide range of

conditions considered, the accuracy of the detachment

diameters predicted using the present model is significantly

improved over existing correlations. The model was also

extended to predict the probability density functions (pdf's)

of detachment diameters by specifying the pdf's of wall

superheat and liquid velocity. The vapor bubble growth rate

during saturated flow boiling was measured using a high speed

cinematography. Based on the experimental data obtained

herein, the vapor bubble radius can be expressed as a function

of time using a power law, where the exponent decreases with

increasing system pressure. The objective of this research is

to understand the fundamentals of the ebullition process in

flow boiling.


xviii














CHAPTER 1

INTRODUCTION

Forced convection boiling, also referred to as flow

boiling, has been used in a variety of engineering

applications for its high heat and mass transfer rates. In

nuclear power applications, flow boiling with water is used to

extract heat from reactors. Also flow boiling can be found in

fossil fuel fired steam generators, the chemical process

industry, refrigeration and air-conditioning industry, and

cooling of electrical distribution facilities. Other

potentially important applications include compact flow

boiling heat exchangers for use in spacecraft and cooling of

microelectronic components.

Due to its engineering importance, boiling heat transfer

has been the focus of extensive research for the past four

decades. However, to date, boiling remains one of the most

controversial subjects in the field of heat transfer. Many

questions raised four decades ago concerning boiling phenomena

remain unanswered (Lienhard, 1988). Current engineering

designs involving boiling phenomena rely heavily on empirical

correlations developed from experimental measurements.

Rohsenow (1952) first suggested that the rate of heat transfer

associated with forced convection boiling is due to two








2

additive mechanisms, that due to bulk turbulence and that due

to ebullition. Based on Rohsenow's conjecture, Chen (1966)

proposed a saturated flow boiling heat transfer correlation

which is simply the sum of the respective macroconvection and

microconvection heat transfer coefficients. The terms macro-

and microconvection respectively denote the contribution due

to heat transfer from bulk turbulent convection and that due

to the ebullition process. The macroconvection heat transfer

coefficient was calculated using a single-phase flow

correlation based on the liquid fraction flowing modified by

an enhancement factor, while the microconvection heat transfer

coefficient was calculated using a pool boiling correlation

modified by a suppression factor.

Chen's (1966) correlation or modified forms of it are

widely used throughout industry despite the fact that they

fail to accurately correlate a wide range of flow boiling heat

transfer data (Gungor and Winterton, 1986). One

characteristic of Chen's correlation is that it predicts the

microconvection contribution to flow boiling heat transfer is

always small compared to macroconvection. In contrast, Mesler

(1977) argued that the microconvection component is dominant.

Staub and Zuber (1966), Frost and Kippenhan (1967), Klausner

(1989), and Kenning and Cooper (1989) have presented flow

boiling heat transfer data which display a strong dependence

on the ebullition process. In addition, experimental data

provided by Koumoutsos et al. (1968) and Cooper et al. (1983)








3

demonstrate that the ebullition process in flow boiling cannot

be adequately modelled with pool boiling correlations. In

order to significantly improve flow boiling heat transfer

predictions over Chen's approach, it is necessary to

understand the mechanisms governing both macro- and

microconvection as well as their relative contribution to the

total heat transfer.

In this work, major efforts have focused on understanding

the physics governing vapor bubble incipience, nucleation site

density, growth and detachment in forced convection boiling.

In order to achieve this goal, a flow boiling facility with

refrigerant R113 was designed and fabricated. The boiling

test section is optically transparent thus allowing for the

visualization of the ebullition process. A CCD camera has

been used to measure nucleation site densities and high speed

cinematography was used to measure vapor bubble growth rates.

Two capacitance-based film thickness sensors were designed and

fabricated to measure the liquid film thickness on the upper

and lower surfaces of the horizontal square test section.

Since the flow boiling facility usually experiences large

temperature variations during operation, the temperature

dependence of the capacitance sensors must be accounted for.

A new and simple method has been developed to account for

temperature when using the film thickness sensor calibration

curve.

Using the current flow boiling facility, experimental








4

evidence was obtained to demonstrate that the heat transfer

contribution due to the ebullition process is significant in

almost all phase of boiling. Experimental data on the

incipience wall superheat, nucleation site density, and vapor

bubble growth rate for saturated flow boiling have been

gathered over a wide range of flow and thermal conditions.

The parametric influence of two-phase flow conditions on the

ebullition process have been analytically investigated.

An analytical model has been developed for the prediction

of vapor bubble departure and lift-off diameters for both pool

and flow boiling. The model was compared against all

experimental data available in the literature, and excellent

agreement has been achieved. Based on this bubble detachment

model, an analytical approach was proposed for predicting

vapor bubble detachment diameter probability density functions

(pdf's) for a specified wall superheat pdf and liquid velocity

pdf.














CHAPTER 2

EXPERIMENTAL FACILITIES



2.1 Flow Boiling Test Loop

A flow boiling facility, shown schematically in Figure 2-

1, was designed and fabricated. Refrigerant R113 was selected

as the boiling liquid in this facility primarily due to its

low latent heat of evaporation and boiling point. A variable

speed model 221 Micropump was used to pump R113 through the

facility. A freon dryer/filter was installed on the discharge

of the pump to filter out alien particles in the liquid and to

prevent the formation of hydrofluoric acid in the refrigerant.

The volumetric flow rate of liquid was monitored with an Erdco

Model 2521 vane type flowmeter equipped with a 4-20 ma analog

output. The flowmeter output was attached to a 500 ohm power

resistor. The voltage across the resistor was recorded with

a digital data acquisition system which will be discussed

later. The flowmeter was calibrated using a volume-time

method. A calibration curve for the volumetric flow rate vs

voltage is displayed in Figure 2-2. The standard deviation of

the experimental data from a third order polynomial least-

squares fit is 0.5%, which is equivalent to the repeatability

of the flow meter claimed by the manufacturer. At the outlet








6

of the flowmeter, five preheaters have been installed to

generate a saturated two-phase mixture. Each preheater


GEAR
PUMP


Figure 2-1. Schematic diagram of flow boiling facility.


consists of a 25 mm ID, 1.2 m long hard copper round pipe

around which 18 gauge nichrome wire has been circumferentially

wrapped. The nichrome wire is electrically insulated from the

copper pipe with ceramic beads. The preheaters are thermally

insulated using a 25 mm thick fiberglass insulation layer.








7

Each of the five preheaters is powered by a 240 volt line

through an adjustable AC autotransformer. The heat loss of

the preheaters has been calibrated as a function of

temperature difference between the outer surface of the

insulation and the ambient. A typical calibration curve is

shown in Figure 2-3. In order to allow the two-phase mixture

generated by the preheaters to be fully developed and smoothly

flow into a square transparent test section, which will be

described shortly, a 1.5 m long and 25 x 25 mm inner dimension

square copper duct has been mounted downstream of the

preheaters. The duct is also thermally insulated using

fiberglass, thus provides an adiabatic developing length for

the two-phase flow. A capacitance-based liquid film thickness

meter, which will be described in detail in section 2.3, was

installed on the either side of the test section to measure

the inlet and outlet liquid film thickness of the two-phase

mixture. Two Viatran model 2415 static pressure transducers

have been installed at the inlet and outlet of the test

section to measure the system pressure with an accuracy of

0.5% of full scale (30 psig). Two type E thermocouple probes

were also located at the same position to measure the bulk

temperature of R113. When the two-phase mixture becomes

saturated, the measured bulk temperature using thermocouples

matches that calculated from the saturation line based on the

measured system pressure to within 0.50C, which is also the

accuracy of the absolute temperature measurement from the








8

thermocouples. A precision Viatran differential pressure

transducer was installed to measure the pressure drop across


0.20

0.18

0.16

0.14

0.12

0.10

0.08

0.06

0.04


0.02 I-


0.00


0 1 2 3 4 5 6 7 8 9


Flow Meter Output (Volts)

Figure 2-2. Calibration curve for flowmeter.


the test section with an accuracy of 0.25% of full scale (120

mmH2O). A throttle valve is located downstream of the test

section, which allows the test section pressure to be adjusted

from atmospheric pressure to 30 psig, which is the maximum


I I I I I I I I


I I I I I I I 1 I

























4


I 3


2


1v



0 10 20 30 40 50

To-T (aC)


Figure 2-3. Calibration curve of heat loss for preheaters.


safe operating pressure of the square pyrex section.

Following the test section, the R113 two-phase mixture

condenses in a shell and tube water cooled heat exchanger to

return to the liquid storage tank.



2.2 Construction of Transparent Test Section

The major difficulties associated with fabricating a

transparent flow boiling refrigerant based test section and












PRYREX GLASS
TEST SECTION


HEATING -
SURFACE
(20mm x 0,127mm)


FLANGE


Figure 2-4. Isometric view of transparent test section.

connecting it with rigid copper pipes are the facts that the

pyrex glass is brittle and small stress concentrations

substantially reduce its safe operating pressure. After

testing many different designs, a satisfactory test section

was eventually fabricated. The fabrication procedures, which

have been detailed by Bernhard (1993), will not be repeated

here. A brief description of the test section is given. The

main body of the flow boiling test section is comprised of a








11

25 x 25 mm ID square pyrex glass tube that is 4 mm thick and

0.457 m long as depicted in Figure 2-4. A 0.13 mm thick and

22 mm wide nichrome strip, used as a heating and boiling

surface, has been adhered to the lower inner surface of the

square tube with epoxy. Six equally spaced 36 gauge type E

thermocouples were located underneath the nichrome strip using

high thermal conductivity epoxy. The mean wall temperature of

the nichrome strip was obtained by averaging the readings from

these six thermocouples. The test section was connected to

the facility with a brass block on either side. Each end of

the nichrome strip was bolted to the block to maintain good

electrical contact. Epoxy was used to seal gaps between the

glass tube and brass blocks. The facility was pressurized

with air to 30 psig and leak-checked prior to introducing

R113. Due to safety considerations, the facility has not been

operated at pressures above 30 psig.



2.3 Development of Capacitance Based Film Thickness Sensors

2.3.1 Introduction

It has been observed that in a horizontal saturated flow

boiling system, vapor-liquid flow is usually in a stratified

or annular flow regime due to the influence of gravity.

Research involving this flow regime requires knowledge of the

liquid film thickness distribution along the wall of a duct.

Many of the techniques used to measure liquid film thickness

and volume fraction were summarized by Hewitt (1978) and Jones








12

(1983). Of these, the only non-intrusive measuring

techniques, applicable to dielectric fluids, which capture the

liquid film thickness and volume fraction with a very rapid

response time are capacitance and radiation absorption

techniques. The implementation of the radiation absorption

technique requires expensive, bulky equipment with which

special safety precautions must be adhered to. In contrast,

the capacitance sensors used to measure liquid film thickness

and volume fraction are compact, safe, and inexpensive and

thus were selected for this research. Ozgu and Chen (1973)

used a capacitance sensor to measure liquid film thickness for

axisymmetric two-phase flow while Abouelwafa and Kendall

(1979), Sami et al. (1980), Irons and Chang (1983), Chun and

Sung (1986), and Gaerates and Borst (1988) used a capacitance

sensor to measure volume fraction. A summary of these

investigations can be found in Delil (1986). All of the

capacitance probes and measuring techniques reported by these

investigators were used only for adiabatic flow; there was

very little mention made of the temperature dependence of

capacitance sensors. Furthermore, the ring-type capacitance

sensor described by Ozgu and Chen (1973) may only be used to

measure symmetric two-phase duct flow, such as vertical up-

flow or down-flow. The sensor is not applicable for use with

horizontal two-phase flow, which is usually asymmetric due to

the gravitational stratification of the phases.

Since a forced convection boiling system undergoes large








13

variations in temperature, the calibration of the capacitance

sensor must account for its temperature dependence in order to

obtain accurate liquid film thickness measurements. Both the

permittivity of the liquid and the material of construction

are temperature dependent. Therefore, either the sensor must

be calibrated over the range of temperatures for which it will

operate or a suitable temperature correction scheme must be

employed when the sensor is calibrated at a fixed temperature.

The latter approach has been successfully used in this work.



2.3.2 Design and Fabrication of Film Thickness Sensor

The liquid film thickness sensor was designed to match

the inner dimension of the test section for a smooth

transition of the flow. Therefore, four Garolite sheets (152

x 38 x 6 mm) were machined and bonded together with Conap

epoxy to form a body which has a 25 x 25 mm inner square cross

section as shown in Figure 2-5. Garolite material was chosen

for the fabrication of the film thickness sensor body because

it has good dielectric properties and is corrosion resistant

to refrigerants. Two parallel grooves, 7.9 mm wide and 3.2 mm

deep were machined on both the outer upper and lower halves of

the sensor body for placement of the capacitance strips. The

distance separating two adjacent grooves is 7.9 mm. This

distance was chosen because Ozgu and Chen (1973) reported the

optimum thickness and distance between the parallel ring

sensors is equal to the film thickness at which the highest















GAROLITE
FLANGE ~-- -




S25 m



BNC --
CONNECTOR -

ALUMINUM LOWER COPPER
SHIELDING CAPACITANCE
STRIPS


15,24 cm

Figure 2-5. Cut-away view of liquid film thickness sensor.

resolution is expected. It shall be demonstrated that the
best resolution is obtained with liquid film thickness of 5 to
6 mm and poor resolution is observed near the centerline.
Copper strips with a thickness of 0.1 mm were bonded into the
grooves with epoxy. An aluminum chassis was fabricated around
the sensor body. The purpose of the chassis is twofold: it
shields extraneous electromagnetic radiation and also
compresses the sensor body so that it is pressure resistant.








15

Four BNC connectors, each of which were connected to a copper

strip using unshielded wire, were electrically insulated and

mounted on the chassis. Two square Garolite flanges (10.2 x

10.2 cm), which were used to mount the film thickness sensor

in the flow boiling facility, were bonded to the two ends of

the body with epoxy. The film thicknesses in the lower and

upper half of the duct were determined from the capacitance

between the respective pair of lower and upper parallel copper

strips. Two identical sensors were fabricated in this manner.



2.3.3 Instrumentation and Calibration

In order to obtain good performance from the film

thickness sensors, a high resolution capacitance meter must be

used to measure the capacitance across the parallel strips.

For this purpose, a Keithley 590 digital CV analyzer, which

has a resolution of 0.1 Ff, an accuracy and repeatability of

0.1% of full scale, and a frequency response of 45 kHz, as

reported by the manufacturer, was used to measure the

capacitance. The analog output of this instrument was

connected to a 12 bit analog to digital converter (A/D) which

will be described in section 2.4.

An accurate analytic relation between the capacitance and

liquid film thickness is extremely difficult to determine due

to the three-dimensional nature of the sensor. Therefore, the

sensor must be calibrated. The difficulty associated

calibrating the sensor for use in a variable temperature








16

environment is that the permittivity of both the solid and

liquid is temperature dependent. A complete calibration is

obtained only when the liquid film thickness and temperature

are varied over the full range. Such a calibration is tedious

and impractical. Therefore, an innovative scheme is proposed

which allows the sensor to be used based on its calibration at

a fixed temperature

A crude model for predicting the capacitance as a

function of film thickness or volume fraction for a known

material permittivity was introduced by Chun and Sung (1986)

by considering the sensor as a network of parallel and series

equivalent plate-type capacitors. This type of modeling was

attempted for the sensor described above. The relative

permittivity of R113 vapor was taken to be unity. The

relative permittivity of liquid R113 as a function of

temperature was determined from the Clausius-Mosotti equation

as reported by Downing (1988):

M+Pp,
et Pp1 (2-1)
SM-Pp,


where et is the liquid relative permittivity, M is the

molecular weight, P is the polarization, and p, is the liquid

density. The temperature dependence of permittivity comes

from the fact that the liquid density varies with temperature.

Published values of permittivity could not be found for

Garolite material. However, Garolite is a composite material








17

manufactured from E-glass and a phenolic based epoxy resin,

and based on the data of Leeds (1972), the relative

permittivity of Garolite was approximated over a range of

temperature of 25 to 200 OC using

eG=4.213+0.0023T (2-2)

where T is temperature in degrees Celsius. Using the crude

model of Chun and Sung (1986) the relative film thickness, 6',

for horizontal stratified flow is shown in Figure 2-6 as a

function of the relative capacitance C* for both the upper and

lower section of the sensor at 25 and 80 OC. Here C* is

defined by

c-c
C*- (2-3)
Cl-C,


where C is the capacitance across the sensor for two-phase

flow, C, is the sensor capacitance for purely liquid flow, and

C, is that for purely vapor flow; all these capacitances are

temperature dependent. For the lower section, 6'=6/h, and for

the upper section, 6'=6/h-1, where 6 is the liquid film

thickness and h=12.7 mm is the distance from the sensor inside

wall to the centerline. The results displayed in Fig. 2.6

reveal that the functional relationship between 6' and C* is

essentially independent of temperature over the 25-80 C range

investigated.

Guided by the fact that the relationship between C* and

6' is not temperature dependent, it was decided to calibrate

















1.0 I I--

v Upper section, 25C
o Upper section, 80C
S0.8-



o 0.6 -



0.4



0.2 Lower section, 250C
Sv Lower section, 80C


0.0 1 1 1
0.0 0.2 0.4 0.6 0.8 1.0

Relative Capacitance C'



Figure 2-6. Prediction of relative film thickness versus
capacitance using model of Chun and Sung
(1986).


the sensor for the stratified flow regime at room temperature

on a bench top under carefully controlled conditions. The

results of the calibration for sensor #1 were tabulated in

terms of 6' and C' and are shown in Figure 2-7. It is noted

that for the lower section of the sensor, 6' was normalized by

0.75h rather than h since the full scale of the measurement

for this section is 0.75h. There was no modification for the
















Meter #1, calibrated at 25 OC
1.0 1 1 -
Upper section
Lower section
0.8


0.6 -



0.4



0.2 -



0.0 I I
0.0 0.2 0.4 0.6 0.8 1.0

Relative Capacitance C'




Figure 2-7. Calibration curve for film thickness sensor.


upper section. It is seen from Figure 2-7 that the resolution

is good when the film thickness is well below the centerline

for the lower section and well above the centerline for the

upper section. There exists a small region near the

centerline where the film thickness can not be resolved.

However, for the current study with saturated forced

convection boiling, this does not pose a severe problem

because the liquid film is always well below the centerline








20

when the two-phase mixture is at saturated conditions.

In order to determine the liquid film thickness from the

calibration curves shown in Figure 2-7, C, and C, must be

determined as a function of temperature when the two-phase

mixture is at a temperature other than the calibration

temperature. The functions were determined after the sensors

had been installed in the facility. To do so, pure liquid

R113 was circulated through the facility and was heated. When

a steady temperature was reached, the capacitances across the

lower and upper pairs of copper strips were recorded. This

procedure was repeated for a series of increments of

temperature while the throttle valve was adjusted to elevate

the system pressure to avoid vapor generation. The results

obtained have been displayed in Figure 2-8. Similar

procedures were followed to obtain the capacitance for pure

vapor as a function of temperature as shown in Figure 2-9.

Because pure vapor flow could only be achieved by

depressurizing the boiling facility after two-phase flow had

been established, it was difficult to obtain the measurement

over a wide range of temperature. In this work only four

different temperatures were obtained. By incorporating the

calibration curves shown in Figures 2-7, 2-8, and 2-9, the

liquid film thickness may be determined as a function of

capacitance and temperature of any two-phase mixture.

In order to evaluate the performance of the liquid film

thickness sensor developed here, a CCD camera was set up for















1.85 1111

v meter #1, upper section
1.80 meter #1, lower section


1.75


S 1.70


S1.65

S1.60 -



1.55


1.50
20 30 40 50 60 70

Temperature (C)

Figure 2-8. Temperature calibration for film thickness
sensor filled with full liquid.

optically measuring the liquid film thickness for stratified

flow. The camera was focused normal to the transparent test

section to avoid optical distortion. A typical picture of

two-phase stratified flow obtained with the CCD camera used

for comparison is displayed in Figure 2-10. Liquid film

thickness is determined from the scale placed on the test















1.75 1 1

v meter #1, upper section
1.70 meter #1, lower section


1.65


) 1.60


1.55
4

o I1.50


1.45 -


1.40 I I i
20 30 40 50 60 70 80

Temperature (C)

Figure 2-9. Temperature calibration for film thickness
sensor filled with full vapor.

section. The length measurement from the pictures is accurate

to 0.1 mm. Since the liquid/vapor interface was wavy in

almost all cases considered, the instantaneous photograph of

the flow structure had to be synchronized with an

instantaneous capacitance and temperature measurement in order

to obtain a reliable comparison. The degree of waviness of













































Figure 2-10. Close-up view of stratified two-phase flow
using CCD camera (flow direction is from left
to right).


the liquid/vapor interface depends on flow conditions. The

two-phase mixture bulk temperature ranged from 50-70 OC. The

use of the digital data acquisition system, which will be

described shortly, greatly facilitated the synchronization

process. The liquid film thickness data measured using the

CCD camera have been compared against that measured by sensor

















25



20 meter #1













0 II
n S








0 5 10 15 20 25
Measured Film Thickness 6 (mm)
CCD Camera



Figure 2-11. Comparison of liquid film thickness measured
with CCD camera and capacitance sensor.


#1 and the results are illustrated in Figure 2-11. It can be

seen that over the entire range of film thickness and

temperature considered, the comparison is good. The average

error based on the data shown in Figure 2-11 is within 2% of

the full scale.



2.4 Data Acquisition System

A digital data acquisition system has been assembled for








25

this investigation, which is used for recording measurements

of pressure, temperature, flow rate, and capacitance for this

investigation. A schematic diagram of the data acquisition

system is displayed in Figure 2-12. The data acquisition

system is comprised of two Acces 16-channel multiplexer cards

(AIM-16) interfaced with one Acces 12-bit 8-channel analog-to-

digital converter (AD12-8), mounted in an I/O slot of a

Northgate PC/AT computer. The AD12-8 has a maximum conversion

speed of 40 kHz and input voltage range of 10 Volts. Each

AIM-16 card is interfaced with one channel of the AD12-8

board. Thus there are 32 different channels available when

using this system. Channel 0 of the AIM-16 has been used to

determine the cold junction temperature using a resistance

temperature device (RTD). The temperature scale factor for

the output of the RTD is 24.4 mV/OC. Each channel of the AIM-

16 has a preamplifier with gains ranging from 0.5 to 1000 and

may be programmed through the computer. The AD12-8 board and

AIM-16 cards were calibrated according to manufacturer's

specifications. Each analog signal from the respective

instrument is connected to one of the 32 channels of the AIM-

16 cards. Appropriate gains were set up for different

channels to achieve maximum resolution. Since two-phase flows

are inherently unstable, all measurements were time-averaged

to obtain repeatable values. Using this system, an average of

500 sampling points were collected over a time period of 30

seconds in order to obtain repeatable measurements. Quick

















RESISTORS


_- 12-MHZ 286-AT
ACCESS AD12-8
A/D INPUT CARD
THERMOCOUPLES TO TEST
FACILITY, (SURFACE PROBES,
INFLOW PROBES, AND HEAT LOSS
PROBES)
VALIDYNE MAGNETIC RELUCTANCE
- DIFF. PRESS, TRANSDUCER


Figure 2-12. A schematic diagram of data acquisition system.


BASIC software routines have been developed for all data

acquisition operations.

A summary of the design operating conditions of the

facility are as follows: mass flux, G=80-350 kg/m2-s; quality,

X=0-0.35; system pressure, P=1.0-2.3 bars; and test section

heat flux, q,=0-40 kW/m2. The operating constraints of the








27

facility are primarily due to the maximum flow rate of the

Micropump, the strength of the pyrex glass, and the

temperature limitation of the E-poxy used in the test section.

All the experiments performed in this investigation have been

confined to the system design conditions.














CHAPTER 3

HEAT TRANSFER AND PRESSURE DROP
IN SATURATED FLOW BOILING



3.1 Introduction

In this section of the investigation, measurements of

heat transfer coefficient with and without boiling are

described which have been obtained for a saturated two-phase

mixture flowing through the test section. The purpose of

these measurements is to elucidate the importance of the

microconvection contribution to the total heat transfer in

flow boiling. The pressure drop and liquid film thickness for

stratified two-phase flow without boiling have also been

measured over a wide range of mass flux, G, and quality, X.

The parametric trends of the heat transfer coefficient and

pressure drop for horizontal two-phase flow are displayed and

compared against those observed for single phase flows.

The total two-phase heat transfer coefficient with and

without boiling is defined by


=qw (3-1)
T,- Tb


where T, is the mean wall temperature, Tb is the two-phase bulk

temperature, and q, is the wall heat flux. Since this work








29

only deals with saturated flow boiling, Tb is equivalent to

the saturation temperature, Tt. In order to sort out the

contribution of heat transfer between macro- and

microconvection during flow boiling, the following

experimental procedure was closely adhered to. The Micropump

and preheaters were adjusted to obtain a fixed G, X, and T,

at the inlet of the test section. The nichrome strip was

heated up gradually until boiling was initiated. During this

process the heat flux and temperature were recorded. The

pressure drop, AP, across the test section and liquid film

thickness 6 at the inlet and outlet of the test section were

also recorded. As heat flux q, was further increased, vapor

bubble generation at the heating surface was generated and

sustained with increasing q,. Further measurements of h2, were

made until q, was increased up to 40 kW/m2. The range of flow

conditions over which measurements were made was G=125-280

kg/m2-s and X=0.04-0.30. As had been expected, it is observed

that for a fixed G, X, and T,, the measured two-phase heat

transfer coefficient, h2, without boiling is independent of

the heat flux, q,. Hence, it was assumed that the non-boiling

two-phase heat transfer is equivalent to that of the

macroconvection heat transfer coefficient in flow boiling and

is herein denoted by hc. As has been discussed in Chapter 1,

Rohsenow (1952) first suggested that the rate of heat transfer

associated with forced convection boiling is due to two

additive mechanisms, that due to bulk turbulence and that due








30

to ebullition. Using Rohsenow's superposition hypothesis, the

heat transfer coefficient attributed to microconvection during

saturated flow boiling may be calculated from

hmC=h2 -hmac. (3-2)





3.2 Experimental Results and Discussions

Prior to discussing the details of the experimental

results, it is necessary to define several parameters. For

two-phase stratified horizontal flow, the mean velocity of the

liquid film may be calculated from

uG(1-X)D (33)


and mean vapor velocity by

GXD
Uv p D (3-4)
p,(D-8)

where u is mean velocity, 6 is liquid film thickness, p is

density and D is the inside dimension of the horizontal square

test section for which only the lower surface is covered with

a liquid film; subscripts t and v denote the liquid and vapor

phases, respectively. The Reynolds number for liquid and

vapor phases are defined by


Re,= PuDa (3-5)
III


and











Rev= pvUPhv (3-6)
Pv


respectively, where Re is Reynolds, Dh is the hydraulic

diameter, and j is the dynamic viscosity.

Microconvective heat transfer coefficients were obtained

for the nucleate flow boiling regime using the methodology

described above. The flow boiling heat transfer data were


organized by plotting h./h2, against hcincs as shown in




Figure 3-1. It is very significant that all the experimental

data have been collapsed into a single curve. Here ATmc,.

denotes the sustaining incipience wall superheat. These data

conclusively demonstrate that microconvection is important in

almost all phases of saturated flow boiling heat transfer and

its contribution becomes dominant at high heat fluxes. This

conclusion distinctly contests most forced convection boiling

heat transfer correlations reported in the literature which

predict that macroconvection is always dominant. The curve

presented in Figure 3-1 may also be viewed as a "flow boiling

curve". As is well known, the conventional heat flux vs wall

superheat plot used for pool boiling cannot collapse the flow

boiling data due to the large variation of macroconvection

heat transfer.

Further consideration was given to the macroconvection














1.0 i 1 i 1

ATn =-8.7 OC (average)
G=1-5-266 kg/m2-s
0.8 X=0.04-0.30
6=1.6-8.8 mm

V7 V





t 0.4 -



0.2



0.0 I I I
1 2 3 4 5 6 7 8

q/(hmacATine,s)

Figure 3-1. Microconvection heat transfer for saturated
forced convection nucleate boiling.


heat transfer component, h,. Figure 3-2 shows h. as a

function of liquid Reynolds number, Ret, and vapor Reynolds

number, Rev. If h., is approximated as a linear function of

Reynolds number over the limited range of data, the standard

deviation based on Re, is 0.165 and that based on Rev is 0.093,

and thus it is seen that h. is better correlated with Rev than

Re,. This result is fundamentally different from that in















Vapor Reynolds Number Re (xlO )

2 4 6 8 10


Liquid Reynolds Number Re, (xlO-)


Figure 3-2.


Macroconvective heat transfer coefficient in
saturated forced convection boiling.


single-phase forced convection and may possibly be due to the

enhanced turbulence caused by strong interfacial waves.

Considering that most flow boiling correlations for h. are

simply modified single-phase heat transfer correlations

applied to the liquid, there remains considerable room for

improved modelling of both hn. as well as h. in flow boiling

heat transfer correlations.


.tI
0



U
4a
a


4.4
U
u
a
h


u
a,
1
42


* hmac vs Re1
V ho vs Re




*ao vv





VV S
vle.Q57


I~iw0 r


1.0 C-


0.5 F-













-4
Vapor Reynolds Number Re (xlO )

2 4 6 8 10


50



40 -


30 -



20 -


AP vs Re,
V AP vs Rev




*0 7q v
v *

Sve


1I




7 1 I
VP j~Vl

I Ii! -4


Liquid Reynolds Number Re, (10- )


Figure 3-3. Pressure drop in horizontal two-phase flow.


Figure 3-3 displays the pressure drop, AP, as a function

of Re, and Rev. In contrast to the case of h,, AP is found to

be better correlated with Re, rather than Rev. This result

suggests that the principle of analogous energy and momentum

transport in incompressible single phase flow may not be

appropriate for stratified two-phase flow in a boiling system

where strong interfacial waves are observed. For two-phase

flow with strong interfacial waves, Andritsos and Hanratty








35

(1987) have provided extensive experimental evidence that the

mean vapor velocity is a controlling parameter on the

interfacial shear stress. Recently, Maciejewski and Moffat

(1992) measured the velocity and temperature distributions in

the near wall region for flow over a flat plate and found that

the strong turbulence intensity in the free stream could

substantially alter the near-wall temperature profile while

the velocity profile maintains a relatively uniform shape.

Therefore, the dissimilarity between heat transfer and

pressure drop observed in this research may be due to the

strong turbulence intensity at the interface which may

influence the temperature profile in a manner significantly

different from that of the velocity profile.

In stratified two-phase flow, the thickness of the liquid

film along the lower surface can be converted to void fraction

a by


=1-- (3-8)
D

Using Zuber and Findlay's (1965) correlation, all the

experimental data obtained in this research were collapsed

into a straight line as shown in Figure 3-4. u, is the

superficial velocity of vapor phase defined by

GX
u-,G (3-9)
Pv

and um is the two-phase mixture velocity which is defined by











(3-10)


U -GX G(X-1)
Pv Pe


It is noted that the observed void fraction for saturated flow

boiling system in this work is always larger than 0.7. Since

the conversion of liquid film thickness to void fraction in

this range has greatly reduced the relative error of the

results, the collapse of the data does not necessarily imply


2 3

Drift Flux um (m/s)


Figure 3-4


Zuber and Findlay's (1965) correlation for
void fraction in horizontal stratified two-
phase flow.








37

that Zuber and Findlay's (1965) correlation captured the

correct physics governing the void fraction distribution in

two-phase flow.



3.3 Conclusions

Measurements of two-phase heat transfer coefficients with

and without boiling have demonstrated that the microconvection

component of heat transfer in saturated flow boiling is

significant in almost all phases of boiling and its

contribution to the total heat transfer becomes dominant as

heat flux increases. The macroconvection heat transfer in

saturated flow boiling with strong interfacial waves is not

well correlated by simply using an analogy between momentum

and heat transport. Therefore, the development of a

significantly improved heat transfer correlation for flow

boiling, which has not been attempted in this study, will

require improved modelling of both the micro- and

macroconvection processes.














CHAPTER 4

INCIPIENCE AND HYSTERESIS



4.1 Introduction and Literature Survey

The development of modern electronics packaging requires

the ability to remove large amounts of heat from

microelectronic chips. The use of nucleate boiling heat

transfer of many dielectric liquids has been investigated for

this purpose due to their high heat transfer rates. Since

dielectric liquids usually have a high wettability on most

solid surfaces, the large overshoot of incipience superheat,

i.e. boiling hysteresis, has prevented their wide

applications. The experimental observation of boiling

incipience and hysteresis for highly wetting liquids is

clearly displayed in Figure 4-1 which was reconstructed from

the experimental data of Kim and Bergles (1988). Figure 4-1

is a typical heat flux vs wall superheat plot. The initiation

incipience point A is the point at which vapor bubbles just

begin appearing on the heating surface with increasing heat

flux. The sustaining incipience point B in Figure 4-1 is the

point just before vapor bubbles disappear from the heating

surface with decreasing heat flux. The overshoot of wall

superheat at the initiation incipience point is generally








39

referred as boiling hysteresis. Incipience and hysteresis in

pool boiling have been the focus of numerous experimental


106






105


103
0.


1


1 10
T -Tat (K)


100


Figure 4-1.


Nucleate pool boiling hysteresis constructed
from the data of Kim and Burgles (1988).


investigations. It has been observed that the sustaining

incipience point for specified liquids and surface conditions

is predictable and is basically independent of the boiling

history (Yin and Abdelmessih, 1976). In contrast, the

initiation incipience point for highly wetting liquids depends

on initial system conditions as well as the history of various


V Increasing Heat Flux
V Decreasing Heat Flux
Pool Boiling with
R113 on Plain Copper
T a=46.4 C


A Initiation point
B Sustaining point


v


104








40

heating, cool-down, and surface drying procedures (You, et

al., 1990; Marto and Lepere, 1982; Bergles and Chyu, 1982).

Recently, the effects of flow on boiling incipience have been

examined by various investigators. In subcooled flow boiling

with highly wetting liquids, such as R113 and F72, mass

velocity showed little effect on boiling incipience (Hino and

Ueda, 1985; Marsh and Mudawwar, 1989). In flow boiling with

water under both subcooled and saturated conditions, Sudo et

al. (1986) and Marsh and Mudawwar (1989) observed a strong

influence of the liquid velocity on boiling incipience. Flow

boiling incipience measurements of R113 at saturated

conditions are not available in the literature.

Numerous models and correlations have been proposed for

the prediction of boiling incipience, which have recently been

reviewed by Brauer and Mayinger (1992). The majority of

models were categorized as being either thermal or mechanical.

Thermal models are those which consider the bubble embryo to

sit at the mouth of a cavity and protrude into a superheated

thermal liquid layer. Once thermal equilibrium at the embryo

interface is exceeded by the superheated liquid, bubble growth

is initiated. Experimental data have verified that for poorly

wetting liquids, such as water, the initiation and sustaining

incipience points almost coincide and these models are useful

for predicting the incipience superheat. Models proposed by

Hsu (1962), Han and Griffith (1965), Bergles and Rohsenow

(1964), Sato and Matsumura (1964), and Davis and Anderson








41

(1966) belong to the thermal category. Mechanical models

consider a stable bubble nucleus resides inside a cavity.

When both the wall and liquid are subcooled, the interface is

required to be concave toward the cavity due to condensation

of vapor. For highly wetting liquids, this kind of concave

interface is possible only with reentrant shaped cavities

because contact angles are usually less than 90 degrees. The

radius of the bubble nucleus for highly wetting liquids is

substantially smaller than that of the cavity mouth.

Therefore, a higher wall superheat is required to initiate

boiling than is required to sustain it. According to

mechanical models, the radius of the interface inside the

cavity is determined by the liquid-solid contact angle, shape

of the cavity, and the degree of subcooling at the wall.

Therefore, the incipience wall superheat is dependent on the

surface conditions, liquid wettability, and the pressure-

temperature history prior to boiling. Incipience models

proposed by Mizukami et al. (1990) and Tong et al. (1990)

belong to the mechanical category.

The difficulties associated with predicting the boiling

incipience points are summarized as follows: 1) lack of

detailed information concerning cavity shapes and sizes on

commercial surfaces, 2) difficulties in determining the

dynamic and static contact angles on a microscale, 3) lack of

knowledge concerning the shape and size of the superheated

thermal layer, especially when two-phase flows are involved,








42

and 4) lack of knowledge of the embryo expansion and recession

process inside a cavity.

In this work, the initiation and sustaining incipience

wall superheats were measured for saturated flow boiling with

refrigerant R113 using the facility described in chapter 2.

The motivation for these measurements is to investigate the

dependence of the initiation and sustaining incipience points

on two-phase flow conditions and the surface heating and

cooling history. The saturated two-phase mixture flowing

through the transparent test section was varied over a range

of G and X at constant pressure. The measurements of the

initial incipience points were obtained by slowly increasing

q, until fully developed nucleate boiling was achieved. Then

measurements of sustaining incipience points were obtained by

gradually reducing the heat flux to return to the two-phase

forced convection regime. Measurements of both initiation and

sustaining points were also performed for variable heating and

cooling cycles at a fixed G and X. A theoretical analysis

which takes into account the hysteresis of the liquid-solid

contact angle and cavity geometry is presented which explains

the incipience process and hysteresis of boiling with highly

wetting liquids.



4.2 Experimental Results

A typical saturated flow boiling plot of q, vs AT, for a

fixed G, X, and 6 is shown in Figure 4-2. A description of








43





40
V Increasing ql from T nm =20'
35 boiling initiated at A.
V Decreasing q,,
boiling sustained at B.
30 A Increasing q% from T r-=63*C
boiling initiated at A'.
0 Increasing qw from Twm=67C
25 boiling reversible. B, A"coincide. 7
T =580C, T =200C
sat room
20- -


15 7A

,o 2f A'
10 I
^r^\ BPA"
5 ABA



0 5 10 15
T --Tsat (C)


Figure 4-2. A typical saturated flow boiling plot of q,
versus AT, for G=180 kg/m2-s, X=0.156, and
6=4.8 mm.


Figure 4-2 is as follows. With a quasi-steady saturated two-

phase mixture flowing through the test section, the heat flux

is increased from zero until the initiation incipience

superheat is reached, which is denoted by point A. Here the

minimum temperature prior to boiling is room temperature,

approximately 20 OC. The heat flux is increased further until

fully developed nucleate boiling is achieved. The heat flux








44

is then decreased until the sustaining superheat is reached,

which is denoted by point B. The heat flux is further reduced

until Tw,-=63 OC. The heat flux is then increased until the

new initiation superheat is reached which is denoted by A'.

Thus is seen that Twn influences the initiation superheat.

The cycle is repeated for Twn=67 OC and A" denotes the

initiation superheat, which coincides with the sustaining

superheat. Thus, hysteresis is fully suppressed provided

Twmn>67 OC. Many experiments were conducted over a variety of

flow conditions to examine the influence of forced convection

on both the initiation and sustaining incipience points.

Figure 4-3 has been prepared for this purpose, where it is

seen that both the initiation and sustaining incipience wall

superheats remain essentially constant (although slightly

scattered) over a wide variety of flow conditions. h., has

been used in Figure 4-3 as a comprehensive parameter to

characterize the bulk turbulence. These results are

consistent with those obtained for subcooled flow boiling with

highly wetting liquids (Hino and Ueda, 1985; Marsh and

Mudawwar, 1989).



4.3 Theoretical Analysis for Boiling Incipience

4.3.1 Boiling Initiation

Effect of Non-condensible Gases. Since non-condensible

gases (usually air) are always trapped in cavities during the

process of a liquid filling over a surface, it is necessary to








45

understand their influence on the initiation of boiling. The

air trapping process has been detailed by Lorenz (1972) and

recently by Tong et al. (1990). Mizukami (1977) investigated

the effect of non-condensible gases on the stability criterion

of the embryo and found that the existence of gas stabilizes

the vapor bubble nucleus but accelerates its nucleation when

the liquid is superheated. However, quantitative


20 -


Vy V


10 -


5
0.


I I I


0


0.5 1.0 1.5

Macroconvection hm (kW/me-C)


2.0


Figure 4-3.


Measured saturated flow boiling incipience
wall superheat.


v Measured Initiation Superheat
V Measured Sustaining Superheat
-- Predicted Sustaining Superheat
Based on Tsat=59 OC, rl=0.66x1O- m





I V I








46

considerations regarding the effect of the gas mass on the

initiation superheat have not been reported in the literature.

It is worthwhile to proceed with such calculations in order to

further understand the nucleating process of a vapor bubble

containing a non-condensible gas.

Consideration is given to an embryo, consisting of a non-

condensible gas and saturated vapor, trapped in a conical

cavity as shown in Figure 4-4. The gas is taken to be air and

the mass is specified. The embryo is initially at static and

thermal equilibrium with its surroundings. Therefore, the

following relations are satisfied,


P+PP,-P 2a (4-1)


T= TV= Tsa (Pv) (4-2)


PgV=mgRTv (4-3)

where r is radius of the liquid/vapor interface, a is surface

tension, R is an engineering gas constant, and the subscripts

v, , and g respectively denote the vapor, liquid, and gas.

The liquid pressure, P,, is typically taken to be the system

pressure. Since air is assumed to be the only non-condensible

gas in the cavity the ideal gas law is obeyed. V is the

volume of an embryo, and for a conical cavity is given by

(Lorenz, 1972)


V=_ r3 (2- (2+cos2 (-)) sin (6-) + cos3 ( ) (4-4)
3 tan(w)









































Figure 4-4. An idealized sketch of a vapor embryo in a
conical cavity.

where 0 is the contact angle and i is the half cone angle of

the cavity. First, consideration is given to the effect of

non-condensible gas on incipience, and for this purpose a

conical cavity with a specified geometry, ri=0.69 pm and *=5,

is considered. The contact hysteresis, which is important for

boiling incipience, has been discussed in detail in the

literature (Johnson and Dettre, 1969; Schwartz and Tejada,








48

1972; Tong, et al., 1990). According to these investigations,

as the liquid/vapor interface gradually moves toward the vapor

phase, a maximum contact angle is reached and is referred to

as the static advancing contact angle, O. Similarly, as the

interface gradually moves toward the liquid phase, a minimum

contact angle, referred to as the static receding contact

angle, 0,, is reached. Assuming the embryo expansion follows

a quasi-equilibrium process, the contact angle 0 lies between

0, and 0, which are determined by liquid wettability and

surface conditions. Based on the data supplied by Tong et al.

(1990), 0,r~2 for R113, while 0, is usually less than 900.

For calculation purposes, here it is assumed that the initial

static contact angle is equal to the static advancing contact

angle and 08,80. Tong et al. (1990) have suggested that when

the cavity is heated, during the first expansion stage the

embryo interface adjusts such that the initial static contact

angle recedes until the static receding contact angle is

reached. Then during the second expansion stage the

liquid/vapor interface moves toward the cavity mouth, with

constant contact angle, 0,. During the first expansion stage,

the contact angle, 0, may be calculated from,


cos(0-9) =-d (4-5)
r

where rd, which remains constant, is the cavity radius at the

initial triple interface. rd may be calculated by specifying

the initial 0, T,, Pt, and mg and solving equations (4-1)

















100

90

80 -

70
v.max
60

50 -- mg =1.0xl0-' kg
m =0.25x10-" k
40 --- m =O.lxlO- kg

30 For Tsat=58 C
8 =2, +=5
sr
20 -

10 I I I I
0.001 0.01 0.1 1 10 100
V (x10-1" mO)


Figure 4-5. Variation of vapor temperature with vapor
embryo volume during expansion inside a
conical cavity.


through (4-5) simultaneously. Once rd is obtained, a Tv vs. V

plot can be constructed by solving equations (4-1) through (4-

5) assuming a quasi-equilibrium expansion process. Figure 4-5

shows three different T, vs. V curves, each for a different mg.

It is seen that as V increases there exists a maximum vapor

temperature, T,nx, which will satisfy equations (4-1) through

(4-5). Based on quasi-equilibrium considerations, it is








50

assumed that T,-T,. When T, exceeds T,,., equation (4-1) will

be violated and vapor bubble growth will be initiated. Thus

the initiation incipience superheat is evaluated from

Ajinc, i=rTvmax- r () (4-6)

It is clear that the initiation incipience superheat increases

with decreasing mg. For the conditions in Figure 4-5, the

assumed value of 0, does not significantly influence the

incipience superheat. Since it is expected that non-

condensible gas will be purged from the cavity during the

vapor bubble departure process, the amount of gas inside the

cavity should decrease as ebullition continues. As the

heating surface is degassed, ATm,i should increase. This trend

has been observed with the current facility. After the test

section is filled with liquid R113, it is necessary to sustain

fully developed nucleate boiling for approximately two hours

in order to obtain a repeatable ATmc,i. This observation is

consistent with those of Griffith and Wallis (1960). They

suggested that fully developed nucleate boiling with water

must be sustained for 1.5 hours to degas conical cavities, and

two hours is required for reentry type cavities. Thus it is

concluded that non-condensible gas trapped in cavities during

liquid flooding of the boiling surface exerts a strong

influence on ATc,i only during the initial stage of boiling.

Provided fully developed nucleate boiling is sustained for a

sufficient period of time, the embryo in active cavities










should consist of pure vapor.


Cavities with Pure Vapor. Mizukami (1990) concluded

that conical cavities with pure vapor can not survive

subcooled conditions and thus are not useful for initiating

boiling. However, even if conical cavities are initially

filled with liquid and can not initiate boiling, they can

become active nucleation sites if vapor is deposited in the

cavity from a neighboring nucleation site as has been

suggested by Calka and Judd (1985).

Mizukami (1990) also pointed out that the most favorable

cavities for surviving subcooled conditions are reentry type

ones. Thus it is likely that boiling is first initiated from

reentry cavities. Now consideration is given to reentry type

cavities, one of which is depicted in Figure 4-6. As is the

case for conical cavities, provided nucleate boiling is

sustained for a sufficient period, a vapor embryo will recede

in the cavity when the surface is cooled. Unlike conical

cavities, a reentry one will allow the liquid vapor interface

of highly wetting liquids to be concave toward the cavity

reservoir. Therefore, the vapor embryo can survive when Pt>Pv.

Following the analysis of Griffith and Wallis (1960) and

Mizukami (1975), for 0,,<4, the maximum curvature that the

liquid vapor interface can achieve is 1/r2, where r2 is the

mouth radius of the cavity reservoir. Thus the initiation

superheat can be obtained from,









































Figure 4-6. An idealized sketch of a vapor embryo in a
reentry cavity.


A Tn i Tsat (p+ ) Tsat (p) (4-7)


For 0,<90, the minimum curvature of the interface is


nsa The minimum curvature determines a minimum vapor
r2

temperature below which the vapor embryo cannot be sustained,












2osinea4
Tv, min=Tsat (Pt 2sinsa
Tz


50

45

40

35

30

25

20

15

10

5

0
0.;


36 0.38 0.40


Figure 4-7.


Variation of minimum vapor temperature and
initiation superheat with cavity reservoir
mouth radius.


T,.n and ATc,i calculated from equations (4-7) and (4-8) over

a range of r2 for R113 at a pressure of 1.45 bars (the no-flow

system pressure of the current facility) are displayed in

Figure 4-7. As shown in Figure 4-3, the measured initiation

incipience superheat for the current facility is approximately


0.32 0.34 0.

r, (gm)


-inci v.min
For T =58 C, a=80
sat sa












I
- I
--


0.42





(4-8)


--


-
r
_I
30








54

14.70 C which corresponds to cavities with r2< 0.31 pm.

Assuming that reentry cavities initiate boiling and r2-0.31

pm, they would be able to sustain their vapor embryos provided

the wall temperature is maintained above 15 OC. The ambient

temperature of the laboratory where the boiling facility is

housed is maintained at 200 C. The fact that the vapor embryo

can be sustained may explain why the initiation incipience

superheats in Figure 4-3 are relatively uniform. The above

analysis suggests that if the heating surface is cooled down

well below 15 OC or if the system is pressurized well above

1.45 bars, AT0c,1 should increase. Experiments have been

performed when the system was pressurized to 2.26 bar with

Twm,=20 OC for about half an hour. ATii measured immediately

following depressurization was found to increase to 19.1 OC,

which is consistent with the above prediction. Further

measurements revealed that AT.i was dropping down toward its

original value as boiling continued but it required several

days for AT,,,i to recover.



4.3.2 Sustaining Incipience Superheat

A vapor bubble departing a cavity which has been active

will likely leave vapor behind at the cavity mouth in the form

of an embryo. Provided sufficient superheat is available to

the embryo, it will readily expand and vapor bubble growth

will again result. This process will continue until the

superheat available to the embryo is insufficient to promote








55

bubble growth. Following bubble departure, the liquid moves

toward the cavity, and the liquid/vapor interface will acquire

an advancing dynamic contact angle, 0,, which is usually

larger than 0,. According to Cole's analysis (1974), for

either conical or reentry cavities with half cone angle, 0,

less than 0,, the maximum curvature the interface can attain

is the reciprocal of the cavity mouth radius, r,. Thus, prior

to bubble growth the interface protrudes the cavity mouth into

the superheated liquid thermal layer. Both the cavity mouth

radius and the temperature profile in the liquid thermal layer

control whether or not bubble growth will result. Assuming

bulk turbulence alone controls the liquid thermal layer, and

the temperature profile is linear in the vicinity of the vapor

embryo, the liquid temperature at the top of the embryo is

given by


T =Tw- (Tw-Ta) (4-9)


where kt is the liquid thermal conductivity. Starting from

the Clayperon equation and the perfect gas approximation for

vapor, Bergles and Rohsenow (1966) derived an equation for the

embryo vapor temperature,


Tv- Tsa TTasRvl n(1+ 20 ) (4-10)
hfg rlp,


They further proposed the criterion for boiling incipience

that T, at the top of the embryo must exceed T,. Based on this








56

criterion, the sustaining superheat ATmc,. (=Tw-Tt) can be

obtained as a function of rl, h., and Tt numerically from

equations (4-9) and (4-10). For constant r,, ATc,, has been

calculated over a range of h, as shown in Figure 4-3. It is

seen that the effect of convection on incipience superheat is

negligible which is in agreement with experimental

observations.


4.3.3 Hysteresis of Boiling Incipience

Hysteresis is the difference between the initial and

sustaining superheats. According to the preceding incipience

analysis, the initiation point depends on the minimum heating

surface temperature, while the sustaining point does not. In

fact, an explanation for incipience hysteresis is provided by

equation (4-8); as T,, decreases, potentially active

incipience nucleation sites are deactivated due to the

collapse of the vapor embryo. Assuming that the heating

surface contains reentry cavities with a large size range,

equation (4-8) predicts that incipience hysteresis will

increase with decreasing Tv,a. Such behavior is exactly what

has been observed with the present R113 flow boiling facility.

It is seen from Figure 4-2 that when T,,n is greater than Tt

and less than the sustaining point, the hysteresis declines

when compared to heating from subcooled conditions.

Furthermore, once fully developed boiling has been established

and T,,mm is maintained above the sustaining point, the boiling








57

process is completely reversible, which indicates that

hysteresis has been suppressed. However, one puzzling result

is that when T,,
may be because the size of the reentry cavities along the

heating surface are fairly uniform, r2~0.31 Am. If this were

the case, only when Tv,<15 OC would ATii be influenced. Such

a test has yet to be conducted. Electron microscope

photographs did not add any further insight.



4.4 Conclusions

Based on the experimental observations and theoretical

analysis presented herein regarding the incipience and

hysteresis associated with saturated forced convection boiling

of R113, two comments are in order:

1) Non-condensible gases trapped in cavities tend to reduce

the initiation superheat, but they will be purged from

cavities after boiling is sustained for a sufficient period.

For highly wetting liquids, the liquid flow could exert a

slight effect on the sustaining incipience superheat but not

on the initiation superheat. Gas-free conical cavities are

not useful for initiating boiling but can become activated

from adjacent nucleation sites. The initiation incipience

likely occurs in reentry cavities.

2) The incipience hysteresis is related to the minimum

heating surface temperature prior to boiling. Once fully

developed boiling has been established, the boiling process is








58

reversible if the heating surface temperature is maintained

above the sustaining point.

3) For the flow conditions investigated herein using R113,

both the initiation and sustaining superheats were not

noticeably influenced by bulk turbulence.














CHAPTER 5

NUCLEATION SITE DENSITY



5.1 Literature Survey

Due to its governing influence on heat transfer, the

nucleation site density has been the focus of numerous

investigations in pool boiling (Clark et al., 1959; Griffith

and Wallis, 1960; Kurihara and Myers, 1960; Gaertner and

Westwater, 1960; Hsu, 1962; Gaertner, 1963; Gaertner, 1965;

Nishikawa et al., 1967; Singh et al., 1976). The general

consensus from these investigations is that the formation of

nucleation sites is highly dependent on surface roughness,

geometry of microscopic scratches and pits on the heating

surface, the wettability of the fluid, the amount of foreign

contaminants on the surface, as well as the material from

which the surface was fabricated. Because of the large number

of variables which are difficult to control, none of these

investigators were successful in developing a general

correlation for nucleation site density. Griffith and Wallis

(1960) suggested that for a given surface the critical cavity

radius, re, is the only length scale pertinent to incipience

provided the wall superheat is uniform. Although they

realized the wall superheat is nonuniform in pool boiling,








60

they used r, to correlate the nucleation site density as

follows,


n=( )m (5-1)
A rc

where n/A is the nucleation site density, Ci and m are

empirically determined constants, and when p, >> p,


S 2Tsat (5-2)
PvhfgAsat


where T, is the saturation temperature, a is the surface

tension, hg is the latent heat of evaporation, and AT=Tw-T.t

is the wall superheat.

Nucleation site density data of Griffith and Wallis

(1960) for pool boiling of water on a copper surface is shown

in Figure 5-1 as a function of ATt. For n/A < 4 cm-2 the

nucleation site density increases smoothly with increasing

AT.. However for n/A > 4 cm2 no correlation exists between

n/A and ATt. Moore and Mesler (1961) used a fast response

thermocouple to demonstrate the heating surface temperature

directly beneath a nucleation site in pool boiling experiences

rapid fluctuations. Recently, Kenning (1992) used

thermochromic liquid crystals to measure the spatial variation

of wall superheat with pool boiling of water on a 0.13 mm

thick stainless steel heater. It was demonstrated that the

wall superheat was very nonuniform and |TIata/T-Ta varied
















12



1 0 V Water on 3/0 Finished Copper O
o0 ] Water on 3/0 Finshed Copper
N- from Griffith and Vallis [1]





6


>-4
U) 4

0
4-)-

CE
2

Z
0


0 5 10 15 20 25 30

Wall Superheat ATsat (OC)


Figure 5-1. Pool boiling nucleation site density data from
Griffith and Wallis (1960).


from 0.25 to 1.5 over a distance of a few mm, where ()

implies a spatial average and (') denotes a spatial variation

from the mean. Using a conduction analysis, it was also shown

that in the presence of ebullition, the spatial non-uniformity

of wall superheat on the surface of "thick plates" may also be

significant depending on geometry and thermal conductivity of








62

the plate. Kenning's (1992) result suggests that a Micro-

length scale which is related to the spatial temperature field

may also be important in characterizing incipience of vapor

bubbles. In addition, according to the suggestion of

Eddington, et al. (1978) and the experimental findings of

Calka and Judd (1985), at low n/A where nucleation sites are

sparsely distributed, neighboring nucleation sites do not

thermally interact. However at large n/A where nucleation

sites are closely packed, thermal interference among

neighboring sites exists. Since it has been demonstrated that

ATU is highly nonuniform in the presence of nucleation sites,

a question arises as to whether at large n/A the average wall

superheat is sufficient for characterizing the local wall

superheat experienced by individual nucleation sites. The

data displayed in Figure 5-1 suggest that is not.

Few similar studies on nucleation site density in flow

boiling have been reported. In one such investigation

Eddington and Kenning (1978) measured the nucleation site

density with subcooled flow boiling of water for a narrow

range of flow conditions. It was suggested that n/A is

related to rc. The objectives of the present work are

twofold: 1) to study the influence of flow and thermal

conditions on the nucleation site density in saturated

convection boiling, and 2) to determine whether or not the

nucleation site density in flow boiling is solely a function

of the critical radius, re, as has been suggested for pool








63

boiling. Measurements of nucleation site density have been

obtained for flow boiling of refrigerant R113 in a 25 x 25 mm

inner square transparent test section. These measurements

have been obtained for an isolated bubble regime with

stratified flow. In general, the vapor/liquid interface was

wavy and periodic slugs of liquid were observed passing

through the test section. No flow regime transitions were

observed. The liquid phase Reynolds number based on the mean

liquid velocity, which is defined by equation (3-5), ranged

from 12,000 to 27,000 and that for the vapor phase as defined

by (3-6) ranged from 24,000 to 80,000. The flow conditions

are characterized by the mass flux, G, liquid film thickness,

6, vapor quality, X, mean liquid velocity, ut, and mean vapor

velocity, uy. Thermal conditions are characterized by the

heat flux, q,, the wall superheat, ATt and saturation

condition (saturation temperature, Tt, or saturation pressure,

Pt). The controllable inputs of the flow boiling facility are

G, X, P, (or Tt), and q,. Of the flow parameters considered,

only two are independent since the mean liquid and vapor

velocities were calculated from equations (3-3) and (3-4),

respectively, based on measured liquid film thickness at a

given G, X, and T,. For the convenience of discussion,

equations (3-3) and (3-4) are rewritten here,

G(1-X)D (3)
u= (5-3)P8


and










GXD
u D (5-4)
pv(D-8)

Therefore, when investigating the influence of one of the flow

parameters on nucleation site density at constant saturation

conditions, only one other flow parameter can be held

constant, which introduced complexities in interpreting the

data. The range of the flow and thermal parameters covered in

these measurements, which to a large extent were limited by

the ability to visualize nucleation sites, are as follows:

G=125-290 kg/m2-s, u,=1.6-4.7 m/s, u,=0.35-0.68 m/s, 6=3.5-9.5

mm, q,=14.0-23.0 kW/m2, ATt=13.0-18.0 C, and T,=55.0-75 OC.


5.2 Optical Facility and Measuring Technique

The nucleation site density was measured optically using

a digital imaging facility shown in Figure 5-2. The facility

consists of a Videk Megaplus CCD camera with a 1320 x 1035

pixel resolution. The CCD camera is equipped with a Vivitar

50 mm macro lens with high magnification and low optical

distortion and a Videk power supply. The output of the CCD

camera is connected to an Epix 4 megabyte framegrabber which

is mounted in an I/O slot of a 386 Zenith computer. The

framegrabber allows for either high resolution (1320 x 1035)

or low resolution (640 x 480) imaging. The images are

displayed on a Sony analog monitor with 1000 lines per inch

resolution, and may also be printed out on an HP Laser Jet III

laser printer or may be saved on floppy disk for future




















ZSONY HIGH RESOLUTION
B/W MONITOR


VIDEK MEGAPLUS
CCD CAMERA
WITH MACRO
LENS I


VIDEK CAMERA
POWER SUPPLY


Z-386/20 COMPUTER

EPIX FRAME GRABBER


Figure 5-2.


A diagram of optical facility for measurement
of nucleation site density.


analysis. Computer software written in Microsoft C has been

developed for the image acquisition and processing. Due to

the strong vapor-liquid entrainment and waviness at the

interface of the two-phase mixture, it is not possible to

obtain a clear view of nucleation sites from the direction

normal to the heating surface. Therefore, the camera was

focused on the boiling surface through the side wall. A 500

Watt light illuminates the heating surface at an appropriate








66

angle from the opposite side wall. An opaque plastic sheet is

placed between the light and the object to diffuse the

incident light. Exposure time and lens aperture are properly

adjusted to obtain a clear image of the nucleation sites. A

typical picture of the nucleation sites is displayed in Figure

5-3 which was taken from the Sony monitor image. In order to

reduce the non-uniformities caused by the heater edge effect,

only nucleation sites in the middle 2/3 of the strip were

counted for measuring purposes. Thus the effective

measurement area was 1.4 cm wide by 2.2 cm. The nucleation

site density measured from an ensemble average of fifty images

was compared against that based on an average of ten images;

identical results were obtained. Therefore, all nucleation

site density measurements reported herein are based on an

ensemble average of ten images. Presumably intermittent sites

are accounted for. As has been demonstrated in Figure 3-2,

hysteresis can be avoided once the fully developed boiling has

been established. The measurements of nucleation site density

here were only made for the fully developed boiling regime

with increasing heat flux and increasing vapor velocity.


5.3 Experimental Results

Nucleation site density measurements were obtained for a

constant heat flux, q,=19.3 kW/m2, and saturation temperature,

T,=58 OC, over a range of flow conditions in which either the

mass flux, G, liquid film thickness, 6, liquid velocity, u,,









































Figure 5-3. A typical photograph of nucleation sites on a
boiling surface (flow direction is from left
to right).

or vapor velocity, U, was maintained constant. The

nucleation site density, n/A, is shown as a function of wall

superheat, ATt, in Figure 5-4. It is seen from Figure 5-4

that the n/A data can not be correlated with AT,. In light

of Figure 5-1 and equations (5-1) and (5-2), the behavior of

n/A with AT, is considered to be anomalous. In order to

demonstrate that the observed behavior is not simply due to














15


qw= 19.3kW/m2, T t=58C
o G=215 kg/m -s
V 6=6.3 mm
10 Au u=0.48 m/s
10 1
Su =3.6 m/s E

V



5 V



0)
(D



A
o 0

0 1 1 1 1
8 10 12 14 16 18 20
Wall Superheat ATsat (OC)



Figure 5-4. Nucleation site density as a function of wall
superheat for constant heat flux and
saturation temperature.

the experimental error, pool boiling nucleation site density

data were obtained using the current facility by filling the

test section with liquid and heating the nichrome strip while

the circulation pump was off. Therefore, the only net flow

was induced by the natural convection currents. The n/A data

for pool boiling are also displayed as functions of wall












Heat Flux qw (kW/m2)

8 10 12 14 16


18 20

Wall


Figure 5-5.


18 20


22 24 26 28 30 32

Superheat ATsat (OC)


Pool boiling nucleation site density as
functions of wall superheat and heat flux.


superheat, AT,, as well as heat flux, qw, in Figure 5.5. It

is seen that n/A increases smoothly with increasing AT, and

q, in a similar fashion to the data shown in Figure 5-1. As

seen from Figure 5-4, parameters other than AT, alone appear

to exert an influence on n/A in flow boiling.

In order to examine the influence of the flow parameters


10 H


Pool Boiling Data, Tsat=57.5 C
V n/A versus ATat
E n/A versus q


I I I I I I I











70

on n/A, measurements of nucleation site density were made

while the saturation temperature was maintained constant. The

nucleation site density is first plotted against vapor

velocity at a fixed heat flux, q,, and fixed liquid film

thickness, 6, as shown in Figure 5-6. It is seen that n/A

decreases markedly with increasing vapor velocity, u,. At a


10 I


5k


Vapor Velocity uv (m/s)


Figure 5-6.


Nucleation site density as a function of vapor
velocity for constant heat flux and liquid
film thickness.


O 6=6.3 mm, q= 19.3 kW/m'
V 6=7.7 mm, qw=14.5 kW/mp
Tsa=58 C





0





V
V7
V

V
V
V


I I I I








71

fixed heat flux and liquid film thickness, the data appear to

fall on a single curve. As the heat flux is increased, the

curve shifts toward higher nucleation site density.

Therefore, when investigating the influence of the flow

parameters on n/A the heat flux will be maintained constant.

Upon examination of equations (5-3) and (5-4) it is possible

that the trend shown in Figure 5-6 is due to either increasing

G or ut instead of increasing u. To sort out whether G, u,,

or u, has a controlling influence on n/A figures 5-7 and 5-8

have been prepared. In Figure 5-7, n/A is shown to increase

with increasing G when u, and q, are fixed, and decreases with

increasing G when S and q, are fixed, and thus it appears that

parameters other than G are controlling n/A. In Figure 5-8

n/A is shown to decrease with increasing u, for a fixed G and

q,. For the case of a fixed 6 and q, n/A also decreases with

increasing ut but the shape of the curve is significantly

different. When comparing Figures 5-6 and 5-8, it appears

that n/A is better behaved when displayed as a function of u,.

Further evidence of this supposition is provided in Figure 5-9

where n/A is displayed as a function of u, for q,=19.3 kW/m2,

T.=58 oC, G=215 kg/m2-s, and u,=0.58 and 0.48 m/s. It is seen

that all of the data approximately fall on a single curve,

thus demonstrating the governing influence of the mean vapor

velocity on nucleation site density. The liquid film

thickness is also shown as a function of u,. Thus, the effect

of liquid film thickness on n/A might also be included in








72







15 ,

o u1=0.48 m/s
Sqw= 19.3 kW/m"
o V 6=6.3 mm
q= 19.3 kW/m"
< 0 T =58 C
10 sat



vO






0~ ~~~~ v-------- I --------
z O
0
0 I I
100 150 200 250 300

Mass Flux G (kg/me-s)


Figure 5-7. Nucleation site density as a function of mass
flux.


Figure 5-9. Therefore, it is necessary to investigate the

influence of 6 on n/A.

In pool boiling, Nishikawa et al. (1967) demonstrated

that the nucleation site density increases with declining

liquid film thickness. Mesler (1976) postulated that the same

behavior should follow for flow boiling and used it to explain
















15

V G=215 kg/mn-s
q= 19.3 kW/m2
S0 5= 6.3 mm
qw= 19.3 kW/m2

S 10- T=58 aC


SV


5 l V
0O




z

0 I






Figure 5-8. Nucleation site density as a function of
liquid velocity.


the measured increase in flow boiling heat transfer

coefficient with declining liquid film thickness for

stratified or annular flow. To the best of the author's

knowledge direct evidence supporting or refuting Mesler's

claim has yet to be presented. To sort out the direct

influence of liquid film thickness on the nucleation site








74







15 I I 14
%=19.3 kW/m", T =58 C
Sn/A 6
Q G=215 kg/m2-s 12
] * u=0.58 m/s
A A u,=0.48 m/s
10
S10 -

8
QC



S5 v
o A 4 a
,4 '-
O 7

2
o


0 II Ii 0
0 1 2 3 4 5

Vapor Velocity uv (m/s)



Figure 5-9. Nucleation site density and liquid film
thickness as functions of vapor velocity.


density, Figure 5-9 suggests that it is necessary to maintain

a fixed u,, q,, and T.. Figure 5-10 shows n/A as a function

of liquid film thickness at uv=3.6 m/s, q,=20.7 kW/m2, and

T,=58 OC. It is seen that n/A indeed increases with declining

film thickness, which tends to support Mesler's claim

regarding n/A as a function of film thickness, provided that

















15




o u =3.6 m/s
Sqw=20.7 kW/m"
0 Tsa=58 oC








.4
(r
0
54->








2 4 6 8

Liquid Film Thickness 6 (mm)



Figure 5-10. Nucleation site density as a function of
liquid film thickness.


u,, q,, and Tt are fixed. However, over the range of film

thickness investigated (3-6 mm) the increase in n/A is only

marginal. Because n/A was obtained using a visualization

technique it was not possible to obtain data for 6<3 mm. As

6-0 the behavior of n/A is uncertain. To determine whether

uy or 6 has stronger influence on n/A, Figure 5-9 is re-








76

examined where it is seen that n/A decreases with declining 6,

which is primarily caused by increasing u,. Therefore, it

appears that u, has a governing influence on the nucleation

site density. An explanation and significance of this finding

will be discussed later.

When testing the influence of thermal conditions on n/A,

it is necessary to control Uy. In Figure 5-11, n/A is shown

as a function of q, for three different values of u, at Te=57

C. It is seen that n/A increases smoothly with increasing

heat flux at a fixed u,. As u, increases, the curves shift

toward decreasing n/A. This trend is consistent with that

observed in Figure 5-6. The n/A data in Figure 5-11 are shown

as a function of AT, in Figure 5-12. The data display

an anomalous behavior similar to that in Figure 5-4.

To examine the dependence of n/A on the critical radius

it was decided that a constant heat flux and vapor velocity

would be maintained, and r, would be controlled by raising the

system pressure. Doubling the system pressure has the effect

of essentially doubling the vapor density and increasing Tt

by only several percent. The nucleation site density was

measured for fixed U, and q, over a range of system pressure

from 1.4 to 2.3 bars which gave a 20 OC increase of saturation

temperature. Figures 5-13 and 5-14 have been prepared from

these measurements. It is seen from Figure 5-13 that n/A

increases with increasing T, while AT, is dropping during this

process. Nevertheless, Figure 5-14 shows that this increase

















15 1

Ts =57 C
sat
D u =2.15 m/s
SA u =2.67 m/s
o V u =3.64 m/s

S10


--4



Ui 5
l.u





i id I i

12 14 16 18 20 22 24

Heat Flux qr (kW/mr)



Figure 5-11. Nucleation site density as a function of heat
flux.


of n/A can be attributed to an increase of 1/r,. Since the

only physically sound explanation for the increase of n/A with

increasing T. is due to a decrease in r,, these data suggest

that r, is an important parameter in characterizing flow

boiling nucleation site density.















15 I I -
Tsat=57 C
0 u =2.15 m/s
SAu =2.67 m/s
SV u =3.64 m/s

10 -










0 D A



10 12 14 16 18 20 22 24
/ \














Figure 5-12. Nucleation site density as a function of wall
superheat.

5.4 Discussion of Results

Although the data shown in Figure 5-14, as well as

theoretical considerations, suggest that r, is an important

parameter for flow boiling nucleation site density, it is by
/


o a

0 I-I--I-II------ -- -------I-
10 12 14 16 18 20 22 24









Waitself insufficient to correlate n/A. In Figure 5-15, allC)

Figure 5-12. Nucleation site density measurements a function of wall
superheat.

5.4 Discussion of Results

Although the data shown in Figure 5-14, as well as

theoretical considerations, suggest that r, is an important

parameter for flow boiling nucleation site density, it is by

itself insufficient to correlate n/A. In Figure 5-15, all

nucleation site density measurements in this work are








79




Wall Superheat ATsat (C)

10 15 20


55 60 65 70 75


Figure 5-13.


Saturation Temperature Tsat (C)



Nucleation site density as
saturation temperature and wall


functions of
superheat.


displayed as a function of r,. All data seem to be collapsed,

but a correlation in the form of equation (5-1) would not be

useful because the slope is too steep. The pool boiling

nucleation site density data also show a similar behavior for

large n/A.

In an attempt to understand this behavior, consideration


n/A vs ATsat n/A vs T at
A uv=3.0 m/s V u =3.0 m/s
A uv=3.5 m/s v u =3.5 m/s
q =17.3 kW/mr


15 -


10 -


0
5(


)


I I I I I















100

Q
V u =3.0 m/s
< V u =3.5 m/s
Sq =17.3 kW/m2




( 10 V


oV


V V7
V



1 --
1 3 5 10
-8
Critical Radius, 1/re (xl 0m-')


Figure 5-14. Nucleation site density as a function of
critical radius.


is first given to the pool boiling analysis of Hsu (1962) in

which it was demonstrated that the nonuniform liquid

temperature field seen by a vapor embryo attempting to grow is

important when considering incipience behavior. If linear

temperature profile is assumed for the liquid layer, a minimum

cavity radius required for incipience may be expressed solely
















100



^All Data for Constant
Saturation Temperature
STsa1=58 C




'5 10 -




o 1
Ssat









a rd

z A
A,

1 3 5 10
Critical Radius, 1/r. (x10lO6m')



Figure 5-15. Nucleation site density as a function of
critical radius.


as a function of wall superheat AT, or heat flux q, provided

the fluid properties are maintained constant. Figure 5-4

demonstrates that AT, or q, alone is insufficient to correlate

the flow boiling n/A data. Bergles and Rohsenow (1964)

applied a similar analysis for flow boiling and the assumption

of a linear temperature profile in the liquid layer also leads








82

to a definition of critical cavity radius which is solely

dependent on AT, or q, for constant fluid properties. One

shortcoming of these analyses is the assumption of a linear

temperature profile in the liquid thermal layer. Certainly,

the strength of heat flux and the intensity of bulk turbulence

will have a strong influence on the shape of the thermal

layer. Nevertheless, the experimental data presented here and

these theoretical analyses suggest that in addition to the

critical radius based on equation (5-2), a length scale

related to the shape of the thermal layer may also be

important in characterizing n/A.

In addition, it is emphasized that r, has been calculated

using equation (5-2) by taking the wall superheat to be the

average value, AT.. As mentioned earlier, Kenning (1992) used

liquid crystal thermography to show that IAat|/Tga could



be as large as 1.5 for pool boiling. In this study the liquid

crystal test section was used in conjunction with a Panasonic

video recorder to record the flow boiling wall temperature

field at conditions of U,=3.0 m/s, qw=18.1 kW/m2, and Tw=58.3

OC. It was found that temperature field was very nonuniform

in both spatial and temporal scales. Therefore, it appears

the average wall superheat is insufficient for characterizing

the local wall superheat experienced by individual sites.

The experimental data of n/A presented here have revealed




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs