DEVELOPMENT OF FUNDAMENTAL EQUATIONS

OF STATE FOR THERMODYNAMIC PROPERTIES

OF HFC REFRIGERANTS

Dissertation
Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy

by

I Made Astina

SCHOOL OF SCIENCE FOR OPEN AND ENVIRONMENTAL SYSTEMS

GRADUATE SCHOOL OF SCIENCE AND TECHNOLOGY

KEIO UNIVERSITY
2003
 THE SUMMARY OF Ph.D. DISSERTATION
Major Student Identification SURNAME, First name
Open and Environ- Number 80162907 I Made Astina
mental Systems
Title
Development of Fundamental Equations of State for Thermodynamic
Properties of HFC Refrigerants
Abstract
Fundamental equations of state for hydrofluorocarbon (HFC) refrigerants were developed due
to several percents of uncertainty of the specific heat in the gaseous phase near saturation and
completely different behavior of the third virial coefficient in the existing equations of state. The
fundamental equations of state expressed in Helmholtz free energy for the purpose of representing
the right thermodynamic consistencies were developed with taking consideration of a physical
background in addition to available experimental information. Development of a computer program
including evolution process and genetic rule as genetic algorithm was conducted for finding high
efficiency of optimization process on the thermodynamic modeling.
The dissertation is divided into eight chapters. In Chapter 1, the present situation on refrigerants
and the thermodynamic properties is introduced, and the latest state of the art on the existing
equations of state is described in advance. It is then closed with the research background and the
research purpose and scope. Chapter 2 explains the thermodynamic relations among physical
properties and the relations with the Helmholtz free energy function on the basis of the
thermodynamics. In the next chapter, literature survey and selection of experimental data on HFC
refrigerants in single phase and at saturation are described. The paradigm and consideration on the
theoretical background that should be considered into the thermodynamic modeling are described in
Chapter 4. Preparation of the derived data for compensating shortage of experimental data and the
optimization procedure and genetic rule for thermodynamic modeling are also explained in this
chapter. Chapter 5 describes thermodynamic formulation of ancillary equations for saturation
properties with simultaneous optimization of four HFC refrigerants to express into a common
functional form. The ancillary equations have functions of temperatures for representing vapor
pressure and saturated-liquid and saturated-vapor densities, respectively. In Chapter 6, development
of the fundamental equations of state, thermodynamic formulation, and their assessment for PVT
and caloric properties in single phase and at saturation are described. Their reliabilities in fluid
phase including region where no experimental data exist are assessed. The discussions, which cover
consistency of the fundamental equations of state with intermolecular potential models,
characteristics of the ideal curves for wide-range thermodynamic surface at extreme pressures and
temperatures, and relation of the optimization procedure to thermodynamic consistency, are
conducted in Chapter 7. Finally, the result of the study is summarized in Chapter 8.
 博 士 論 文 要 旨

開放環境科学 専攻 学籍番号 80162907 イー マデ アスティナ

主論文題名
HFC 系冷媒の熱力学性質に関する基礎状態方程式の開発
(内容の要旨)
ハイドロフルオロカーボン(HFC)系冷媒については既にヘルムホルツ状態式が開発されてい
る。しかしながら飽和付近の気相域での比熱の不確かさが数パーセント以上であること、第 3 ビ
リアル係数が状態式毎に全く異なる挙動を示すことなどの問題が指摘されている。本研究では、
実測値情報に加えて物理的な背景を考慮に入れながら、理論的にも正しい熱力学性質を表現する
ことを目標として、HFC 系冷媒のヘルムホルツ自由エネルギーを表現する新たな熱力学状態方
程式を開発している。このヘルムホルツ状態式開発においては、最適化の効率に配慮して、進化
プロセスおよび遺伝アルゴリズムを用いたプログラムの開発も行っている。論文の構成は以下の
8 章である。
第 1 章では、冷媒および熱力学性質に関する現在の状況を紹介し、更に HFC 系冷媒の熱力学
状態式に関しても最新および最先端の紹介をしている。そして、研究背景および研究目的を述べ
ている。
第 2 章では、熱物性に関する熱力学関係式を紹介し、熱力学状態式の基礎事項からヘルムホル
ツ状態式の開発に関する本研究の取り組み方を述べている。
第 3 章では、HFC 系冷媒の実測値情報の評価および選定方法について説明し、気相、液相、
および飽和域において選択した実測値を紹介している。
第 4 章では、実測値に加えて考慮すべき理論背景に関する本研究のパラダイムについて述べて
いる。分子間ポテンシャルモデルの挙動および気相の比熱の挙動に関して、ビルアル状態式を用
いた扱いについて紹介する。飽和蒸気圧力、飽和液体密度および飽和蒸気密度に関しても実測値
情報の不足を補うための相関式を開発し状態式開発の入力データとして導入している。さらに遺
伝子操作などの熱力学モデリング最適化手法についても述べている。
第 5 章では、具体的な飽和状態の相関式開発について述べている。4 種類の HFC 系冷媒に関
して、共通な関数形で表現できるように最適化しながら相関式を開発している。すなわち、飽和
蒸気圧力相関式、飽和液体密度相関式、飽和蒸気密度相関式である。相関式の信頼性に関する評
価についても述べる。
第 6 章では、ヘルムホルツ状態式の開発について説明し、導かれたヘルムホルツ状態式を紹介
し、その信頼性の評価に関して述べている。
第 7 章では、本研究で開発したものを含む全ての熱力学モデリングに関して、第 2 および第 3
ビリアル係数、音速および比熱の挙動について比較した結果を紹介する。さらに実測値が存在し
ない高温高圧域を含む広い範囲の理想特性曲線(Ideal Curves)の挙動についても述べている。
第 8 章は本論文の結果である。
 Content

Page

Content ……………………………………………………………………………. i

Preface…………...…………………………………………...…………………….. vii

List of Symbols…………………………………………………………………….. ix

List of Figures……………………………………………………………………… x

List of Tables……………………………………………………………………….. xv

Chapter 1

Introduction……………….……………………………………………………… 1

1.1 Refrigerants and Thermodynamic Properties…………………………..…….. 1

1.2 Existing Equations of State for HFC Refrigerants……………………………. 4

1.2.1 Consistency Relation with Intermolecular Potential Energy…………… 5

1.2.2 Consistency of Ideal Curves………………………..…………..………. 8

1.3 Research Background………………………………………………….……... 10

1.4 Research Purpose and Scope.…………….……………………………..…… 11

Chapter 2

State of the Art on Thermodynamic Model for Pure Fluids………………….… 15

2.1 Thermodynamic Consequences and Thermodynamic Properties………….…. 15

2.2 Thermodynamic Equations of State……………………………..……….…… 18

2.2.1 Analytical Equation of State……………………………………….…… 18

2.2.2 Semi-Empirical Equation of State..……………………………….……. 21

i
 2.2.3 Empirical Equation of State..………………………..…………….……. 22

2.3 Thermodynamic Property Relations to Helmholtz Free Energy Function…..... 23

2.4 Thermodynamic Property Calculation……………....………………….…….. 25

2.4.1 Practical Calculation Strategy for Thermodynamic Properties ….……... 25

2.4.2 Iteration Strategy and Ancillary Equations………………….…….……. 30

Chapter 3

Existing Data and Data Selection for Thermodynamic Modeling……………… 32

3.1 Literature Survey……………………………………………………………… 32

3.2 Experimental-Data Selection in Single Phase….……………………………... 34

3.2.1 Data for R-32…………………………………….……….…………….. 35

3.2.2 Data for R-125……………………………………….…….…….……... 36

3.2.3 Data for R-134a………………………………………….….…….……. 37

3.2.4 Data for R-152a…..…………………………………………..….……... 38

3.3 Experimental-Data Selection at Saturation……...……………….…..…….…. 39

Chapter 4

Data Preparation and Modeling Procedure…………………………………….. 42

4.1 Rational Paradigm and Consideration……………………………………….. 42

4.2 Principal Consideration of Thermodynamic Modeling..………...…………… 44

4.2.1 Mathematical Aspect……………………………………………………. 45

4.2.2 Thermodynamic Aspect..……………………………………………….. 45

4.3 Data Preparation on Development of Fundamental Equations of State……… 48

4.3.1 Preparation of PVT Property at Saturation…….………….……………. 48

ii
 4.3.2 Preparation of Gaseous Caloric Properties………….……………..…… 49

4.3.3 Preparation of Virial Coefficients……………………………….…..….. 50

4.4 Weighting Factor and Regression……………………………….……….……. 50

4.5 Thermodynamic Modeling…………….……………………….………….…. 52

4.6 Optimization Method….………………….…………………….…….………. 55

4.6.1 Optimization Procedure………..………….………………….………… 56

4.6.2 Genetic Rules for Operators………………………………….………… 60

4.6.3 Convergence Ability……………………………………………………. 63

4.6.4 Statistical Inference View……………………………………………….. 67

Chapter 5

Ancillary Equations for Saturation Properties of HFC Refrigerants……..…… 69

5.1 Development and Consideration……………………………………………… 69

5.2 Thermodynamic Formulation of Ancillary Equations………………………… 70

5.2.1 Formulation of Vapor-Pressure Equation..………………………….…… 71

5.2.2 Formulation of Saturated-Liquid Density Equation…………….………. 71

5.2.3 Formulation of Saturated-Vapor Density Equation..…………….……… 72

5.3 Ancillary Equations for Saturation Properties……………………….………... 72

5.3.1 Vapor-Pressure Equations..……………………………….…….……….. 73

5.3.2 Saturated-Liquid Density Equations..…………………….…….……….. 74

5.3.3 Saturated-Vapor Density Equations..……………………………………. 75

5.4 Assessment of Ancillary Equations..…………………..……………………… 75

5.4.1 Assessment of Vapor-Pressure Equations..……………………………… 76

5.4.2 Assessment of Saturated-Liquid Density Equations..……...…..…….….. 79

iii
 5.4.3 Assessment of Saturated-Vapor Density Equations..………..…..….…… 83

Chapter 6

Fundamental Equations of State for HFC Refrigerants……………………… 87

6.1 Important Parameters..……………………………………………………… 87

6.2 Thermodynamic Formulation of Fundamental Equations of State....………. 88

6.2.1 Formulation of Ideal-Gas Part..………………………….….………… 88

6.2.2 Formulation of Residual Part..………………………………………… 89

6.3 Fundamental Equations of State..…………………………………………… 93

6.3.1 Ideal-Gas Part..……………………..…………………….…………… 93

6.3.2 Residual Part..…………………..………………………..……………. 94

6.4 Assessment of Ideal-Gas Property...………………….……….…………….. 97

6.5 Assessment of PVT Property………………………………….…………….. 101

6.5.1 PVT Property in Single Phase..…………………….…….…………… 101

6.5.2 PVT Property at Saturation……………..…………………….………. 107

6.6 Assessment of Caloric Properties……………………………….…………... 115

6.6.1 Specific-Heat Assessment…...…………………………………..….… 115

6.6.2 Speed-of-Sound Assessment…...………………………………..….… 118

6.6.3 Saturated-Liquid Specific Heat Assessment…………………….….… 122

6.7 Assessment of Critical Parameters..…………………....……….…..……… 123

6.8 Validity Range and Accuracy………………………………………..……… 124

Chapter 7

Discussion……………………………………………………………..…………. 127

iv
7.1 Gaseous Specific Heat and Intermolecular Potential Behaviors..….…..……. 127

7.2 Extrapolation Behavior of Fundamental Equations of State……….……..…. 133

7.3 Ideal Curves of Fundamental Equations of State………………..….……….. 136

7.4 Reliability of Fundamental Equations of State………………….…..………. 139

7.5 Correlation of Virial Coefficients and Ideal Curves………………………… 144

7.6 Thermodynamic Consistency and Optimization Procedure……..….………. 147

Chapter 8

Conclusion………………………………………………………………...……… 149

List of Author’s Papers……….…………………………………………………. 151

References…………………………………………………………………………. 152

Appendix A Thermodynamic Property Relations………...…………………. 163

A.1 Derivation of Relations for Thermodynamic Properties..……………………. 163

A.1.1 Pressure and Compressibility……………………………………….…. 163

A.1.2 Specific Entropy…………………………………………………..…… 164

A.1.3 Specific Internal Energy……………………………………….………. 165

A.1.4 Isochoric Specific Heat……………………………………..…………. 165

A.1.5 Specific Enthalpy………………………………………………………. 165

A.1.6 Isobaric Specific Heat………………………………………………….. 165

A.1.7 Speed of Sound………………………………………..…………….…. 166

A.1.8 Virial Coefficients……………………………………..………………. 166

A.1.9 Vapor-Pressure Relations at Saturation………………….…………….. 167

v
 A.1.10 Specific Heat Along Saturation….………………….………………... 169

A.1.11 Joule-Thomson Coefficient………………………….……………….. 170

A.1.12 Fugacity………………………………………….…………………… 170

A.1.13 Isothermal Compressibility………………………….……………….. 171

A.1.14 Volume Expansivity………………………………….………………. 171

A.2 Basic Relations of Specific Heat…………………………………..…………. 172

Appendix B Fundamental Equation of State and Its Derivatives……..……. 175

B.1 Dimensionless Helmholtz Free Energy for Fundamental Equation of State… 175

B.2 Mathematical Operation for Ideal-Gas Part………………………..…..…….. 175

B.2.1 Integration Framework………………………………………………… 176

B.2.2 Derivatives of Ideal-Gas Part……………………….…………………. 177

B.2.3 Planck-Einstein Term……………………………….…………………. 178

B.3 Mathematical Operation for Residual Part………………..……………..…… 178

B.3.1 Separation of Residual Part…………………………………………… 179

B.3.2 Derivatives of Residual Part…………………………………………… 179

B.4 Convention of Reference State for Fundamental Equation of State…….…… 180

Appendix C Statistical Analysis Results……………………….……………... 182

Appendix D Thermodynamic Properties for Computer Program

Verification………………………………………………………. 199

Appendix E Pressure-Enthalpy Diagram…………………….………..……... 207

vi
 Preface

This dissertation was made to complete the partial fulfillment of the requirements

for the degree of doctor of philosophy in doctoral course of Graduate School of Science

and Technology at Keio University. In accordance with the title, this study firstly

explored the existing equations of state for HFC refrigerants and investigated the

thermodynamic consistencies among the equations of state. The dissertation describes

not only on consistency, assessments, and formulation, but also on development

procedure and implementation strategy of the fundamental equations of state.

Thermodynamic properties play important role on thermal design and analysis.

They are essentially used for thermodynamic calculation both for scientific and

engineering interests. Environmental issues have strongly forced technologies to operate

without any negative impact. Increasing efficiency on thermal systems may contribute

to reduce the negative impacts. Reliable thermodynamic properties are needed to

develop high-efficient thermal systems. The dissertation is hoped to enrich and

contribute on thermodynamic properties especially for thermodynamic modeling.

The dissertation was completed with any help whether directly or indirectly to this

research. In this chance we would like to thank,

Prof. Haruki Sato, who works hard to advice us to complete the Ph.D. research. It

agrees with my initial knowledge condition on thermophysical properties that is

very less so that all goes well due to his patient to bring the research in progress.

Prof. Akira Nagashima, Prof. Takahiko Tanahashi, and Prof. Masahiko Uematsu,

for their constructive critics that are very useful in revising this dissertation.

vii
 Monbunkagakusho (Japanese Ministry of Education, Science, and Culture) for the

scholarship as long as this study.

Japanese Embassy at Jakarta for all valuable helps to prepare travel documents for

pursuing the study and recommendation of the scholarship.

Bandung Institute of Technology and Indonesian Ministry of National Education

that have been permitted us to release the job for pursuing the doctoral course.

Prof. Osamu Kawaguchi, for support us to pursue the doctoral course.

Dr. Eric W. Lemmon, Physical and Chemical Properties Division, National Institute

of Standards and Technology (U.S.A.), for valuable helps as long as this research.

Inpex Foundation that always motivates his alumni to continue the doctoral course

in Japan.

Our colleagues at Department of Mechanical Engineering, Bandung Institute of

Technology, Indonesia, who always motivate us to complete the Ph.D. degree.

All students who work researches at the Energy Utilization and Fluid Properties

Laboratory in Keio University for their helps as long as this study.

Foreign students who are studying at Keio University from various countries for the

valuable helps.

Our family who always be patient to wait the graduating from this course for

returning home again and their praying for the success.

I Made Astina

January 2004

viii
 List of Symbols

Notations

a : specific Helmholtz free energy p : pressure

α : reduced Helmholtz free energy PVT : pressure-volume-temperature
B : second virial coefficient property
β : volume expansivity R : gas constant
C : third virial coefficient ρ : mass density
D : fourth virial coefficient s : specific entropy
c : specific heat T : temperature
δ : reduced density τ : inverse reduced temperature
exp : exponent with a natural base u : specific internal energy
f : fugacity µ : Joule-Thomson coefficient
G : gas phase V : volume
g : specific Gibbs free energy v : specific volume
h : specific enthalpy w : speed of sound
L : liquid phase ϕ : fugacity coefficient
M : molar mass x : mass ratio of vapor in
N : data number two-phase
n : mole number Z : compressibility factor
κ : isothermal compressibility

Subscripts Superscripts
p : process at constant pressure ' : saturated-liquid state
v : process at constant volume " : saturated-vapor state
c : critical point r : residual part
s : saturation state o : ideal-gas part
t : triple point out : rejected data
0 : reference state

ix
 List of Figures

Page
Fig. 1.1 Example for behaviors of second and third virial coefficients from
several existing equations of state…………………………………….. 5
Fig. 1.2 Example for specific heat consistency in the gaseous phase from
several existing equations of state………………..……………………. 6
Fig. 1.3 Example for ideal curves from several existing equations of state….… 9
Fig. 2.1 Phase diagram for ideal fluid and real substance……...………………. 17
Fig. 2.2 Schematic calculation strategy for (a) case of known density and
temperature and (b) case of known pressure and temperature.……….. 26
Fig. 2.3 Schematic calculation strategy for (a) case of known entropy and
pressure and (b) case of known entropy and temperature….…….…… 27
Fig. 2.4 Schematic calculation strategy for (a) case of known enthalpy and
pressure and (b) case of known enthalpy and temperature …………… 27
Fig. 2.5 Schematic calculation strategy for (a) case of known saturation
temperature and (b) case of known vapor pressure………….……..…. 28
Fig. 2.6 Strategy for searching solution in iteration process…………..……..… 30
Fig. 3.1 Points of experimental data of R-32 selected for the development….… 36
Fig. 3.2 Points of experimental data of R-125 selected for the development.….. 37
Fig. 3.3 Points of experimental data of R-134a selected for the development…. 38
Fig. 3.4 Points of experimental data R-152a selected for the development……. 39
Fig. 4.1 Common paradigm for determination of thermodynamic properties….. 43
Fig. 4.2 Paradigm for developing fundamental equation of state ……………… 43
Fig. 4.3 Basic principle of development for thermodynamic modeling……….. 52
Fig. 4.4 Difference of single-fluid and multi-fluid optimizations……………… 53
Fig. 4.5 Basic Strategy on development of optimization procedure…………… 57
Fig. 4.6 Main genetic optimization procedure……………………..…………… 58
Fig. 4.7 Flowchart for generation of competitor……………….………………. 59
Fig. 4.8 Recombination and mutation operators…………………..……………. 60

x
Fig. 4.9 Mapping fitness onto normal distribution density probability (NDDP). 62
Fig. 4.10 Convergence process of the main population in the optimization
process…………………………………………………………………. 65
Fig. 4.11 Example for generation competitor………..………………..…………. 66
Fig. 4.12 Statistical inference analyses from F-test (left) and t-test (right)
distributions…………………………….…………………………….... 68
Fig. 5.1 Deviation of measurements from the vapor-pressure equation for
R-32…………………………………………….……………………… 76
Fig. 5.2 Deviation of measurements from the vapor-pressure equation for
R-125.…………………………………………….…………………… 77
Fig. 5.3 Deviation of measurements from the vapor-pressure equation for
R-134a……………………………………………….………………… 78
Fig. 5.4 Deviation of measurements from the vapor-pressure equation for
R-152a………………………………………………..……………….. 79
Fig. 5.5 Deviation of measurements from the saturated-liquid density equation
for R-32…………………………………………………………..……. 80
Fig. 5.6 Deviation of measurements from the saturated-liquid density equation
for R-125………………………………………………………………. 80
Fig. 5.7 Deviation of measurements from the saturated-liquid density equation
for R-134a…………………………………………………………….. 81
Fig. 5.8 Deviation of measurements from the saturated-liquid density equation
for R-152a…………………………………………………………….. 82
Fig. 5.9 Deviation of measurements from the saturated-vapor density equation
for R-32..……………………………………………………………… 84
Fig. 5.10 Deviation of measurements from the saturated-vapor density equation
for R-125………………………………………………….………….. 84
Fig. 5.11 Deviation of measurements from the saturated-vapor density equation
for R-134a…………………………………………………………….. 85
Fig. 5.12 Deviation of measurements from the saturated-vapor density equation
for R-152a…………………………………………………………….. 85
Fig. 6.1 Deviation of ideal-gas isobaric specific heat from the new

xi
 fundamental equation of state for R-32………………………………. 98
Fig. 6.2 Deviation of ideal-gas isobaric specific heat from the new
fundamental equation of state for R-125……………………………… 99
Fig. 6.3 Deviation of ideal-gas isobaric specific heat from the new
fundamental equation of state for R-134a…………………..………… 100
Fig. 6.4 Deviation of ideal-gas isobaric specific heat from the new
fundamental equation of state for R-152a……………………..……… 100
Fig. 6.5 Deviation of density in gaseous phase from the new fundamental
equation of state for R-32……………………………………..……… 102
Fig. 6.6 Deviation of density in liquid phase from the new fundamental
equation of state for R-32……………………………………..……… 103
Fig. 6.7 Deviation of density in gaseous phase from the new fundamental
equation of state for R-125……………………………………………. 103
Fig. 6.8 Deviation of density in liquid phase from the new fundamental
equation of state for R-125………………………………………….…. 104
Fig. 6.9 Deviation of density in gaseous phase from the new fundamental
equation of state for R-134a………………………………………..… 105
Fig. 6.10 Deviation of density in liquid phase from the new fundamental
equation of state for R-134a.………………………………….……….. 106
Fig. 6.11 Deviation of density in gaseous phase from the new fundamental
equation of state for R-152a…………………………………………… 106
Fig. 6.12 Deviation of density in liquid phase from the new fundamental
equation of state for R-152a……………………………………..……. 107
Fig. 6.13 Deviations of vapor pressure and saturation temperature from the new
fundamental equation of state for R-32………………………………. 108
Fig. 6.14 Deviations of vapor pressure and saturation temperature from the new
fundamental equation of state for R-125………………………………. 109
Fig. 6.15 Deviations of vapor pressure and saturation temperature from the new
fundamental equation of state for R-134a…………………………….. 110
Fig. 6.16 Deviations of vapor pressure and saturation temperature from the new
fundamental equation of state for R-152a…………………….………. 111

xii
Fig. 6.17 Deviations of saturated-liquid and saturated-vapor densities from the
new fundamental equation of state for R-32………………….……….. 112
Fig. 6.18 Deviations of saturated-liquid and saturated-vapor densities from the
new fundamental equation of state for R-125……………….………… 112
Fig. 6.19 Deviations of saturated-liquid and saturated-vapor densities from the
new fundamental equation of state for R-134a…………….………….. 113
Fig. 6.20 Deviations of saturated-liquid and saturated-vapor densities from the
new fundamental equation of state for R-152a…………….………….. 114
Fig. 6.21 Deviations of isobaric and isochoric specific heats in the liquid phase
from the new fundamental equation of state for R-32………………… 116
Fig. 6.22 Deviations of isobaric and isochoric specific heats in the liquid phase
from the new fundamental equation of state for R-125……………….. 117
Fig. 6.23 Deviations of isobaric and isochoric specific heats in the liquid phase
from the new fundamental equation of state for R-134a……………… 117
Fig. 6.24 Deviations of isobaric and isochoric specific heats in the liquid phase
from the new fundamental equation of state for R-152a……….……… 118
Fig. 6.25 Deviation of speed of sound from the new fundamental equation of
state for R-32………………………….……………………..………… 119
Fig. 6.26 Deviation of speed of sound from the new fundamental equation of
state for R-125…………………………..……………………………. 120
Fig. 6.27 Deviation of speed of sound from the new fundamental equation of
state for R-134a……………….……………………………………… 120
Fig. 6.28 Deviation of speed of sound from the new fundamental equation of
state for R-152a………………..……………………………………… 121
Fig. 6.29 Deviations of saturated-liquid specific heats from the four
fundamental equations of state………………………………………… 122
Fig. 7.1 Comparison of second and third virial coefficients and gaseous
specific heat consistency for R-32…………………………………….. 128
Fig. 7.2 Comparison of second and third virial coefficients and gaseous
specific heat consistency for R-125…………………………………… 129
Fig. 7.3 Comparison of second and third virial coefficients and gaseous

xiii
 specific heat consistency for R-134a………………………………….. 130
Fig. 7.4 Comparison of second and third virial coefficients and gaseous
specific heat consistency for R-152a………………………………….. 131
Fig. 7.5 Caloric and PVT properties for R-32 from the new fundamental
equation of state………………………………………………………. 134
Fig. 7.6 Caloric and PVT properties for R-125 from the new fundamental
equation of state……………………………………………………….. 134
Fig. 7.7 Caloric and PVT properties for R-134a from the new fundamental
equation of state………………………………………………………. 135
Fig. 7.8 Caloric and PVT properties for R-152a from the new fundamental
equation of state……………………………………………………….. 135
Fig. 7.9 Comparison of ideal curves for R-32….…………….……………….. 137
Fig. 7.10 Comparison of ideal curves for R-125………………………………… 137
Fig. 7.11 Comparison of ideal curves for R-134a……………………………….. 138
Fig. 7.12 Comparison of ideal curves for R-152a…………………………….… 138
Fig. 7.13 Second and third virial coefficient behaviors at extreme
temperatures…………………………………………………………… 145
Fig. A.1 Maxwell criterion on p-v diagram……………………………………... 168
Fig. E.1 Pressure-enthalpy diagram for R-32 calculated from the new
fundamental equation of state…………………………………………. 208
Fig. E.2 Pressure-enthalpy diagram for R-125 calculated from the new
fundamental equation of state………………………………………… 209
Fig. E.3 Pressure-enthalpy diagram for R-134a calculated from the new
fundamental equation of state…………………………………………. 210
Fig. E.4 Pressure-enthalpy diagram for R-152a calculated from the new
fundamental equation of state…………………………………………. 211

xiv
 List of Tables

Page
Table 2.1 Analytical Equations of State…………………………….…………. 19
Table 2.2 Relations between Thermodynamic Properties and the Helmholtz
Free Energy function α (δ ,τ ) and Its Derivatives…..……………… 23
Table 3.1 Selected Experimental PVT Data at Saturation…….……………….. 40
Table 4.1 Interval Temperature Rule for Generating PVT Property at
Saturation……………………………………………………………. 49
Table 4.2 Virial Coefficient Relations for Derived-Data Preparation…………. 50
Table 4.3 Combinatory Estimations for Calculations and Times……………… 64
Table 5.1 Critical Parameters of HFC Refrigerants…………………………… 73
Table 5.2 Numerical Coefficients of Vapor-Pressure Equation (Eq. 5.1)……… 73
Table 5.3 Numerical Coefficients of Saturated-Liquid Density Equation
(Eq. 5.2)……………………………………………………………. 74
Table 5.4 Numerical Coefficients of Saturated-Vapor Density Equation
(Eq. 5.3)…………………………………………………………….. 75
Table 6.1 Main Properties of HFC Refrigerants………………………………. 88
Table 6.2 Objective Functions in Multi-Property Regression………………… 91
Table 6.3 Numerical Parameters of Each of the Ideal-Gas Parts……………… 94
Table 6.4 Numerical Coefficients and Constants for Each of the Residual Parts 95
Table 6.5 Calculated Critical Parameters for Four HFC Refrigerants…………. 124
Table 6.6 Validity Ranges of the Four Fundamental Equations of State……….. 124
Table 7.1 Summary of Assessment for Equations of State of HFC Refrigerants 142
Table B.1 Convention for Reference State of Thermodynamic Properties……... 181
Table C.1 Deviations of Measurements from the Vapor-Pressure Equations…... 182
Table C.2 Deviations of Measurements from the Saturated-Liquid Density
Equations…………………………………………………………… 184
Table C.3 Deviations of Measurements from the Saturated-Vapor Density
Equations……………………………………………………………. 186

xv
Table C.4 Deviations of Measurements from the Fundamental Equation of
State for R-32………………………………………………………. 187
Table C.5 Deviations of Measurements from the Fundamental Equation of
State for R-125………………………………………………………. 190
Table C.6 Deviations of Measurements from the Fundamental Equation of
State for R-134a……………………………………………………... 193
Table C.7 Deviations of Measurements from the Fundamental Equation of
State for R-152a…………………………………………………….. 196
Table D.1 Thermodynamic Properties of R-32 in Single Phase..………………. 199
Table D.2 Thermodynamic Properties of R-32 at Saturation…………………… 200
Table D.3 Thermodynamic Properties of R-125 in Single Phase.……………… 201
Table D.4 Thermodynamic Properties of R-125 at Saturation………………….. 202
Table D.5 Thermodynamic Properties of R-134a in Single Phase.…………….. 203
Table D.6 Thermodynamic Properties of R-134a at Saturation………………… 204
Table D.7 Thermodynamic Properties of R-152a in Single Phase……………… 205
Table D.8 Thermodynamic Properties of R-152a at Saturation………………… 206

xvi
 Chapter 1
Introduction

1.1 Refrigerants and Thermodynamic Properties

Historically, refrigerants cannot be separated from chlorofluorocarbons (CFCs)

that had been used widely since two years after their invention in 1928. The CFCs had

been used in diverse applications such as refrigerants for air-conditioning and

refrigeration systems, blowing agents for plastic foams, solvent for microelectronic

circuit or dry cleaning, aerosol propellants for personal hygiene products and pesticides,

etc. (Bhatti 1999). Theory linking the CFCs to the destruction of stratosphere ozone

layers shielding the earth from the harmful effects of the sun’s ultraviolet rays was

firstly reported in 1974 (Molina and Rowland 1974). Research results of the British

scientists in period of 1978 to 1984 have contributed on issues of ozone hole over

Antarctica. In October, the beginning of spring in the southern hemisphere, they

detected less ozone than that had been detected during the past 20 years. October ozone

levels were about 35 percent lower than the average for the 1960s. Their disturbing

findings were firstly reported in 1984 (Farman et al. 1985).

International society attention to the ozone depletion can be seen from adoption of

the Montreal Protocol on 16 September 1987 for phasing out substances that destroy the

ozone layer. This protocol was subsequently adjusted and amended by the Protocol

Committee at several meeting places, i.e., London in 1990, Copenhagen in 1992, Vienna

in 1995, Montreal in 1997, and Beijing in 1999. The effect of a substance on ozone

depletion is given with an indicator so called as ozone-depleting potential (ODP).

Refrigerants of CFCs and hydrochlorofluorocarbons (HCFCs) have certain amount of

1
the ODP due to their contained chlorine atoms.

Environmental issues on warming of the earth have challenged us to find the best

solution. Temperature of the earth is determined by a balance between incoming energy

from the sun and outgoing energy radiated from the earth’s surface and atmosphere.

Ultraviolet and visible radiations from the sun pass through the earth’s atmosphere and

strike the earth’s surface. Certain constituents of the atmosphere such as hydrocarbons

(HCs), CFCs, HCFCs, hydrofluorocarbons (HFCs), CO2, water vapor, and many other

chemical gases absorb the infrared radiation and trap it in the atmosphere. The effect of

these constituents is named as greenhouse effect. The greenhouse effect is indicated

with an index to quantify the relative capability of different chemicals to absorb infrared

radiation so called as global-warming potential (GWP). International society attention

on the global warming was indicated by adoption of the Kyoto Protocol that regulates

emission of greenhouse gases into atmosphere on 11 December 1997.

Effect of refrigerants on global warming are apparently insignificant in

comparison of the effect released by fueled power plants exhausting CO2 as long as their

operation. As consequence of CO2 effecting on the global warming and correlation of

CO2 emission relating to energy consumption, the effect of refrigerants such as HCs,

CFCs, HCFCs, and HFCs, can be critically divided into direct and indirect effects, i.e.,

effect of gases into the atmosphere and effect of inefficient energy consumption of the

air-conditioning and refrigeration systems. The right judgment is that both effects on

global-warming potential should be considered.

Since leakage of refrigerants from air-conditioning and refrigeration systems to

atmosphere is relatively small, the indirect effect is more crucially to global-warming

potential and improvement of the efficiency becomes very important than the direct

effect to global warming. One of efforts to improve efficiency of the thermal system is

2
to provide accurate thermodynamic calculations. By preparing reliable thermodynamic

properties for alternative refrigerants with zero-ozone depleting potential, the effort to

find high-efficient systems of friendly air-conditioning and refrigeration machines can

successfully achieve the target.

Among thermodynamic equations of state for fluids, an equation of state that can

accurately derive all the thermodynamic properties in whole fluid range is so called as

fundamental equation of state. The fundamental equation of state is represented in

Helmholtz free energy and it can be applied to calculate the thermodynamic properties

of both pure and mixture fluids as academic background for energy utilization or

environment science and engineering. Recently, most available standard thermodynamic

tables in those fields were calculated from the fundamental equations of state. They

represent most common fluids such as water, nitrogen, oxygen, CO2, ammonia, HCs,

CFCs, HCFCs, HFCs, etc.

Development of equations of state was started more than one hundred years ago.

In accordance with human interests to understand air, an ideal gas model was then found.

Understanding of air properties was accomplished by extending the ideal-gas model to

real-gas model. The models to represent the properties whether in the liquid phase or at

saturation also followed the real-gas model. On the other hand, fundamental equations

of state have been developed since several decades ago, after when high-speed

computers are available.

As response of the negative effects of the refrigerants, the efforts for either

development of equations of state or measurements have been conducted for

environmentally acceptable refrigerants. Recently, five HFC refrigerants of R-32

(difluoromethane), R-125 (pentafluoroethane), R-134a (1,1,1,2-tetrafluoroethane),

R-143a (1,1,1-trifluoroethane), and R-152a (1,1-difluoroethane) were considered as

3
alternative refrigerants. The Japanese Association of Refrigeration in cooperation with

the Japan Association of Japan Flon Gas responded to the refrigerant issue by compiling

the research results on thermophysical properties with their publication on

thermophysical properties for R-134a and R-123 in 1990. On the other hand,

International Energy Agency (IEA) has organized a working group on thermophysical

properties of environmentally acceptable refrigerants as Annex 18 since March 1990.

The Annex 18 established several International Standard equations of state for R-32

(Tillner-Roth and Yokozeki 1997), R-125 (Piao and Noguchi 1998), R-134a

(Tillner-Roth and Baehr 1994), and R-143a (Lemmon and Jacobsen 2000). Generally,

the equations are valid from the triple point to the highest limited temperature and

pressure including liquid and gaseous phases and saturation. Concerning on issues of

the energy and environment, reliable thermodynamic properties especially for HFC

refrigerants in their application ranges play important role on accurate thermodynamic

calculation for the development of highly efficient air-conditioning and refrigeration

machines.

1.2 Existing Equations of State for HFC refrigerants

As previously described, several existing equations of state were proposed for

HFC refrigerants including International Standard equations of state. Among the

existing equations of state are represented in Helmholtz free energy as previously called

as fundamental equation of state, and others are represented in pressure as so-called as

modified Benedict-Webb-Rubin (MBWR) equation of state. Detailed explanation of

both equation types is given in Section 2.2. However, the investigations of the existing

equations of state indicated that some remaining problems are still contained in the

equation. Several studies concerning the consistencies of the existing equations of state

4
to virial equations of state for HFC refrigerants were reported in elsewhere (Narukawa

et al. 2000; Mizuoka 2001; Matsuda et al. 2002; Kojima and Sato 2002). More

advanced investigations were conducted in this preliminary study. These problems are

completely described in the next section to give more comprehensive information of the

existing equations of state for HFC refrigerants.

1.2.1 Consistency Relation with Intermolecular Potential Energy

Interesting in consistency relations of fundamental equation of state to

intermolecular energy can be revealed from the second and third virial coefficients. In

principle, the virial coefficients can be derived from the equation of state. However, the

reliable values are still difficult to be judged since these properties are very difficult to

be measured.

0 0.2
Nicola et al. (2002) Tillner-Roth and Yokozeki (1997)
-0.2 Outcalt and McLinden (1995)
Yokozeki et al. (1998) 0.1
C , dm 6·mol -2

Yokozeki et al. (1998)

B , dm 3·mol -1

-0.4 Vasserman and Fominsky (2001)

0
-0.6 Nicola et al. (2002)
Tillner-Roth and Yokozeki (1997)
-0.1
-0.8
R-32 Outcalt and McLinden (1995) Vassermann and Fominsky (2001)
R-32
-1 -0.2
130 170 210 250 290 330 370 410 450 490 530 130 170 210 250 290 330 370 410 450 490 530 570
T, K T, K
0 0.2
Yokozeki et al. (1998) Vasserman and Fominsky (2001) R-125
-0.2 Piao and Noguchi (1998)
0.1
Sunaga et al. (1998)
B , dm 3·mol -1

C , dm 6·mol -2

-0.4
Outcalt and McLinden (1995)
0 Yokozeki et al. (1998)
-0.6 Piao and Noguchi (1998)
-0.1
-0.8
R-125 Sunaga et al. (1998) Outcalt and McLinden (1995)
-1 -0.2
150 190 230 270 310 350 390 430 470 510 550 150 190 230 270 310 350 390 430 470 510 550 590
T, K T, K

Fig. 1.1 Example for behaviors of second and third virial coefficients from several
existing equations of state. Left figures: second virial coefficients; Right figures: third

5
virial coefficients.

In addition, the shortage and inconsistency of thermodynamic relationships among

experimental data cause inaccurate equation of state. Larger degrees of freedom for

thermodynamic modeling may cause error progression that can trap the model into the

systematic error of input data. Intermolecular potential theory is rational to be

considered to assess the existing equations of state. Thermodynamic consistencies are

adequate to reveal the validity of universal principles of thermodynamics. These all are

interesting to be investigated in this study.

1.1
0.9 Outcalt and McLinden (1998)
1.6 MPa 2
1 Lemmon and Jacobsen (2000)
0.8 Tillner-Roth and Yokozeki (1997) 1 Li et al. (1999) 1.0
cv , kJ·kg -1 ·K-1
cv , kJ·kg -1 ·K-1

0.9
0.5 0.5
0.7 0.25 0.8
0.1
0.6 Piao et al. (1998)
0.01 0.7
Matsuda et al. (2002) Kojima and Sato (2002)
0.5 0.001 MPa 0.6
Outcalt and McLinden (1995) R-32 0.0001 MPa R-143a
0.4 0.5
125 175 225 275 325 375 150 200 250 300 350 400
T, K T, K
0.90 1.3
2 MPa Tillner-Roth (1996) 1.75 MPa
0.85 1.2
0.80 1 1.1 1
Sunaga et al. (1998) Span and Wagner (2003)
cv , kJ·kg -1 ·K-1

0.5
cv , kJ·kg -1·K-1

0.75
0.5 1
0.70 Mizuoka (2001) 0.25
0.25 Kojima and Sato (2002) 0.9 0.1
0.65 0.1
0.8 Outcalt and McLinden (1996)
0.60
0.55 Piao and Noguchi (1998) 0.7 0.0001 MPa
0.0001 MPa
0.50 0.6 R-152a
R-125
0.45 0.5
150 200 250 300 350 400 150 200 250 300 350 400
T, K T, K

Fig. 1.2 Example for specific heat consistency in the gaseous phase from several
existing equations of state.

Concerning consistency relation of equation of state and intermolecular potential,

the second and third virial coefficients from several existing equations of state for HFC

refrigerants were assessed. Figure 1.1 shows comparisons of the virial coefficients. It is

not very significantly different among of the second virial coefficients derived from

6
different equations of state. Their behaviors are still in acceptable range. The good

consistencies of the second virial coefficients are not only for R-32 and R-125 but also

for R-134a, R-143a, and R-152a. On the other hand for the third virial coefficients, it

can be noticed that significant differences are indicated at low temperatures as shown in

the right-side figures of Fig. 1.1. The inconsistent behaviors are not only for R-32 and

R-125 but also for R-134, R-143a, and R-152a.

Whether unreliable behavior of virial coefficients affecting on specific heat in

gaseous phase is interesting to be revealed in advance. Four examples for the HFC

refrigerants are shown in Fig. 1.2. As shown in the figure, significant different values

are performed at states near saturation. In the examples, virial equations of state were

used for comparison, i.e., equations of Matsuda et al. (2002) for R-32, and Kojima and

Sato (2002) for R-125 and R-143a, and Mizuoka (2001) for R-152a. Regarding derived

specific heat values, deviation of the International Standard equations of state for R-32

and R-143a from virial equations of state are more than 5 % in the worst case. It is also

occurring in other equations of state for HFC refrigerants. Caloric properties in the

gaseous phase are very important due to the compressors operating there. These

inconsistencies may be contributed by the limited number input data near saturation as

pointed out by Matsuda et al. (2002, 2003).

These investigation results suggest that the theoretical intermolecular background

is enable to be considered to improve the reliability of an equation of state for low

temperatures, especially in the region near saturation. These behaviors are supposed as

consequence of higher degree of freedom for thermodynamic model and scarce data

near the saturation both PVT and caloric data used for thermodynamic modeling.

Consequently, intermolecular potential behavior is rationally considered as one of

parameters on development of fundamental equations of state. Several intermolecular

7
potential models such as the hard-sphere potential, the Kihara potential, the

Lennard-Jones potential, the square-well potential, the Stockmayer potential, etc., have

been adequate models to analyze thermodynamic properties especially in gaseous phase.

In practice, virial coefficients were derived from intermolecular potential models and

empirically determined by fitting to experimental data. Possibility of these efforts to be

considered on thermodynamic modeling for improving reliability of fundamental

equation of state is interesting to investigate in advance.

1.2.2 Consistency of Ideal Curves

Ideals curves are considered to judge whether the fluid surface of an equation of

state is reasonable or not. Reasonable behavior of thermodynamic surface of a fluid is

necessary to be assessed for verifying the equation of state to adequately represent

thermodynamic properties in wide-range fluid. Detailed definitions for the ideal curves

are given as follows,

Ideal curve:

Z =1 (1.1)

Boyle curve:
 ∂Z 
  = 0 (1.2)
 ∂ρ  T

Joule inversion curve:

 ∂Z 
  =0 (1.3)
 ∂T  ρ

Joule-Thomson inversion curve:

 ∂Z 
  =0 (1.4)
 ∂T  p

Based on the above definitions, ideal curves can be derived from the equations of

state. Figure 1.3 shows the examples for the ideal curves that were derived from the

8
International Standard equations of state for HFC refrigerants. The equation of state for

R-32 (Tillner-Roth and Yokozeki 1997) suffers from inability to represent reasonable

Joule inversion curve. The same problem also occurs in the equation of state for R-134a

(Tillner-Roth and Baehr 1994). The equation of state for R-125 (Piao and Noguchi

1998) performs lack of the ideal curves. It is not only the Joule inversion curve, but also

the other curves that cannot be derived as right form like those derived from the other

equations of state. The equation of state for R-143a (Lemmon and Jacobsen 2000) is

one of the International Standard equations of state for HFC refrigerants can represent

the ideal curves without any lack.

1000 1000
Joule inversion curve R-32 Joule inversion curve R-125

100 Joule-Thomson inversion curve 100 Joule-Thomson inversion curve

Ideal curve Ideal curve

p /pc

p /pc

10 10
Boyle curve
Saturation line
1 1 Saturation line
Boyle curve
0.1 0.1
0.1 1 10 0.1 1 10
T /T c T/T c
1000 1000
Joule inversion curve R-134a Joule inversion curve R-143a
100 Joule-Thomson inversion curve 100
Joule-Thomson inversion curve
Ideal curve
p/p c

p /pc

10 10 Ideal curve
Saturation line Saturation line
1 1
Boyle curve Boyle curve
0.1 0.1
0.1 1 10 0.1 1 10
T /T c T /T c

Fig. 1.3 Example for ideal curves from several existing equations of state. The ideal
curves were calculated from the International Standard equations of state: R-32
(Tillner-Roth and Yokozeki 1997); R-125 (Piao and Noguchi 1998); R-134a
(Tilner-Roth and Baehr 1994); and R-143a (Lemmon and Jacobsen 2000).

Considering the reasonable behavior of ideal curves given by Span and Wagner

(1997) for simple fluid of nitrogen and carbon dioxide, unreasonable ideal curves will

affect on ability of the property representation at extreme temperatures and pressures.

9
The assessment of ideal curves is important to confirm the goodness of the structural

terms of the equation of state for wide-range of thermodynamic surface in the fluid

phase. Finding equation of state with reasonable behavior of these curves is one of

targets in this study.

1.3 Research Background

There are two aspects of background of the study. The first aspect is to develop

reliable fundamental equation of state for HFC refrigerants as consequence of the

existing equations of state still remaining several problems. The first problem often

occurring in the existing equations of state is the inability to represent third virial

coefficient, which does not agree with the virial coefficients derived from

intermolecular potential model. The third virial coefficient is quantity correlated from

intermolecular potential energy among three-molecule cluster. Behavior of

intermolecular potential energy relates to caloric properties, and approach of fluid as

molecular cluster is reasonable for gaseous phase. Therefore, the inconsistent relation of

fundamental equation of state to intermolecular potential model is suspected to affect on

reliability of specific heat in gaseous phase. This suspicion is sustained by second

problem. The second problem is inconsistency indicated with very large discrepancy

among specific heats derived from the equations of state in comparison of specific heat

from accurate virial equation of state. It is more than 5 % in the worst discrepancy in

case of the discrepancy indicated in relative deviation.

The third problem is that the existing equations of state for HFC refrigerants

cannot represent reasonable ideal curves as shown in previous section. This problem

may cause the existing equations of state cannot provide thermodynamic representation

in the region where experimental data are not available. Advanced investigation of the

10
existing equation of state for HFC refrigerants in preliminary study indicated that

inconsistent behaviors of third virial coefficients, unreasonable behavior of Joule

inversion curve, and large discrepancy of specific heat in the gaseous phase are still

occurring. International Standard equations of state for HFC refrigerants also suffer

from these improper behaviors. The efforts to improve the equations of state are very

important to be conducted in advance.

As consequence of importance in the practical application for air-conditioning and

refrigeration systems, it is desired to develop fundamental equations of state for HFC

refrigerants with suitable considerations to reduce the recent remaining problems.

Reliable input data at any state covering wide range are also needed. It would not be

always satisfied experimentally. An alternative way for developing reliable fundamental

equation of state would be based not only on experimental data but also on theoretical

background such as virial coefficients derived from the intermolecular potential energy.

This alternative is introduced by considering various theoretical aspects including

comprehensive analysis of mathematical and physical aspects.

The second aspect of the background is for developing procedure. Actually,

several procedures were reported for thermodynamic modeling (Hust and McCarty

1967; Wagner 1973; Span et al. 1998; Setzmann and Wagner 1989), but it is only

limited information. The optimization procedure plays important role on development

process of thermodynamic models. The procedure for thermodynamic modeling of this

study was originally developed.

1.4 Research Purpose and Scope

The aim of this study is to develop fundamental equation of state for wide-range

validity and consistent with theoretical backgrounds. Intermolecular potential theory is

11
one of considerations that are considered in development of fundamental equation of

state to improve thermodynamic consistencies. More comprehensive considerations of

thermodynamic relations and mathematical modeling are involved to develop

thermodynamic equation of state. They are expected to represent experimental data

accurately and theoretical background consistently.

Ideal curves are assessed to avoid unreasonable behavior in the region where no

experimental data are available. Ancillary equations for saturation properties are also

needed to develop fundamental equation of state and to give initial values for its

implementation when known properties are not density and temperature. Scarce

experimental data at saturation for low temperatures influence in low reliability of the

ancillary equation at low temperatures if development of the ancillary equations is

conducted on the basis of experimental data only. The ancillary equations are developed

to make a good agreement of values from the fundamental equation of state and the

saturation measurements. Nevertheless, the ancillary equations are possible to calculate

properties directly.

The present research will constrain four HFC refrigerants of R-32, R-125, R-134a,

and R-152a as object fluids on development of thermodynamic model. An optimization

procedure with including biological evolution process, genetic principle, and

probabilistic process was developed as an auxiliary tool for the development. The

optimization procedure was used, not only to develop the fundamental equations of state,

but also to develop ancillary equations for vapor pressure and saturated-liquid and

saturated-vapor densities of the four HFC refrigerants. Simultaneous optimization was

applied to establish a same functional form for each group of the ancillary equations.

The dissertation is divided into eight chapters in accordance with the research

purpose and scope. The scheme of each chapter can be summarized as,

12
Chapter 1: Introduction.

This first chapter gives general information on the research including refrigerants

and thermodynamic property issues, the exploration of the existing equations of

state for HFC refrigerants, the research background, and the research purpose and

scope.

Chapter 2: State of the Art on Thermodynamic Model for Pure Fluids.

This chapter mainly describes thermodynamic-law consequences and

thermodynamic properties, equations of state with validity covering gaseous phase,

saturation, and liquid phase for pure fluids, and strategy for implementation of the

fundamental equations of state in practical applications.

Chapter 3: Existing Data and Data Selection for Thermodynamic Modeling.

Surveying of experimental data is described and it is then followed by selection of

experimental data for modeling.

Chapter 4: Data Preparation and Development Procedure.

Data preparation has contribution on thermodynamic modeling. This preparation

relates to existing data and modeling consideration. The modeling paradigm and

consideration are firstly described, and then is followed by data preparation.

Weighting factor and regression are also introduced to commence the development

procedure. The development procedure covers development principle, optimization

procedure, demonstration of optimization process, and analyses of statistical

inference view on the optimization procedure.

Chapter 5: Ancillary Equations for Saturation Properties of HFC refrigerants.

The ancillary equations consisting of equations for vapor pressure and

saturated-vapor and saturated-liquid densities are developed with considering their

harmonization behavior with respect to fundamental equation of state and

13
experimental data. The ancillary equations with common structural form for the four

HFC refrigerants (R-32, R-125, R-134a, and R-152a) are developed with the same

genetic optimization procedure. Their assessments are also described here including

consistent assessment to the fundamental equations of state.

Chapter 6: Fundamental Equations of State for HFC Refrigerants.

This chapter describes thermodynamic formulation and development of fundamental

equations of state for the four HFC refrigerants as result of this study. Assessments

and comparisons are represented to give more comprehensive information of the

fundamental equations of state.

Chapter 7: Discussion.

Discussions include reliability assessments and consistency of the fundamental

equations of state with intermolecular potential energy, their extrapolation behaviors

to represent thermodynamic properties, and characteristics of ideal curves.

Comparisons to other existing equations of state are also included in this discussion.

Chapter 8: Conclusion.

The study is summarized in this chapter.

14
 Chapter 2
State of the Art on Thermodynamic Model for Pure Fluids

Thermodynamic properties have become a field that has been developing in

accordance with engineering and science interests. They play important role on

preparing reliable information for thermodynamic calculations. The main explanation is

state-of-the-art information of thermodynamic equation of state with validity covering

gaseous and liquid phases and saturation for pure fluids. Thermodynamic relations and

implementation strategy of fundamental equation of state are also described.

2.1 Thermodynamic Consequences and Thermodynamic Properties

Thermodynamic laws have given in great consequences on progress of

engineering sciences including great contributions on thermodynamic properties.

Thermodynamic quantities are only stated in equilibrium as consequence of the zeroth

law. Temperature is a thermodynamic quantity to substantiate the zeroth law. In thermal

equilibrium with each other, they have a common intensive property so called as

temperature. Consequently, systems that are not in thermal equilibrium with others have

different temperatures.

The first law has made a correlation between mechanical energy and thermal

energy. Pressure can be understood in purely mechanical terms as force that acts

perpendicularly to a known area, or mechanical energy can be formulated as work in

fluid mechanics. Definition of work in mechanical term can be used in thermodynamics.

Thermodynamic properties of the pressure, volume, and thermal energy have

contributed to thermodynamic properties in advance. These all essentially help the effort

15
to measure and substantiate other thermodynamic properties.

Thermal heat was understood from the experience and temperature definition.

Phenomena and consideration of measuring the thermodynamic properties were

considered to introduce sensible and latent heats. Specific heats were introduced to give

precise physical meaning. Heat capacity transferred at constant pressure is called as

isobaric specific heat and the heat capacity transferred at constant volume is called as

isochoric specific heat.

Entropy is a thermodynamic quantity that was introduced to substantiate the

second law. Combination of the zeroth, first and second laws, contributes on relations of

thermodynamic properties. Statistically, canonical ensemble principle also bridges

understanding of the thermodynamic relations. Expression for internal energy and

entropy of system on the basis of molecular model can be derived from statistical

thermodynamics.

Thermodynamic properties have been developed and consistently substantiate

thermodynamic law. In accordance with matter phases, thermodynamic properties were

developed from the ideal-gas concept to gaseous, liquid and solid phases. Engineering

interests have stimulated to investigate the gaseous and liquid phases including

saturation state as what all results can be seen recently.

Historically, in 1662 Boyle conducted experiments on air and deduced that volume

of a constant gaseous mass is inversely proportional to pressure. Charles in 1787

investigated effect of temperature, and then Gay-Lussac found that volume is linearly to

temperature at constant pressure in 1802. Gathering with Dalton’s law for partial

pressure, those results suggest a relation that can be written in Eq. (2.1). This relation

can be expressed in PVT diagram as illustrated in Fig. 2.1(a).

16
 RT
p = (∑ ni ) (2.1)
V

S+L
c.p.

Pressure
S
Pressure

G G
L+ G
S+G e
ur
e tur
at Volu era
Vol pe
r me e mp
ume m T
Te

(a) Ideal Fluid (b) Real Fluid

Fig. 2.1 Phase diagram for ideal fluid and real substance. (G) Gas phase; (L) Liquid
phase; (S) Solid phase.

van der Waals in 1873 postulated an equation based on consideration of theoretical

background. The model was assumed as molecules in hard spheres with infinitesimal

attractive interaction at infinite long. Virial equation with more rigorous statistical

thermodynamic analysis then followed it. Investigations of fluids were not only in

gaseous phase but also in liquid phase and at saturation. It was developed to represent

thermodynamic properties not only PVT property but also caloric properties of specific

heat and speed of sound. The valid range was expanded from one phase to several

phases. The solid phase is likely to lagging from the investigation compared with the

other phases. Thermal properties especially for solid material of metal and non-metal

are adequately interested. These all goes as consequence of engineering interests to

handle the materials.

Generally, phase diagram for a substance can be illustrated as shown in Fig.

2.1(b). This figure shows phase transformation of a compound. Heating solid-phase

17
substance may change it to liquid or gaseous phases. At pressure lower than the triple

point, the substance can change to gaseous phase so called as sublimation. On the other

hand, at the higher pressure, it melts and changes to liquid phase so called as melting.

When temperature rising at constant pressure, the phase change of substance from the

liquid phase to gaseous phase is called as evaporation. Heating at temperature higher

than critical point (c.p.) may cause the liquid substance changing directly to the gaseous

phase without passing through transition transformation (two-phase, L+G). Converse

process will undergo if the substance is cooled.

2.2 Thermodynamic Equations of State

Thermodynamic equations of state with wide-range validity for fluid phase

including gaseous phase, saturation, and liquid phase have been commenced since

several decades ago. The equations of state may be categorized into several groups such

as analytical equation of state, semi-empirical equation of state, and empirical equation

of state. Actually, several types of them are cubic equation of state, corresponding

equation of state, extended corresponding equation of state, modified Benedict-

Webb-Rubin (MBWR) equation of state, and Helmholtz equation of state.

2.2.1 Analytical Equation of State

Several equations of state that can be grouped into analytical approaches are cubic

equation of state, corresponding equation of state, and extended corresponding state.

Naturally, the main requirement of the equations is to represent experimental data as

accurate as possible. Pressure can be expressed as sum of two terms, i.e., repulsive

pressure and an attractive pressure like substantiation of van der Waals equation.

Various equations of state were proposed and among them were declared with

18
development including this principle. General form for analytical equation of state can

be written with 6 parameters consisting of b, λ, Φ, Θ, µ, and η. These parameters

excluding b are parameters introduced to express a relation of analytical equation of

state to be simple only and not have physical meaning. It can be written as,
RT  V − Φ  Θ(V − η )
p=  − (2.2)
V − b  V  (V − b)(V 2 + λV + µ )

By taking Z as compressibility, the equation can be converted into a polynomial form as,

N 1 Z 3 + N 2 Z 2 + N 3 Z + N 4 + N 5 Z −1 = 0 (2.3)

R 3T 3 R 2T 2 3
T3 R 2T 2
where N1 = 3
; N2 = 2
(λ − b ) − R 3
; N3 = 2
(Φ − λ ) + RT (µ − λ b + Θ / p ) ;
p p p p p

N4 =
RT
(Φλ − µ ) + Θη ; N 5 = Φµ
p p

Table 2.1 Analytical Equations of State

No. Equation Year Φ* Θ λ µ η

1 van der Waals 1873 0 a 0 0 b
2 Clausius 1880 0 a /T 2c 0 b
3 Redlich-Kwong 1949 0 a /T 0.5
b 0 b
4 Scott 1971 −d a c−b − bc b
5 Soave 1972 0 a b 0 b
6 Carnahan-Starling 1972 Φ1 a /T 0.5
b 0 b
7 Peng-Robinson 1976 0 a 2b −b 2
b
8 den Santis-Goroni-Marelli 1976 Φ1 a b 0 b
9 Harmans-Knapp 1980 0 a cb (1 − c)b 2 b
10 Schmidt-Wanzel 1980 0 a ub wb 2
b
11 Patel-Teja 1982 0 a b+c − bc b
12 Heyen 1983 0 a/2 b/2 − bc / 2 b
13 Trebble-Bishnoi 1987 0 a b+c − bc − d 2
b
14 Al-Shafei-Mecarik 1997 Φ1 a /(1 − c) b(1 + c) /(1 − c) 0 b
15 Zhang-Sato-Watanabe 1997 Φ2 a c 0 b
 b 
2 3   3
*
Φ1 = V − (V − b)1 +
b
+
b
− 1 − b  ; Φ 2 = V − (V − b) 8V + 3b  ; and a, b, c
3   64V 3 
 4V 4V 64V     8V − 5b 

are substance specific parameters.

19
 If Θ is not function of V and/or has a certain value and Φ is zero, the equation

form becomes cubic equation. More cubic equation types were summarized in reference

(Poling et al. 2000). For both Θ and Φ are not function of V and/or have a certain value,

the equation form becomes as quadric equation. On the other hand, in case of Φ as

function of V, the equation cannot be categorized as a polynomial equation. It is neither

the cubic equation nor the quadric equation such as the equations that can be found

elsewhere (Carnahan and Starling 1972; de Santis et al. 1976; Al-Shafei-Mecarik 1997;

Zhang et al. 1997).

Corresponding equation of state is an equation developed with corresponding

state principles. If this equation is extended with non-conformal compound to acquire

an accurate model, it is called as extended corresponding equation of state. The

corresponding state principles commonly involve principle of dimensionless function

and dimensionless variable to develop relation of several fluids in an equation.

Molecular condition for lowest level corresponding state principle is on when only two

characteristic properties such critical temperature and critical pressure are used to make

the state condition dimensionless (Pitzer 1939; Guggenheim 1945). Development of the

corresponding equation of state contributes on prediction for properties of many fluids

from few of the known properties (Leland and Chappelear 1968).

Corresponding equation of state was treated through mapping Helmholtz free

energy and compressibility factor for complex substances (Rowlinson and Watson 1969).

Shape factor parameters were taken for modifying scaling factors of temperature and

pressure to improve its accuracy. Taking two independent variables, the corresponding

parameters can be defined as,

Tj vj ρo
fj = ; hj = = (2.4)
To vo ρj

20
 Tc , j vc , j
f j (v j , T j ) = θ j ( v j , T j ) ; h j (v j , T j ) = φ j (v j , T j ) (2.5)
Tc,o vc ,o

where θj and φj are shape parameters; subscript c, o, and j are critical parameter,

reference substance, and corresponding substance.

The correlation of the corresponding of states can be expressed in residual

Helmholtz free energy as,

A r ( ρ j , T j ) = Aor ( ρ o , To ) = Aor ( ρ j h j , T j / f j ) (2.6)

On the other hand, the θj and φj should be fitted to experimental data. This way was

used for saturation boundary properties of several refrigerants and R-134a was selected

as a reference fluid (Huber and Ely 1994).

2.2.2 Semi-Empirical Equation of State

Semi-empirical equation of state was developed from expansion of virial equation

of state. Equations in this group have been developed since several decades ago. The

equations explicitly describe pressures against temperature and density so called as

MBWR equation of state. The MBWR is abbreviated from modified Benedict Webb

Rubin in accordance with historical development of the equation and the names of the

authors who had introduced it. It is most commonly used for developing equation of

state with wide-range application and was firstly applied for Nitrogen (Jacobsen and

Stewart 1973). The equation can be generalized as,

p = ∑a i ρ i + exp( − ρ 2 / ρ c2 ) ∑ a j ρ 2 j − m (2.7)
i j

where ai coefficient exception a1 = RT , other aj coefficients are expressed as

polynomial functions of temperature. The above MBWR model was widely used to

establish equations of state for refrigerants (Younglove and Ely 1987; Younglove and

McLinden 1994; Oucalt and McLinden 1995; etc). Piao and Noguchi (1998) proposed

21
other form of the MBWR model to establish the equation of state with shorter term and

expressions of pressure, temperature and density in reduced form.

2.2.3 Empirical Equation of State

Empirical approach for representing thermodynamic properties in wide-range

validity can be given in Helmholtz free energy. The Helmholtz free energy was firstly

used as the basis of characteristic function to represent properties of water including

vapor and liquid (Keenan et al. 1969). More rigorous ideal-gas law was firstly included

in the Helmholtz equation of state by dividing into an ideal-gas part and a residual part

(Schmidt and Wagner 1985). This expression was used for several compounds by other

researchers.

Recently, tendency to use Helmholtz equation of state continuously increases. The

Helmholtz equation of state can be found elsewhere (Jacobsen et al. 1992; Kamei et al.

1995; Span and Wagner 1996; Tillner Roth and Yokozeki 1997; Li et al. 1999; Span et

al. 2000; Lemmon and Jacobsen 2000; etc). This equation of state is so called as

fundamental equation of state.

A fundamental equation of state is a mathematical formulation to represent

thermodynamic properties of either pure or mixture fluids. It should be capable to

reproduce experimental data in accordance with uncertainties of the experimental data

over wide ranges of temperatures and pressures, including liquid phase, saturation, and

gaseous phase. Explicit representation in Helmholtz free energy is more valuable for

implementation of the fundamental equation of state. All the thermodynamic properties

are easily derived from the Helmholtz free energy function by mathematical

differentiation operation only. Surveying of various literatures from the oldest to the

newest one on thermodynamic fundamental equation of state and considering

22
mathematical programming aspects and flexibility of this work for future development,

a general form of Helmholtz free energy function as a basis of fundamental equation of

state is written as,

Ideal-gas part

α o (δ , τ ) = ln δ + N io ln τ + ∑ N ioτ
i
η oi
{ (
+ ∑ N io ln 1 − exp − η ioτ
i
)} (2.8)

Residual part

{
α r (δ ,τ ) = ∑ N i δ dl τ tl exp − m d (δ − s d ) te − mt (τ − st ) te
i
} (2.9)

Detailed expression and physical relations of the above fundamental equation of state

are described in Appendix B.

2.3 Thermodynamic Property Relations to Helmholtz Free Energy Function

Relations of the thermodynamic properties to Helmholtz free energy function play

important role on development and implementation of fundamental equation of state.

Explicit representation in the Helmholtz free energy function gives several features due

to accuracies and easiness of implementation of the equation of state. Precise work to

derive the right relations was conducted and detailed work is given in Appendix A. The

relations are summarized as listed in Table 2.2.

Table 2.2 Relations between Thermodynamic Properties and the Helmholtz Free Energy
Function α (δ ,τ ) and Its Derivatives

Property Relations

Compressibility factor Z = 1 + δα δr

Fugacity f (δ ,τ ) = p exp(α r + δα δr − ln(1 + δα δr ))

Ideal-gas specific heat

c op (τ ) / R = 1 − τ 2αττo = 1 + cvo (τ ) / R

Isobaric specific heat r 2

(1 + δα δr − δτα δτ )
c p (δ ,τ ) / R = cv (δ ,τ ) / R +
(1 + 2δα δr + δ 2α δδ
r
)

23
 Table 2.2 (continued)
Property Relations

Isochoric specific heat cv (δ ,τ ) / R = −τ 2 (α ττo + α ττr )

Isothermal compressibility
{
κ (δ ,τ ) ρRT = 1 + 2δα δr + δ 2α δδ
r
}−1

Joule-Thomson coefficient − (δα δr + δ 2α δδ

r r
+ δτα δτ )
µ (δ ,τ ) Rρ = r 2
(1 + δα δr − δτα δτ ) + (1 + 2δα δr + δ 2α δδ
r
) c v (δ ,τ ) / R

Pressure p(δ ,τ ) /( ρRT ) = 1 + δα δr

Second virial coefficient B (τ ) ρ c = lim α δr

δ →0

Specific internal energy u (δ ,τ ) /( RT ) = τ (α τo + α τr )

Specific enthalpy h(δ ,τ ) /( RT ) = τ (α τo + α τr ) + 1 + δα δr

Specific entropy s (δ ,τ ) / R = τ (α τo + α τr ) − (α o + α r )

Specific Gibbs free energy g (δ ,τ ) /( RT ) = 1 + α o + α r + δα δr

Speed of sound (1 + δα δr − δ τα δτ r 2
)
w 2 (δ ,τ ) M /( RT ) = 1 + 2δα δr + δ 2α δδ
r
+
c v (δ ,τ ) / R

Third virial coefficient r

C (τ ) ρ c2 = lim α δδ
δ →0

Saturated-liquid specific heat cv (δ ' ,τ s ) (1 + δ 'α δr '−δ 'τ sα δτr

')
c s ' (δ ' , δ " ,τ s ) / R = + ×
R (1 + 2δ 'α δr '+δ ' 2 α δδ
r
')
 r r 1 dp s (δ ' , δ " ,τ s ) 
1 + δ 'α δ '−δ 'τ sα δτ '− 
 R ρ cδ ' dT 

Saturated-vapor specific heat cv (δ " ,τ s ) (1 + δ "α δr '−δ "τ sα δτ

r
")
c s " (δ ' , δ " ,τ s ) / R = + r 2 r
×
R (1 + 2δ "α δ "+δ " α δδ " )
 r r 1 dp s (δ ' , δ " ,τ s ) 
1 + δ "α δ "−δ "τ sα δτ "− 
 Rρ c δ " dT 

Vapor pressure p s (δ ' , δ " ,τ s ) ρ ' ρ ''  δ ' r r 

= '  ln + α '−α " 
RTs ρ − ρ ''  δ " 

Volume expansivity (
β (δ ,τ )T = 1 + δα δr − τ δα δτ
r
) (
÷ 1 + 2δα δr + δ 2α δδ
r
)
Abbreviations:
 ∂α   ∂α   ∂ 2α   2   2 
 , α ττ =  ∂ α  , α δτ =  ∂ α , δ = ρ , τ = c
T
αδ =   , ατ =   , α δδ =  2   ∂τ 2   ∂δ∂τ 
 ∂δ τ  ∂τ  δ  ∂δ τ  δ   ρ c T

24
2.4 Thermodynamic Property Calculation

As indicated in the relations of thermodynamic properties from Helmholtz free

energy function, if a state with density and temperature is known, all other

thermodynamic properties can be easily calculated. On the other hand, unless density

and temperature are known, understanding of the thermodynamic behavior and

mathematical strategy are needed to calculate thermodynamic properties. Most of these

calculations need iteration process due to complex relation of the free energy function

with respect to its independent variables. Mathematical strategies without including

physical behavior of thermodynamic properties may make instable calculation process.

Therefore, adequate strategy is needed to establish an effective calculation procedure.

2.4.1 Practical Calculation Strategy for Thermodynamic Properties

Various possibilities in thermodynamic calculations commonly relate to

engineering interests. Two properties of a fluid are usually known and other properties

would like to be derived. Especially for case of a state at saturation, only one property is

known, other properties can be derived. Several thermodynamic properties of a state

may be previously known such as density-temperature, pressure-temperature,

pressure-entropy, enthalpy-temperature, entropy-temperature, etc. Recently, most

available tables for standard thermodynamic properties were calculated from

fundamental equation of state.

Most cases of the calculation are necessary to judge a state of the known

properties whether in liquid or gaseous phases or at saturation. Judgments for several

cases are illustrated in Figs. 2.2 to 2.4. Especially for a state at saturation, the

calculation strategies are illustrated in Fig. 2.5. In case of density and temperature are

known, other properties for the state in either gaseous or liquid phases can be calculated

25
directly from relations in Section 2.3. For the state at saturation, it should obey the

saturation strategy described in the next paragraph. The judgment for the state in

thermodynamic surface is illustrated in Fig. 2.2(a). On the other hand, for case of

pressure and temperature are known, it can be solved in accordance with illustration in

Fig. 2.2(b). This case needs iteration strategy to find density in case of the state in

gaseous or liquid phases. Initial trial value is needed in iteration process to find the right

density. For the state at saturation, the calculation obeys the saturation strategy.

T >T c p >p c T >T c

Gas c.p. ρ >ρ'
ρ <ρ"
Temperature

T <T c
T <T c c.p.
Pressure

p c >p >p s
T<T c
Gas
Liquid
Liquid
p <p s
Liquid+Gas ρ ' >ρ >ρ" T<T c
T <T c

Density Temperature

(a) (b)

Fig. 2.2 Schematic calculation strategy for (a) case of known density and temperature
and (b) case of known pressure and temperature.

In cycle calculation, pressure and entropy of a fluid from outlet side of thermal

system are usually known for isotropic process such as turbine or compressor, while

other properties would like to be calculated. In case of the state in single phase, density

and temperature should firstly be calculated to yield the same known pressure and

entropy. Initial trial values for the density and temperature, and the state can be

determined by considering Fig. 2.3(a). Initial values of density and temperature can be

determined from behavior of entropy along saturation line to get a stable iteration. Other

26
way, it can be solved by mathematical iteration methods for two unknown properties

with two property relations. The calculation follows the saturation strategy for case of

the state at saturation.

start from saturation T <T c

point to find ρ and T
c.p. so that the same known s >s"
Gas p can be obtainied at
p <p c isentropic line Gas
s"<s Liquid+Gas

Entropy
search s on saturation line T >T c
Pressure

T <T c c.p.
Liquid+Gas Liquid
s' <s <s"
p <p c and s<s'
Liquid T <T c
s <s'
p <p c and s" >s >s'
s" max s' min
Temperature
Density

(a) (b)

Fig. 2.3 Schematic calculation strategy for (a) case of known entropy and pressure and
(b) case of known entropy and temperature.

p <p c Gas
p >p c Gas T <T c , h >h"
h"<h
T >T c
c.p. T <T c c.p.
h' <h <h"
Liquid+Gas
Enthalpy
Enthalpy

p <p c Liquid+Gas
h" >h >h'
T <T c
p <p c h <h'
h<h' Liquid
Liquid

Pressure Temperature

(a) (b)

Fig. 2.4 Schematic calculation strategy for (a) case of known enthalpy and pressure and
(b) case of known enthalpy and temperature.

27
 If entropy and temperature are known, it can be solved in accordance with strategy

as shown in Fig. 2.3(b). For the state in single phase, density is firstly calculated to find

a result as same as the known entropy. On the other hand, the calculation obeys the

saturation strategy in case of the state at saturation.

As the previous cases, when enthalpy and pressure are known, the state is firstly

judged as shown in Fig. 2.4(a). Density and temperature are solved from the enthalpy

and pressure relations for the state in single phase. On the other hand, the calculation is

conducted in accordance with the saturation strategy in case of the state at saturation.

For case of enthalpy and temperature are known and the other thermodynamic

properties would be calculated, the state should be checked at entropy and temperature

as shown in Fig. 2.4(b). The calculation is conducted in accordance with the saturation

strategy in case of the state at saturation. In contrary, density is firstly determined to

yield enthalpy as same as the known enthalpy if the state is in single phase.

p c.p.
Gas c.p. T
adjust ρ' Gas adjust T
adjust ρ
" adjust ρ'
Temperature

Liquid
Pressure

adjust ρ"

Liquid+Gas Liquid+Gas Liquid

Density Density

(a) (b)

Fig. 2.5 Schematic calculation strategy for (a) case of known saturation temperature and
(b) case of known vapor pressure.

28
 Finally, it is explained on the case for a state either at saturation or in two-phase.

Properties for the state in two-phase can be determined after the properties of the state at

saturation were firstly determined. For the state at saturation, whichever known

properties can be solved in accordance with physical behavior described either in Fig.

2.5(a) or (b) depending the cases. If the saturation temperature is known, other

saturation properties of vapor pressure and saturated-liquid and saturated-vapor

properties can be determined as illustration in Fig. 2.5(a). The saturated-liquid and

saturated-vapor densities at the saturation temperature are determined to satisfy a

saturation requirement, i.e., vapor-pressure values calculated from the pressure relation

at saturated-liquid and saturated-vapor states, and vapor pressure from the Maxwell

criterion relation yield the same value. Maxwell criterion would be satisfied when

energies of both areas in saturation dome indicate same value.

Unless the saturation temperature is known, three properties of saturated-liquid

and saturated-vapor densities and saturation temperature are firstly calculated so that the

saturation requirement can be satisfied and the known property is as same as values at

the three properties that were found. All of the three properties should be adjusted to

obtain their right values as illustrated in Fig. 2.5(b).

Thermodynamic properties for the state in two-phase can be calculated in

accordance with arm-lever rule for vapor (gas) and liquid fractions after the

saturated-liquid and saturated-vapor properties were calculated either case of known

temperature or not. In principle, the properties with states in the two-phase can be

calculated if two properties are known. One property is selected for calculation of the

properties at saturation, and the other is used to complement the property for

determining a mass fraction of vapor and liquid in the two-phase if the mass fraction is

not known yet.

29
2.4.2 Iteration Strategy and Ancillary Equations

The three ancillary equations for vapor pressure and saturated-liquid and

saturated-vapor densities are usually needed in the iteration strategy of thermodynamic

calculation from the fundamental equation of state. The implementation of Helmholtz

free energy function to calculate other thermodynamic properties for cases of the known

properties are not independent variables of Helmholtz free energy function, initial trial

values for either vapor pressure, saturation temperature, saturated-liquid density, or

saturated-vapor density are usually needed in the iteration process depending on the

cases. Very different initial values such as the density of liquid phase for finding density

in gaseous phase may cause the iteration process resulting incorrect value or over

consumption time. Furthermore, the ancillary equations are also needed on development

of fundamental equation of state. Difference from the implementation that does not need

accuracy, the development is that requires high-accurate ancillary equations for the

saturation properties.

Next Step

Target

Initial Step Starting of Trial Value

Fig. 2.6 Strategy for searching solution in iteration process.

As described previously, most calculations excluding density and temperature as

two known properties need an iteration strategy to get satisfied results. The strategies

30
purely based on mathematical methods may fail to give satisfied result. Alternatively,

including thermodynamic behavior in an iteration procedure is very helpful to avoid

failure calculation. The mathematical method that can be implemented in programming

is Newton-Raphson method.

Thermodynamic behavior can be included in the iteration procedure whether the

unknown independent variables are increased or decreased so that satisfied result can be

yielded. Several thermodynamic behaviors may be included in the iteration procedures

i.e., consequence of increasing density, temperature should be decreasing at isobaric

line; consequence of increasing temperature, vapor pressure should be increasing at

saturation; consequence of increasing density, pressure should be increasing at

isothermal line; consequence of increasing density, entropy should be decreasing at

saturation; etc.

The strategy for searching a satisfied solution can be illustrated in Fig. 2.6. Step

for both increment and decrement should be designed to make step values decreasing as

searching process so that very small error can be achieved. It is set as a specific fraction

of previous step when change from increment to decrement and from decrement to

increment. After adding and subtracting the step values have to change an error value to

be insignificant value. Whether a trial value is added or subtracted should refer to the

thermodynamic behavior. For example of a state in gaseous or liquid phases, effect of

density on pressure is that increasing density affects on increasing pressure for a

constant temperature. If a calculated pressure value is higher than the known pressure,

density is decreased or conversely. Compared with pure mathematical methods, this

simple strategy enable to calculate thermodynamic properties from thermodynamic

model needing an iteration process. Combination with this strategy and mathematical

method is possible to be applied to make stable iteration.

31
 Chapter 3
Existing Data and Data Selection for Thermodynamic Modeling

Experimental data are bases of thermodynamic modeling. Reliable input data on

the thermodynamic modeling contribute on accuracy of the model. This chapter

describes a short description of the existing data and detail of the selected data of HFC

refrigerants consisting of R-32, R-125, R-134a, and R-152a for thermodynamic

modeling of this study.

3.1 Literature Survey

Literature survey of available thermodynamic data for the fluids is needed to know

whether the data is adequate or not to develop the model. The available data can be

compiled in accordance with the information needed for thermodynamic modeling.

Assessments of the available data are very important to judge the information given in

the literatures. As described in the previous chapter that several equations of state for

HFC refrigerants were developed before this study, the literature survey is to

complement the information if new data or not reported data are found.

Since the previous works on determination of International Standard equations of

state by International Energy Agency (IEA) Annex-18 and previous study on

development of the equations of state, more information of available data for HFC

refrigerants were reported by the time of this development. Additionally, IEA-Annex 18

that has worked since 1990 to determine International Standard equations of state

especially for HFC refrigerants stimulates thermophysical societies over the world to

participate and share thermodynamic data. Most available data in the time were

32
collected in the literature study including the unpublished data.

Information of available data for R-32 is very adequate given in report of

International Standard equation of state for R-32 (Tillner-Roth and Yokozeki 1997). Not

only published data but also unpublished data were included in the report. Based on the

situation on development of equation of state for R-32 in this study, several

experimental data such as Kunimoto et al. (1994) and Kubota et al. (1995) were not

included even though they were published in that time. Both these data are the same and

their accuracies are not good. Several data such as the speed of sound by Pires and

Guedes (1999) and the PVT and vapor pressure by Matsuda et al. (2002) complement

the previous data. Especially for the ideal-gas isobaric specific heat, Yokozeki et al.

(1998) reported reanalysis data that are significant different values at higher

temperatures than the data derived and used on the development for the International

Standard equation of state.

Most of available data for R-125 were collected as long as development for the

International Standard equation of state (Piao and Noguchi 1998) in accordance with

situation in the time. Several additional data as chronological measurements are then

completed the data. Data of the PVT and vapor pressure were reported by Matsuda et al.

(2003) to enrich the available data. Speed-of-sound measurements in the gaseous phase

by Kojima and Sato (2002) and Grigiante et al. (2000) added a list of the available data.

Zhao et al. (2001) reported isobaric specific heat in the liquid phase. Reanalysis data by

Yokozeki et al. (1998) also contribute on the available data for the ideal-gas isobaric

specific heat.

Compared with the R-32 and R-152, measurements of properties for R-134 are

more intensely done. Report of International Standard equation of state for R-134a

(Tillner-Roth and Baehr 1994) gave information of the available data in the time.

33
Several sets of the data were found to complement the previous data. Hozumi et al.

(1997) reported a set of speed-of-sound measurements in the gaseous phase.

Additionally, Defibaugh and Moldover (1997) reported the PVT data in the liquid phase

and at the saturated-liquid state. Yokozeki et al. (1998) reported reanalysis data of the

ideal-gas isobaric specific heat to complement the available data.

The last one fluid as object of the study is R-152a. Even though the International

Standard equation of state was not reported, two papers on the equation of state giving

information of available data for R-152a are papers of Outcalt and McLinden (1996)

and Tillner-Roth (1996). In both these papers, information of the available data for

R-152a was given. Several data are added to complement the information here.

Speed-of-sound measurement in the gaseous phase by He et al. (2002) and PVT data in

the liquid phase by Defibaugh and Moldover (1997) enrich list of the available data.

Reanalysis data by Yokozeki et al. (1998) contribute on the available data for the

ideal-gas isobaric specific heat.

Most of the available data were collected and considered on the data preparation

for development of the fundamental equations of state. The selected data for the

development are described in Section 3.2. The collected data are simultaneously listed

to represent deviation of the available data from the fundamental equations of state as

given in Appendix C. Compilations were given in accordance with original sources for

grouping of the data whether they are included in the gaseous or liquid phases. Number

of data, ranges of pressures and temperatures are included to assess the fundamental

equations of state.

3.2 Experimental-Data Selection for Single Phase

Most of the thermodynamic modeling used experimental data as a basic data in the

34
development. Assessments of experimental data were previously performed to find a

reliable set for thermodynamic modeling. All data were prepared in the same standard

reference. Data with older temperature standard than ITS-90 were converted in ITS-90.

Sets of data are then selected as input data on the modeling.

3.2.1 Data for R-32

Experimental PVT data for the superheated-vapor phase were chosen from 408

points of de Vries (1997) and 144 points of Matsuda et al. (2002). For the liquid phase,

408 points of de Vries (1997), 120 points of Magee (1997), and 75 points of Matsuda et

al. (2002) were chosen. Additionally, 43 points of Holste et al. (1993) at higher

pressures than 25 MPa were chosen to compensate the limited data at higher pressures.

Data for the supercritical region were chosen from 240 points of de Vries (1997) (by

means of Burnet and vibrating-tube measurements), 14 points of Magee (1996), and 9

points of Holste et al. (1993) at pressures higher than 25 MPa. Based on a consistency

assessment for the data of Holste et al. (1993) against the data of Magee and of de Vries,

density values of Holste et al. (1993) were adjusted by + 0.15 % to obtain consistent

results with other data.

All of the basic data for isochoric specific heat in the liquid phase by Lüddecke

and Magee (1996) were selected as input data. Speed-of-sound data consist of 111

points of Takagi (1993), 304 points of Pires and Guedes (1999), and 30 points of

Grebenkov et al. (1995) for the liquid phase, and 67 points of Hozumi et al. (1994) for

the gaseous phase. Distribution of the selected data is shown in Fig. 3.1. It appears that

scarce of the data at lower temperatures and low pressure both in the liquid and gaseous

phases.

35
 100

10

1
p , MPa
0.1

0.01
R-32
0.001
100 150 200 250 300 350 400 450
T, K

Fig. 3.1 Points of experimental data of R-32 selected for the development. (◊) PVT; (∆)
cv; (○) w.

3.2.2 Data for R-125

Distribution of the selected experimental data is shown in Fig. 3.2. PVT data in the

superheated-vapor phase consist of 54 data points of Zhang et al. (1995) at temperature

higher than the critical point, 72 points of Boyes and Weber (1995), 385 points of de

Vries (1997), and 63 points of Matsuda et al. (2003). PVT data in the liquid phase

consisted of 77 points of Magee (1996), 274 points of de Vries (1997), 20 points of

Defibaugh and Morrison (1992), 152 points of Duarte-Garza et al. (1997), and 20 points

of Matsuda et al. (2003). PVT data in the supercritical region consist of 305 points of de

Vries (1997), 34 points of Duarte-Garza et al. (1997), and 31 points of Defibaugh and

Morrison (1992).

All isochoric specific heat data in the liquid phase of Lüddecke and Magee (1996)

were used. Speed-of-sound data in the gaseous phase consist of 149 points of Gillis

(1997), 55 points of Kojima and Sato (2002), and 69 points of Grigiante et al. (2000).

Additionally, 167 points for speed of sound in the liquid phase of Takagi (1996) were

selected as input data.

36
 100

10

p , MPa
1

0.1

0.01
R-125
0.001
150 200 250 300 350 400 450 500
T, K

Fig. 3.2 Points of experimental data of R-125 selected for the development. (◊) PVT;
(∆) cv; (○) w.

3.2.3 Data for R-134a

PVT data in the liquid phase were selected from Hou et al. (1992) of 357 points

and Tillner-Roth and Baehr (1993) of 282 points. PVT data in the superheated-vapor

phase consist of 341 points of Tillner-Roth and Baehr (1993), 24 points of Tillner-Roth

and Baehr (1992), and 53 points of Weber (1989). On the other hand, PVT in the

supercritical region were selected with 70 points of Tillner-Roth and Baehr (1992), 126

points of Tillner-Roth and Baehr (1993), 43 points of Hou et al. (1992), and 6 points of

Weber (1989).

Caloric properties as input data covers speed of sound, isochoric specific heat, and

saturated-liquid specific heat. All experimental isochoric specific heats for the liquid

phase of Magee (1992) were selected as input data. Speed-of-sound data cover liquid

and gaseous phases. Speed-of-sound measurements in the liquid phase were selected

from 206 points of Guedes and Zollweg (1992), 226 points of Takagi (1996), and 48

points of Grebenkov et al. (1995). For the gaseous phase, the data consist of 23 points of

Hozumi et al. (1997), and 94 points of Goodwin and Moldover (1990). Distribution of

37
points for the selected data is illustrated in Fig. 3.3.

100

10

1
p , MPa

0.1

0.01
R-134a
0.001
150 200 250 300 350 400 450 500
T, K

Fig. 3.3 Points of experimental data of R-134a selected for the development. (◊) PVT;
(∆) cv; (○) w.

3.2.4 Data for R-152a

The main input data are experimental data covering PVT and caloric properties.

The PVT data for the liquid phase consist of 26 points of Geller et al. (1979), 126 points

of Magee (1998), 289 points from Tillner-Roth and Baehr (1993). The PVT data in the

superheated-vapor phase consist of 20 points of Dressner and Bier (1993), 319 points

from Tillner-Roth and Baehr (1992), 28 points of Tillner-Roth and Baehr (1993). On the

other hand, the PVT data in the supercritical region consist of 27 points of Blanke and

Weiss (1992), 8 points of Dressner and Bier (1993), and 8 points of Magee (1998), 44

points of Tillner-Roth and Baehr (1992), and 88 points of Tillner-Roth and Baehr

(1993).

The caloric properties for the modeling consist of isochoric specific heat and

speed of sound. All points for isochoric specific heat in the liquid phase of Magee

(1998) including 66 points of derived data for the saturated-liquid specific heat were

used. Speed of sound in the gaseous phase were selected with 44 points of Hozumi et al.

38
(1993) and 148 points of Gillis (1997), and the data for the liquid phase consist of 51

points of Beliajeva et al. (1994) and 50 points from Grebenkov et al. (1995).

Distribution of the experimental data points is illustrated in Fig. 3.4.

100

10

1
p , MPa

0.1

0.01
R-152a
0.001
150 200 250 300 350 400 450 500
T, K

Fig. 3.4 Points of experimental data R-152a selected for the development. (◊) PVT; (∆)
cv; (○) w.

3.3 Experimental-Data Selection at Saturation

These data group consist of saturation-PVT and caloric properties. Even though

there are less saturation data for speed of sound and specific heat properties, not all of

these data are considered as input data on the development. Saturated-liquid specific

heat data only are considered on the development. All derived data for saturated-liquid

specific heat data of R-125 of Lüddecke and Magee (1996) were used to fit the

fundamental equation of state. Derived data for saturated-liquid specific heat of R-134a

by Magee (1992) were also introduced on the modeling. Additionally, all data for

isochoric specific heat of R-152a of Magee (1998) were also used. Saturation-PVT data

consist of vapor pressure, saturated-liquid density, and saturated liquid density. These

data were used to develop accurate ancillary equations for saturation properties besides

39
for developing the fundamental equations of state. The selected data for the

development are listed in Table 3.1.

Consequence of scarcely experimental data existing at low temperatures, an

alternative way to reduce inaccurate representation of saturation properties at lower

temperatures is needed. The extrapolated saturation properties were determined by

iterative process as long as development of the fundamental equations of state. Each

ancillary equation for saturation properties with specified structural equation for HFC

refrigerants were developed on the development process for each of the fundamental

equations of state.

Table 3.1 Selected Experimental PVT Data at Saturation

Selected Points
Fluid Source
ps ρ' ρ"
R-32 Defibaugh et al. (1994) - - 28
de Vries (1997) 139 - -
Holcomb et al. (1993) - 21 20
Kuwabara et al. (1995) - 13 -
Magee (1996) - 13 10
Weber and Goodwin (1993) 27 - -
Weber and Silva (1994) 17 - -
R-125 Boyes and Weber (1995) 29 - 10
de Vries (1997) 98 - -
Fukushima et al. (1995) - - 14
Kuwabara et al. (1995) - 9 8
Magee (1996) - 7 -
Weber and Silva (1994) 103 - -
Widiatmo et al. (1994) - 25 -
R-134a Aoyama et al. (1996) - 4 7
Baehr and Tillner-Roth (1991) 35 - -
Basu and Wilson (1989) - 9 -

40
 Table 3.1 Selected Experimental PVT Data at Saturation State (continued)
Selected Points
Fluid Source
ps ρ' ρ"
R-134a Goodwin et al. (1992) 57 - -
Hou et al. (1992) - 5 -
Magee and Bowley (1991) 19 - -
Niesen et al. (1994) - - 12
Weber (1989) - - 5
Yokoyama and Takahashi (1991) - 21 -
R-152a Blanke and Weiss (1992) - 19 -
Duarte-Garza and Magee (1999) 21 - -
Higashi et al. (1987) 40 6 11
Holcomb et al. (1993) - 33 33
Kamei et al. (1989) - - 24
Magee (1998) - 9 -
Silva and Weber (1993) 38 - -

After the final fundamental equations of state were accomplished, their

extrapolated data were used for reducing scarce data to develop a same functional form

of ancillary equations for vapor pressure, saturated-liquid density, and saturated-vapor

density as discussed in Chapter 5. This method is expected to obtain the ancillary

equations that can be independently used for calculating PVT properties at saturation

and giving the values close to values from fundamental equations of state for

temperature range from the triple point to the critical point. Simultaneous optimization

with genetic optimization procedure was applied to develop final the ancillary equations

as final functional form. This means to find reliable structural form of the ancillary

equations, to develop the ancillary equations for direct calculations of saturation

properties, and to prepare the ancillary equations for implementation of the fundamental

equations of state.

41
 Chapter 4
Data Preparation and Modeling Procedure

Data preparation and development procedure are two aspects that play

important role for most modeling work. Consequence of development paradigm and

consideration have significant effects on preparation for input data, both of these are

firstly described and are then followed by data preparation especially for derived data to

complement the experimental data. The optimization procedure is described, including

the flowchart and rules for each genetic operator of the optimization procedure. The

optimization process is demonstrated and the statistical inference view is finally

described to close this chapter.

4.1 Rational Paradigm and Consideration

All thermodynamic properties are needed to develop fundamental equation of state,

as consequence of the equation should be able to represent the properties well.

Experimental work has any shortage causing that not all measurable thermodynamic

properties can be accurately measured. This fact appears that the measurable properties

have a limited range and there are region where experimental data are not available as

described in previous chapter. It has challenged us on thermodynamic modeling

especially for developing the fundamental equation of state with limited available

experimental data.

In case of simple equation for thermodynamic properties with a narrow validity

range, it is not significant problem on preparing parameters and input data for the

modeling. However, development of fundamental equation of state will be in confusing,

42
and unsuitable paradigm will affect on the result and may bring a fatal error.

Interdependent relation of Helmholtz fundamental equation of state and thermodynamic

properties can be visualized in a schematic block as shown in Fig. 4.1. This paradigm

might have caused remaining consistencies in the fundamental equation of state for

HFC refrigerants. Typical characteristics of refrigerants have lower temperatures. In

contrast, experimental work at low temperatures have often arisen various problems so

that accurate input data for the modeling in this range are difficult to be prepared.

Experimental Thermodynamic Data Modeling

All Thermodynamic Properties Helmholtz Equation of state

Fig. 4.1 Common paradigm for determination of thermodynamic properties.

Thermodynamic relations among experimental and derived properties including

the intermolecular potential theory and ideal gas law are considered. Representation of

thermodynamic properties within the experimental uncertainties is also required as

criteria in the development. Simply, a paradigm on thermodynamic modeling for

fundamental equation of state in this study can be illustrated as shown in Fig. 4.2.

Experimental Data Theoretical Background

Common Parameters Basic Parameters

Reproducibility Consistencies

Reliable Helmholtz
Equation of State

Fig. 4.2 Paradigm for developing fundamental equation of state in this study.

43
 The square-well or Stockmayer potential models can be considered for

representing the second and third virial coefficients, and specific heats in the gaseous

phase can be derived from the accurate virial equation of state. Representation of

experimental data will reveal the accuracies of the equation, and derived data with

theoretical background will keep consistencies of the equation to thermodynamic

relations.

Assessment and involvement of theoretical background are necessary to judge

reliability of thermodynamic properties in case of those states cannot be compared with

limited experimental data. Virial coefficients cannot be measured directly so that data

derived from the experimental data should be assessed from theoretical viewpoint. A

viewpoint of the intermolecular potential theory was thoroughly considered. Reasonable

consideration for involving virial coefficients was introduced. Development of

fundamental equations of state with this new paradigm is expected to improve their

reliability.

4.2 Principal Consideration of Thermodynamic Modeling

Comprehensive considerations are needed to make reliable thermodynamic model,

not only accurate representation of available experimental data but also rational and

reasonable from theoretical aspects. For example, the behaviors of simple equations

such as vapor pressure and saturated-liquid and saturated-vapor densities in critical

point such as non-analytic behavior have been interested topic since several years ago

(Wagner 1973; Iglesias-Silva et al. 1987). Similarly, fundamental equation of state

should be comprehensively assessed from the theoretical aspects. More complex

functional form for the fundamental equation of state was criticized from accuracy

44
aspects (Setzmann and Wagner 1989; Span et al. 2000).

4.2.1 Mathematical Aspect

Thermodynamic modeling has to consider mathematical views especially for

degrees of freedom of the model from constraints of parameters and input data. It may

insufficiently develop a reliable model due to inadequate mathematical considerations.

In case of the equation is developed with a property quantity represented either one or

more variables, the development becomes very simple. A set of the property data

includes their independent variables, and regression process can be easily conducted.

Ancillary equations for saturation properties and isobaric specific heat equation of ideal

gas are included in this group.

On the other hand, for a case of many dependent properties should be represented

in an equation, more precise mathematical relation among the properties should be

correlated. Development of equation of state becomes more difficult problem.

Thermodynamic modeling for fundamental equation of state can be categorized into this

group. Considering the precise relation of thermodynamic properties given in Section

2.3, development of the fundamental equation of state needs many property quantities

so that Helmholtz free energy function and its derivatives can be adequately included in

regression process.

4.2.2 Thermodynamic Aspect

Consequence of shortages and inconsistencies of the experimental data still arising

and causing problems in thermodynamic modeling, an alternative way was considered

to eliminate in this study. Accompanying theoretical contributions are supposed as an

essential way for developing reliable equations of state. Importance of considering virial

45
equation of state to develop fundamental Helmholtz equation of state since it lies in its

rigorous theoretical foundation by which virial coefficients appear not only as empirical

constants but also as representation of the intermolecular potential energy. The virial

coefficients are related analytically to the intermolecular potential energy of clusters of

molecules. The second virial coefficient arises from the interaction between a pair of

molecule; the third virial coefficient depends on the interaction of clusters of three

molecules, etc.

As a fundamental equation of state of the thermodynamic properties for a fluid,

the equation of state should be able to represent not only measurements of

thermodynamic properties but also the properties in specific region where no

experimental data are available. The properties should be reasonably represented

according to theoretical background of the superheated-vapor phase, liquid phase,

supercritical region, and saturation. A statistical thermodynamic analysis for forces

between molecules has already confirmed an exact relation of the virial coefficients and

intermolecular potential energy. Therefore, no rational reason can be accepted if the

fundamental equation of state cannot reliably derive the virial coefficients.

Although theories are commonly invented from empirical facts, the theories were

also examined from various aspects. Therefore, consideration theory for developing

equation of state is a rational thinking. Theoretical background can be used to improve

the reliability of an equation of state at low temperatures, especially in the region near

saturation. Several intermolecular potential models such as the hard-sphere potential,

the Kihara potential, the Lennard-Jones potential, the square-well potential, the

Stockmayer potential, etc., can be considered when developing an equation of state. The

optimum selection of a model for a specified fluid will help reveal its proper behavior.

In practice, virial coefficients were derived from intermolecular potential models

46
and empirically determined by fitting to experimental data. The intermolecular potential

model can be determined from speed-of-sound measurements with or without PVT

measurements. Detail of discussion about the hard-core-square-well potential to obtain

the virial equation of state from speed-of-sound data was exploited (Gillis and

Moldover 1997). The virial coefficients for HFC refrigerants were examined and

deduced with the Stockmayer potential (Yokozeki et al. 1998). A set of molecular

constants from the hard-core-square-well potential for HFC refrigerants was derived

from speed-of-sound data (Gillis 1997). These efforts were focused on finding a new

potential model for representing thermophysical properties with very high accuracy. A

practical modification parameter of the square-well and Stockmayer potential models

was tried to improve thermodynamic representation (Kojima and Sato 2002). This

modification made it possible to predict the thermodynamic properties in the gaseous

phase based only on a limited number of speed-of-sound measurements, and the

potential models can be applied without any change but with the new parameter added

to temperature.

Other empirical approaches to obtain virial coefficients can be determined by

considering reliable experimental data of the gaseous speed-of-sound and PVT

properties. This approach was conducted for determining the virial coefficients and a

simple empirical equation was proposed to represent gaseous properties (Zhang et al.

1995). The same equation form was used in series of researches concerning on

consistent equation of state for HFC refrigerants (Narukawa et al. 2000; Mizuoka 2002;

Matsuda et al. 2002; etc.). It is interested to note here that these approaches keep third

virial coefficient, but the second virial coefficients were adjusted to make accurate

representation of gaseous thermodynamic properties.

Regarding research progress on the theoretical and empirical approaches for the

47
gaseous phase, and consistencies of existing fundamental equation of state, unified

relation of the intermolecular potential energy and equation of state are apparently

required to reduce the inconsistencies. Therefore, assessments of the second and third

virial coefficients are needed to adjust the primary and secondary differentiations of the

free energy function. Reasonable mathematical and physical backgrounds were also

confirmed for determination of the third virial coefficient. It is important to reduce the

degrees of freedom of the equation. Relations of the thermodynamic properties from

Helmholtz free energy function indicate that secondary differentiation of the free energy

function with respect to reduced density are contained in relations of the third virial

coefficient, speed of sound, and isobaric specific heat. Not only the speed of sound and

isobaric specific heat but also the third virial coefficient with background of

intermolecular potential energy should be considered in the modeling due to these

coefficients directly control the secondary and ternary differentiations of the equation.

4.3 Data Preparation on Development of Fundamental Equation of State

Besides using experimental data for the development as described in Chapter 3,

the critical constraints and derived data of several properties were also involved. The

derived data were calculated from specific equations such as properties of vapor

pressure, saturated-liquid and saturated-vapor densities, and virial coefficients. Detailed

rules and purposes of the derived data are described in the next section.

4.3.1 Preparation of PVT Property at Saturation

Data of vapor pressure and saturated-vapor and saturated-liquid densities at the

same temperatures are significant to introduce the Maxwell criterion for saturation

behavior, besides introducing PVT data at saturation on the modeling. Consequence of

48
unavailable reliable data for the properties at the same temperature, these properties are

prepared by calculating from three ancillary equations. Using these equations, the data

were generated from the triple point to the critical point.

Changing of gradient along the saturation line from the triple point to the critical

point is responded with variation of interval of input data to perform a good fit from a

regression process. Interval of temperatures for generating the data were made with

smaller interval at near the critical point, and then followed by larger interval close to

the triple point. Detailed rule for the interval is described in Table 4.1. These data were

also used to prepare input data for PVT property at saturation.

Table 4.1 Interval Temperature Rule for Generating PVT Property at Saturation

Temperature range Interval and data number

Tc - T1 0.01 K, 10 - 19 points
T1 - T2 0.1 K, 10 - 19 points
T2 - T3 0.5 K, 10 points
T3 - T4 1 K, 10 points
T5 - Tt 2 K, data number depends on difference
of Tc - Tt of the fluid

4.3.2 Preparation of Gaseous Caloric Properties

Reliable data for isochoric specific heat in gaseous phase are difficult to be found

in practice. These data were calculated from the accurate virial equations of state.

Considering discrepancy of the specific heats often arise in the gaseous phase near

saturation, higher intensity data were introduced on this region. Rule for the derived

data is that the data are generated for temperatures from the triple point to the highest

valid range of the virial equation of state especially for states near the saturation.

Interval temperature is made in 3 K and several states for the same temperatures. The

49
first density values are selected from the states that close to saturation, and next density

values are 0.98 from the previous values like geometric series. These few data were

used because specific heat properties near ideal gas are strongly controlled by the

ideal-gas specific heat equation.

4.3.3 Preparation of Virial Coefficients

Second and third virial coefficients are used as input data on development of

fundamental equation of state. The virial coefficients for each HFC refrigerant were

derived from several equations with sources listed in Table 4.2. These coefficients were

also used to compose terms in virial equations of state for calculating the gaseous

specific heat (Section 4.3.2). Each virial coefficient was calculated from the triple point

to the higher temperatures (600 K). Most virial coefficients with sources listed in the

table are approximation of the theoretical analysis and experimental data.

Table 4.2 Virial Coefficient Relations for Derived-Data Preparation

Fluid Sources
R-32 Matsuda et al. 2002
R-125 Kojima and Sato 2002
R-134a Narukawa et al. 2000
R-152a Mizuoka 2001

4.4 Weighting Factor and Regression

Weighting factor is a constant factor given to a data point to adjust their deviations

from the model in accordance with the expected uncertainties, data number, and data

distribution. Theoretically, the weighting factor relates to uncertainty of the input data.

If measurement of three properties (x, y, and z) are assigned in a model with

independent variables x and y, and dependent variables z, deviation of the z property can

50
be given as,

∆z = z exptl − z calc (4.1)

According to Gaussian propagation error formulation, the experimental uncertainty of a

measurement for a state is given as,

  ∂∆z  2  ∂∆z 
2
 ∂∆z 
2 
σ exptl
2
=  σ2 + σ2 + σ2 (4.2)
  ∂x  x  ∂y  y  ∂z  z 
 

The above formulation cannot be absolutely implemented in practice. This

disadvantage may be caused by reliable information of experimental uncertainties and

distribution of data. A multiplying factor or dividing factor for the experimental

uncertainty is usually introduced to modify the above relation so that it is applicable.

This way is likely theoretical-sound way, but the factor is also very flexible to be

introduced in order to find a best model that can represent thermodynamic properties

with uncertainties close to the experimental uncertainties. Therefore, it is not strictly

different with introducing an absolute weighting factor.

Input data are fitted to a model with regression method. Most regression can be

written in Y=AX form with unknown matrixes of X and A with size in accordance with

data-point number (N) and unknown variables (m). Chi-square (χ2) of the regression can

be calculated as,
2
N  m 
χ 2 = ∑  y i − ∑ ai , j x j  (4.3)
i =1  j =1 

Introducing weighting factor (w) and/or standard deviation (σ) due to uncertainties of

the input data, the chi-square becomes weighted least squares with relation as,
2 2
Ny m ai , j  N  m 
χ = ∑  i − ∑
2
x j  = ∑ wi  y i − ∑ ai , j x j  (4.4)
σ
i =1 i σ
j =1 i  i =1  j =1 

It can be written in a matrix by multiplying the all elements of matrix A and Y with

square root of weighting factor, or dividing the all elements of matrix A and Y by

51
standard deviation. The χ2 will be minimum at condition as,

AT Y = AT AX (4.5)

At this condition, the X can simply be solved in accordance with algebra.

4.5 Thermodynamic Modeling

Various optimization methods have been developed in accordance with necessities

for practical problems. The problems may be able to be solved into simple to complex

ways depending on correlations of their parameters. Many optimization problems can

be solved with mathematical optimal criteria only. If a problem with relation function

between the variables and objective function is unknown, it is forced to experiment

either on the real object or on the scale model. The relation can be established as free as

possible to vary the independent variables and have access to the measuring instruments

at which as the dependent variable or quality that can be measured.

Weighting factor Data Functional Form Data

Regression Regression

Trial functional form Solution Trial weighting factor Solution

STOP STOP

(a) Variant functional form (b) Certain functional form

Fig. 4.3 Basic principle of development for thermodynamic modeling.

Especially for thermodynamic modeling, the problems can be solved in principle

as shown in Fig. 4.3. This agrees with real thermodynamic problems to develop

empirical equations for thermodynamic properties. It may introduce constant values for

weighting factors to find functional form as illustrated in Fig. 4.3(a). Improvement by

iteration of weighting factor may be introduced for a certain functional form to obtain

52
uniform deviations of the equation from the input data or to find uncertainties of the

equation similar with the experimental uncertainty. In a simple case such as

development of ancillary equation for saturation properties and equation for ideal-gas

isobaric specific heat, the way as shown in Fig. 4.3(a) is effective to find a satisfied

result. In development of fundamental equation of state, however, combination of ways

in Figs. 4.3(a) and 4.3(b) plays important role to find a satisfied solution.

Single-fluid optimization of functional form is a process to find an optimal

equation for a fluid. Optimization of ideal-gas isobaric specific heat equation and

fundamental equations of state for HFC refrigerants were separately conducted for each

fluid. These equation developments can be expressed as shown Fig. 4.4(a). Fitness of

model can be defined as sum of the weighted least squares. The fitness can be calculated

after the model was established from the regression.

Fluid Data

Trial functional form Regression Fitness

(a) Single-fluid optimization

Fluid Data 1

Regression Fitness 1
Fluid Data 2
Trial functional form Regression Fitness 2 Total Fitness
Fluid Data n

Regression Fitness n

(b) Multi-fluid optimization

Fig. 4.4 Difference of single-fluid and multi-fluid optimizations

On the other hand, simultaneous multi-fluid optimization is not very strictly

different single-fluid optimization procedure. It is parallel of several single-fluid

optimizations and the fitness is the total fitness of models in accordance with fluid

53
number involved as shown in Fig. 4.3(b). It usually consumes more times due to several

regressions should be executed. The simultaneous optimization was applied to develop

vapor pressure, saturated-liquid and saturated-vapor density equations of the four HFC

refrigerants as described in Chapter 5.

Various methods have been developed in various fields to find an optimal

functional form as solution of a specified problem. Statistical inferences can be used as

an alternative way for selecting terms or structural forms of a mathematical model than

a conventional method that is trial and error for all possible combinations (conventional

combinatory method). Among the methods with statistical inference are forward

selection, backward selection, and stepwise selection. Statistical inferences often used

are t-test and F-test. Both tests are operated in accordance with a significant level for

judgment. Although the methods work in accordance with statistical inferences, they

also have target to find solution close to the conventional combinatory method.

Random strategies are the most methods in which the parameters are varied

according to probabilistic random instead of deterministic rules. Decision to use random

strategies in optimization whenever the deterministic strategies cannot achieve a desired

success. Searching by gradient of the objective function has convergent difficulties at

point where the partial derivatives are discontinues. On the other sides, random

strategies are often supposed to be essentially more costly.

Engineering approaches to solve engineering and science have significantly

accomplished satisfied results. One is the method that is developed with basis of

evolutionary thinking on biology so called as evolutionary computation. The second is

the method that is developed with consideration of biological neuron system to simulate

computational systems. This second method can be categorized into a group of

biological neuron computation.

54
 This study explored the evolutionary computation to implement in development of

thermodynamic models. The general form of terms for the model is firstly determined.

Linear coefficients of the terms are fitted in regression and parameters for the general

term are selected from the predicted terms so called gene pool. Simple problems such as

ideal-gas specific heat and ancillary equations for saturation properties can be

developed with the development procedure as shown in Fig. 4.3(a) only. However,

complex problems such as develop residual part of fundamental equation of state that

should be fitted to multi properties and including gaseous and liquid phases and

saturation was developed with both development procedure shown in Figs. 4.3(a) and

4.3(b).

4.6 Optimization Method

The optimization method was originally developed with considering biological

evolutionary process, understanding of biological genetic, and probabilistic random

process. It was simulated in a computer program with object-oriented programming

language C++. This procedure includes the evolutionary computation group. Recently,

four avenues of optimization procedures can be included in the evolutionary

computation. There are evolutionary programming, evolution strategy, genetic

algorithm, and genetic programming.

Evolutionary programming has been begun since 1960. Lawrence J. Fogel

commenced the evolutionary programming when he conducted a research for a less

eco-centric approach to artificial intelligence in 1960 (Fogel 1994). The evolution

strategy was firstly applied for hydrodynamic problems like shape optimization of a

band pipe, drag minimization of joint plate, and structure optimization of two-phase

flashing nozzle. This method was firstly developed as a joint development of Biernet,

55
Rechenberg, and Schwefel in 1960. The genetic algorithm was initially motivated for

design and implementation of robust adaptive systems capable of dealing with an

uncertain and changing environment by John Holland in 1967. Holland then formally

introduced genetic algorithms at University of Michigan (U.S.A.) in the 1970.

Lastly, genetic programming is derivation of genetic algorithm. Significant

difference is that genetic programming represents a solution in lisp expression, but

genetic algorithm represents a solution either in bit-string or in order-string. The

modification of representation is a response of difficulties may be arisen to solve the

problem. Lisp expression was firstly applied for solving the Prisoner’s Dilemma (Fujiko

and Dickson 1987). Application for weights in neural network was solved with lisp

expression so that complex behavior can be built (de Garis 1990). More intensive works

for solving large number of optimization and machine-learning tasks were solved with

lisp expression (Koza 1991).

Similar with biological words, main differences among them can be explained.

Genetic algorithms stress chromosomal operator, evolution strategy emphasize behavior

changes at the level of individual, evolutionary programming stresses behavioral change

at the level species, and genetic programming stresses on DNA (deoxyribunule acid) to

change behavior of chromosome. The ways of researchers to solve the problems depend

on their problems and experiences. These have caused various rules and optimization

procedures that have been proposed in various fields.

4.6.1 Optimization Procedure

An optimization procedure was accomplished in this study by including genetic

rules and evolutionary process. By considering a population of a species and their

evolutionary from a generation to a next generation and a migration process of species

56
from a habitat to another habitat, thus interactions of these processes to a population of

a habitat are considered by introducing competitors as representation of individuals

infiltrating into the population. Simply, the optimization was developed from departing

the basic strategy as illustrated in Fig. 4.5.

Competitor 1st Generation Competitor

Competitor 2nd Generation Competitor

Competitor 3rd Generation Competitor

Competitor 4th Generation Competitor

Competitor 5th Generation Competitor

Competitor 6th Generation Competitor

Competitor 7th Generation Competitor

Time Progress

Fig. 4.5 Basic Strategy on development of optimization procedure.

Two-parent marriage is interpreted by introducing a recombination operator. In

reproductive process, whether it mutates or not, may be caused by several factors. It is

modeled by introducing mutation operator and the causing factors are compensated by

mutation rate. Population size is set constant and two parents allow reproducing two

offspring. Each offspring is exposed to an environment system to obey the selection

process as shown in Fig. 4.6.

Numerous competitors can be generated according to specified rules. The best of

competitors is compared with an offspring and better one is then selected as parent for a

57
next generation. If individuals are not satisfied yet, regeneration process is repeated

again. Several rules and combinations for the genetic operators may be developed to

find a powerful and efficient tool for developing thermodynamic model.

START

General Gene Pool

Parent Parent Parent Parent Parent Parent

Recombination Recombination
Competitor Competitor Competitor

Mutation Mutation Mutation Mutation

Competitor
Offspring Offspring Offspring Offspring

Survival Survival Survival Survival

Selection

Parent Parent Parent Parent Parent Parent

Satisfied Satisfied

STOP

Fig. 4.6 Main genetic optimization procedure.

Considering better individuals can be reproduced from better genes, two types of

gene pool consisting of a general gene pool and a selected-gene pool are introduced.

The general gene pool is a collection of genes predicted possible to be a solution, and

the selected-gene pool is a collection of genes in the individuals at every generation in

the main population. A qualified individual consisting of survival individuals in the

environment is also involved. Therefore, the competitors may be initialized from the

three alternatives to infiltrate improvement effect on the main population.

58
 General Gene Pool Selected Gene Pool Qualified Individual

Parent Parent Parent Parent Parent Parent

Recombination Recombination Recombination

Mutation Mutation Mutation Mutation Mutation Mutation

Offspring Offspring Offspring Offspring Offspring Offspring

Sorting the Parent & Offspring and Select Half

better individual number for selection process

Individual Individual Individual Individual Individual Individual

Repeat

the best one selected as competitor

Fig. 4.7 Flowchart for generation of competitor.

This procedure is adopted from strategy for reproducing better breeder.

Competitor generator was developed in accordance with a rule near reproductive

breeder process such as illustrated in Fig. 4.7. Relationship of parent and child are not

concerned in the process. If a qualified child may be recombined a qualified previous

parent. The process is started with initializing several individuals from three alternatives

in the procedure. The genetic operators are also included in the procedure. Sorting

parents and offspring are introduced to select qualified individuals in accordance with

their fitness. Ratio of selected individual and rejected individuals are set constants, i.e.,

a half is rejected and the other half is selected to generate new population for the

process in finding the best one as selected competitor. Certain number for looping and

population size can be set in accordance with gene pool and chromosome sizes.

Evaluation criteria of optimization depend on specific problems that are being

59
currently solved. Correlation illustrating their behavior can be called as objective

function, and the value indicating these criteria is called fitness. This value may be

minimization or maximization targets. Suitability of input data in thermodynamic

modeling is indicated with minimum values of fitness.

4.6.2 Genetic Rules for Operators

Recombination operator is a tool with function to simulate biological reproductive

process into optimization process. Rules for the recombination and mutation are

illustrated in Fig. 4.8. The recombination process is that genes are selected from their

parent in accordance with probabilistic random process. All of genes of both parents are

collected in gene pool without deleting for replica gene, and two individuals are

randomly generated from this gene pool. As consequence is that one or more genes may

not be appeared in both their offspring and a gene from both parents with high

probability may appears in both offspring. This process undergoes in stochastic process

so that it changes instantaneously.

Two parents

After Probabilistic Crossover After Gene Mutation

Two offspring

(a) Recombination Operator (b) Mutation Operator

Fig. 4.8 Recombination and mutation operators.

Mutation is an operator taken from biology phenomenon that occurs in replica and

reproductive regeneration due to error (failure). Actually, three different deviations of

the copy from the original are usually distinguished in biology. The mutation is set in

accordance with gene mutation in biological genetics. A particular gene is changed and

60
may cause a deviation effects on organisms. Mutation is introduced after introducing

recombination operator. Naturally, individual as result of recombination process

whether it mutates or not, it depends on biological condition. Indicator for the mutation

in the optimization procedure was developed in accordance with a sampling of normally

distributed random. By taking value of mutation probability pm, which is called as

mutation rate, and initial order-string is ai, after introducing mutation operator, its state

may change or not, in mathematical correlation may be written as,

a if ξ i > p m
a m ,i =  i (4.6)
 f(ai ) if ξ i ≤ p m

where ξ i ∈ [0, 1] denotes a uniform random variable of sampling indicator; and f (ai )

is a function that generates mutation of ai in accordance with a defined rule.

Natural selection processes are very difficult to define precisely. Darwin, founder

of biology evolutionary theory, very informally introduced the term for selection.

Natural selection is no independent force of nature; it is a result of the competition of

natural organisms for resources. The selection means here as selecting of survival

individuals to be parent for a next generation. Preservation of favorable operator

emphasizes a selection mechanism using a probabilistic distribution of survival

individual in accordance with their fitness. More fitness individuals have higher chances

selected to be parent.

The selection is described as given in Fig. 4.9. Left side of the figure is an

example for fitness distribution of individuals in the selection process, and the right side

is probabilistic normal distribution density for 0 of mean and 1 of standard deviation. As

the fitness for thermodynamic models is commonly given in sum of weighted least

squares, thus mapping of the individual fitness in selection process is that smaller values

of fitness have larger probability values. The ranges of the probabilities for each

61
individual were assigned to make proportional with respect to their fitness.

140 0.45
0.4
120
0.35
100
0.3
Minimum Fitness

NDDP
Fitness

80 0.25
60 0.2
0.15
40 Maximum Fitness
0.1
20 0.05
0 0
1 2 3 4 5 6 7 8 9 10 -5 -4 -3 -2 -1 0 1 2 3 4 5
Individual Deviation

Fig. 4.9 Mapping fitness onto normal distribution density probability (NDDP).

The conversion rule was developed as following. Firstly, all survival individuals

into selection process are sorted in accordance with their fitness from better fitness

(smaller fitness values) to larger fitness and numbering from one, two, and so fourth as

Fit1, Fit2, Fit3, …and Fitn for population size of n. Since their characteristic is that better

individuals have smaller fitness value, thus the higher probability should be assigned to

the smaller fitness. The fitness for each individual is conversed into inverse-fit scale

(InvFit1, InvFit2, InvFit3,…and InvFitn) as,

n
∑ Fit j − Fit i
j =1
InvFiti = n
(4.7)
∑ Fit j
j =1

The inverse-fit scale is firstly normalized as,

InvFiti
InvFiti = n
(4.8)
∑ InvFit j
j =1

These values are adjusted to the selected probabilistic distribution density by taking its

maximum probability is Pbmax and its minimum probability is zero. The adjusted

probability scale Pb is proportionally assigned as,

62
 Pbmax
Pbi = n
InvFiti (4.9)
∑ InvFit j
j

i
Probability range of the first individual (Fit1) is assigned from Pbmax to ( Pbmax − ∑ Pbi ) .
j =1

i −1 i
The next individual with Fiti is assigned from ( Pbmax − ∑ Pbi ) to ( Pbmax − ∑ Pbi ) . The
j =1 j =1

i −1
last individual (Fitn) is assigned from ( Pbmax − ∑ Pbi ) to 0.
j =1

4.6.3 Convergence Ability

An illustration is given for objective function of ideal-gas isobaric specific heat

equation. Conventional combinatory method evaluating all possibilities of alternatives

for the equation is used as a comparison. The general pool consists of 30 genes and each

individual has chromosome with 4 genes. The mutation rate is set at 0.075. The main

population as a long process is shown in Fig. 4.10 and an example for generating

competitor is shown in Fig. 4.11. The equation of ideal-gas isobaric specific heat can be

written as Eq. (4.10) when all terms are included.

c op
= N1 X 1 + N 2 X 2 + N 3 X 3 + N 4 X 4 + N 5 X 5 + L + N 30 X 30 (4.10)
R

The X is representation of terms numbered from 1 to 30. The N is representation of

coefficients also numbered from 1 to 30. The X is selected in accordance with the

genetic optimization procedure, and the N is fitted to input data after the selected X

were determined. For the N that are not selected is automatically set to zero, just the N

of the selected X are fitted in the regression procedure.

All individuals existing after survival process is the individual with the same

fitness and gene-order after third regeneration, i.e., fitness of 0.000144266 and the

genes of 10, 21, 22, and 25. This is the same result as the conventional combinatorial

63
optimization procedure to find the global optimum. However, t-test of statistical

inference giving significant level for existing terms in the global optimal is 10th term at

11.5 %, 21st term at 20.5 %, 22nd term at 14.5 %, and 25th term at 8 %. Overall F-test for

the global optimal solution is less than 0.001 % for the significant level. This result

proven that for case of optimization procedure with statistical inference with setting of a

significant level for t-test at 5 %, the optimization will not able to find an optimal global

solution. The result agrees with constraints of stepwise selection that operates with basis

of statistical inference as confirmed by Setzmann and Wagner (1989) when they

compared their method with the stepwise selection for vapor-pressure equation.

Table 4.3 Combinatory Estimations for Calculations and Times

Individual Pool Combination Time1 Time2
18 30 8.64932 × 107 10 days 13.7 yrs
13 3
18 50 1.80535 × 10 5.724 × 10 yrs 2.862 × 103 yrs
18 100 3.06645 × 1019 9.724 × 109 yrs 4.862 × 1012 yrs
18 200 1.86134 × 1025 5.902 × 1015 yrs 2.951 × 1018 yrs
18 250 1.2144 × 1027 3.851 × 1017 yrs 1.925 × 1020 yrs
18 500 4.37166 × 1032 1.386 × 1023 yrs 6.931 × 1025 yrs
18 750 7.16916 × 1035 2.273 × 1026 yrs 1.137 × 1029 yrs
18 1000 1.33912 × 1038 4.246 × 1028 yrs 2.123 × 1031 yrs
4 30 2.7405 × 104 274 sec 1.586 days
4 50 2.30300 × 105 0.64 hrs 13.333 days
4 100 3.92122 × 106 0.4538 days 226.9 days
4 200 6.4685 × 107 7.4867 days 10.256 yrs
4 250 1.58883 × 108 18.389 days 25.191 yrs
4 500 2.57303 × 109 297.8 days 4.080 × 102 yrs
4 750 1.30784 × 1010 4.147 yrs 2.074 × 103 yrs
4 1000 4.14171 × 1010 13.133 yrs 6.567 × 103 yrs
1
One evaluation = 0.01 sec. (cpo, ps, etc.: 0.01 - 0.05 sec.)
2
One evaluation = 5 secs. (residual part of equation of state: 5 - 7 secs.)

64
 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

10 12 27 4 2 14 1 15 2 10 22 20 2 12 22 1 Initial
2.75177 5.65481 0.00399058 0.65453 parents
Recombination

15 2 10 14 15 13 14 10 20 12 22 1 22 2 12 10
0.0728181 0.273893 0.257718 0.021044
Mutation
15 2 10 14 27 13 14 10 20 12 22 1 22 25 21 10
0.0728181 0.23539 0.257718 0.000144266 Offspring

2 15 10 14 15 21 22 10 22 20 1 21 21 10 15 22 Competitor
0.00337221 0.000145255 0.0136336 0.000145255
Survival
2 15 10 14
0.00337221
15 21 22 10
0.000145255
22 20 1 21
0.0136336
22 25 21 10
0.000144266 I
Selection
22 25 21 10 15 21 22 10 22 25 21 10 15 21 22 10
0.000145255
Parents
0.000144266 0.000144266 0.000145255
Recombination
10 21 22 15 21 25 22 15 25 21 10 22 15 22 10 25
0.000145255 0.13577 0.000144266 0.278816
Mutation
10 21 22 15 21 25 22 3 25 21 10 22 15 22 10 25
0.000145255 0.112028 0.000144266 0.278816
Offspring

21 10 22 25 21 25 10 22 22 25 21 10 21 25 10 21
Competitor
0.000144266 0.000144266 0.000144266 0.000144266
Survival
21 10 22 25
0.000144266
21 25 10 22
0.000144266
25 21 10 22
0.000144266
21 25 10 21
0.000144266 II
Selection

Fig. 4.10 Convergence process of the main population in the optimization process.

Implementation of the conventional combinatorial optimization procedure needs

very much evaluation as shown in Table 4.3. In case of the conventional combinatorial

optimization procedure is applied, it needs 27405 evaluations. On the other hand, the

genetic procedure gives convergent result for global optimum at the third generation. As

each generation in the genetic procedure is 100 evaluations, thus the total evaluation is

200 evaluations.

65
 Initial Competitors

23 30 10 6 13 24 18 15 10 15 14 27 22 10 15 2 22 20 25 21 1 2 27 10
1.54742 2.27426 0.276913 0.0271861 0.182563 68.4547

23 30 24 13 6 30 24 18 10 15 13 30 6 15 30 18 15 14 22 10 15 10 2 27
0.784302 37.9328 0.239827 7.67431 0.277165 0.00337221

15 17 22 10 15 10 2 27
23 7 24 13 6 30 24 18 10 15 13 30 16 15 30 18
0.281328 0.00337221
0.626731 37.9328 0.239827 3.46795

Sort and Select

15 10 2 27 22 10 15 2 22 20 25 21 10 15 13 30 10 15 14 27 15 17 22 10
0.00337221 0.0271861 0.182563 0.239827 0.276913 0.281328

15 10 2 27 15 2 10 27 15 2 27 10 22 14 27 10 27 15 10 2 15 10 22 27
0.00337221 0.00337221 0.00337221 0.279463 0.00337221 0.279716

15 10 2 5 15 2 10 27 15 2 10 27 22 14 27 10 15 10 22 27
27 15 10 2
0.8622161 0.00337221 0.00337221 0.279463 0.279716
0.00337221

Sort and Select

15 2 10 27 15 10 2 27 15 2 10 27 27 15 10 2 22 10 15 2 22 20 25 21
0.00337221 0.00337221 0.00337221 0.00337221 0.0271861 0.182563

2 27 15 10 15 27 10 2 2 15 10 27 27 15 2 10 27 15 2 10 27 15 10 2
0.00337221 0.00337221 0.00337221 0.00337221 0.00337221 0.00337221

1 27 15 10 15 27 10 2 2 15 10 27 27 15 2 10 27 15 2 10 27 15 10 2
0.00337221 0.00337221 0.00337221 0.00337221 0.00337221 0.00337221

Sort and Select

15 2 10 27 15 10 2 27 15 2 10 27 27 15 10 2 1 27 15 10 15 27 10 2
0.00337221 0.00337221 0.00337221 0.00337221 0.00337221 0.00337221

Fig. 4.11 Example for generation competitor.

Consequence of the procedure involving probabilistic and random process as

engine of genetic operator, it should be noted that the optimization process is stochastic

66
process. It always gives different apparent result as long as optimization procedure, and

the convergent result may be at third generation, forth generation or other. Different

than the combinatorial optimization procedure ensuring 100 % to find global optimum,

the genetic procedure with probability of finding global optimum is less than 100 %.

This depends on gene pool and chromosome sizes. As shown in Table 4.3 is that the

combinatorial optimization procedure needs very long time for large gene pool size. It

looks no realistic way if the optimization process should be run more than several years.

Considering the combinatorial optimization procedure is not suitable for majority of the

problems with large predicted terms, the genetic procedure can help the works to find

qualified functional forms for thermodynamic empirical models.

4.6.4 Statistical Inference View

Statistical inference was originally applied in the methods such as forward

selection, backward selection, and stepwise selection methods. It should be noted that

these methods was proposed due to very much evaluation that should be conducted in

the conventional combinatory optimization (Draper and Smith 1981). Since the

evolutionary computation were proposed to avoid impossible optimization process with

whether random strategy optimization and conventional combinatory optimization, this

means both statistical and engineering approaches were proposed for the same propose.

Simply, significance of terms and total terms in equations are often concluded

from statistical analyses. However, these analyses cannot work alone. They need

practical values of the significant level for the judgments as shown in Fig. 4.12.

Significant level for either t-test or F-test is ratio of other area of the set to the total area

of their cumulative probability. This set will affect on the result of the equation. Method

for evaluation is also variation such as F-test for examples overall F-test, partial F-test,

67
or ratio F-test.

Although the t-test is not included for judgment in the optimization process, they

were evaluated as given in demonstration of the convergence process. As described in

Section 4.6.3, value for the significant level affects on the procedure to find the global

optimum. Other disadvantage is the evaluation of t-test enforces us to select the

regression method that consumes more time.

1 0.45
F (2,10)
0.9 F (3,35) F (6,10 0.4 t(∞)=N(0,1)
0.8 F (4,10) Judgment 0.35 t(8)
Judgment Judgment
0.7
0.3
prob(t)

F (3,5)
prob(n1,n2,F)

0.6
0.25 t(2)
0.5 Set Set Set
0.2
0.4
0.15
0.3
t(1)
0.2 0.1

0.1 0.05
0 0
0 1 2 3 4 -5 -4 -3 -2 -1 0 1 2 3 4 5
F t

Fig. 4.12 Statistical inference analyses from F-test (left) and t-test (right) distributions.

Especially for judgment of term number in thermodynamic model, the F-test may

be included. However, set of the significant level will affect on the term number.

Satisfied term number in thermodynamic model should able to represent

thermodynamic properties in accordance with experimental uncertainty of the data.

Therefore, the significant level should be adjusted to meet the thermodynamic

requirements such as uncertainty representation. Different quantities of thermodynamic

require different significant level values. This was demonstrated in development of

ancillary equations for saturation properties in Chapter 5.

68
 Chapter 5
Ancillary Equations for Saturation Properties of HFC Refrigerants

Saturation properties of vapor pressure, saturated-vapor and saturated-liquid

densities are accurately represented as function of temperature only besides they can be

represented in relations of fundamental equation of state. Ancillary equations for

saturation properties are required for development and implementation of fundamental

equation of state. The ancillary equations that can accurately represent the properties

and should keep a good connection of the properties in gaseous and liquid phases are

needed on the development. On the other hand, accuracy is not very important when

values of the properties are used as trial values for calculating thermodynamic

properties from fundamental equation of state.

The aim of the present study is to prepare empirical equations that are reliable to

reproduce saturation properties with the values close to the fundamental equation of

state. Nevertheless it can be used either to calculate thermodynamic properties directly

or to have a function as ancillary equations for initializing the trial values for

implementation of the fundamental equation of state. The ancillary equations are

simultaneously established from four HFC refrigerants in accordance with the genetic

optimization procedure.

5.1 Modeling Consideration

In principle, implementation of the optimization method for selecting terms in

case of simultaneous optimization for multi fluids is not strictly different from that for

single fluid. Regression process for a trial-selected term in the optimization method

69
needed for multi-fluid simultaneous optimization are several times of the single-fluid

optimization method. Therefore, the optimization process is accumulation of the

single-fluid optimization in accordance with fluid number. Detailed explanation for each

case was described in Section 4.6.

Each of the ancillary equations was developed in iteration process to make a good

connection between single-phases and two-phase as long as development of

fundamental equation of state. They need extrapolation data for the range where

available experimental data are scarce. These data are derived from the fundamental

equation of state. After developing all fundamental equations of state for the four HFC

refrigerants, reliable ancillary equations with the same functional form are then

established to find a good consistency with the fundamental equations of state. Optimal

solution for the ancillary equations was established in accordance with the genetic

optimization procedure.

5.2 Thermodynamic Formulation of Ancillary Equations

Ancillary equations for saturation properties are given as functions of temperature

and the validity is a range from the triple point to the critical point. They consist of the

vapor pressure, saturated-liquid density and saturated-vapor density equations. The

equations play important role on development and implementation of fundamental

equation of state. The equations are needed on the development to produce the vapor

pressures and saturated-liquid and saturated-vapor densities for the same temperatures

from the triple to the critical point. In practice, experimental data satisfying the

requirements are difficult to be found. Therefore, the ancillary equations are repeatedly

revised on the basis of the best of temporal fundamental equations of state as long as the

development.

70
5.2.1 Formulation of Vapor-Pressure Equation

There are various functional forms that were applied for vapor-pressure equation.

The oldest form is Antoine equation that was proposed in 1888. Most of the

vapor-pressure equations are developed from approaches to represent high accuracy of

vapor-pressure measurements with high accuracy and to satisfy behavior of critical

point (Ambrose et al. 1972; Wagner 1973; Iglesias-Silva et al. 1987; Shaver et al. 1991).

Corresponding states were also used to develop the equation for predicting vapor

pressure (Thomson et al. 1982). By considering empirical vapor-pressure equation with

including non-analytic of critical point, the general form of the equations can be written

as,
t t2i
p   T  1i  T 
ln s  = ∑ N i  c  1 −  (5.1)
p
 c i T
   T c 

where t1 and t2 are predicted parameters of the equation; T and Tc are temperature and

critical temperature.

5.2.2 Formulation of Saturated-Liquid Density Equation

As a similar way with the vapor-pressure equation, saturated-liquid density

equation was developed with either empirical or corresponding state approaches. Most

of the corresponding state included critical parameters and acentric factors (Lyckman et

al. 1965; Racket 1970; Spencer and Danner 1972; Hankinson and Thomson 1979).

Empirical approaches may consider rectilinear diameter rule to represent the

saturated-liquid density from the triple point to the critical point (Hall and Eubank 1976;

Iglesias-Silva and Hall 1997). Corresponding state was also used to represent the

saturated-liquid density (Shaver et al. 1992). The saturated-liquid density equation was

71
developed with general functional form as,
t
 ρ'   T i
    (5.2)
 ρ − 1 = ∑i N i 1 − T 
 c   c 

where t is predicted parameters of the equation; T and Tc are temperature and critical

temperature.

5.2.3 Formulation of Saturated-Vapor Density Equation

As a different way from that for the saturated-liquid density, saturated-vapor

density equation is usually developed as an empirical equation. Both non-analytical

behavior and rectilinear density diameter can be considered in the equation. The

saturated-vapor density equation was formulated by taking a functional form as,

t
 ρ"   T i
ln  = ∑ N i 1 −
 

 (5.3)
 ρc  i  Tc 

where t is predicted parameters of the equation; T and Tc are temperature and critical

temperature.

5.3 Ancillary Equations for Saturation Properties

In comparison with the multi-property optimization of fundamental equation of

state, development of ancillary equation for the saturation properties is very simple.

Critical parameters used to reduce parameters of each equation are listed in Table 5.1.

These values also used for the development of four fundamental equations of state in

this study.

Besides the genetic optimization procedure, statistical inference was also used to

revealing the result. The F-test was calculated in accordance with an evaluation method

using partial F-test that ratio of the overall F-test for regression with different term

number. Significant level for the F-test was adjusted to make uncertainty of the ancillary

72
equations agreeing with uncertainty of the experimental data. Evaluation can be based

on the null hypothesis. The t-test was calculated to reveal this indicator from statistical

inference aspect. Predicted terms for the ancillary equations were introduced with

numerical exponents in terms from 0.1 to 10 with steps of 0.1.

Table 5.1 Critical Parameters of HFC Refrigerants*
Refrigerant R-32 R-125 R-134a R-152a
Tc, K 351.255 339.165 374.083 386.41
pc, MPa 5.780 3.6175 4.048 4.516
-3
ρ c , kg·m 424 568 509 368
*
Critical parameters were selected from several literatures: R-32 and R-125 from
Kuwabara et al. (1995); R-134a from Aoyama et al. (1996); and R-152a from Higashi et
al. (1987). Critical pressures are results of extrapolation of vapor-pressure equations.

5.3.1 Vapor-Pressure Equations

Simultaneous optimization of vapor-pressure equations for four HFC refrigerants

was conducted in accordance with the genetic optimization procedure. Total of effective

terms in equations was determined at significant level for F-test less than 0.5 %. The

equations consist of 4 terms and have a same functional form as,

1 .5 2 .2 4 .5
ps T  T  T  T  Tc  T  Tc  T  (5.1)
ln = N 1 c  1 −  + N 2 c  1 −  + N3  1 −  + N4  1 − 
pc T  T c  T  T c  T  T c  T  T c 

Numerical coefficients for each refrigerant are listed in Table 5.2.

Table 5.2 Numerical Coefficients of Vapor-Pressure Equation (Eq. 5.1)

Fluid N1 N2 N3 N4

R-32 -7.462789 1.886120 -1.644908 -2.876626

R-125 -7.513614 1.923414 -2.116034 -3.569228
R-134a -7.650442 1.995461 -2.206406 -3.662602
R-152a -7.425969 1.872947 -1.713556 -3.218394

Most of the terms in the vapor-pressure equations are assessed from statistical

73
inferences for each fluid. Each term in the vapor-pressure equations has significant level

for t-test less than 5 % expect fourth term of the equation for R-32 at 9 %, first term of

the equation for R-125, first and fourth terms of the equation for R-134a at 7 %, and

first and fourth terms of the equation for R-152a at 7 %.

5.3.2 Saturated-Liquid Density Equations

Saturated-liquid density equations of four HFC refrigerants were simultaneously

established in accordance with the genetic optimization procedure. Total of effective

terms in the equations was determined at significant level for F-test less than 2.5 %. The

equations consist of 4 terms and have a same functional form as,

0 .1 0 .2 0 .3 1 .4
ρ'  T   T   T   T 
− 1 = N 1 1 − 
 + N 2 1 − 
 + N 3 1 − 
 + N 4 1 − 
 (5.2)
ρc  Tc   Tc   Tc   Tc 

Numerical coefficients depending on refrigerants are listed in Table 5.3.

Table 5.3 Numerical Coefficients of Saturated-Liquid Density Equation (Eq. 5.2)

Fluid N1 N2 N3 N4

R-32 1.1074151 -4.166318 5.580605 0.5412373

R-125 0.3097022 -1.704846 3.635670 0.6226037
R-134a 0.3963097 -1.940054 3.827410 0.6387256
R-152a 0.7908671 -3.158724 4.735456 0.5621207

Result of t-test for each term of the saturated-liquid density equations has modus

of significant level values larger than the vapor-pressure equations. The modus values

are less than 10 % except all terms of the R-32 at 15 %, third term of the R-125 at 12 %,

the third and fourth terms of the R-134a at 12 to 14 %, and third and forth terms of the

R-152a at 11 %.

74
5.3.3 Saturated-Vapor Density Equations

Saturated-vapor density equations of four HFC refrigerants were simultaneously

developed with genetic optimization procedure. Total of effective terms in the equations

was determined at significant level for F-test less than 5 %. The equations consist of 5

terms and have a same functional form as,

0 .1 0 .3 0 .8 2 .4 6 .1
ρ"  T   T   T   T   T 
ln = N 1  1 − 
 + N 2  1 − 
 + N 3  1 − 
 + N 4  1 − 
 + N 5  1 − 

(5.3)
ρc  Tc   Tc   Tc   Tc   Tc 

Numerical coefficients for each refrigerant are listed in Table 5.4.

Table 5.4 Numerical Coefficients of Saturated-Vapor Density Equation (Eq. 5.3)

Fluid N1 N2 N3 N4 N5

R-32 -0.07414063 -1.405255 -5.349276 -14.20536 -57.69994

R-125 0.01630809 -1.578731 -4.937218 -15.81130 -61.86889
R-134a 0.03883292 -1.676962 -4.990296 -16.12274 -63.48467
R-152a -0.006995113 -1.533727 -5.137617 -14.49613 -59.19053

Each term in the saturated-vapor density equations has significant level for t-test

less than 10 % expect second term of the equation for R-32 at 15 %, second, third and

fifth terms of the equation for R-125 at 11 to 14 %, second term of the equation for

R-134a at 16 %, and second term of the equation for R-152a at 16 %.

5.4 Assessment of Ancillary Equations

Comparisons of experimental data with values calculated from the ancillary

equations of this study are very useful to judge the reliability. The values from

fundamental equations of state that were developed in the study (Chapter 6) are

compared to disclose consistencies of the ancillary equations. Saturation condition

constraints including Maxwell criterion are simultaneously included to calculate the

75
saturation properties from the fundamental equation of state. The values from the

ancillary equations serve as baselines to represent deviations of the saturation

properties.

5.4.1 Assessment of the Vapor-Pressure Equations

Sets of experimental data are used to disclose accuracies of the vapor-pressure

equations. Deviations of vapor pressures from the vapor-pressure equation for the four

HFC refrigerants are shown in Figs. 5.1 to 5.4. Those figures show deviations of vapor

pressures for R-32, R-125, R-134a, and R-152a, respectively. On the other hand,

statistical results for experimental data from the vapor-pressure equations are

represented in Tables C.1 (Appendix C). Most reliable experimental data for moderate

and high temperatures can be accurately reproduced. It is noted here that data of

Tillner-Roth (1996) is not experimental, but calculated data from analysis of the

evaporation heat and vapor pressure.

0.3
R-32
0.2
Deviation of ps , %

0.1
0
-0.1
-0.2
-0.3
125 150 175 200 225 250 275 300 325 350
T, K

Fig. 5.1 Deviation of measurements from the vapor-pressure equation for R-32. (□) de
Vries (1997); (◊) Magee (1996); (∆) Defibaugh et al. (1994); (+) Holcomb et al. (1993);
(×) Fu et al. (1995); (—○—) Tillner-Roth (1996); (----) The new fundamental equation
of state.

76
 Deviation of vapor pressure data from the vapor-pressure equation for R-32 is

shown in Fig. 5.1. Good agreement values with the fundamental equation of state can be

reproduced, except for the values near critical temperatures. Most experimental data can

be represented with uncertainty of 0.05 %. The data of de Vries (1997) can be

represented from the lower temperatures to critical point. The data at low temperature

by Magee (1996) are also adequately reproduced. Especially for lower temperatures, the

values can be reproduced with acceptable deviation from the fundamental equation of

state. Larger deviations at lower temperatures are reproduced in case of comparison

with derived data by Tillner-Roth (1996).

Typical experimental vapor-pressure data for R-125 have larger variances, not

only near critical point but also intermediate higher and lower temperatures. The vapor

pressure equation for R-125 can represent the data mostly in deviation of 0.1 % as

shown in Fig. 5.2. Deviation of the equation to reproduce values from the fundamental

equation of state is less than 0.05 %.

0.3
R-125
0.2
Deviation of ps , %

0.1
0
-0.1
-0.2
-0.3
170 195 220 245 270 295 320 345
T, K

Fig. 5.2 Deviation of measurements from the vapor-pressure equation for R-125a. (□)
de Vries (1997); (◊) Magee (1996); (∆) Boyes and Weber (1995); (+) Duarte-Garza et al.
(1997); (×) Sagawa et al. (1994); (—○—) Tillner-Roth (1996); (----) The new
fundamental equation of state.

77
 Furthermore, the vapor pressure equation for R-125 is assessed with derived data.

The derived data of Tillner-Roth (1996) at lower temperatures may be contained worse

uncertainties. These results can be understood from the discrepancy of the existing data.

For low temperatures, large discrepancy for data of Magee (1996) to de Vries (1997) is

larger than 0.1 % of relative deviation in worst case. However, small discrepancies with

respect to the values from the fundamental equation of state can be performed.

Revealing of these all, the vapor-pressure equation is still reliable to calculate the

properties for engineering interests.

0.3
R-134a
0.2
Deviation of ps , %

0.1
0
-0.1
-0.2
-0.3
150 175 200 225 250 275 300 325 350 375
T, K

Fig. 5.3 Deviation of measurements from the vapor-pressure equation for R-134a. (∆)
Magee and Bowley (1992); (∆) Boyes and Weber (1995); (×) Goodwin et al. (1992); (□)
Weber (1989); (◊) Baehr and Tillner-Roth (1991); (—○—) Tillner-Roth (1996); (----)
The new fundamental equation of state.

Vapor pressures of R-134a are accurately represented by the vapor-pressure

equation as more detailed illustration in Fig. 5.3. It is not only very small deviation of

vapor pressures from the fundamental equation of state but also from most reliable

experimental data. Derived data of Tillner-Roth might not be good to consider due to

high uncertainty of data used in the calculation method. All data of Baehr and

Tillner-Roth (1991) are well reproduced including the critical region. Representation of

78
the data by Goodwin et al. (1992) is slightly inaccurate due to uncertainty of the data.

Most reliable experimental data can be reproduced with uncertainty less than 0.05 %.

0.3
R-152a
0.2
Deviation of ps , %

0.1
0
-0.1
-0.2
-0.3
125 150 175 200 225 250 275 300 325 350 375 400
T, K

Fig. 5.4 Deviation of measurements from the vapor-pressure equation for R-152a. (□)
Duarte-Garza and Magee (1996); (∆) Higashi et al. (1987); (+) Holcomb et al. (1993);
(×) Kamei et al. (1989); (◊) McLinden (1989); (-) Silva and Weber (1993); (—○—)
Tillner-Roth (1996); (●) Blanke and Weiss (1992); (----) The new fundamental equation
of state.

The last of the vapor-pressure equations is an equation for R-152a. Assessment

result of the equation is shown in Fig. 5.4. In comparison with the other vapor pressure

equations, its uncertainty is slightly larger than the others. The available experimental

data may cause this disadvantage of larger discrepancy values from the lowest

temperature to highest temperature of the available experimental data. At low

temperatures, deviation of the vapor pressures with respect to values from the

fundamental equation of state is less than 0.05 %. However, a good agreement with the

data of Duarte-Garza and Magee (1996) is performed.

5.4.2 Assessment of the Saturated-Liquid Density Equations

Deviations of saturated-liquid densities from the saturated-liquid density equations

as a simultaneous optimization result for the four HFC refrigerants are shown in Figs.

79
5.5 to 5.8. Statistical results are listed in Table C.2 (Appendix C). It differs from the

vapor pressures that the saturated-liquid densities can be generally reproduced with

lower accuracies. These accuracies relate to uncertainty of the experimental data.

Saturated-liquid density is more difficult than the vapor pressure to be measured.

0.9
R-32
0.6
Deviation of ρ ', %

0.3
0
-0.3
-0.6
-0.9
130 155 180 205 230 255 280 305 330 355
T, K

Fig. 5.5 Deviation of measurements from the saturated-liquid density equation for R-32.
(□) Magee (1996); (○) Fukushima (1995); (×) Holcomb et al. (1995); (+) Kuwabara et
al. (1995); (----) The new fundamental equation of state.

0.9
R-125
0.6
Deviation of ρ ', %

0.3
0
-0.3
-0.6
-0.9
170 195 220 245 270 295 320 345
T, K

Fig. 5.6 Deviation of measurements from the saturated-liquid density equation for
R-125. (◊) Magee (1996); (×) Takahashi et al. (1994); (∆) Kuwabara et al. (1995); (□)
Fukushima et al. (1995); (----) The new fundamental equation of state.

Experimental data were used to assess the saturated-liquid density equation for

R-32. Graphical result of the assessment result is shown in Fig. 5.5. Especially for data

80
with higher temperatures and near the critical point are reproduced with larger

uncertainties. The uncertainty of 0.2 % can be reproduced for intermediate and lower

temperatures. Good agreement values with the fundamental equation of state are also

performed. The data of Magee (1996) at low temperatures can be well reproduced.

Especially for the experimental data with temperatures by the critical point, larger

deviations are appeared.

0.9
R-134a
0.6
Deviation of ρ ', %

0.3
0
-0.3
-0.6
-0.9
150 180 210 240 270 300 330 360 390
T, K

Fig. 5.7 Deviation of measurements from the saturated-liquid density equation for
R-134a. (□) Yokoyama and Takahashi (1991); (∆) Hou et al. (1992); (○) Higashi (1994);
(----) The new fundamental equation of state.

Representation ability of the saturated-liquid density equation for R-125 is shown

in Fig. 5.6. A good agreement result with the values from the fundamental equation of

state can be performed, except for temperatures near the critical point. Most

experimental data are reproduced with deviation less than 0.2 %. The data of Magee

(1996) at lower temperatures can be well reproduced. The data of Kuwabara et al.

(1995) for most the data near the critical point can be adequately represented.

Saturated-liquid density of R-134a can be represented as result shown in Fig. 5.7.

Deviation of the density with respect to most reliable experimental data is performed

within 0.2 %, except slightly larger for the density with temperatures near the critical

81
point. The deviation from the fundamental equation of state is acceptable values even

though in the region near the critical point is still inaccurate representation. The data of

Hou et al. (1992) are well reproduced in wide range of temperatures. Yokoyama and

Takahashi (1991) are also adequately reproduced. In accordance with usual uncertainty

at critical point, the large discrepancies are also appeared at the critical region.

0.9
R-152a
0.6
Deviation of ρ ', %

0.3
0
-0.3
-0.6
-0.9
150 180 210 240 270 300 330 360 390
T, K

Fig. 5.8 Deviation of measurements from the saturated-liquid density equation for
R-152a. (□) Geller et al. (1979); (×) Magee (1998); (◊) Blanke and Weiss (1992); (∆)
Holcomb et al. (1993); (+) Kamei et al. (1989); (----) The new fundamental equation of
state.

Saturated-liquid density equation of R-152a was assessed with several set

experimental data and the fundamental equation of state as shown in Fig. 5.10.

Deviation of the density with respect to most reliable experimental data is reproduced

within 0.2 %, except slightly larger for the density with temperatures near the critical

point. The deviation from the values from the fundamental equation of state is also

acceptable values. Especially for low temperatures, the equation represents the densities

in mean of values of the experimental data and close values to data of Magee (1998).

The data of Blanke and Weiss (1992) can be reproduced more accurately than other

available data for high temperatures.

82
5.4.3 Assessments of the Saturated-Vapor Density Equations

The third group of the ancillary equations is saturated-vapor density equations.

These assessment results are shown in Figs. 5.9 to 5.12. Their statistical results are

represented in Table C.3 (Appendix C). Different than the vapor pressures and

saturated-liquid densities, experimental data for saturated-vapor density for each of

HFC refrigerants are very scarce and difficult to be find for lower temperatures.

The saturated-vapor density equation for R-32 was assessed by comparison with

experimental data and the fundamental equation of state. Experimental data near critical

temperatures are reproduced with lower accuracies. Few of the experimental data are

available at lower temperatures. A set of data by Holcomb et al. (1995) is adequately

reproduced. A good agreement of the densities can be achieved with the values from the

fundamental equation of state. This representation can be excused due to limited

experimental data that may be caused by difficulty of measurement. Assessment result

of the equation is shown in Fig. 5.9. Consequence of comprehensive assessments in the

fundamental equation of state included in the development, agreement with the values

of the fundamental equation of state indicates that the saturated-vapor equation can be

applied for low temperature even though the experimental data are not available.

Several sets of experimental data were used to assess the saturated-vapor density

equation for R-125. The reliable data can be represented within 0.2 % of deviation.

Good agreement values with the fundamental equation of state can be achieved. The

deviation of density at temperatures near the critical point slightly increases. More

detailed result is shown in Fig. 5.10. Similar situation for these data with other fluid, it

is very scarce data at low temperatures. Moreover, the data with temperatures lower

than 270 K is not available. However, the equation is also reliable to be applied for low

83
temperatures due to a good agreement with the values from the fundamental equation of

state.

2.4
R-32
1.8
Deviation of ρ" , %

1.2
0.6
0
-0.6
-1.2
-1.8
-2.4
130 160 190 220 250 280 310 340
T, K

Fig. 5.9 Deviation of measurements from the saturated-vapor density equation for R-32.
(◊) Defibaugh et al. (1994); (○) Kuwabara et al. (1995); (□) Holcomb et al. (1995); (----)
The new fundamental equation of state.

2.4
R-125
1.8
Deviation of ρ" , %

1.2
0.6
0
-0.6
-1.2
-1.8
-2.4
170 190 210 230 250 270 290 310 330
T, K

Fig. 5.10 Deviation of measurements from the saturated-vapor density equation for
R-125. (◊) Boyes and Weber (1995); (□) Kuwabara et al. (1995); (∆) Fukushima et al.
(1995); (----) The new fundamental equation of state.

The saturated-vapor density equation of R-134a can represent density of most

reliable experimental data with deviation less than 0.5 %, except slightly larger for the

density data with temperatures near the critical point. Although the measurements of

this fluid have been started earlier than R-32, the data is also scarcely available. A good

84
agreement with the fundamental equation of state indicates that it satisfies the Maxwell

criterion at saturation. The deviation from the fundamental equation of state is still

acceptable as shown in Fig. 5.11, not only at lower temperatures but also for higher

temperatures without any exception including critical region.

2.4
R-134a
1.8
Deviation of ρ" , %

1.2
0.6
0
-0.6
-1.2
-1.8
-2.4
165 190 215 240 265 290 315 340 365 390
T, K

Fig. 5.11 Deviation of measurements from the saturated-vapor density equation for
R-134a. (□) Niessen et al. (1994); (∆) Aoyama et al. (1996); (◊) Weber (1989); (----)
The new fundamental equation of state.

2.4
R-152a
1.8
Deviation of ρ" , %

1.2
0.6
0
-0.6
-1.2
-1.8
-2.4
150 175 200 225 250 275 300 325 350 375
T, K

Fig. 5.12 Deviation of measurements from the saturated-vapor density equation for
R-152a. (□) Kamei et al. (1989); (◊) Holcomb et al. (1993); (∆) Higashi et al. (1987);
(×) Wang et al. (1992); (----) The new fundamental equation of state.

The last one of the saturated-vapor density equations is an equation for R-152a.

85
This equation represents the density deviation with slightly large as shown in Fig. 5.12.

Discrepancy of experimental data causes development of model to be difficult. It is not

only for developing fundamental equation of state, but also for developing this ancillary

equation. Among available experimental data, just a set of data until low temperatures

of Kamei et al. (1989) is available. Since these data can be accurately fitted by neither

the fundamental equation of state nor the saturated-vapor density equation, the ancillary

equation can reproduce mean values of the fundamental equation and experimental data.

The deviation from the fundamental equation of state is still in acceptable values and the

adequate accuracy can be maintained.

86
 Chapter 6
Fundamental Equations of State for HFC Refrigerants

Among thermodynamic equations of state with validity from the triple point to

a certain highest temperature and pressure covering liquid and gaseous phases and

saturation, Helmholtz equation of state is the most reliable model. The Helmholtz

equation of state is represented in Helmholtz free energy to give the state of

thermodynamic properties in terms of density and temperature for fluids. In this chapter,

the fundamental equations of state are introduced by four Helmholtz equations of state

for HFC refrigerants (R-32, R-125, R-134a, and R-152a) as results of this development.

Assessments of the Helmholtz equations of state are described, including their

accuracies for reproducing PVT and caloric properties in the gaseous and liquid phases

and saturation. Critical parameters were also assessed from critical constraints to derive

their values from the Helmholtz equations of state.

6.1 Important Parameters

Main properties of fluids such as molecular mass, gas constant, critical density,

triple point, and critical temperature are often used to reduce relations of

thermodynamic properties. Dimensionless of thermodynamic properties is usually

applied to avoid unit errors in both thermodynamic development and implementation of

the fundamental equations of state. The gas constant that was used to reduce

thermodynamic relations of thermodynamic properties is 8.314472 J·mol-1·K-1 (Mohr

and Taylor 1999). Critical density and temperature were used to reduce independent

variables (density and temperature) of the fundamental equation of state to be

87
dimensionless. The main properties for the HFC refrigerants are listed in Table 6.1. The

triple point constrains the lowest validity range for the fundamental equations of state.
Table 6.1 Main Properties of HFC Refrigerants*
-3
Refrigerant M, kg·mol-1 Tc, K pc, MPa ρ c , kg·m Ttr, K
R-32 0.052023 351.255 5.780 424 136.34
R-125 0.120022 339.165 3.6175 568 172.54
R-134a 0.102031 374.083 4.048 509 169.85
R-152a 0.066050 386.41 4.516 368 154.56
*
Critical parameters are the same as values used for the ancillary equations (Table 5.1).
Triple temperatures: R-32 and R-125 (Lüddecke and Magee 1996); R-134a (Magee
1992); and R-152a (Magee 1998).

6.2 Thermodynamic Formulation of Fundamental Equation of State

Complexity of tasks on development of fundamental equation of state is caused by

input data, property relations, and wide-range thermodynamic surface included in

optimization. Thermodynamic formulation for the fundamental equation of state is

made it suitable with steps and constraints of the development. Ideal-gas law is included

in the fundamental equation of state by dividing it into an ideal-gas part and a residual

part. Three steps are needed in the development, i.e., the first is to establish an ideal-gas

part, the second is to establish a residual part, and the last is to fit the fundamental

equation of state to the reference value selected as a reference state. The last step

contributes on values of entropy, internal energy or enthalpy calculated from the

fundamental equation of state. Different values selected for the reference state will not

influence on the quantity of change of the properties from a state to other state.

6.2.1 Formulation of Ideal-Gas Part

Ideal-gas isobaric specific heat of a fluid can be expressed as a function of

temperatures. General form of this function is usually composed with introducing

88
polynomial terms and Planck-Einstein terms that both are function of temperatures.

Both terms were considered on development of ideal-gas isobaric specific heat. The

form can be written as,

te

Φ=τ [ ]
ηo
tl  η o 2 exp(−η oτ ) 
τ 2  (6.1)
(
 1 − exp(−η oτ ) 2 
  )
where tl, ηo, and te are predicted parameters of terms for the equation. The isobaric

specific heat equation is optimized with selecting terms and fitting the equation to the

ideal-gas isobaric specific heat data. General form of the equation can be written as,
c op
= ∑ N 'j Φ j (6.2)
R j

Therefore, the optimization function becomes as,

2
 c op 
'
∑ Wi  − ∑ N j Φ j  (6.3)
i  R j 
i

The ideal-gas part can be established from integral operation of the ideal-gas isobaric

specific heat equation and consistence with ideal-gas law. Detailed framework of this

operator is given in Appendix B. General form of this result can be written as,
o n
{
α o (δ , τ ) = ln δ + N io ln τ + ∑ N ioτ ηi + ∑ N io ln 1 − exp − η ioτ
i i
( )} (6.4)

Since the equation was established empirically, the Planck-Einstein terms are not related

to the fundamental frequency of each fluid at ideal-gas state.

6.2.2 Formulation of Residual Part

The residual part is composed of several terms. General form for each term of the

residual part is determined as,

(
Ψ = δ dl τ tl exp − m d (δ − s d )d e − mt (τ − s t )te ) (6.5)

where dl, tl, md, sd, de, mt, st, and te are predicted parameters of the terms for the equation.

89
Therefore, the residual part can be written as,

α r (δ ,τ ) = ∑ N i Ψi (6.6)
i

By taking this form, all the existing fundamental equations of state for pure fluids not

only HFC refrigerant but also other fluid such as nitrogen, carbon dioxide, etc. can be

represented with the above general form.

Total of sum of weighted least squares for each of thermodynamic properties can

be considered to judge quality of the residual part. Each relation of the sum of weighted

least squares are called as objective function of regression. Although a certain number

of the objective functions are included in the optimization, a thermodynamic model

would be established in this development. This agrees with thermodynamic relation that

all thermodynamic properties relate to Helmholtz free energy.

The objective functions are simultaneously optimized by multi-property regression

to establish the residual part. Non-linear relations such as speed of sound, isobaric

specific heat, and saturated-liquid specific heat can be linearized as constants that can be

evaluated in iteration process to find an accurate result. Initial values for the constants

are firstly evaluated from the temporal best solution in the optimization process.

As shown in Table 6.2, independent parameters (density and temperature) and

dependent parameters are introduced as input data in accordance with quantities of

thermodynamic properties. All thermodynamic properties are possibly included in

development of fundamental equation of state. The properties included in development

of the residual parts of this study were described in Section 4.3. PVT properties both in

single phase and at saturation can be represented as the relations shown in first, second

and third rows. The forth row is the relation for saturation with consideration of

Maxwell criterion. A set of PVT data at saturation line is needed from the triple point to

90
the critical point to consider the Maxwell criterion. It is similarly needed for PVT at

saturation i.e., the second and third relations in the table. Especially for the relation of

the Maxwell criterion, the same temperatures for PVT data at saturated liquid and

saturated vapor are needed.

Table 6.2 Objective Functions in Multi-Property Regression

Data Relations for sum of weighted least squares for each property
2
p, ρ , T  p 
∑ Wi  − 1 − δα δr 
i  ρRT i

p s , ρ ' , Ts 2
 ps 
∑ Wi  − 1 − δ ' α δr '
i  ρ ' RT s i

p s , ρ " , Ts 2
 ps 
∑ Wi  − 1 − δ "α δr "
i  ρ " RTs i

ps , ρ' , ρ", Ts 2
 p  1 1 δ' 
∑ Wi  s  −  − ln − (α r '−α r " ) , Maxwell relation for saturation
i  RTs  ρ " ρ '  δ " i

2
B, T
∑ Wi  Bρ c − δlim
→0
α δr 

i i
2
C, T
∑ Wi Cρ c − δlim r 
2
α δδ
i →0  i
2
cv , ρ , T c
( 
∑ Wi  v + τ 2 αττo + α ττr  )
i R i

c p , ρ,T 2
 c p c (δ ,τ ) (1 + δα δr − δτα δτ
r 2 
)
∑ Wi  − v − r 2 r 
i  R R (1 + 2δα δ + δ α δδ )  i

w, ρ , T 2
 w2 M (1 + δα δr − δ τα δτ r 2
)
∑ Wi  − 1 − 2δα δr − δ 2α δδ
r
− 
i  RT c v (δ , τ ) / R  i

cs ' , ρ' , ρ", Ts 2

 c ' c (δ ' ,τ s ) (1 + δ ' α δr '−δ 'τ sα δτ
r
')  r r dp s (δ ' , δ " ,τ s ) 
∑ Wi  s − v − 1 + δ ' α δ '−δ 'τ sα δτ '− 
i  R R r 2 r
(1 + 2δ ' α δ '+δ ' α δδ ' )  Rρ c δ ' dT 
i

91
 Table 6.2 Objective Functions in Multi-Property Regression (continued)
Data Relations for sum of weighted least squares for each property

cs ", ρ' , ρ", Ts 2

 c " c (δ " ,τ s ) (1 + δ "α δr "−δ "τ sα δτ
r
")  r r dp s (δ ' , δ " ,τ s ) 
∑ Wi  s − v − 1 + δ "α δ "−δ "τ sα δτ "− 
i  R R r 2 r
(1 + 2δ "α δ "+δ " α δδ " )  Rρ c δ " dT 
i

ρ c , Tc
[ ] ∂p
2
Wi 0 − 1 − 2δα δr − δ 2α δδ
r
, = 0 at the critical point
i ∂v

ρ c , Tc
[ ] ∂2 p
2
Wi 0 − 2 − 6δα δr − 6δ 2α δδ
r
− δ 3α δδ
r
, = 0 at the critical point
i ∂v 2
2
u1 , ρ1 , T1 ,  u u o r o r 
∑ Wi  2 − 1 − τ 2 (α τ 2 + α τ 2 ) + τ 1 (α τ 1 + α τ 1 )
i  RT2 RT1 i
u2 , ρ 2 , T2

2
h1 , ρ1 , T1 ,  h h o r r o r r 
∑ Wi  2 − 1 − τ 2 (α τ 2 + α τ 2 ) − (1 +δ 2α δ 2 ) + τ 1 (α τ 1 + α τ 1 ) + (1 + δ 1α δ 1 )
i  RT2 RT1 i
h2 , ρ 2 , T2

2
s1 , ρ1 , T1 , s − s o r o r o r o r 
∑ Wi  2 1 − τ 2 (α τ 2 + α τ 2 ) + (α 2 + α 2 ) + τ 1 (α τ 1 + α τ 1 ) − (α 1 + α 1 )
i  R i
s2 , ρ 2 , T2

Abbreviations:

 ∂α   ∂α   ∂ 2α   2   2 
 , α ττ =  ∂ α  , α δτ =  ∂ α , δ = ρ , τ = c
T
αδ =   , ατ =   , α δδ =  2   ∂τ 2   ∂δ∂τ 
 ∂δ τ  ∂τ  δ  ∂δ τ  δ   ρc T

The second and third virial coefficients are listed as the relations in fifth and sixth

rows. Both these properties have independent variables of temperature only. Isochoric

specific heat, isobaric specific heat, and speed of sound have the objective function as

given in seventh, eighth and ninth rows, respectively. These relations can be expanded

to saturation either saturated-liquid or saturated-vapor states as implemented in the PVT

relation. The tenth and eleventh rows represent the relation of saturation specific heat

along saturation line when the properties are evaluated at saturated-liquid and

saturated-vapor states, respectively. More detailed meaning of this specific heat is

described in Appendix A (Section A.2).

92
 It should be noted that internal energy, enthalpy, and entropy are the

thermodynamic properties that cannot be measured. These properties can be derived

from thermodynamic relations. Since the relations for these properties contain integral

constants, thus their values depend on the reference state. Therefore, including these

properties by directly considering their relations into regression cannot be conducted.

Alternatively, their regression relations can be derived from the change of their

properties from one state to another state. The relations are listed in the three last rows

of regression relation in Table 6.2.

6.3 Fundamental Equations of State

The four fundamental equations of state for HFC refrigerants were established

with a reference state of the thermodynamic properties for specific entropy and enthalpy

of 1 kJ·kg-1·K-1 and 200 kJ·kg-1 at the saturated liquid of 273.15 K. This reference state

was selected in accordance with a convention recommended by International Institute of

Refrigeration. Each of the fundamental equations of state is represented into an

ideal-gas part and a residual part as previously described in their formulation.

6.3.1 Ideal-Gas Part

Each ideal-gas part of the four fundamental equations of state was established by

integration of ideal-gas isobaric specific heat equation. Each of the specific heat

equations was established by selecting terms with the genetic optimization procedure

and fitting to reanalysis data of Yokozeki et al. (1998). The ideal-gas parts for the four

HFC refrigerants were then established from the specific heat equations. Each of the

ideal-gas parts is given in Eqs. (6.7) to (6.10) for R-32, R-125, R-134a, and R-152a,

respectively. Their numerical parameters are listed in Table 6.3

93
 6
{ (
α o (δ , τ ) = ln δ + N 0o + N 1oτ + N 2o ln τ + ∑ N io ln 1 − exp − η ioτ
i =3
)} (6.7)

5
{
α o (δ , τ ) = ln δ + N 0o + N 1oτ + N 2oτ 0.25 + N 3o ln τ + ∑ N io ln 1 − exp( −η ioτ )
i=4
} (6.8)

δ  5
{ (
α o (δ , τ ) = ln  + N 0o + N 1oτ + N 2oτ −0.25 + ∑ N io ln 1 − exp − η ioτ )} (6.9)
τ  i =3

δ  5
{
α o (δ , τ ) = ln  + N 0o + N 1oτ + N 2oτ −0.5 + N 3oτ 0.25 + ∑ N io ln 1 − exp − η ioτ ( )} (6.10)
τ  i =4

Table 6.3 Numerical Parameters of Each of the Ideal-Gas Parts

Parameter R-32 R-125 R-134a R-152a

N 0o -8.25383436 13.7947897 10.7849779 -9.50813507

N1o 6.35191800 9.23166907 8.61297741 6.81206878
N 2o 2.99966040 -27.8731735 -24.3754838 -7.28591604
N 3o 3.12115048 11.9902671 7.4517850 6.74113010
η 3o 4.55977726 - 4.10383034 -
N 4o 0.99942209 7.02844573 -4.23923951 1.97815203
η 4o 2.16478825 4.90712643 2.56152868 1.75374115
N 5o 2.41272106 4.58663536 6.44573983 5.88082631
η 5o 12.3468714 2.08081818 2.08460736 4.36015034
N 6o 3.05543496 - - -
η 6o 5.87790203 - - -

6.3.2 Residual Part

Each residual part of the four fundamental equations of state was established with

the genetic optimization procedure and multi-property regression. Development results

of the residual parts are given in Eqs. (6.11) to (6.14) for R-32, R-125, R-134a, and

R-152a, respectively. Their numerical coefficients and constants are listed in Table 6.4.

The residual parts for R-32 consists of 18 terms, R-125 consists of 17 terms, R-134a

consists of 18 terms, and R-152a consists of 17 terms.

6 11 15 18
α r (δ , τ ) = ∑ N i δ di τ ti + ∑ N i δ di τ ti exp(−δ ) + ∑ N i δ di τ ti exp(−δ 2 ) + ∑ N i δ di τ ti exp(−δ 3 ) (6.11)
i =1 i =7 i =12 i =16

94
 8 11 14 17
α r (δ ,τ ) = ∑ N i δ di τ ti + ∑ N i δ di τ ti exp(−δ ) + ∑ N i δ di τ ti exp(−δ 2 ) + ∑ N i δ di τ ti exp(−δ 3 ) (6.12)
i =1 i =9 i =12 i =15

9 12 15 18
α r (δ ,τ ) = ∑ N i δ di τ ti + ∑ N i δ di τ ti exp(−δ ) + ∑ N i δ di τ ti exp(−δ 2 ) + ∑ N i δ di τ ti exp(−δ 3 ) (6.13)
i =1 i =10 i =13 i =16

8 11 14 17
α r (δ ,τ ) = ∑ N i δ di τ ti + ∑ N i δ di τ ti exp(−δ ) + ∑ N i δ di τ ti exp(−δ 2 ) + ∑ N i δ di τ ti exp(−δ 3 ) (6.14)
i =1 i =9 i =12 i =15

Table 6.4 Numerical Coefficients and Constants for Each of the Residual Parts

i Ni di ti

R-32
0
1 2.118687504 × 10 1 0.5
2 -4.531096272 × 100 1 1.125
3 1.442455906 × 100 1 1.625
-1
4 2.053905615 × 10 3 0.875
-1
5 -1.311674766 × 10 3 1.5
6 1.022271660 × 10-2 4 1.75
7 4.873981873 × 10-1 1 1.75
8 -1.062212807 × 100 1 2.75
9 -4.542050579 × 10-3 5 0.25
10 -6.933347471 × 10-4 5 3.75
11 -3.510306669 × 10-2 6 1
12 -5.606161492 × 10-2 1 6.5
13 8.849625092 × 10-2 2 2.5
14 -1.850757607 × 10-2 5 7.5
15 7.878071487 × 10-3 6 7.5
16 -3.384114830 × 10-2 2 11
17 1.641979240 × 10-4 2 16
18 -1.459171940 × 10-3 8 13
R-125
0
1 1.516288222 × 10 1 0.5
2 -1.495980500 × 100 1 0.75
3 -1.289396503 × 100 1 2.25
0
4 1.472951954 × 10 2 0.5
5 -2.229764361 × 100 2 0.875
6 1.020820106 × 100 2 2
7 -9.616958808 × 10-3 3 3
8 4.141425224 × 10-2 4 0.5

95
 Table 6.4 (continued)

i Ni di ti

9 1.462174898 × 10-1 3 4
10 -6.564863711 × 10-2 6 2
11 -9.183197274 × 10-2 4 3.25
12 -2.903433855 × 10-2 2 9.5
13 -1.743433568 × 10-2 4 4.5
14 -8.774064976 × 10-4 4 10.5
15 -5.106483618 × 10-3 3 25
16 3.524259467 × 10-3 5 5
17 4.980228501 × 10-4 7 28

R-134a
1 1.832124209 × 100 1 0.5
2 -2.940698861 × 100 1 1.125
3 5.156071823 × 10-1 1 3.25
4 2.756965911 × 10-1 2 0.5
5 1.225264939 × 100 2 1.875
6 -6.486749497 × 10-1 2 2.75
7 -9.286738053 × 10-1 3 1.625
8 3.920381291 × 10-1 3 2.125
9 1.056692108 × 10-1 4 1.125
10 -7.586523371 × 10-1 1 3.75
11 -1.198140136 × 100 2 1.5
12 -2.878260390 × 10-1 3 0.75
13 -9.723032379 × 10-2 2 9
14 5.307113358 × 10-2 3 8.5
15 -4.681610582 × 10-2 4 5.5
16 -9.604697902 × 10-3 4 32
17 6.668035048 × 10-3 5 23
18 2.361266290 × 10-3 6 31
R-152a
1 1.753847317 × 100 1 0.5
2 -4.049760759 × 100 1 1.125
3 -2.277389257 × 10-1 1 2.875

96
 Table 6.4 (continued)

i Ni di ti

4 7.087751950 × 10-1 2 0.875

5 -5.528619502 × 10-1 2 1.875
6 -3.025046686 × 10-2 3 0.5
7 1.396289974 × 10-1 3 1.875
8 1.121238954 × 10-4 4 4
9 1.181005890 × 100 1 1.25
10 1.535785579 × 100 2 2
11 7.468363045 × 10-1 3 2.75
12 -1.252266405 × 10-1 1 6
-2
13 -3.898223986 × 10 2 9
-2
14 -7.260588801 × 10 3 6
15 -2.659302250 × 10-3 3 22
16 4.210849329 × 10-3 4 20
17 2.015953966 × 10-4 5 32

All thermodynamic properties can be calculated from the fundamental equations

of state when two properties are known for single phase or a property is known for

saturation state. Detailed explanation on calculation strategy for the thermodynamic

properties from fundamental equation of state was described in Section 2.4. Examples

for calculation results calculated from each of the fundamental equations of state are

listed in Appendix D. Pressure-enthalpy diagrams calculated from each the fundamental

equation of state were plotted as results given in Appendix E.

6.4 Assessment of Ideal-Gas Properties

Accuracy assessments were revealed from deviation of available data with respect

to property values from fundamental equations of state established in this study. The

97
property values from the existing equations of state including International Standard

equation of state were considered in the assessments. Deviations of ideal-gas isobaric

specific heat cpo from the fundamental equations of state are shown in Figs. 6.1 to 6.4

for R-32, R-125, R-134a, and R-152a, respectively.

The fundamental equation of state for R-32 can accurately reproduce the

reanalysis cpo data of Yokozeki et al. (1998) for temperatures 120 K to 1000 K with

deviation less than 0.1 % as shown in Fig. 6.1. The International Standard equation of

state (Tillner-Roth and Yokozeki 1997) represents with higher deviations for

temperatures higher than 400 K. It is also larger discrepancy with the MBWR equation

of state (Outcalt and McLinden 1995) at higher temperatures. Several data of cpo derived

from speed-of-sound measurements are also well represented. This result indicates that

ideal-gas representation of the fundamental equation of state is not only reliable at the

temperature range where the experimental data are available but also at the range where

experimental data are not available.

0.6
R-32
0.4
Deviation of cp o, %

0.2
0
-0.2
-0.4
-0.6
100 250 400 550 700 850 1000
T, K
Fig. 6.1 Deviation of ideal-gas isobaric specific heat from the new fundamental
equation of state for R-32. (▲) Hozumi et al. (1994); (--×--) Outcalt and McLinden
(1995); (--□--) Sato et al. (2002); (∆) Sun et al. (1997); (--○--) Tillner-Roth and
Yokozeki (1997); (♦) Yokozeki et al. (1998).

Deviation of cpo from the fundamental equation of state for R-125 is shown in Fig.

98
6.2. All reanalysis data cpo of Yokozeki et al. (1998) from 120 K to 1000 K can be

represented with deviation less than 0.1 %. Data of cpo derived from speed-of-sound

measurements by Gillis (1997) are represented within uncertainty of 0.17 %. On the

other hand, the International Standard equation of state (Piao and Noguchi 1998)

represents cpo with large discrepancies not only at higher temperatures but also at lower

temperatures. MBWR equation of state by Outcalt and McLinden (1995) also

reproduces the property with large discrepancies.

0.6
R-125
0.4
Deviation of cp , %
o

0.2
0
-0.2
-0.4
-0.6
100 250 400 550 700 850 1000
T, K
Fig. 6.2 Deviation of ideal-gas isobaric specific heat from the new fundamental
equation of state for R-125. (□) Gillis (1997); (○) Grigiante et al. (2000); (---) Outcalt
and McLinden (1995); (—) Piao and Noguchi (1998); (--×--) Sunaga et al. (1998); (♦)
Yokozeki et al. (1998).

The fundamental equation of state for R-134a accurately represents the cpo as

shown in Fig. 6.3. Data derived from speed-of-sound data by Sato et al. (2002) are

represented within uncertainty of 0.4 %, and cpo data of Goodwin and Moldover (1997)

are represented within 0.3 %. All reanalysis cpo data of Yokozeki et al. (1998) from 120

K to 1000 K can be represented with deviation less than 0.1 %. On the other hand the

International Standard equation of state (Tillner-Roth and Baehr 1994) represents the

specific heat with large discrepancies not only with at higher temperatures but also at

lower temperatures. Similar large discrepancies appear for the values derived from the

99
equations of state (Huber and Ely 1992, 1994).

0.6
R-134a
0.4

Deviation of cp , %
o 0.2
0
-0.2
-0.4
-0.6
100 250 400 550 700 850 1000
T, K

Fig. 6.3 Deviation of ideal-gas isobaric specific heat from the new fundamental
equation of state for R-134a. (◊) Yokozeki et al. (1998); (---) Piao et al. (1991); (—)
Tillner-Roth and Baehr (1991); (--+--) Huber and Ely (1992, 1994); (○) Sato et al
(2002); (□) Goodwin and Moldover (1990).

0.6
R-152a
0.4
Deviation of cp o, %

0.2
0
-0.2
-0.4
-0.6
100 250 400 550 700 850 1000
T, K

Fig. 6.4 Deviation of ideal-gas isobaric specific heat from the new fundamental
equation of state for R-152a. (∆) Gillis (1997); (+) Hozumi et al. (1993); (◊) Yokozeki et
al. (1998); (---) Tillner-Roth (1996) and Span and Wagner (2003); (—) Oucalt and
McLinden (1996).

Deviation of cpo from the fundamental equation of state for R-152a is shown in Fig.

6.4. All reanalysis data cpo of Yokozeki et al. (1998) from 120 K to 1000 K can be

reproduced with deviation less than 0.1 %. Data of cpo derived from speed-of-sound

100
measurements by Gillis (1997) are represented within uncertainty of 0.17 %. On the

other hand, other equations of state (Tillner-Roth 1996; Span and Wagner 2003; Outcalt

and McLinden 1996) reproduce cpo with large discrepancies not only at higher

temperatures but also at lower temperatures.

All the above assessment results confirm that each of the fundamental equations of

state can represent cpo values well. Consequence of their behaviors is very good

agreements with reanalysis data of Yokozeki et al. (1998), these indicates that the

fundamental equations of state have good consistency with the theory. Additionally,

adequate representation of the data derived from the speed-of-sound measurements that

can also be performed, confirm that the fundamental equations of state are reliable to

reproduce ideal-gas properties.

6.5 Assessment of PVT Property

The PVT assessments include single phase and saturation. Supercritical region may

be included either in gaseous or liquid phases. Strictly, the phase of every state can be

known from the figures since both deviations for the same states are expressed in

variations with respect to either pressure or temperature.

6.5.1 PVT Property in Single Phase

PVT properties both in the gaseous and liquid phases were assessed on

development of the four fundamental equations of state. Assessment results for PVT

properties of R-32 from the fundamental equation of state are shown in Figs. 6.5 and 6.6.

Very accurate pressures can be reproduced at low pressures in the gaseous phase, and

similar accuracy of the densities for the same range can be reproduced. Representation

for the liquid phase can be performed with an uncertainty of 0.06 % in density.

101
 On the other hand, PVT properties of R-32 in supercritical region are reproduced

with slightly higher deviation especially for the critical region as partially shown in Figs.

6.5 and 6.6. The uncertainties are 0.2 % and 0.4 % in densities for the superheated-

vapor phase and supercritical region, respectively. Typical uncertainties for the

superheated-vapor phase are less than 0.01 % in pressure at low densities. This

uncertainty behavior is similar with uncertainty of the experimental PVT data. Most

data of de Vries (1997) can be accurately represented. It is strictly different uncertainties

for the data of Defibaugh et al. (1994) can be reproduced.

1 Tc
R-32
0.5
0
-0.5
Deviation of ρ , %

-1
200 250 300 350 400 450
T, K
1
R-32 pc
0.5
0
-0.5
-1
0.01 0.1 1 10 100
p , MPa

Fig. 6.5 Deviation of density in gaseous phase from the new fundamental equation of
state for R-32. (◊) de Vries (1997); (×) Defibaugh et al. (1994); (+) Matsuda et al.
(2002); (○) Qian et al. (1993).

Deviations of experimental PVT data from the fundamental equation of state for

R-125 are shown in Figs. 6.7 and 6.8. Deviations of densities for numerous sets of

experimental data in the gaseous phase are very small at low temperatures, and slightly

increases at temperatures close to the critical point as shown in Fig. 6.7. In a range of

temperatures below 330 K and pressures below 40 MPa, the maximum deviation is

0.15 % in density. The deviation slightly increases at the states near the critical point.

102
 1
0.5 R-32
0
Deviation of ρ , %
-0.5 Tc
-1
130 180 230 280 330 380 430
T, K
1
0.5 R-32
0
-0.5 pc
-1
0 20 40 60 80
p , MPa

Fig. 6.6 Deviation of density in liquid phase from the new fundamental equation of state
for R-32. (◊) de Vries (1997); (×) Magee (1994); (+) Hostle et al. (1994); (□) Defibaugh
et al. (1994); (∆) Matsuda et al. (2002).

1 R-125
0.5
0
-0.5 Tc
Deviation of ρ , %

-1
250 300 350 400 450 500
T, K
1
R-125
0.5 pc
0
-0.5
-1
0.01 0.1 1 10 100
p , MPa

Fig. 6.7 Deviation of density in gaseous phase from the new fundamental equation of
state for R-125. (○) de Vries (1997); (∆) Boyes and Weber (1995); (×) Oguchi et al.
(1996); (+) Ye et al. (1995); (*) Zhang et al. (1995); (◊) Duarte-Garza et al. (1997).

Figure 6.8 illustrates deviations of density measurements in liquid phase of R-125.

Accurate representation is performed by the fundamental equation of state at lower

temperatures, but slightly larger deviations in the critical region. Data of de Vries (1997)

103
are represented with an uncertainty of 0.1 % at lower pressures and temperatures, while

slightly increase at temperatures near the critical temperature. Data of Duarte-Garza et

al. (1997) are also well reproduced at higher temperatures.

1
Tc R-125
0.5
0
-0.5
Deviation of ρ , %

-1
170 200 230 260 290 320 350 380 410 440
T, K
1
R-125
0.5
0
-0.5 p c
-1
0 20 40 60 80
p , MPa

Fig. 6.8 Deviation of density in liquid phase from the new fundamental equation of state
for R-125. (◊) Magee (1996); (×) de Vries (1997); (□) Duarte-Garza et al. (1997).

Deviations of densities in the supercritical region from the fundamental equation

of state for R-125 are shown in Figs. 6.7 and 6.8. Especially for density at high

pressures can be reproduced with an uncertainty of 0.1 %. However, the states near the

critical point are reproduced with larger deviations. This behavior agrees with usual

thermodynamic behavior in the critical region.

Accuracies of PVT property from the fundamental equation of state for R-134a in

the single phase are shown in Figs. 6.9 and 6.10. The density deviations in the

superheated-vapor phase are shown in Fig. 6.9 for pressures lower than the critical

pressure. The deviations in the liquid phase are shown in Fig. 6.10 for temperatures

lower than the critical temperature. On the other hand, the deviations in the supercritical

104
region are shown in Fig. 6.9 for pressures higher than the critical pressure and in Fig.

6.10 for temperatures higher than the critical temperature.

1 pc
0.5 R-134a
0
-0.5
Deviation of ρ , %

-1
0.01 0.1 1 10 100
p , MPa
1
R-134a
0.5
0
-0.5 Tc
-1
200 250 300 350 400 450
T, K

Fig. 6.9 Deviation of density in gaseous phase from the new fundamental equation of
state for R-134a. (∆) Zhu et al. (1992); (□) Zhang et al. (1995); (×) Tillner-Roth and
Baehr (1992); (◊) Dresner and Bier (1992).

The densities of R-134a can be represented with an uncertainty of 0.1 % for the

superheated-vapor phase, 0.08 % for the liquid phase, and 0.3 % for the supercritical

region. As shown in these figures, data of Tillner-Roth and Baehr (1992,1993) and

Magee (1992) are also well reproduced.

Deviations of experimental PVT data from the fundamental equation of state for

R-152a are shown in Figs. 6.11 and 6.12. Most reliable experimental data are accurately

reproduced for the states in the gaseous and liquid phases. An adequate uncertainty for

the supercritical region can also be represented as shown in Fig. 6.11 for pressure higher

than the critical pressure and Fig. 6.12 for temperatures higher than the critical

temperature.

105
 1 pc R-134a
0.5
Deviation of ρ , % 0
-0.5
-1
0.1 1 10 100
p , MPa
1 Tc R-134a
0.5
0
-0.5
-1
150 200 250 300 350 400 450
T, K

Fig. 6.10 Deviation of density in liquid phase from the new fundamental equation of
state for R-134a. (◊) Magee (1992); (×) Tillner-Roth and Baehr (1993); (□) Blanke et al.
(1995); (∆) Morisson and Ward (1991).

1
pc R-152a
0.5
0
Deviation of ρ , %

-0.5
-1
0.01 0.1 1 10 100
p , MPa
1 R-152a
0.5 Tc
0
-0.5
-1
250 300 350 400 450
T, K

Fig. 6.11 Deviation of density in gaseous phase from the new fundamental equation of
state for R-152a. (◊) Tillner-Roth and Baehr (1992); (□) Dressner and Bier (1993).

PVT data of Tillner-Roth and Baehr (1992) in the superheated-vapor phase can be

reproduced by the fundamental equation of state for R-152a with a density uncertainty

of 0.15 % as shown in Fig. 6.11 for pressures lower than the critical pressure. On the

other the PVT data for the liquid phase of Tillner-Roth and Baehr (1993) are reproduced

106
with a density uncertainty of 0.07 % as shown in Fig. 6.12 for temperatures lower than

the critical temperature.

1
pc R-152a
0.5
0
Deviation of ρ , %

-0.5
-1
0.1 1 10 100
p , MPa
1
Tc R-152a
0.5
0
-0.5
-1
150 225 300 375 450
T, K

Fig. 6.12 Deviation of density in liquid phase from the new fundamental equation of
state for R-152a. (◊) Magee (1998); (×) Tillner-Roth and Baehr (1993); (□) Dressner
and Bier (1993); (∆) Defibaugh and Moldover (1997).

6.5.2 PVT Property at Saturation

The PVT property at saturation derived from the fundamental equations of state

should satisfy requirement of the saturation state. The requirement are that vapor

pressures evaluated from the pressure relation at saturated-vapor state, the pressure

relations at saturated-liquid state, and the Maxwell relation at saturation state for the

same temperatures should yield the same values. Assessments of these properties are

very important to verify reliability of the fundamental equations of state at saturation.

The existing equations of state are also included in these assessments to give more

information of the ability of the fundamental equations of state established in this study.

The fundamental equation of state for R-32 can reproduce reliable vapor pressures

with an uncertainty of 0.07 % and reliable saturation temperatures with an uncertainty

107
of 0.03 K as shown in Fig. 6.13. Vapor-pressure data of Weber and Goodman (1993) at

lower temperatures are reproduced with deviation less than 0.1 %. The data of de Vries

(1997), Matsuda et al. (2002), Qian et al. (1993), and Sato et al. (1994) are also well

represented. In comparison of vapor pressures from existing equations of state appears

that large discrepancies occur at low temperatures. However, measurements in this

temperature range are scarcely found so that it is difficult to judge reliability of the

equations of state.

0.6
0.4 R-32
0.2
ps , %

0
-0.2
-0.4
Deviations

-0.6
0.1
R-32
0.05
Ts , %

0
-0.05
-0.1
130 160 190 220 250 280 310 340
T, K

Fig. 6.13 Deviations of vapor pressure and saturation temperature from the new
fundamental equation of state for R-32. (◊) Defibaugh et al. (1994); (○) de Vries (1997);
(–) Fu et al. (1995); (+) Qian et al. (1993); (♦) Malbrunot et al. (1968); (*) Matsuda et al.
(2002); (□) Sato et al. (1994); (×) Weber and Goodwin (1993); (∆) Weber and Silva
(1994); (■) Zhu et al. (1993); (--○--) Tillner-Roth (1996); (▬▬) Tillner-Roth and
Yokozeki (1997); (----) Outcalt and McLinden (1995); (--+--) Vasserman and Fominsky
(2001).

Vapor pressures of R-125 can be reproduced with an uncertainty of 0.1 %, and an

uncertainty of the saturation temperatures is 20 mK. Larger percentages of the

deviations are shown especially for low temperatures, even though the absolute

deviations are small enough as shown in Fig. 6.14. An agreement result with

measurement of vapor pressure is very difficult at low temperatures near the triple point.

108
From comparison with the values from other equations of state, it appears smaller

deviation at higher temperatures and slightly larger deviation at low temperatures. In

comparison of the equations of state for R-32, the equations of state for R-125 perform

similar discrepancies at low temperatures.

0.6
R-125
0.4
0.2
ps , %

0
-0.2
Deviations

-0.4
-0.6
0.14
R-125
0.07
Ts , %

0
-0.07
-0.14
170 200 230 260 290 320 350
T, K

Fig. 6.14 Deviations of vapor pressure and saturation temperature from the new
fundamental equation of state for R-125. (◊) de Vries (1997); (●) Fukushima (1995); (∆)
Tsvetkov et al. (1996); (○) Weber and Silva (1994); (+) Oguchi et al. (1996); ( ∗ ) Magee
(1996); (■) Boyes and Weber (1995); (×) Sagawa et al. (1994); (□) Tillner-Roth (1996);
(-----) Piao and Noguchi (1998); (──) Sunaga et al. (1998); (▬▬) Vasserman and
Fominsky (2001).

Figure 6.15 shows deviations of vapor pressures and saturation temperatures for

R-134a from the fundamental equation of state. The most reliable measurements are

represented with uncertainties of 0.05 % (0.8 kPa) in vapor pressure and 0.003 % (14

mK) in saturation temperature. The data of Magee and Bowley (1992) at low

temperatures can be reproduced with higher accuracies in comparison with other

fundamental equations of state such as equations of Tillner-Roth and Baehr (1994) and

Span and Wagner (2003). The data of Baehr and Tillner-Roth (1991) can also be

accurately reproduced. This result also indicates large discrepancy values among the

109
equations of state. The derived data of Tillner-Roth (1996) for low temperature, which is

one of theoretically calculated data, can be reproduced with larger deviation but smaller

deviation than other equations of state. The discrepancy at low temperatures is similar

discrepancy with the equations of state for R-32 and R-125.

1
R-134a
0.5
ps , %

0
-0.5
Deviations

-1
0.1
R-134a
0.05
Ts , %

0
-0.05
-0.1
150 225 300 375
T, K

Fig. 6.15 Deviations of vapor pressure and saturation temperature from the new
fundamental equation of state for R-134a. (□) Magee and Bowley (1992); (×)
Fukushima et al. (1990); (◊) Baehr and Tillner-Roth (1991); (+) Baroncini et al. (1990);
(○) Weber (1989); (∆) Tillner-Roth (1996); (——) Tillner-Roth and Baehr (1994); (----)
Span and Wagner (2003).

Vapor pressure deviations of experimental data from the fundamental equation of

state for R-152a are shown in Fig. 6.16. Most reliable data for the vapor pressure can be

reproduced with an uncertainty of 0.1 % and saturation temperature within 0.015 %.

Adequate accuracies are performed at low temperatures, even though most equations of

state performed larger deviations at this temperature range. Data of Duarte-Garza and

Magee (1999) with low temperature range are well represented. For higher temperatures,

data of Sato et al. (1987) are accurately reproduced. From this result, it can be judged

that data of Zhao et al. (1992) have larger uncertainties. The fundamental equation of

state performs vapor pressures close to the vapor pressures calculated from the Outcalt

110
and McLinden (1996). Large discrepancies of the vapor pressures not only at low

temperatures but also at higher temperatures are indicated in comparison of the vapor

pressures from the equation of state by Tillner-Roth (1996).

1
0.5
ps , %

0
-0.5
R-152a
Deviations

-1
0.06
0.03
Ts , %

0
-0.03
R-152a
-0.06
150 200 250 300 350 400
T, K

Fig. 6.16 Deviations of vapor pressure and saturation temperature from the new
fundamental equation of state for R-152a. (◊) Higashi et al. (1987); (+) Zhao et al.
(1992); (∆) Duarte-Garza and Magee (1999); (×) Sato et al. (1987); (○) Holcomb et al.
(1993); (------) Tillner-Roth (1996); (―) Outcalt and McLinden (1996); (□) Silva and
Weber (1993).

The fundamental equation of state for R-32 can reproduce most reliable

measurements of the saturated-liquid and saturated-vapor densities within uncertainties

of 0.2 % and 0.7 % as shown in Fig. 6.17, respectively. Data of Defibaugh et al. (1994)

are one of the saturated-vapor densities that are available with range to low

temperatures can be reproduced with average of deviation of 0.3 %. For the

saturated-vapor densities near the critical point, the deviation trends divergence but the

values from the fundamental equation of state are still in mean values of the

measurements. The saturated-liquid densities with lower temperatures of Magee (1996)

are reproduced with average of absolute deviation of 0.08 %. Generally, a good

agreement can be performed among the equations of state for a range from intermediate

111
to lower temperatures but larger discrepancy for temperatures near the critical point.

0.6
R-32
0.3
ρ' , %

0
-0.3
Deviations

-0.6
2
R-32
1
ρ" , %

0
-1
-2
130 175 220 265 310 355
T, K

Fig. 6.17 Deviations of saturated-liquid and saturated-vapor densities from the new
fundamental equation of state for R-32. (□) Defibaugh et al. (1994); (+) Fukushima et al.
(1995); (–) Higashi (1994); (♦) Holcomb et al. (1993); (●) Kuwabara et al. (1995); (×)
Magee (1996); (○) Malbrunot et al. (1968); (∆) Sato et al. (1994); (▲) Widiatmo et al.
(1994); (▬▬) Tillner-Roth and Yokozeki (1997); (----) Outcalt and McLinden (1995);
(--+--) Vasserman and Fominsky (2001).

0.8
R-125
0.4
ρ' , %

0
-0.4
Deviations

-0.8
2
R-125
1
ρ" , %

0
-1
-2
170 200 230 260 290 320 350
T, K

Fig. 6.18 Deviations of saturated-liquid and saturated-vapor densities from the new
fundamental equation of state for R-125. (×) Fukushima et al. (1995); (○) Kuwabara et
al. (1995); (◊) Widiatmo et al. (1994); (□) Magee (1996); (*) Higashi (1994); (∆)
Takahashi et al. (1994); (▲) Boyes and Weber (1995); (-----) Piao and Noguchi (1998);
(───) Sunaga et al. (1998); (▬▬) Vasserman and Fominsky (2001).

112
 Saturated-liquid and saturated-vapor densities of R-125 can be reproduced from

the fundamental equation of state within uncertainties of 0.15 % and 0.5 %, respectively,

for most temperatures except the range near the critical temperature. Graphical

representations for the saturation properties are given in Fig. 6.18. The saturated-liquid

densities such as data of Widiatmo et al. (1994) and Fukushima et al. (1995) are

accurately reproduced. In the range near the critical point, adequate agreement is also

performed in comparison of the available experimental data. Very scarce data of the

saturated-vapor density at low temperatures are usually found as a problem on

assessment for saturated-vapor phase. Since, insignificant different values are

performed in comparison to the other equations of state and in addition to the

fundamental equation of state are assessed from various aspects, these all strongly

confirm good representation reliability of the saturation properties at low temperatures.

0.8
R-134a
0.4
ρ' , %

0
-0.4
Deviations

-0.8
4
R-134a
2
ρ" , %

0
-2
-4
150 185 220 255 290 325 360
T, K

Fig. 6.19 Deviations of saturated-liquid and saturated-vapor densities from the new
fundamental equation of state for R-134a. (◊) Yokoyama and Takahashi (1991); (+)
Maezawa et al. (1990); (○) Higashi (1994); (□) Hou et al. (1992); (●) Niessen et al.
(1994); (∆) Fukushima et al. (1990); (×) Takahashi et al. (1994); (——) Tillner-Roth
and Baehr (1994); (----) Span and Wagner (2003).

113
 The fundamental equation of state for R-134a can reproduce most reliable

measurements of the saturated-liquid and saturated-vapor densities with uncertainties of

0.2 % and 0.5 %, respectively. Saturated-liquid densities of Maezawa et al. (1990) and

Hou et al. (1992) for lower temperatures are well represented. Compared with other

equations of state (Tillner-Roth and Baehr 1994; Span and Wagner 2003), it can

reproduce these measurements with higher accuracies. Deviations of the densities near

critical point slightly increase as shown in Fig. 6.19. This result agrees with

experimental uncertainties at the critical region that are usually worse than other region.

0.6
R-152a
0.3
ρ' , %

0
-0.3
Deviations

-0.6
4
R-152a
2
ρ" , %

0
-2
-4
150 225 300 375
T, K

Fig. 6.20 Deviations of saturated-liquid and saturated-vapor densities from the new
fundamental equation of state for R-152a. (∆) Wang et al. (1992); (○) Holcomb et al.
(1993); (■) Blanke and Weiss (1992); (+) Higashi et al. (1987); (×) McLinden (1989);
(◊) Kamei et al. (1990); (-----) Tillner-Roth (1996); (——) Outcalt and McLinden
(1996).

Saturated-liquid and saturated-vapor densities of R-152a can be reproduced from

the fundamental equation of state within uncertainties of 0.1 % and 1 %, respectively,

for most temperatures except the range near the critical temperature as shown in Fig.

6.20. Especially for low and intermediate temperatures, measurements of the

saturated-liquid densities can represented accurately. Very large discrepancies of the

114
saturated-liquid densities appear for temperatures near the critical point, not only for the

measurements but also for the values derived from the other equations of state. For this

fluid, very scarce data of the saturated-vapor density at low temperatures are also found

as a problem on assessment for saturated-vapor phase. However, the saturated-vapor

densities such as data of Holcomb et al. (1993) and Kamei et al. (1990) have

discrepancy among them. Since fitting of the fundamental equation of state was also

considered other parameters that are also targeted to maintain the thermodynamic

consistencies, this result can be judged as an optimal solution that was accomplished for

the reliability of the saturation properties.

6.6 Assessment of Caloric Properties

Caloric properties were assessed on development of the four fundamental

equations of state. The assessments include specific heat and speed-of-sound properties.

Since reliable experimental data for specific heat in gaseous phase are scarce, the

assessments from experimental data are limited to the liquid phase. On the other hand,

the speed-of-sound assessments include the gaseous and liquid phases.

6.6.1 Specific-Heat Assessment

Isobaric and isochoric specific heat values were calculated from each of the

fundamental equations of state to assess them with the experimental data. Unfortunately,

the available data for the gaseous phase are scarcely found in practice. Therefore,

illustrations of deviations of these properties are given for the liquid phases only.

Deviations for cv and cp measurements from the fundamental equation of state for

R-32 in the liquid phase as shown in Fig. 6.21. The cv data of R-32 by Lüddecke and

Magee (1996) can be reproduced within an uncertainty of 0.6 %. The cp data of Yomo et

115
al. (1994) are reproduced with deviations of less than 1.5 %. However, the cp cannot be

represented with small bias. Other equations of state also represent the cp with bias.

Even though assessment result for the cp in the gaseous phase is not shown here, the

data were assessed as shown in Appendix C (Table C.4). Unfortunately, Kunimoto et al.

(1994) reported cp data in the gaseous phase (the same data as Kubota et al. (1995)

which cannot be represented properly. Similar results were also shown in comparison of

the values calculated from other existing equations of state. Their bias is more than

10 %. Perhaps the cp data in the gaseous phase have a highly systematic error.

1.5
Liquid phase, R-32
0.75
cp , %

0
-0.75
Deviations

-1.5

1
Liquid phase, R-32
0.5
%

0
cv ,

-0.5
-1
140 170 200 230 260 290 320 350
T, K

Fig. 6.21 Deviations of isobaric and isochoric specific heats in the liquid phase from the
new fundamental equation of state for R-32. (◊) Lüddecke and Magee (1996); (×) Yomo
et al. (1994).

Isochoric and isobaric specific heats of R-125 in liquid phase are shown in Fig.

6.22. Measurements of Lüddecke and Magee (1996) are reproduced mostly within

1.2 %. Although isobaric specific heats were not fitted in the development, the data of

Zhao et al. (2001) can be represented within an uncertainty of 0.5 %.

116
 1
Liquid phase, R-125
0.5

cp , %
0
-0.5
Deviations

-1
2
1
cv , %

0
-1
Liquid phase, R-125
-2
200 225 250 275 300 325 350
T, K

Fig. 6.22 Deviations of isobaric and isochoric specific heats in the liquid phase from the
new fundamental equation of state for R-125. (◊) Lüddecke and Magee (1996); (○)
Zhao et al. (2001).

2
1
cp , %

0
-1 Liquid phase, R-134a
Deviations

-2

2
Liquid phase, R-134a
1
cv , %

0
-1
-2
150 200 250 300 350 400
T, K

Fig. 6.23 Deviations of isobaric and isochoric specific heats in the liquid phase from the
new fundamental equation of state for R-134a. (∆) Magee (1992); (×) Saitoh et al.
(1990).

Assessment results for specific heat of R-134a are shown in Fig. 6.23. On the

other hand, specific heats are reproduced with an uncertainty of 0.8 % for isochoric

specific heat in the liquid-phase, and an uncertainty of 1.0 % for cp in the liquid phase.

117
 6
3
cp , %
0
-3
Deviations

Liquid phase, R-152a

-6
1.6
0.8
cv , %

0
-0.8 Liquid phase, R-152a
-1.6
150 225 300 375 450
T, K

Fig. 6.24 Deviations of isobaric and isochoric specific heats in the liquid phase from the
new fundamental equation of state for R-152a. (◊) Magee (1998); (×) Kubota et al
(1987); (○) Nakagawa et al. (1993); (∆) Porichanski et al. (1982); (□) Sato and
Watanabe (1994).

The fundamental equation of state represents specific heat for R-152a as shown in

Fig. 6.24. The cv data of Magee (1998) are reproduced with an uncertainty of 0.9 %.

These data have bias of 0.4 % as illustrated in the figure, but most of the data can be

reproduced with deviation less than 1.2 %. The cp data of this fluid such as data

Porinchanski et al. (1982) with wide-range but it cannot accurately reproduced. The

result may be consequence of uncertainty of the experimental data.

6.6.2 Speed-of-Sound Assessment

Deviations of speed-of-sound measurements for R-32 from the fundamental

equation of state are given in Fig. 6.25. Very high-accurate representation can be

performed for the gaseous phase and slightly low accuracy can be performed for the

liquid phase. Most reliable data can be reproduced with an uncertainty of 0.01 %. In

contrary, the fundamental equation of state reproduces most reliable speed-of-sound

118
values in the liquid phase with an uncertainty of 0.7 %. The speed-of-sound

measurements in the gaseous phase of Hozumi et al. (1996) are reproduced with high

accuracy and then are followed by the measurements of Sun et al. (1997) with slightly

low accuracy. On the other hand, among the speed-of-sound measurements in the liquid

phase appear the larger deviations of among them. Nevertheless, their uncertainty can

be acceptable with values larger than the speed of sound in the gaseous phase as

performed by the fundamental equation of state.

2
1
0
Deviation of w , %

-1
R-32 Liquid phase
-2
0.14
0.07
0
-0.07 Gaseous phase
R-32
-0.14
230 250 270 290 310 330 350 370
T, K

Fig. 6.25 Deviation of speed of sound from the new fundamental equation of state for
R-32. (◊) Takagi (1993); (□) Grebenkov et al. (1994); (×) Grebenkov et al. (1995); (∆)
Pires and Guedes (1999); (○) Hozumi et al. (1996); (+) Sun et al. (1997).

Deviations of speed of sound from the fundamental equation of state for both

gaseous and liquid phases of R-125 are shown in Fig. 6.26. The data of Kojima and Sato

(2002) can be reproduced with maximum deviation of 0.02 % and an uncertainty of

0.012 %. The data of Gillis (1997) in a wider range of temperatures and pressures are

reproduced within maximum deviation of 0.08 % and an uncertainty of 0.044 %,

respectively. On the other hand, an uncertainty for the speed of sound in the liquid phase

is larger than the speed of sound in the gaseous phase. The speed-of-sound values can

119
be reproduced within an uncertainty of 0.8 %.

2
Liquid phase, R-125
1
Deviations of w , %

0
-1
-2
0.08
Gaseous phase, R-125
0.04
0
-0.04
-0.08
200 250 300 350 400 450
T, K

Fig. 6.26 Deviation of speed of sound from the new fundamental equation of state for
R-125. (○) Takagi (1996b); (+) Grebenkov et al. (1995); (×) Hozumi et al. (1996); (□)
Grigiante et al. (2000); (◊) Gillis (1997).

3 Liquid phase, R-134a

2
1
0
Deviation of w , %

-1
-2
-3
0.1
Gaseous phase, R-134a
0.05
0
-0.05
-0.1
150 200 250 300 350 400
T, K

Fig. 6.27 Deviation of speed of sound from the new fundamental equation of state for
R-134a. (□) Guedes and Zollweg (1992); (+) Grebenkov et al. (1995); (×) Takagi
(1996a); (∆) Goodwin and Moldover (1990); (○) Hozumi et al. (1997).

The fundamental equation of state for R-134a can reproduce speed-of-sound

measurements for both the gaseous and liquid phases as shown in Fig. 6.27. Most

120
reliable speed-of-sound measurements are reproduced within uncertainties of 0.02 % for

the gaseous phase and 0.5 % for the liquid phase. The speed-of-sound measurements in

the gaseous phase such as Goodwin and Moldover (1990) and Hozumi et al. (1997) are

accurately represented. The measurements in the liquid phase can be reproduced with

lower accuracies as the data of Takagi (1996a) and Guedes and Zollweg (1992).

2
Liquid phase, R-152a
1
Deviations of w , %

0
-1
-2
0.1
0.05
0
-0.05 Gaseous phase, R-152a
-0.1
225 275 325 375 425
T, K

Fig. 6.28 Deviation of speed of sound from the new fundamental equation of state for
R-152a. (×) Grebenkov et al. (1995); (□) Gillis (1997); (○) Beliajeva et al. (1994); (◊)
Hozumi et al. (1993).

The fundamental equation of state for R-152a can reproduce speed-of-sound

measurements as shown in Fig. 6.28. Most reliable data for speed-of-sound

measurements are reproduced with uncertainties of 0.02 % for the gaseous phase and

1 % for the liquid phase. The measurements of Hozumi et al. (1993) are accurately

represented, and then are followed by Gillis (1997). In contrast, the speed of sound

measurements in the liquid phase such as the data of Grebenkov et al. (1995) and

Beliajeva et al. (1994) are reproduced with slightly low accuracy.

121
6.6.3 Saturated-Liquid Specific Heat Assessment

In principle, there are three types of specific heat quantities in accordance with

heating process i.e. isobaric process, isochoric process, and isothermal process.

Saturated-liquid specific heat is the specific heat in accordance with isothermal process

at saturation that is evaluated at saturated-liquid state. Therefore, this property is neither

isobaric specific heat nor isochoric specific heat at the saturation. Physical meaning is

that a quantity of thermodynamic properties relates close to evaporation process, which

is evaluated in accordance with the thermodynamic interests such as more detailed

explanations given in Appendix A (Section A.2).

1.6
0.8
0
Deviations of cs' , %

-0.8 R-32 and R-152a

-1.6
130 180 230 280 330
1.6
0.8
0
-0.8 R-125 and R-134a
-1.6
160 190 220 250 280
T, K

Fig. 6.29 Deviations of saturated-liquid specific heats from the four fundamental
equations of state. (+) R-32 data of Lüddecke and Magee (1996); (×) R-125 data of
Lüddecke and Magee (1996); (∆) R-134a data of Magee (1992); (□) R-152a of Magee
(1998).

Sets of data for saturated-liquid specific heats of HFC refrigerants were derived

from two-phase specific heat data (Magee 1992; Lüddecke and Magee 1996; Magee

1998). These properties were accomplished in accordance with the properties can be

simply assessed to the fundamental equations of state. Comparisons of these data to the

122
four fundamental equations of state are shown in Fig. 6.29.

Although derived saturated-liquid specific heats for R-32 of Lüddecke and Magee

(1996) that were not fitted on the thermodynamic modeling, they are reproduced with an

average absolute deviation less than 0.4 %. The derived data for saturated-liquid

specific heat of R-125 by Lüddecke and Magee (1996) were used on the modeling.

Those of the data can be reproduced within an uncertainty of 1 %. Saturated-liquid

specific heats of R-134a derived by Magee (1992) can be reproduced with an

uncertainty of 0.6 % and most the deviations less than 0.9 %. The data of R-152a are

reproduced with an uncertainty of 0.7 % and all deviations are less than 1 %.

6.7 Assessment of Critical Parameters

In principle, the independent critical parameters of the fundamental equation of

state can be obtained by primary and secondary partial differentiation of pressure with

respect to volume in accordance with the critical constraint conditions, i.e., both of the

differential values are set to zero. The critical temperature and density were derived

from the four fundamental equations of state by introducing these two constraint

conditions, resulting in the values of the critical temperature and density. At these values,

the critical pressure can be calculated.

The results for the four fundamental equation of state are represented in Table 6.5.

As shown in the table, all critical parameter values can be derived in accordance with

values introduced on the modeling as listed in Table 6.1. These results are consequences

of the critical constraints that were fitted as development of the fundamental equations

of state. The good agreements of the calculated values confirm that the selected values

for the thermodynamic modeling can be accepted due to the most parameters introduced

on the modeling are also well represented.

123
 Table 6.5 Calculated Critical Parameters for Four HFC Refrigerants

Parameter R-32 R-125 R-134a R-152a

-3
ρ c , kg·m 424 568 509 368
Tc , K 351.255 339.165 374.083 386.41
pc , MPa 5.78003 3.61746 4.04806 4.51603

6.8 Validity Range and Accuracy

Detailed accuracies of each of the four fundamental equations of state with respect

to available data are listed in Tables C.4 to C.7 (Appendix C) for R-32, R-125, R-134a,

and R-152a, respectively. This information is expected to give more detailed abilities of

the fundamental equations of state.

Since unavailable experimental data to assess reliability in advance for very high

temperatures and pressures, validity of the fundamental equations of state are

recommended by considering limit of input data used as input data on the

thermodynamic modeling. The validity ranges with their acceptable uncertainties are

listed Table 6.6. These applicable ranges were determined based on the range of

available experimental limit data.

Table 6.6 Validity Ranges of the Four Fundamental Equations of State

Fluid Pressure range Temperature range
R-32 Triple point to 72 MPa Triple point to 450 K
R-125 Triple point to 70 MPa Triple point to 480 K
R-134a Triple point to 70 MPa Triple point to 460 K
R-152a Triple point to 55 MPa Triple point to 450 K

The fundamental equation of state for R-32 has uncertainties as 0.07 % in density

for the liquid phase, 0.1 % in pressure for the superheated-vapor phase, 0.35 % in

pressure for the supercritical region, 0.07 % in vapor pressure, 0.2 % in saturated-liquid

124
density, 0.7 % in saturated-vapor density, 0.01 % in speed of sound for the gaseous

phase, 0.7 % in speed of sound for the liquid phase, 1 % in isobaric specific heat for the

liquid phase, and 0.6 % in isochoric specific heat for the liquid phase. These

uncertainties would be yielded in the above validity range.

Estimated uncertainties of the fundamental equation of state for R-125 are 0.06 %

for vapor pressure, 0.15 % in density for the superheated-vapor and liquid phases, 0.5 %

in density for the supercritical region, 0.01 % in speed of sound for the gaseous phase,

1 % in speed of sound for the liquid phase, 0.6 % in isobaric specific heat for the liquid

phase, and 1.2 % isochoric specific heat for the liquid phase. The uncertainties of

saturation properties are 0.05 % in vapor pressure, and 0.25 % and 0.5 % in

saturated-liquid and saturated-vapor densities, respectively. The uncertainties can be

performed in the above validity range.

The fundamental equation of state for R-134a has been estimated as following

uncertainties. The estimated uncertainties are 0.1 % in density for the superheated-vapor

and liquid phases and 0.4 % in the supercritical region, 0.02 % in speed of sound for the

gaseous phase, 0.5 % in speed of sound for the liquid phase, 0.8 % in isochoric specific

heat for the liquid phase, and 1 % in isobaric specific heat for the liquid phase. The

uncertainties of saturation properties are 0.05 % in vapor pressure, and 0.25 % and

0.5 % in saturated-liquid and saturated-vapor densities, respectively. The uncertainties

can be performed in the above validity range.

Finally, the last is the fundamental equation of state for R-152a. The estimated

uncertainties are 0.1 % in density for the saturated-vapor phase, 0.1 % in density for the

liquid phase, 0.4 % in density for the supercritical region, 0.02 % in speed of sound for

the gaseous phase, 1 % in speed of sound for the liquid phase, 1 % in isochoric specific

heat for the liquid phase, and 2 % in isobaric specific heat for the liquid phase. The

125
uncertainties of saturation properties are 0.1 % in vapor pressure, and 0.1 % and 1 % in

saturated-liquid and saturated-vapor densities, respectively. These uncertainties can be

realized in the above validity range.

All of the four fundamental equations of state are not recommended to calculate

caloric properties in the critical region. Theoretically, specific heat values whether

isobaric specific heat or isochoric specific heat should be toward infinite values at the

critical point, and speed-of-sound values at the critical point should be toward zero. Up

to now these critical behaviors cannot be included in the fundamental equation of state.

Therefore, there is a spot area of thermodynamic surface for caloric properties where the

caloric properties cannot be calculated from the fundamental equation of state. These

area are predicted in |T - Tc| < 3 K and |ρ - ρc| < 25 kg/m3. The right way to calculate the

caloric properties in the critical region is using scale-law and crossover equations of

state. Both these equations were developed and proved for including the critical

phenomena in thermodynamic models (Levelt-Sengers 1970; Levelt-Sengers et al.

1976; Kiselev and Sengers 1993; Kiselev 1998; Kiselev et al. 1998).

126
 Chapter 7
Discussion

The intermolecular potential behaviors, thermodynamic consistency in the

gaseous phase, and characteristics of ideal curves derived from the fundamental

equations of state are discussed in this chapter. Behavior of the fundamental equation of

state at extreme temperatures and pressures and the optimization procedure are also

discussed. Not only the fundamental equations of state of this study are discussed, but

also other existing equations of state for HFC refrigerants are involved in the

discussion.

7.1 Gaseous Specific Heat and Intermolecular Potential Behaviors

The consistencies of the four fundamental equations of state with intermolecular

potential behaviors were investigated with comparisons of second and third virial

coefficient values. Values of specific heat in the gaseous phase at lower temperatures

were assessed in comparison of the other equations of state. Most results indicate that

significantly different isochoric specific heat values appear in the gaseous phase near

saturation. Reliable caloric properties in this region are very important for

thermodynamic calculation of refrigeration cycles. These all results are shown in Figs.

7.1 to 7.4 for R-32, R-125, R-134a, and R-152a, respectively.

As shown in those results appear that the most existing equations of state

including the four fundamental equations of state can reproduce a good agreement for

behaviors of the second virial coefficients, although the values are still significantly

different. On the other hand, third virial coefficients and gaseous specific heat should be

127
noticed here. These properties are interesting to be discussed in advance. Considering

the third virial coefficient derived from intermolecular potential model, the right

behavior is the third virial coefficient increasing when temperature decreasing to certain

temperature and then the virial coefficient sharply decreasing when temperature

decreasing again.

Third virial coefficient values for R-32 were derived from several existing

equations of state including the International Standard equation of state. The

International Standard equation of state (Tillner-Roth and Yokozeki 1997) cannot allow

the reasonable behavior of the third virial coefficients. On the other hand, the

fundamental equation of state of this study can reasonably represent the third virial

coefficient in the application ranges and it can provide reliable specific heat values that

are close to the values derived from basis of accurate virial coefficients.

0
Nicola et al. (2002)
-0.2 R-32
0.9
-1

This work Yokozeki et al. (1998) This Work 1.6 MPa

B , dm ·mol

-0.4
3

Vasserman and Fominsky (2001)

-0.6
Tillner-Roth and Yokozeki (1997)

0.8 1
Tillner-Roth and Yokozeki (1997)
-0.8
R-32 Outcalt and McLinden (1995) 0.5
-1
cv , kJ·kg ·K

-1 0.7
-1

0.25
130 170 210 250 290 330 370 410 450 490 530
T, K 0.1
0.6
0.2
0.01
Tillner-Roth and Yokozeki (1997)
Outcalt and McLinden (1995) Matsuda et al. (2002)
0.1
Yokozeki et al. (1998)
-2

0.5
0.001 MPa
C , dm ·mol

Outcalt and McLinden (1995)

0
6

Nicola et al. (2002)

This work 0.4
-0.1
125 175 225 275 325 375
R-32 Vassermann and Fominsky (2001)
-0.2 T, K
130 170 210 250 290 330 370 410 450 490 530 570
T, K

Fig. 7.1 Comparison of second and third virial coefficients and gaseous specific heat
consistency for R-32.

128
 Similar results for third virial coefficient were obtained for the fundamental

equations of state for R-125 as shown in Fig. 7.2. The International Standard equation of

state (Piao and Noguchi 1998) also reproduces the third virial coefficient with large

discrepancies at low temperatures. The reasonable behavior can be represented by the

fundamental equation of state of this study. It is also able to provide reliable gaseous

specific heat values that are close to the values calculated based on accurate virial

coefficients. The specific heats of Sunaga et al. (1998) have large discrepancy from the

other equations of state.

0
Yokozeki et al. (1998) 0.9
-0.2 Vasserman and Fominsky (2001) 2 MPa
0.85
-1

Outcalt and McLinden (1995)

B , dm ·mol

-0.4
This work Outcalt and McLinden (1995)
3

0.8 1
-0.6
Piao and Noguchi (1998) Sunaga et al. (1998)
-0.8 0.75 This work
0.5
Sunaga et al. (1998)
R-125
-1
cv , kJ· kg ·K

-1 0.7 0.25
-1

150 190 230 270 310 350 390 430 470 510 550
T, K 0.65
0.1 Kojima and Sato (2002)

0.2 0.6
Vasserman and Fominsky (2001) R-125
Piao and Noguchi (1998)
Piao and Noguchi (1998)
0.1 0.55
Sunaga et al. (1998)
-2

0.0001 MPa
C , dm ·mol

0 0.5
6

Yokozeki et al. (1998)

Span and Wagner (2003) R-125
-0.1 This work
0.45
Outcalt and McLinden (1995) 150 200 250 300 350 400
-0.2
T, K
150 190 230 270 310 350 390 430 470 510 550 590

T, K

Fig. 7.2 Comparison of second and third virial coefficients and gaseous specific heat
consistency for R-125.

In comparison with the previous results (R-32 and R-125), the International

Standard equation of state for R-134a (Tillner-Roth and Baehr 1994) can derive the

third virial coefficients near the values from intermolecular potential model. However, it

129
is still incorrect representation as shown in Fig 7.3. A very good representation can be

reproduced by the fundamental equation of state as result of this study. The reanalysis

values of Yokozeki et al. (1998) were also accurately reproduced. Reliable specific heat

values in the gaseous phase can be provided and their values are close to the values

calculated from accurate virial equation of state.

0
Huber and Ely (1992) Huber and Ely (1994)

-0.2
1.1
Piao et al. (1991) This work
Yokozeki et al. (1998)
-1
B , dm ·mol

-0.4
3

Tillner-Roth and Baehr (1994)

1
-0.6 Tillner-Roth and Baehr (1994) 2 MPa

-0.8
This work 0.9
R-134a 1
-1
cv , kJ·kg-1 ·K-1

150 190 230 270 310 350 390 430 470 510 550 0.5 Piao et al. (1991)
T, K 0.8
0.2
Huber and Ely (1994) 0.25
Huber and Ely (1992)

Piao et al. (1991)

0.1 0.7 0.1 Huber and Ely (1992)
-2

Yokozeki et al. (1998)

C , dm ·mol
6

0
0.6 0.0001 MPa
Tillner-Roth and Baehr (1994) Huber and Ely (1994)

-0.1 R-134a
This work
0.5
R-134a 150 250 350 450
-0.2
150 190 230 270 310 350 390 430 470 510 550 590 T, K
T, K

Fig. 7.3 Comparison of second and third virial coefficients and gaseous specific heat
consistency for R-134a.

Especially for the existing equations of state of R-134a excluding the equation of

state by Huber and Ely (1994), it is not significant different values of the specific heat in

the gaseous phase. The large discrepancy and strictly wrong behavior of the values

derived from the Huber and Ely (1994). This is supposed that the equation of state were

assessed on the experimental data only, without plotting the thermodynamic surface to

judge the thermodynamic surface. This opinion is sustained by the fact that no isochoric

130
specific heats in the gaseous phase were introduced on the development. Even though

they adopted a functional form of the Hemholtz equation of state that was success for

representing a certain fluid, without assessing caloric surface, it cannot be ensured to be

able to derive the reasonable thermodynamic surface.

0
Tillner-Roth (1996) 1.3
-0.2
-0.4
This work Tamatsu et al. (1993)
-0.6 1.2 1.75 MPa
Tillner-Roth (1996)
-1
B , dm ·mol

-0.8
-1
3

Span and Wagner (2003) 1

-1.2 1.1
Outcalt and McLinden (1996)
-1.4
-1.6 Span and W agner (2003) 0.5
1 Tamatsu et al. (1993)
-1.8 R-152a
0.25
-2
-1
cv , kJ·kg ·K
-1

150 200 250 300 350 400 450 500 550 600 This work
0.9
T, K 0.1

1
Tillner-R oth (1996): infinite va lues 0.8
Ta m a tsu e t a l. (1993)
0.5 Outcalt and McLinden (1996)
O utc a lt and M cL inde n (1996)
0.7
-2

Y okozeki et al. (1998)

C , dm ·m ol

0.0001 MPa
0
6

Spa n a nd W a gne r (2003) 0.6

-0.5 R-152a
This w ork
0.5
R -152a
150 200 250 300 350 400
-1
T, K
150 250 350 450
T,K

Fig. 7.4 Comparison of second and third virial coefficients and gaseous specific heat
consistency for R-152a.

Third virial coefficient values for R-152a were derived from several existing

equations of state to revealing their behavior as results shown in Fig. 7.4. Although most

of the equations of state change their values sharply down at a certain temperature when

temperature decreasing, large discrepancy values are indicated at lower temperatures.

Similar discrepancy for specific heat in the gaseous phase near saturation is indicated.

Since the fundamental equation of state was developed with consideration of the second

and third virial coefficients, the specific heat values can be reproduced with the values

close to the values calculated from the equation with accurate virial coefficients.

131
 These all results confirm that effects of third and higher order virial coefficients on

caloric properties near saturation are very significant. Since the virial equation of state

has a theoretical background, these values are more believed as reliable values than

those obtained from other empirical equations of state only based on experimental data.

This study is the first study that concerns on the third virial coefficients on development

of fundamental equations of state and consistencies of thermodynamic properties in the

gaseous phase.

Virial equations of state are often successfully used to represent thermodynamic

properties in gaseous phase. In practice, the equations are expressed in a truncated form

of serial polynomial relation of density. On the other hand fundamental equations of

state have relations to virial coefficients without limiting number in series of its

coefficients. Analysis from the fitting works of virial equations of state points out that

the virial coefficients depend on series number of coefficients and wide range of density

when the equation is developed. Therefore, determination of reliable third virial

coefficients is difficult in practice. Since fundamental equations of state have universal

relations among thermodynamic properties, consistency relations among

thermodynamic properties should be existed in their behaviors in addition to accuracy

for representing measurable thermodynamic properties should be performed well.

Consequently, accuracies and consistencies among the thermodynamic properties are

essentially considered. Based on this viewpoint, it can be judged that the new

fundamental equations of state can give reliable thermodynamic properties including

third virial coefficients.

The investigation results on the existing equations of state including

internationally recommended ones indicate that all of them cannot satisfy all the

judgment criteria of this study. The development results of fundamental equations of

132
state verify that adding input parameters can eliminate the lacks in the existing

equations of state. Virial coefficients as parameters that have not been introduced on

other modeling than ours are very dilemmatic in preparation of the reliable data as input

data. Reliable coefficients are very difficult to be derived from both theoretical and

experimental frameworks. Strategy involving the virial coefficients until the third

coefficient in the optimization procedure would contributes an optimal equation of state

with reasonable both theoretical and empirical approaches. An improvement of the

reliability and avoiding the model missed by systematic error of data can be conducted

by this way.

Considering the above results, reevaluating tasks for the recent equations of state

for HFC refrigerants to be International Standard equations of state are very significant.

Reevaluating International Standard equations of state by adding new several criteria as

proposed in development of fundamental equations of state in this study should be

conducted. These efforts may be bring the development work with more comprehension

from the universal principle.

7.2 Extrapolation Behavior of Fundamental Equations of State

Since the four fundamental equations of state are not theoretical equations of state,

their surface behaviors should be carefully assessed especially in the critical region and

the region where experimental data are not available. Assessments cover isobaric lines

of caloric properties and isothermal lines of PVT property. Caloric properties can

change drastically, approaching either zero or infinity, and it is no longer possible to

distinguish the gas from liquid at the critical point. All of the extrapolation results are

shown in Figs. 7.5 to 7.8 for R-32, R-125, R-134a, and R-152a, respectively.

133
 4.8 1.6
4.3 R-32 R-32
10
3.8 1.4
7.5
cp , kJ -1·kg -1·K-1

cv , kJ -1 ·kg -1 ·K -1
3.3 1.2 10 100 MPa
15
2.8 15
2.3 1 5
3
7.5 30
1.8 1 1.5
100 MPa 0.8 0.5
1.3 3 5
1 1.5 0.25
0.25 0.5 0.1
0.8 o
0.6 0.01
cp o
cv
0.3 0.4
125 225 325 425 525 125 225 325 425 525
T, K T, K
1600 100
R-32 R-32
1400 450 K
10
1200
280 K
100 MPa 1
1000
w , m·s -1

70 0.1 220 K
800
0.001

p , MPa
50
0.01

185 K
600 0.01
0.10

30
0.25
0.5
1.0

400 15 160 K
1.5

0.001
3.0

10
200 140 K
7.5 0.0001
0
125 225 325 425 525 0.00001
T, K 0.0001 0.001 0.01 0.1 1 10 100 1000 10000
3 -1
v, m ·kg

Fig. 7.5 Caloric and PVT properties for R-32 from the new fundamental equation of
state.

3.0 1.1 70 MPa

R-125 R-125 15
2.5 1.0 10
5 7
5
0.9
cv , kJ·kg -1·K-1

3.5
cp , kJ·kg -1·K-1

2.0 3
7 2
1
0.8
10 0.75 cvo
1.5 15 100 MPa 0.7 0.25
2 3 3.5 0.1
1.0 0.7 1.5 0.6
0.25
0.1 cp o 0.01
0.01 0.5
0.5
150 200 250 300 350 400 450 500 550 150 200 250 300 350 400 450 500 550
T, K T, K
1400 100
R-125 R-125
1200 T = 500 K
10
1000
285
1
w , m·s -1

p , MPa

800 100 MPa 250

225
600 70 0.1
50 205
0.01

35
0.1

400 25
0.25

185
0.01
0.75

15 172.52 K
1.5

200 7
5 0.001
0
0.0001 0.001 0.01 0.1 1 10 100
150 200 250 300 350 400 450 500 550
T, K v , m3·kg-1

Fig. 7.6 Caloric and PVT properties for R-125 from the new fundamental equation of
state.

134
 3.4 1.4
3.2 R-134a R-134a
5
3 100 MPa
2.8 1.2 5
2.6 4 7
4
cp , kJ·kg -1 ·K-1

cv , kJ·kg -1 ·K-1
7
2.4
2.2 1 2.5
10 1.5
2 1
1.8 15 100 MPa 0.5
1.6 0.8
0.1
1.4 2.5 o
1.2 1
1.5 0.01
cv
1 0.1 0.5
0.6
0.8 cp
o
0.6
0.4 0.4
150 200 250 300 350 400 450 500 550 150 200 250 300 350 400 450 500 550
T, K T, K
1600 100
R-134a R-134a
1400 500 K
10
1200 340

1000 1 290
w , m·s -1

260

p , MPa
800 100 MPa 0.1 240
70 220
600 50 205
0.01
0.01

35 190
0.1

400 25
0.5

15 180
0.001
1
1.5

200 7 10 170 K
2.5

5
0 0.0001
150 200 250 300 350 400 450 500 550 0.0001 0.001 0.01 0.1 1 10 100
T, K v , m3·kg-1

Fig. 7.7 Caloric and PVT properties for R-134a from the new fundamental equation of
state.

6.0 1.9
5.5 R-152a R-152a
5.0 1.7 100 MPa
7
4.5 5 7 70
1.5 50
cp , kJ·kg -1·K-1

cv , kJ·kg -1 ·K-1

10
4.0
3.5 4
1.3 3
5 2
3.0 10 1
1.1 0.5
2.5 15 MPa
0.25
2.0 3
25 35
70 100 0.9 0.15
2
1.5 1
0.7
0.01
0.1 0.25
0.5 cv o
1.0 0.01 cp o

0.5 0.5
150 200 250 300 350 400 450 500 550 150 200 250 300 350 400 450 500 550
T, K T, K
1800 100
R-152a 500 K R-152a
1600 10
1400
1
1200 275
w , m·s -1

250
1000 0.1
p , MPa

230
100 MPa 210
800 70 0.01 190
0.01

50
600
0.1

35
175
0.25

25 0.001
0.5

400 15 165
1

7 10
200 0.0001 154.56
3

5
0 0.00001
150 200 250 300 350 400 450 500 550 0.0001 0.001 0.01 0.1 1 10 100 1000
T, K
v , m3·kg-1

Fig. 7.8 Caloric and PVT properties for R-152a from the new fundamental equation of
state.

135
 Figure 7.5 shows isobaric lines of isobaric specific heat, isochoric specific heat

and speed of sound, and isothermal lines of PVT property that were derived from the

fundamental equation of state for R-32. Similar result can be represented for the

fundamental equation of state for R-125 as shown in Fig. 7.6. The fundamental

equations of state for R-134a and R-152a of this study are also able to derive the right

behaviors of the thermodynamic surface as results shown in Figs. 7.7 and 7.8. Values of

the specific heats of the states toward the critical point increase extremely. The

speed-of-sound values also appear decreasing when the states toward the critical point.

It should be noted that the fundamental equations of state are not assigned to calculate

to caloric properties at the critical region. Additionally, the input data near critical

region is not available so that they were not introduced on the development. Rationally,

these results confirm the fundamental equations of state can reproduce both PVT and

caloric properties well.

7.3 Ideal Curves of Fundamental Equations of State

The ideal curves are functions to judge reasonable fluid surface of fundamental

equations of state if the reasonable behaviors can be derived. The ideal curves are the

curves consisting of relation of the ideal-gas law and its derivatives as given in Section

1.2.2. These ideal curves were not introduced as input data, but they were assessed as

long as development of the fundamental equations of state. The curves were

investigated to reveal behavior of the equations of state for the states at very extreme

pressures and temperatures. Comparisons with other equations of state especially for

International Standard equations of state are compiled as results shown in Figs. 7.9 to

7.12.

Figure 7.9 demonstrates that the selection of the structural form for the

136
fundamental equation of state for R-32 is reasonable to represent the thermodynamic

properties over wide range including the extrapolated range or the range where

experimental data are not available. The International Standard equation of state

(Tillner-Roth and Yokozeki 1997) cannot represent reasonable Joule inversion curve.

1000
Joule inversion curve R-32
100 Joule-Thomson inversion curve
Ideal curve
p /pc

10
Saturation line
1
Boyle curve
0.1
0.1 1 10
T /T c
Fig. 7.9 Comparison of ideal curves for R-32. (—) This work; (----) International
Standard equation of state (Tillner-Roth and Yokozeki 1997).

1000
Joule inversion curve R-125
100 Joule-Thomson inversion curve
Ideal curve
p /pc

10
Boyle curve
1 Saturation line

0.1
0.1 1 10
T /T c

Fig. 7.10 Comparison of ideal curves for R-125. (—) This work; (----) International
Standard equation of state (Piao and Noguchi 1998).

Comparison of ideal curves for two equations of state for R-125 is given in Fig.

7.10. The International Standard equation of state (Piao and Noguchi 1998) cannot

represent ideal curves, not only unreasonable Joule inversion curve but also the other

curves in right forms. The fundamental equation of state from this work performs a

137
rational Joule inversion curve and the other curves.

1000
Joule inversion curve R-134a
100 Joule-Thomson inversion curve
Ideal curve
p/pc

10
Saturation line
1
Boyle curve
0.1
0.1 1 10
T /T c

Fig. 7.11 Comparison of ideal curves for R-134a. (—) This work; (----) International
Standard equation of state (Tillner-Roth and Baehr 1994).

1000
Joule inversion curve
100 Joule-Thomson inversion curve
Ideal curve
p /pc

10
Boyle curve
1 Saturation line
R-152a
0.1
0.1 1 10
T /T c

Fig. 7.12 Comparison of ideal curves for R-152a. (—) This work; (----) Equation of
state (Tillner-Roth 1996).

Comparison of ideal curves for R-134a is shown in Fig. 7.11. The International

Standard equation of state (Tillner-Roth and Baehr 1994) cannot represent reasonable

Joule inversion curve. It demonstrates that the selection of the structural form for the

fundamental equation of state is reasonable to represent the thermodynamic properties

over wide range including the extrapolated range or the range where experimental data

are not available.

138
 Figure 7.12 shows comparison of characteristics of ideal curves for two

fundamental equations of state for R-152a. Equation of state (Tillner-Roth 1996)

represents unreasonable Joule inversion curve. This Joule inversion curve cannot

gradually change its slope to down. All reasonable ideal curves derived from the

fundamental equation of state for R-152a indicate that the selection of right structural

functional form was suitable for representing thermodynamic surface at high pressures

and temperatures as shown in the figure.

Structural form of the equation of state may significantly effect on the

characteristics of ideal curves of equation of state. The smooth curves may indicate that

the equation can represent all thermodynamic properties over wide-range of fluid phase

in rational uncertainties for the properties. In accordance with assessments were

conducted previously, the four fundamental equations of state can be applied to

calculated ideal-gas properties from 120 K to 1000 K. Extrapolations of caloric

properties is reasonably represented in wide range and reasonable characteristics of all

ideal curves derived from the four fundamental equations of state can be reproduced.

Considering these behaviors, the four fundamental equations of state may represent

right thermodynamic surface for temperatures around 2500 - 3200 K and pressures

around 350 - 1000 MPa depending on the fundamental equations of state as shown in

the previous figures. Thermodynamic properties at these ranges cannot be assessed due

to limited of theoretical and experimental data.

7.4 Reliability of Fundamental Equations of State

Study of development of the four fundamental equations of state for HFCs

confirmed that adding parameters in the thermodynamic modeling can eliminate the

lacks from inconsistent relations to other properties. In addition to bases of experimental

139
data, intermolecular potential background, and reasonable mathematical and physical

aspects were considered in this study. Targets involved in the development are a unified

relation of intermolecular potential model, reasonable characteristics of ideal curves,

and reasonable gaseous specific heat values at low temperatures.

The specific heat values play important role for thermodynamic analyses, as the

accurate properties are very essential information in thermodynamic cycle calculation.

Experimental systematic error due to shortage of experimental data may trap the

equation of state if the development relies in experimental bases only. This caused the

inconsistent behaviors of the fundamental equations of state for HFC refrigerants.

Regarding the previous paradigm were used on development of the existing equations

of state, it can be revealed that unreasonable behaviors of third virial coefficients from

the most existing equations of state may be caused of insufficient theoretical

background.

Since the fundamental equations of state were established by including unified

relation principle of intermolecular potential for determination of third virial coefficient

in the validity range, their degrees of freedom can be reduced to eliminate the

inconsistencies. Assessment results of the new fundamental equations of state for R-32,

R-125, R-134a, and R-152a confirm that unified relation of intermolecular potential

model, suitable values of gaseous specific heat, and reasonable characteristics of ideal

curves could be performed.

As being fundamental equations of state for thermodynamic properties, the

equations should be able to revealing and predicting all thermodynamic relation among

the properties well. A unified relation between the intermolecular potential energies and

the fundamental equations of state should be consistently maintained, besides the

rational relations of the fundamental equations of state to other equations or equations

140
of state accurately representing in specific regions such as gaseous phase, liquid phase

or critical region should be considered in the development. By referring relations of

Helmholtz free energy function to other properties, it can be revealed that the reliable

fundamental equations of state should have rational behaviors of the free energy

function, and among its derivatives until the second derivatives, whether against density

or temperature, or both density and temperature, have crucial effects on thermodynamic

properties.

More accurate information about behaviors of fundamental equations of state can

be obtained from assessments. These assessments are very important to know their

reliabilities and accuracies. Learning from the previous involved parameters and the

currently appeared disadvantages, any essential improvement is confirmed in this study.

The existing equations of state are thoroughly investigated and then followed by

hypothesis according to comprehensive considerations from thermodynamic and

modeling aspects. Considering consistency correlations of the fundamental equations of

state to the intermolecular potential energy, and to other equations of state that have

theoretical background covering a specific region, and additional criteria are involved in

this study.

Comparisons of the existing equations of state for HFC refrigerant are precisely

examined according to several criteria strictly correlating to the focuses of the present

study. The existing equations of state for R-143a were also included even though the

fundamental equation of state for this fluid was not developed in this study.

Nevertheless, more information of the equations of state for HFC refrigerants can be

pointed out. The criteria include the second and the third virial coefficients, consistency

with the derived properties from the accurate virial equation of state in gaseous phase,

characteristics of ideal curves, and accuracy representation of the thermodynamic

141
properties. Summary of the results are qualitatively given in Table 7.1.

Table 7.1 Summary of Assessment for Equations of State of HFC refrigerants

Equation of State
Cd Sp. Ht.e Curv.f Accur.g
Fluid Author
Outcalt and McLinden (1995)b × × × ○
Tillner-Roth and Yokozeki (1997)*,a × × □ ○
R-32
Vasserman and Fominsky (2001)a ∆ × × ○
This work ○ ○ ○ ○
Outcalt and McLinden (1995)b ∆ × × ○
*,b
Piao and Noguchi (1998) × × × ○
R-125 Sunaga et al. (1998)a × × □ ○
Vasserman and Fominsky (2001)a × ∆ × ○
This work ○ ○ ○ ○
a
Huber and Ely (1994) × × □ ○
b
Huber and Ely (1992) ○ ∆ × ○
R-134a Piao et al. (1991) c
∆ ○ × ∆
Tillner-Roth and Baehr (1994) *,a ○ ○ □ ○
This work ○ ○ ○ ○
Lemmon and Jacobsen (2000)*,a × × ○ ○
a
Li et al. (1998) × × □ ○
R-143a
Outcalt and McLinden (1997)b ∆ × × ○
Piao et al. (1998)b × ∆ × ○
Outcalt and McLinden (1996)b ∆ × □ ○
c
Tamatsu et al. (1993) ∆ × × ∆
R-152a
Tillner-Roth (1996)a × × □ ○
This work ○ ○ ○ ○
* a
International Standard equation of state recommended by IEA-Annex-18; Helmholtz
equation of state; b MBWR equation of state; c virial equation of state; d behavior of
third virial coefficients; e behavior of specific heat in gaseous phase; f Ideal curves; g
accuracy of the equation of state; (○) acceptable; (∆) satisfied; (□) unreasonable
characteristics of ideal curves; and (×) not satisfied.

This study used references for the second virial coefficient, the third virial

coefficient, and the truncated virial equations of state as described in Section 4.4.3.

Qualitative judgments for the virial coefficient form are based on consistent change and

142
form of the curves in the validity range. The qualification for comparison of the virial

equations of state and the fundamental equations of state are judged by revealing of

isochoric specific heats in gaseous phase. In comparison of speed-of-sound values are

not found significant discrepancies in the gaseous phase.

As shown in Table 7.1, most the equations of state both Helmholtz and MBWR

types can represent experimental data in acceptable accuracies, but equations of state

with virial form by Piao et al. (1991) for R-134a and by Tamatsu et al. (1993) for

R-152a, which are less satisfied due to worse representation in the liquid phase and at

the saturation especially for the caloric properties of experimental data. In generally, a

fundamental equation of state for R-134a by Tillner-Roth and Baehr (1994) qualitatively

obeys the above criteria. It is the best one among the existing equation of state, except it

cannot reasonably represent Joule inversion curve. Most of them suffer from

characteristics of ideal curves excluding the fundamental equation of state for R-143a

by Lemmon and Jacobsen (2000). These results confirm that the previous considerations

of assessment and development used for the existing equations of state are still

inadequate.

Reasonable behaviors of thermodynamic properties may indicate the reliability of

fundamental equations of state, besides their abilities to represent reliable experimental

data. Consistent with thermodynamic relations, the fundamental equations of state

appear strictly relating to virial coefficients. Since most thermodynamic properties can

be derived from Helmholtz free energy until secondary derivatives, considering the

second and the third virial coefficient behaviors on thermodynamic modeling are

reasonable to reduce degrees of freedom of thermodynamic model.

Consistent behaviors of the fundamental equations of state with respect to specific

equations in specific region such as virial equation of state for the gaseous phase are

143
rational to be considered. As successful representation of virial equations of state in

gaseous phase at low densities and having theoretical backgrounds, it may be essential

to be involved on the modeling. Moreover, this way can eliminate scarcities of specific

heat data in the gaseous phase.

7.5 Correlation of Virial Coefficients and Ideal Curves

Virial coefficients are thermodynamic quantities relating to intermolecular

potential energy. Observing relation of virial equation of state, it can be revealed that

higher order of the coefficients will highly affect on other thermodynamic properties at

higher density. Considering the isobaric line of specific heat with respect to temperature,

large discrepancy of specific heat in the gaseous phase is highly possible that it is

caused by inconsistent behavior of higher-order virial coefficient derived from the

equations of state.

Ideal curves were derived from the existing equations of state and the comparison

was conducted to disclose correlations of the curves to intermolecular potential models

especially for virial coefficients. For comparison in the valid range of the fundamental

equations of state can be summarized as listed in Table 7.1. Since tendency value of the

second and third virial coefficient is an asymptotic value to zero for very higher

temperatures from the critical temperature and the states at lower pressure and higher

temperatures behave ideal gas and have low density, it can be revealed that the both

virial coefficients insignificantly affect on the region. These behaviors agree with facts

of assessment of fundamental equations of state. Although their second of third virial

coefficients are reasonable behaviors but it is not ensured that they have reasonable

characteristics of the ideal curves. An equation of state can derive adequate third virial

coefficients but it also failures to represent the reasonable ideal curves. The equations of

144
state agree with this behavior such as shown in works of Huber and Ely (1992) and

Tillner-Roth and Baehr (1994). Conversely, even reasonable characteristics of the ideal

curves can be derived from an equation of state, it cannot be ensured that the right third

virial coefficients can be derived such as the fundamental equation of state by Lemmon

and Jacobsen (2000).

Furthermore, correlation of virial coefficients is revealed from plot the second and

third virial coefficients at extreme temperature in the region range of the ideal curves.

As shown in Fig. 7.13, discrepancy behaviors both the second and third virial

coefficients are performed by several equations of state. In case of intermolecular

behaviors are extrapolated to extreme temperatures, the second virial coefficients should

increase with increment to be smaller at extreme temperatures. On the other hand the

third virial coefficients should decrease with decrements to be smaller at extreme

temperatures.

0.1 0.025
R-134a
0.08 R-125
R-143a 0.02
0.06 R-152a
0.04
0.015
C , dm6·mol-2
B , dm3·mol-1

0.02
0 0.01 R-134a
R-125 R-32
-0.02
0.005 R-152a
-0.04
R-143a
-0.06 0
-0.08 R-32

-0.1 -0.005
400 1400 2400 3400 400 1400 2400 3400
T, K T, K

Fig. 7.13 Second and third virial coefficient behaviors at extreme temperatures. (---)
International Standard equations of state for R-32 (Tillner-Roth and Yokozeki 1997),
R-125 (Piao and Noguchi 1998), R-134a (Tillner-Roth and Baehr 1994) and R-143a
(Lemmon and Jacobsen 2000); (▬) This work.

145
 An interesting case of the second virial coefficients is that the International

Standard equation of state for R-32 represents the values decreasing after reaching at a

certain high temperature (around 1500 K). However, various cases can be noted for the

third virial coefficients at these extreme temperatures. The International Standard

equations of state for R-125 and R-134a and the new fundamental equation of state for

R-125 of this study perform the third virial coefficients with values increasing after

decreasing to a certain temperature. Only the International Standard equation of state for

R-143a and the new fundamental equation of state for R-152a perform consistent values

of the third virial coefficients at extreme temperatures. The International Standard

equation of state for R-32 performs the third virial coefficients decrease continuously,

but the values change to negative values. Although the several new fundamental

equations of state cannot perform rational behavior from intermolecular aspect at

extreme pressure and temperature, they all can reasonably represent their ideal curves.

At extreme temperatures, kinetic energy of molecule is very large. This can be

agreed due to the energy has function of temperatures such as described in statistical

thermodynamics. Such as pointed out for characteristics of ideal curves for nitrogen and

carbon dioxide (Span and Wagner 1997), the reasonable characteristics of ideal curves

can be used to judge the fundamental equations of state to reproduce thermodynamic

properties at extreme temperatures and pressures. Therefore, behavior of the third virial

coefficients is not significant to affect on thermodynamic properties. Interaction of

among molecules is very weak in comparison of the kinetic energy due to the condition

at extreme temperatures. In observation of characteristics of ideal curves and behaviors

of third virial coefficients at extreme temperatures, it is supposed that the intermolecular

potential energy does not contribute to the thermodynamic properties for the state at

146
outer range of the ideal-gas states. The approaches for fluid as molecular cluster would

not be appropriate there.

These facts proven that thermodynamic modeling by introducing intermolecular

potential behavior on development of fundamental equation of state is inadequate to

ensure that reasonable characteristics of ideal curves. Since introducing the ideal curves

as modeling parameters on the development have any difficulty, the assessments after

the equations of state established in the development process is essential to be

conducted. Therefore, more tasks on the development should be conducted not only to

good fit data modeling to the model but also various optimization targets are conducted.

All ideal curves can be derived from the fundamental equations of state of this

study. These reasonable characteristics of the ideal curves indicate the suitable structural

fundamental equations of state and uncertainties of the equations harmonically represent

thermodynamic properties for each property in overall fluid phase. Although the

experimental data may not cover all of the thermodynamic surfaces, the fundamental

equations of state assigned are valid. The characteristics of ideal curves, therefore, are

reasonable to consider in the development. Unfortunately, none of ways can be made to

directly consider on modeling so that it should be assessed in the optimization

procedure.

7.6 Thermodynamic Consistency and Optimization Procedure

Optimization procedure plays important role to find a result in accordance with

objective function used in the optimization process. On the other hand, consistencies

indicate that characteristics relating to the model resulted from the optimization process

for representing thermodynamic properties. The consistencies of the study include virial

coefficients, characteristics of ideal curves, and gaseous specific heat. Since

147
optimization procedure has target to make an efficient process, the procedure only

contributes to efficient optimization process and not relating to thermodynamic

consistency. The consistencies can be required when a satisfied result can be obtained

and the consistency criteria were included in the objective function or assessed in the

development process.

In accordance with the above rational considerations, consistencies don’t directly

relate the optimization procedure due to the ideal curves excluding on the regression

process. Therefore, analysis by Span and Wagner (1997) for equations of state with

optimization algorithm that can make consistency of ideal curves should be inquired

again, even though which the analysis was investigated simple compounds. It shows

that International Standard equations of state for R-32 (Tillner-Roth and Yokozeki 1997)

and R-134a (Tillner-Roth and Baehr 1994) also developed with the same optimization

algorithm cannot derive reasonable Joule inversion curve. This fact agrees the present

opinion for the optimization procedure relating to efficiency of the optimization process,

and the consistencies relating to the objective function in the optimization procedure.

The second and third virial coefficients in validity range of the fundamental

equations of state were introduced on the development. Satisfied result can be achieved

due to the objective function including the consistencies of virial coefficients. As

discussed in Section 7.5, assessments of consistency behavior of thermodynamic

properties are simultaneously conducted in the development process. Therefore,

optimization method play important role to fit the model.

148
 Chapter 8
Conclusion

Fundamental equations of state for four HFC refrigerants consisting of R-32,

R-125, R-134a, and R-152a were established with the reasonable second and third virial

coefficients having theoretical background of intermolecular potential theory. Ideal

curves were involved in assessment to find acceptable the fundamental equations of

state whose structural functions can represent reasonable behaviors and avoid trapping

of the systematical experimental error. The fundamental equations of state can provide

the rational thermodynamic properties even at very low temperatures, near saturation in

gaseous phase, and in the region beyond the region where experimental data are

available. Gaseous specific heats near saturation are reasonably represented without any

thermodynamic inconsistency. The fundamental equation of state has reliable third virial

coefficient making it possible to derive reliable thermodynamic properties near

saturation in gaseous phase even at low temperatures that are useful to precisely design

refrigeration and air-conditioning systems.

The typical estimated uncertainties for the four fundamental equations of state are

0.05 - 0.1 % in density for the superheated-vapor and liquid phases, 0.2 - 0.5 % in

density for the supercritical region, 0.01 - 0.02 % in speed of sound for the gaseous

phase, 0.5 - 1 % in speed of sound for the liquid phase, 0.6 - 2 % in isochoric specific

heat for the liquid phase, and 0.5 - 1 % in isochoric specific heat for the liquid phase.

The uncertainties of saturation properties are 0.05 - 0.1 % in vapor pressure, and 0.25 -

0.5 % and 0.5 - 1.2 % in saturated-liquid and saturated-vapor densities, respectively.

Those all uncertainties can be achieved in a wide range of their validity. The

149
fundamental equations of state are valid from the triple point up to pressures and

temperatures until highest input data used, i.e. 70 MPa and 450 K for R-32, 72 MPa and

480 K for R-125, 70 MPa and 460 K for R-134a, and 55 MPa and 450 K for R-152a.

Investigation results for the properties calculated at extreme temperatures from the

fundamental equations of state suggested that rational characteristics of ideal curves are

not significantly influenced by the intermolecular potential behavior, although structural

functional form of the fundamental equation of state may influence in the ideal curves.

The ancillary equations for saturation properties of the HFC refrigerants were

established with considerations of experimental data and consistent relation with the

fundamental equations of state. The equations consist of vapor pressure, saturated-liquid

density and saturated-vapor density equations. Non-analytic behavior at the critical

point is included in the ancillary equations. These equations play important role on

development of the fundamental equation of state, in addition to directly calculate the

saturation properties for engineering interests. Uncertainties of most of the vapor-

pressure equations are predicted in range of 0.03 to 0.08 %. Slightly better uncertainties

are performed in comparison of the fundamental equations of state. The estimated

uncertainties for saturated-liquid density equations are in range of 0.08 to 0.2 % and for

saturated-vapor densities are in range of 0.5 to 1 %.

The optimization procedure that was developed in this study was successfully

applied to develop fundamental equations of state for the four HFC refrigerants. The

procedure was also applied to simultaneously develop ancillary equations for saturation

properties so that the same structural equation can be obtained.

150
 List of Author’s Papers
The following papers are the papers written by the author when enrolls in the doctoral
course.

Journal Papers

1. I M. Astina and H. Sato, A Fundamental Equation of State for 1,1,1,2-

Tetrafluoroethane with Intermolecular Potential Energy Background and Reliable
Ideal-Gas Properties, Fluid Phase Equilibria (in press, 2004).
2. I M. Astina and H. Sato, A Rational Fundamental Equation of State for
Pentafluoroethane with Theoretical and Experimental Bases, International. J.
Thermophysics (in press, 2004).
3. I M. Astina and H. Sato, A Rational Helmholtz Fundamental Equation of State for
Difluoromethane with an Intermolecular Potential Background, International J.
Thermophysics 24(4), 963-990 (2003).

Conference Papers
International Conference Papers
1. I M. Astina and H. Sato, A Rapid Genetic Optimization Technique for Rational
Thermodynamic Modeling Having Reliable Third Virial Coefficient, 15th
Symposium on Thermophysical Properties, June 2003, Boulder, U.S.A.
2. I M. Astina and H. Sato, Development of Fundamental Equation of State for
HFC125 with Theoretical and Experimental Bases, Asian Conference on
Refrigeration and Air Conditioning, Dec. 2002, Kobe, Japan.
3. I M. Astina and H. Sato, Method for Developing Thermodynamically Reliable
Fundamental Equations of State and an Equation of State for HFC Refrigerant, R32,
2002 European Conference on Thermophysical Properties, Sept. 2002, London,
U.K.

National Conference Papers

1. I M. Astina and H. Sato, A New Viewpoint for Development of Helmholtz
Fundamental Equation of State, The 24th Japan Symposium on Thermophysical
Properties, Oct. 2003, Okayama, Japan (in Japanese).
2. I M. Astina and H. Sato, Development of Thermodynamic Equations of State for
HFC32 and HFC125 with Background of Intermolecular Potential, The 14th Japan
Symposium on Air Conditioning and Refrigeration, Nov. 2002, Okayama, Japan (in
Japanese).

Other
I M. Astina and H. Sato, State of the Art on Thermodynamic Modeling for HFC
Refrigerants: The Recent Challenge to Develop Fundamental Equations of State, CD
Proc. 2nd Doctoral Conference Asia Pacific Rim Universities, Mexico City, Mexico
(2003).

151
 References

Abbott, M. M., Fluid Phase Equil. 37, 29-62 (1987).

Al-Shafei, M. O., and K. Mecarik, Int. J. Refrig. 20(1), 38-48 (1997).
Ambrose, D., J. F. Counsell, and A. J. Davenport, J. Chem. Thermodyn. 2, 283-294
(1972).
Angeline, P. J., Proc. 1998 IEEE Int. Conf. Evolutionary Computation, 84-89 (1998).
Aoyama, H., G. Kishizawa, H. Sato, and K. Watanabe, J. Chem. Eng. Data 41(5),
1046-1051 (1996).
Arita, K., T. Tomozawa, Y. Nakagawa, and Y. Yoshida, Fluid Phase Equil. 63, 151-156
(1991).
Arnaud, D., S. Macaudiere, and L. Niveau et S. Wosinski, Proc. 18th Int. Cong. Refrig.,
664-668 (1991).
Assael, M. J., J. P. M. Trusler, and T. F. Tsolakis, Thermophysical Properties of Fluids:
an Introduction to their Prediction, Imperial College Press, London (1998).
Atmar, W., IEEE Trans. Neural Networks 5(1), 130-147 (1994).
Baehr, H. D., and R. Tillner-Roth, J. Chem. Thermodyn. 23, 1063-1068 (1991).
Baroncini, C., R. Camporese, G. Giuliani, G. Latini, and F. Polonara, High Temp. High
Pres. 25, 459-464 (1993).
Basu, R. S., and D. P. Wilson, Int. J. Thermophys. 10(3), 591-603 (1989).
Bates, D. M., and D. G. Watts, Non Linear Regression Analysis and Its Applications,
John Wiley & Sons, New York (1988).
Beckermann, W., and F. Kohler, Fortschr.-Ber. VDI 19(67), Duesseldorf, FRG,
VDI-Verlag (1993).
Beckermann, W., and F. Kohler, Int. J. Thermophys. 16(2), 455-464 (1995).
Beliajeva, O. V., A. J. Grebenkov, T. A. Zajatz, and B. D. Timofejev, Dokl. Akad. Nauk
SSSR 38 (1994).
Beliajeva, O. V., A. J. Grebenkov, T. A. Zajatz, and B. D. Timofejev, Int. J. Thermophys.
20(6), 1689-1697 (1999).
Bethea, R. M., B. S. Duran, and T. L. Boullion, Statistical Methods for Engineers and
Scientists, Second Edition, Marcel Dekker, New York (1985).
Bhattacharya, G. K., and R. A. Johnson, Statistical Concepts and Methods, John Wiley &
Sons, New York (1977).
Bhatti, M. S., ASHRAE Trans: symposia CH99-21-3, 1186-1206 (1999).

152
Blanke, W., and R. Weiss, Fluid Phase Equil. 80, 179-190 (1992).
Blanke, W., G. Klingenberg, and R. Weiss, Int. J. Thermophys. 16(5), 1143-1153 (1995).
Bouchot, C., and D. Richon, Proc. Int. Conf. "CFCs, The day after", Joint Meeting of
IIR Commissions B1, B2, E1 and E2, 517-524, Padova (1994).
Bouchot, C., and D. Richon, Int. Electronic J. Physico-Chem. Data 3, 1-16 (1997).
Boyes, S. J., and L. A. Weber, J. Chem. Thermodyn. 27, 163-174 (1995).
Carnahan, N. F., and K. E. Starling, J. Chem. Phys. 51(2), 635-636 (1969).
Carnahan, N. F., and K. E. Starling, AIChE J. 8(6), 1184-1189 (1972).
Chae, H. B., J. W. Schmidt, and M. R. Moldover, J. Phys. Chem. 94(25), 8840-8845
(1990).
Defibaugh, D. R., and G. Morrison, Fluid Phase Equil. 80, 157-166 (1992).
Defibaugh, D. R., G. Morrison, and L. A. Weber, J. Chem. Eng. Data 39(2), 333-340
(1994).
Defibaugh, D. R., and M. R. Moldover, J. Chem. Eng. Data. 42(1), 160-168 (1997).
de Garis, H., Proc. 1990 Int. Joint Conf. Neural Networks, 194-197, Washington (1990).
Deiters, U. K., and K. M. De Reuck, Fluid Phase Equil. 161, 205-219 (1999).
DeJong, K. A., Computational Intelligence Imitating Life, Edited by Zurada, J. M.,
Marks II, R. J., and Robinson, C. J., IEEE Press, 125-134 (1994).
den Santis, R., F. Goroni, and L. Marrelli, Ind. Eng. Chem. Fundam. 15(3), 183-188
(1976).
de Vries, B., Deutschen Kälte- und Klimatechnischen Vereins Nr. 55, Hannover
University (1997).
Döering, R., H. Buchwald, and C. von Eynatten, Oesterreichisher Kälte- und
Klimatechnischer Verein (1994).
Draper, N. R., and H. Smith, Applied Regression Analysis, Second Edition, John Wiley
and Sons, New York (1981).
Dressner, M., and K. Bier, Fortschr.-Ber. VDI 3 (332), (1993).
Duarte-Garza, H. A., C. E. Stouffer, K. R. Hall, J. C. Hostle, K. N. Marsh, and B. E.
Gammon, J. Chem. Eng. Data 42(4), 745-753 (1997).
Duarte-Garza, H. A., and J. W. Magee, Int. J. Thermophys. 20(5), 1467-1481 (1999).
Farman, J. C., B. G. Gardiner, and J. D. Shanklin, Nature 315, 207-210 (1985).
Fisher, G. D., and T. W. Leland, Ind. Eng. Chem. Fundam. 9(4), 537-544 (1970).
Fisher, M. E., and G. Orkoulus, Physical Review Letters 85(4), 696-697 (2000).
Fogel, D. B., and J. W. Atmar, Biological Cybernetics 63, 111-114 (1990).
Fogel, D. B., IEEE Trans. Neural Networks 5(1), 3-14 (1994).
Fogel, J. L., Computational Intelligence Imitating Life, Edited by Zurada, J. M., Marks
II, R. J., and Robinson, C. J., IEEE Press, 135-146 (1994).

153
Fransson, E., A. Barreau, and J. Vidal, J. Chem. Eng. Data 37(4), 521-525 (1992).
Fröba, A. P., S. Will, and A. Leipertz, Int. J. Thermophys. 22(4), 1021-1033 (2001).
Fröberg, Carl-Erik, Numerical Mathematics: Theory and Computer Application, The
Benjamin/Cummings Publishing Company, California (1985).
Fu, Y. D., L. Z. Han, and M. S. Zhu, Fluid Phase Equil. 111, 273-286 (1995).
Fujiko, C., and J. Dickinson, Proc. 2th Int. Conf. Genetic Algorithms, Cambridge,
236-240 (1987).
Fujimine, T., H. Sato, and K. Watanabe, Int. J. Thermophys. 20(3), 911-922 (1999).
Fujiwara, K., S. Nakamura, and M. Noguchi, J. Chem. Eng. Data 43(1), 55-59 (1998).
Fukushima, M., Trans. JAR 8 (1), 65-75 (1991).
Fukushima, M., Trans. JAR 10(1), 87-93 (1993).
Fukushima, M., N. Watanabe, and T. Kamimura, Trans. JAR 7(2), 85-95 (1990a).
Fukushima, M., N. Watanabe, and T. Kamimura, Trans. JAR 7(3), 243-256 (1990b).
Fukushima, M., S. Ohotoshi, and T. Miki, Proc. 19th Int. Cong. Refrig. IVa, 207-214,
Hague (1995).
Geller, V. Z., E. G. Porichanskii, P. I. Svetlichnyi, and Y. G. Elkin, Kholod. Tekh. Tekhnol.
29, 43-48 (1979).
Gillis, K. A., Int. J. Thermophys. 18(1), 73-135 (1997).
Gillis, K. A., and M. R. Moldover, Int. J. Thermophys. 17(6), 1305-1324 (1996).
Giuliani, G., S. Kumar, F. Polanara, and P. Zazzini, Proc. Int. Conf. "CFCs, The day
after", Joint Meeting of IIR Commissions B1, B2, E1 and E2, 525-531, Padova
(1994).
Giuliani, G., S. Kumar, P. Zazzini, and F. Polonara, J. Chem. Eng. Data 40(4), 903-908
(1995).
Goodwin, A. R. H., and M. R. Moldover, J. Chem. Phys. 93(4), 2741-2753 (1990).
Goodwin, A. R. H., D. R. Defibaugh, and L. A. Weber, Int. J. Thermophys. 13(5),
837-854 (1992).
Grebenkov, A. J., O. V. Beliajeva, T. A. Zajatz, B. D. Timofejev, Proc. 4th Asian
Thermophys. Props. Conf., Tokyo, 311-314 (1995).
Grebenkov, A. J., Y. G. Kotelevsky, V. V. Saplitza, O. V. Beliajeva, T. A. Zajatz, and B.
D. Timofeev, Proc. Int. Conf. "CFCs, The day after", Joint Meeting of IIR
Commissions B1, B2, E1 and E2, 419-429, Padova (1994).
Grigiante, M., G. Scalabrin, G. Benedetto, R. M. Gavioso, and R. Spagnolo, Fluid Phase
Equil. 174, 69-79 (2000).
Guedes, H. J. R., and J. A. Zollweg, Int. J. Refrig. 15(6), 381-385 (1992).
Guggenheim, E. A., J. Chem. Phys. 13(7), 253-261 (1945).
Gunther, D., and F. Steimle, Int. J. Refrig. 20(4), 235-243 (1997).

154
Hall, K. R., and P. T. Eubank, Ind. Eng. Chem., Fundam. 15(4), 323-333 (1976).
Hankinson, R. W., and G. H. Thomson, AIChE J. 23(4), 653-663 (1979).
Harmans, A., and H. Knapp, Ind. Eng. Chem. Fundam. 19, 291-294 (1980).
He, M. G., Z. G. Liu, and J. M. Yin, Int. J. Thermophys. 23(6), 1599-1615 (2002).
Heyen, G., 2nd World Cong. Chem. Eng., 41-46, Montreal (1981).
Higashi, Y., Int. J. Refrig. 17(8), 524-531 (1994).
Higashi, Y., and T. Ikeda, Fluid Phase Equil. 125, 139-147 (1996).
Higashi Y., M. Ashizawa, Y. Kabata, T. Majima, M. Uematsu, and K. Watanabe, JSME
Int. J. 30, 1106-1112 (1987).
Hirschfelder, J. O., and W. E. Roseveare, J. Phys. Chem. 43, 15-35 (1939).
Holcomb, C. D., V. G. Niesen, L. J. Van Poolen, and S. L. Outcalt, Fluid Phase Equil.
91, 145-157 (1993).
Holcomb, C. D., and L. J. Van Poolen, Fluid Phase Equil. 100, 223-239 (1994).
Holste, J. C., H. A. Duarte-Garza, and M. A. Villamanan-Olfos, ASME Winter Annual
Meeting, 93-WA/HT-60, 1-6, New Orleans (1993).
Hou, H., J. C. Holste, B. E. Gammon, and K. N. Marsh, Int. J. Refrig. 15(6), 365-371
(1992).
Hozumi, T, T. Koga, H. Sato, and K. Watanabe, Int. J. Thermophys. 14(4), 739-762
(1993).
Hozumi, T., H. Sato, and K. Watanabe, J. Chem. Eng. Data 39(3), 493-495 (1994).
Hozumi, T., H. Sato, and K. Watanabe, Int. J. Thermophys. 17(3), 587-595 (1996).
Hozumi, T., H. Sato, and K. Watanabe, J. Chem. Eng. Data 42(3), 541-547 (1997).
http://unfccc.int/resource The Climate Change Convention and its Kyoto Protocol.
http://www.beyonddiscovery.org Beyond Discovery series.
http://www.nist.gov/public_affairs/chemical.htm NIST Standards are Elementary.
http://www.unep.org/ozone The Montreal Protocol on Substances that Deplete Ozone.
Huber, M. L., and J. F. Ely, Int. J. Refrig. 15(6), 393-400 (1992).
Huber, M. L., and J. F. Ely, Int. J. Refrig. 17(1), 18-31 (1994).
Hust, J. G., and R. D. McCarty, Cryogenics 7, 200-206 (1967).
Iso, A., and M. Uematsu, Physica A 156, 454-466 (1989).
Ichikawa, T., T. Hozumi, H. Sato, and K. Watanabe, Proc. 17th Japan Symp.
Thermophys. Props., 267-270, Tsukuba (1996).
Iglesias-Silva, G. A., and K. R. Hall, Fluid Phase Equil. 131, 97-105 (1997).
Iglesias-Silva, G. A., J. C. Holste , P. T. Eubank, K. N. Marsh, and K. R. Hall, AIChE J.
33, 1550-1556 (1987).
Jacobsen, R. T., and R. B. Stewart, J. Phys. Chem. Eng. Data 2, 757-922 (1973).
Jacobsen, R. T., S. G. Penoncella, and E. W. Lemmon, Fluid Phase Equil. 80, 45-56

155
 (1992).
Japanese Association Refrigeration - Japan Association Japan Flon Gas,
Thermophysical Properties of Environmentally Acceptable Fluorocarbons:
HFC-134a and HCFC-123, Tokyo (1990).
Kabata, Y., S. Tanikawa, M. Uematsu, and K. Watanabe, Int. J. Thermophys. 10(3),
605-615 (1989).
Kabata, Y., Y. Higashi, M. Uematsu, and K. Watanabe, Trans. JAR 5(1), 97-108 (1988).
Kamei, A., C. C. Piao, H. Sato, and K. Watanabe, ASHRAE Trans. 96(1), 141-149
(1990).
Kamei, A., S. Beyerlein, and R. T. Jacobsen, Int. J. Thermophys. 16(5), 1155-1164
(1995).
Kanungo, A., T. Oi, A. Popowicz, and T. Ishida, J. Phys. Chem. 91(15), 4198-4203
(1987).
Keenan, J. H., F. G. Keyes, P. G. Hill, and J. G. More, Steam Tables, International
Edition, John Wiley & Sons, New York (1969).
Kerninghan, B. W., and D. M. Ritchie, The C Programming Language, 2nd Ed.,
Prentice Hall PTR, Englewood Cliffs (1988).
Kincaid, J. F., and H. Eyring, J. Phys. Chem. 43, 37-47 (1939).
Kiselev, S. B., Fluid Phase Equil. 147, 7-23 (1998).
Kiselev, S. B., and J. V. Sengers, Int. J. Thermophys. 14(1), 1-32 (1993).
Kiselev, S. B., J. C. Rainwater, and M. L. Huber, Fluid Phase Equil. 150-151, pp.
469-478 (1998).
Klomfar, J., J. Hruby, and O. Sifner, Int. J. Thermophys. 14(4), 727-738 (1993).
Kojima, T., and H. Sato, CD Proc. European Conf. Thermophys. Props., (2002).
Kojima, T., M S Thesis, Keio University (2001).
Koza, J. R., Proc. 4th Int. Conf. Genetic Algorithms, 37-44, Morgan Kaufmann (1991).
Kraft, K., and A. Leipertz, Int. J. Thermophys. 15(3), 387-399 (1994).
Kubota, H., Chemistry Express 2(7), 397-402 (1987).
Kubota, H., T. Yamashita, Y. Tanaka, and T. Makita, Int. J. Thermophys. 10(3), 629-637
(1989).
Kubota, H., T. Sonani, and Y. Kunimoto, Fluid Phase Equil. 104, 413-419 (1995).
Kubota, M., H. Sato, and K. Watanabe, Proc. 27th Japanese Joint Conf. Air Cond. and
Refrig. 173-176, (1993).
Kunimoto, Y., H. Kubota, Y. Tanaka, S. Matsuo, and T. Sotani, Proc. 15th Japan Symp.
Thermophys. Props. 149-152 (1994).
Kuwabara, S., H. Aoyama, H. Sato, and K. Watanabe, J. Chem. Eng. Data 40(1),
112-116 (1995).

156
Leach, J. W., P. S. Chappelear, and T. W. Leland, AIChE J. 14(4), 568-576 (1968).
Lee, B. I., and M. G. Kesler, AIChE J. 21(3), 510-527 (1975).
Lee, John F., Francis. W. Sears, and D. L. Turcotte, Statistical Thermodynamics, Second
Edition, Addison-Wesley Publishing Co., Reading (1973).
Leland, T. W., and P. S. Chappelear, Ind. Eng. Chem. 60(7), 15-43 (1968).
Lemmon, E. W., and R. T. Jacobsen, J. Phys. Chem. Ref. Data 29(4), 521-552 (2000).
Levelt-Sengers, J. M. H., W. L. Greer, and J. V. Sengers, J. Phys. Chem. Ref. Data 5(1),
1-51 (1976).
Levelt-Sengers, J. M. H., Ind. Eng. Chem. Fundam. 9( 3), 470-480 (1970).
Li, J., R. Tillner-Roth, H. Sato, and K. Watanabe, Int. J. Thermophys. 20(6), 1639-1651
(1999).
Lüddecke, T. O., and J. V. Magee, Int. J. Thermophys. 17(4), 823-849 (1996).
Lyckman, E. W., C. A. Eckert, and J. M. Prausnitz, Chem. Eng. Science. 20, 703-706
(1965).
Maezawa, Y., H. Sato, and K. Watanabe, J. Chem. Eng. Data 35(3), 225-228 (1990).
Magee, J. W., Int. J. Refrig. 15(6), 372-380 (1992).
Magee, J. W., Int. J. Thermophys. 17(4), 803-822 (1996).
Magee, J. W., Int. J. Thermophys. 19(5), 1381-1395 (1998).
Magee, J. W., Int. J. Thermophys. 19(5), 1397-1420 (1998).
Magee, J. W., and J. B. Howley, Int. J. Refrig. 15(6), 362-364 (1992).
Malbrunot, P. F., P. A. Meunier, G. M. Scatena, W. H. Mears, K. P. Murphy, and J. V.
Sinka, J. Chem. Eng. Data 13(1), 16-21 (1968).
Matsuda, T., M. Morishima, and H. Sato, 15th Symp. Thermophys. Props., Boulder
(2003).
Matsuda, T., N. Morita, and H. Sato, CD Proc. European Conf. Thermophys. Props.,
(2002).
McLinden, M. O., Scientific Assessment of Stratospheric Ozone: Volume II. Global
Ozone Research and Monitoring Project-Report 20, (1989).
McLinden, M. O., E. W. Lemmon, and R. T. Jacobsen, Int. J. Refrig. 21(4), 322-338
(1998).
Mohr, J. P., and B. N. Taylor, J. Phys. Chem. Ref. Data 28, 1713-1852 (1999).
Mears, W. H., R. F. Stahl, S. R. Orfeo, R. C. Shair, L. F. Kells, W. Thompson, and H.
McCann, Ind. Eng. Chem. 47, 1449-1454 (1955).
Mizuoka, A., M S Thesis, Keio University (2001).
Molina, M. J., and F. S. Rowland, Nature 249, 810-812 (1974).
Mühlenbein, H., and D. Schlierkamp-Voosen, Computational Intelligence Imitating Life,
Edited by Zurada, J. M., Marks II, R. J., and Robinson, C. J., IEEE Press, 182-193

157
 (1994).
Nagel, M., and K. Bier, Int. J. Refrig 19(4), 264-271 (1995).
Nagel, M., and K. Bier, Int. J. Refrig. 18(8), 534-543 (1996).
Nakagawa, S., T. Hori, H. Sato, and K. Watanabe, J. Chem. Eng. Data 38(1), 70-74
(1993).
Nakamura, S., K. Fujiwara, and M. Noguchi, J. Chem. Eng. Data 42(2), 334-338
(1997).
Narukawa, K., A. Mizuoka, and H. Sato, 14th Symp. Thermophys. Props., Boulder,
(2000).
Nicola, G. D., F. Polonara, and R. Stryjek, J. Chem. Eng. Data 47(4), 876-881 (2002).
Niesen, V. G., L. J. Van Poolen, S. L. Outcalt, C. D. Holcomb, Fluid Phase Equil. 97,
81-95 (1994).
Nishimura, A., Z. Y. Qian, H. Sato, and K. Watanabe, Proc. 13th Japan Symp.
Thermophys. Props., 57-60 (1992).
Ogawa, K., T. Kojima, and H. Sato, J. Chem. Eng. Data 46(5), 1082-1085 (2001).
Oguchi, K., A. Murano, K. Omata, and N. Yada, Int. J. Thermophys. 17(1), 55-64
(1996).
Oguchi, K., K. Amano, T. Namiki, and N. Umezawa, Int. J. Thermophys. 20(6),
1667-1676 (1999).
Otto, J., and W. Thomas, Int. J. Heat Mass Transfer 1, 41-47 (1964).
Outcalt, S. L, and M. O. McLinden, Int. J. Thermophys. 16(1), 79-89 (1995).
Outcalt, S. L, and M. O. McLinden, J. Phys. Chem. Ref. Data 25(2), 605-636 (1996).
Outcalt, S. L, and M. O. McLinden, Int. J. Thermophys. 18(6), 1445-1463 (1997).
Patel, N. C., and A. S. Teja, Chem. Eng. Sci. 37(3), 463-473 (1982).
Peng, D. Y., and D. B. Robinson, Ind. Eng. Chem. Fundam. 15(1), 59-64 (1976).
Penoncello, S. G., R. T. Jacobsen, K. M. de Reuck, A. E. Elhasan, R. C. Williams, and E.
W. Lemmon, Int. J. Thermophys. 16(3), 781-790 (1995).
Piao, C. C., and M. Noguchi, J. Phys. Chem. Ref. Data 27(4), 775-806 (1998).
Piao, C. C., H. Sato, and K. Watanabe, ASHRAE Trans. 96, 132-140 (1990).
Piao, C. C., H. Sato, and K. Watanabe, Trans. JSME B 57, 4182-4189 (1991).
Piao, C. C., K. Fujiwara, and M. Noguchi, Fluid Phase Equil. 150-151, 303-312 (1998).
Pires, P. F., and H. J. R. Guedes, J. Chem. Thermodyn. 31, 55-69 (1999).
Pitzer, K. S., J. Chem. Phys. 7, 583-590 (1939).
Polling, B. E., J. M. Prausnitz, and J. P. O’Connell, The Properties of Gases and Liquids,
Fifth Edition, McGraw-Hill, New York (2000).
Porichanski, E. G., O. P. Ponomareva, and P. I. Svetlichny, Izv. Vyssh. Uchebn. Zaved.
Energ. 3, 122-125 (1982).

158
Press, W. H., S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery, Numerical Recipes
in C: the Art of Scientific Computing, Cambridge University Press, New York
(1992).
Preston-Thomas, H., Metrologia 27, 3-10 (1990).
Qian, Z. Y., A. Nishimura, H. Sato, and K. Watanabe, JSME Int. J. Ser. B 36(4), 665-670
(1993).
Qian, Z. Y., H. Sato, and K. Watanabe, Fluid Phase Equil. 78, 323-329 (1992).
Rackett, H. G., J. Chem. Eng. Data 15(4), 514-517 (1970).
Redlich, O., and J. N. S. Kwong, Chem. Rev. 44, 233-244 (1949).
Renon, H., C. A. Eckert, and J. M. Prausnitz, Ind. Eng. Chem. Fundam. 6(1), 54-58
(1967).
Roy, B. N., Fundamentals of Classical and Statistical Thermodynamics, John Wiley &
Son, Chichester (2002).
Rowlinson, J. S., and I. D. Watson, Chem. Eng. Science. 24, 1565-1574 (1969).
Rusly, R. L., J. Chem. Thermodyn. 23, 1152-1161 (1991).
Russel, H., D. R. V. Golding, and D. M. Yost, J. Am. Chem. Society 66, 16-20, (1944).
Sagawa, T., H. Sato, and K. Watanabe, High Temp. High Pres. 26, 193-201 (1994).
Saitoh, A., S. Nakagawa, H. Sato, and K. Watanabe, J. Chem. Data 35(2), 107-110
(1990).
Sato, H., and K. Watanabe, High Temp. High Pres. 26, 691-700 (1994).
Sato, H., M. Uematsu, K. Watanabe, and M. Okada, Fluid Phase Equil. 36, 167-181
(1987).
Sato, H., T. Kojima, and K. Ogawa, Int. J. Thermophys. 23(3), 787-799 (2002).
Sato, T., H. Sato, and K. Watanabe, J. Chem. Eng. Data 39(4), 851-854 (1994).
Schmidt, R., and W. Wagner, Fluid Phase Equil. 19, 175-200 (1985).
Schmidt, G., and H. Wenzel, Chem. Eng. Sci. 35, 1503-1512 (1980).
Schwefel, H. P., Numerical Optimization of Computer Models, Translated by staff of
John Wiley & Sons, John Wiley & Sons (1981).
Schwefel, H. P., Computational Intelligence Imitating Life, Edited by Zurada, J. M.,
Marks II, R. J., and Robinson, C. J., IEEE Press, 116-124 (1994).
Scott, R., Physical Chemistry, An Advanced Treastise Vol. 3A, Liquid State, Academic
press, New York (1971).
Shaver, R. D., R. L. Robinson, and K. A. M. Gasem, Fluid Phase Equil. 64, 141-163
(1991).
Shaver, R. D., R. L. Robinson, and K. A. M. Gasem, Fluid Phase Equil. 78, 81-98
(1992).
Sove G., Chem. Eng. Sci. 27, 1197-1203 (1972).

159
Sen, A., and M. Srivastava, Regression Analysis, Theory, Methods and Applications,
Spring-Verlag, NewYork (1990).
Setzmann, U., and W. Wagner, Int. J. Thermophys. 10(6), 1103-1126 (1989).
Setzmann, U., and W. Wagner, J. Phys. Chem. Ref. Data 20(6), 1061-1155 (1991).
Shinsaka, K., N. Gee, and G. R. Freeman J. Chem. Thermodyn. 17, 1111-1119 (1985).
Silva, A. M., and L. A. Weber, J. Chem. Eng. Data 38(4), 644-646 (1994).
Span, R., and W. Wagner, J. Phys. Chem. Ref. Data 25(6), 1509-1596 (1996).
Span, R., and W. Wagner, Int. J. Thermophys. 18(6), 1415-1443 (1997).
Span, R., H. J. Collman, and W. Wagner, Int. J. Thermophys. 19(2), 491-500 (1998).
Span, R., E. W. Lemmon, R. T. Jacobsen, W. Wagner, and A. Yokozeki, J. Phys. Chem.
Ref. Data 29(6), 1361-1433 (2000).
Span, R., and W. Wagner, Int. J. Thermophys. 24(1), 1-39 (2003).
Span, R., and W. Wagner, 24(1), 111-162 (2003).
Spears, W. M., K. A. De Jong, T. Bäck, D. B. Fogel, and H. de Garis, Proc. 1993
European Conf. Machine Learning (1993).
Spencer, C. F., and R. P. Danner, J. Chem. Eng. Data 17(2), 236-241 (1972).
Stroustrup, B., The C++ Programming Language, Special Edition, Addison-Wesley,
Massachusetts (2000).
Strom, K. H. U., and U. B. Gren, J. Chem. Eng. Data 38(1), 18-22 (1993).
Sun, L. Q., Y. Y. Duan, L. Shi, M. S. Zhu, and L. Z. Han, J. Chem. Eng. Data 42(4),
795-799 (1997).
Sunaga, H., R. Tillner-Roth, H. Sato, and K. Watanabe, Int. J. Thermophys. 19(6),
1623-1635 (1998).
Takagi, T., High Temp. High Pres. 25, 685-691 (1993).
Takagi, T., J. Chem. Eng. Data 41(5), 1061-1065 (1996a).
Takagi, T., J. Chem. Eng. Data 41(6), 1325-1328 (1996b).
Takagi, T., J. Chem. Eng. Data 42(6), 1129-1132 (1997).
Takagi, private communication (1999).
Takahashi, K., Y. Takaishi, and K. Oguchi, Proc. 15th Japan Symp. Thermophys. Props.,
141-144 (1994).
Takahashi, M., N. Shibasaki-Kitakawa, C. Yokoyama, and S. Takahashi, J. Chem. Eng.
Data 40(4), 900-902 (1995).
Takahashi, M., N. Shibasaki-Kitazawa, and C. Yokoyama, Int. J. Thermophys. 20(2),
435-443 (1999).
Tamatsu, T., H. Sato, and K. Watanabe, Int. J. Refrig. 16(5), 347-352 (1993).
Thomson, G. H., K. R. Brobst, and R. W. Hankinson, AIChE J. 28(4), 671-676 (1982).
Tillner-Roth, R., Int. J. Thermophys. 16(1), 91-100 (1995).

160
Tillner-Roth, R., Int. J. Thermophys. 17(6), 1365-1385 (1996).
Tillner-Roth, R., and A. Yokozeki, J. Phys. Chem. Ref. Data 26(5), 1273-1328 (1997).
Tillner-Roth, R., and H. D. Baehr, J. Chem. Thermodyn. 24, 413-424 (1992).
Tillner-Roth, R., and H. D. Baehr, J. Chem. Thermodyn. 25, 277-292 (1993).
Tillner-Roth, R., and H. D. Baehr, J. Phys. Chem. Ref. Data 23(5), 657-729 (1994).
Trebble, M., and P. Bishnoi, Fluid Phase Equil. 35, 1-18 (1987).
Tsvetkov, O. B., A. V. Kletski, Y. A. Laptev, A. J. Asambaev, and I. A. Zausaev, Int. J.
Thermophys. 16(5), 1185-1192 (1995).
Van Poolen, L. J., C. D. Holcomb, and V. G. Niesen, Fluid Phase Equil. 129, 105-111
(1997).
Vasserman, A. A., and D. V. Fominsky, Int. J. Thermophys. 22(4), 1089-1098 (2001).
van der Waals, J. D., Doctoral Dissertation, Leiden (1873).
Voigt, H. M., and H. Mühlenbein, Proc. 1997 IEEE Int. Conf. Evolutionary
Computation, 172-177 (1995).
Wagner, W., Cryogenics 13, 470-482 (1973).
Wang, J., Z. G. Liu, L. C. Tan, and J. M. Yin, Fluid Phase Equil. 80, 203-211 (1992).
Weber, L. A., Int. J. Thermophys. 10(2), 617-627 (1989).
Weber, L. A., and A. R. H. Goodwin, J. Chem. Eng. Data 38(2), 254-256 (1993).
Weber, L. A., and A. M. Silva, J. Chem. Eng. Data 39(4), 808-812 (1994).
Weber, L. A., and D. R. Defibaugh, Int. J. Thermophys. 15(5), 863-880 (1994).
Weber, L. A., Int. J. Refrig., 17(2), 117-122 (1994).
Weber, L. A., and D. R. Defibaugh, J. Chem. Eng. Data 41(6), 1477-1480 (1996).
Widiatmo, J. V., H. Sato, and K. Watanabe, J. Chem. Eng. Data 39(2), 304-308 (1994).
Wilson, D. P., and R. S. Basu, ASHRAE Trans. 94(2), 2095-2118 (1988).
Wilson, L. C., W. W. Wilding, G. M. Wilson, R. L. Rowley, V. M. Felix, and T.
Chisholm-Carter, Fluid Phase Equil. 80, 167-177 (1992).
Yada, N., M. Uematsu, and K. Watanabe, Trans. JAR 5, 107-115 (1988).
Yamamoto, R., S. Matsuo, and Y. Tanaka, Int. J. Thermophys. 14(1), 79-90 (1993).
Yang, C. N., and C. P. Yang, Phys. Rev. Letters 13(9), 303-305 (1964).
Yamamura, T., X. Zhao, and H. Sato, Proc. 20th Japan Symp. Thermophys. Props.,
213-216, Tokyo (1999).
Ye, F., M.S. Thesis, Keio University (1994).
Ye, F., H. Sato, and K. Watanabe, J. Chem. Eng. Data 40(1), 148-152 (1995).
Yokoyama, C., and S. Takahashi, Fluid Phase Equil. 67, 227-240 (1991).
Yokozeki, A., H. Sato, and K. Watanabe, Int. J. Thermophys. 19(1), 89-127 (1998).
Yokozeki, A., H. Sato, and K. Watanabe, Int. J. Thermophys. 20(1), 141-148 (1999).
Yomo, M., H. Sato, and K. Watanabe, High Temp. High Pres. 26, 267-272 (1994).

161
Younglove, B. A., and J. F. Ely, J. Phys. Chem. Ref. Data 16(4), 577-798 (1987).
Younglove, B. A., and M. O. McLinden, J. Phys. Chem. Ref. Data 23(5), 731-779
(1994).
Zhang, H. L., H. Sato, and K. Watanabe, J. Chem. Eng. Data 40(4), 887-890 (1995).
Zhang, H. L., H. Sato, and K. Watanabe, J. Chem. Eng. Data 41(6), 1401-1408 (1996).
Zhang, H. L., H. Sato, and K. Watanabe, Int. J. Refrig. 20(6), 421-440 (1997).
Zhang, H. L., S. Tada, H. Sato, and K. Watanabe, Fluid Phase Equil. 150-151, 333-341
(1998).
Zhao, X., S. Matsueda, and H. Sato, Thermophys. Props. Transfer Processes New
Refrigerants, IIR, Commission B1, Paderborn (2001).
Zhao, Z. Y., J. M. Yin, and L. C. Tan, Fluid Phase Equil. 80, 191-202 (1992).
Zhu, M. S, Y. D. Fu, and L. Z. Han, Fluid Phase Equil. 80, 149-156 (1992).
Zhu, M. S., J. Li, and B. X. Wang, Int. J. Thermophys. 14(6) 1221-1227 (1993).
Zhu, M. S., J. Wu, and Y. D. Fu, Fluid Phase Equil. 80, 99-105 (1992).

162
 Appendix A

Thermodynamic Property Relations

A.1 Derivation of Relations for Thermodynamic Properties

Relations of fundamental equation of state to thermodynamic properties can be derived
from basic definitions of thermodynamics, thermodynamic laws, and mathematical
operations. The thermodynamic properties are used to represent their extensive and
intensive properties according to their self-nature behavior. All derivations of
thermodynamics properties are represented in intensive properties due to these
representations have more advantages in practice. The fundamental equation of state
gives the state of functions in terms of their natural independent variables. It can be
represented by two independent variables for pure fluid. Most common choice for
fundamental equation of state is Helmholtz free energy as a function of temperature and
density. Both independent variables are reduced in dimensionless parameters as,
vc ρ Tc
δ= = τ= (A.1)
v ρc T

where v : specific volume; vc : critical specific volume; ρ : density; ρ c : critical

density; T : absolute temperature; and Tc : critical absolute temperature.
The Helmholtz free energy is reduced as,
a
α= (A.2)
RT

where a : specific Helmholtz free energy; R : gas constant with suitable unit to a.
Ideal-gas law is included into the fundamental equation of state by dividing it into an
ideal-gas part and a residual part so that it can be written as,
α = α o +α r (A.3)

A.1.1 Pressure and Compressibility

Basic formulation of pressure relates to Helmholtz free energy for fixed systems is
written as,
 ∂a 
p = −  (A.4)
 ∂v T

Since ρ = 1 / v , thus

163
  ∂a 
p = ρ 2   (A.5)
 ∂ρ  T
Converted into the reduced independent variables,
 ∂α 
p = RTρ  δ  (A.6)
 ∂δ T

As the ideal-gas part can be written as α o = ln δ + f (τ ) (see Section B.2), the pressure
becomes,

  ∂α r  
p = RTρ 1 + δ  

 (A.7)
 
  ∂δ τ 

Its derivatives becomes

  ∂α r   2 r  
 +δ 2 ∂ α
 ∂p 
  = RT 1 + 2δ    (A.8)
 ∂δ   ∂δ 2 
 ∂ρ T   τ  τ 

 ∂2 p    r   2 r   3 r  
  = RT 2δ  ∂α  + 4δ 2  ∂ α  +δ 3  ∂ α   (A.9)
 ∂ρ 2  ρ   ∂δ   ∂δ 2   ∂δ 3 
 T  τ  τ  τ 

  ∂α r   2 r  
 ∂p 
  = Rρ 1 + δ   −τ δ  ∂ α   (A.10)
 ∂T  ρ  ∂δ   ∂τ ∂δ 
  τ  τ 

Compressibility:
p  ∂α r 
= 1 + δ  
 (A.11)
RTρ  ∂δ τ

A.1.2 Specific Entropy

Basic definition of specific entropy for fixed systems is,
 ∂a 
s = −  (A.12)
 ∂T T

Reduced independent variables are substituted, it then becomes

 ∂RTα   ∂α   ∂α 
s = −  = − R α + T  = − R α − τ  (A.13)
 ∂T δ  ∂T δ  ∂τ δ

  ∂α o 
s
R
= τ  
 ∂τ
  ∂α r
 +
  ∂τ


 
(
 − α o +α r ) (A.14)
 δ  δ 

164
A.1.3 Specific Internal Energy
Basic definition:
u = a + Ts (A.15)

  ∂α o   ∂α r    
 − α o − α r = τ   ∂α   ∂α r 
o
= α + τ   
u  +   + 
  ∂τ  (A.16)
RT  ∂τ    ∂τ   ∂τ  
 δ  δ   δ  δ 

A.1.4 Isochoric Specific Heat

Basic definition:
 ∂u 
cv =   (A.17)
 ∂T  v

cv ∂   ∂α o   ∂α r
 +


  2 o
 = −τ 2   ∂ α   ∂ 2α r
 +




= −τ 2 (A.18)
R ∂τ   ∂τ   ∂τ
δ 

δ
   ∂τ 2   ∂τ 2  
 v  δ  δ 

A.1.5 Specific Enthalpy

Basic Relation:
h = u + pv (A.19)

  ∂α o   r    
 +  ∂α  + 1 + δ  ∂α 
r
= τ   
h u pv  
= +     ∂δ  (A.20)
 ∂τ   
RT RT RT
 δ  ∂τ δ    τ 

A.1.6 Isobaric Specific Heat

Transformation of thermodynamic relations that are given in general derivations of
several types of specific heat from the basic definition gives a relation as,
 ∂p   ∂v 
c p = cv + T     (A.21)
 ∂T  v  ∂T  p

Since
∂v ∂T ∂p ∂v ∂v ∂p
= −1 =− (A.22)
∂T ∂p ∂v ∂T ∂p ∂T

Thus
2 2 2
 ∂p   ∂p   ∂p 
     
 ∂T v  ∂T v T  ∂T  v
c p = cv − T = cv − T = cv + 2 (A.23)
 ∂T   ∂T  ρ  ∂p 
     
 ∂ v p  ∂ v p  ∂ρ T

165
 2
  
1 + δ  ∂α   2 r
 − δτ  ∂ α
r

  ∂δ   ∂τ∂δ 
cp cv   τ  
= + (A.24)
R  
1 + 2δ  ∂α   2 r 
R
 +δ 2 ∂ α
r
 
  ∂δ   ∂δ 2  
  τ  τ 

A.1.7 Speed of Sound

Basic relation:
 ∂p 
w =   (A.25)
 ∂ρ  s
In accordance with dimensional analysis, in case for the speed of sound in m/s2 and
pressure in N/m2, thus density should be in kg/m3 (mass base). Iso-entropy correlation
can be accomplished from analyzing of the constant composition PVT system. Here
h=h(s,p) and u=u (s,v), and using Maxwell relation, it can be obtained as,
 ∂s  ∂s 
cv = T   and c p = T   (A.26)
 ∂T  v ∂
 T p
cp  ∂s   ∂T 
=    (A.27)
cv  ∂T  p  ∂s  v

cp  ∂s   ∂v    ∂T   ∂p    ∂v   ∂p 
=          =    (A.28)
cv  ∂p T  ∂s T   ∂v  s  ∂T  s   ∂p T  ∂v  s

 ∂p   ∂p  c p  ∂p 
  = −v 2   = −v 2   (A.29)
 ∂ρ  s  ∂v  s cv  ∂v T
2
 ∂p   ∂p  T  ∂p 
  =   + 2   (A.30)
 ∂ρ  s  ∂ρ T ρ cv  ∂T  v
2
w2 1  ∂p  1  ∂p  1  ∂p 
=   =   +   (A.31)
RT RT  ∂ρ  s RT  ∂ρ T Rρ 2 cv  ∂T  v

2
  ∂α r   2 r 
1 + δ   − δ τ  ∂ α 
  ∂δ ∂τ 
w 2   ∂α r



 ∂ 2α r     ∂δ τ  
= 1 + 2δ  ∂δ  + δ 2  2 
 + (A.32)
RT   ∂δ τ   ∂ 2α o   2 r 
  τ
− τ 2   +  ∂ α  
2   2 
 ∂τ δ  ∂τ δ 

The above R is a gas constant in mass-based unit. It may be in kJ/(kg·K) or J/(g·K).

A.1.8 Virial Coefficients

General correlation of virial coefficients can be derived from polynomial form of virial
equation of state. The virial equation of state is written as,
Z = 1 + Bρ + Cρ 2 + Dρ 3 + Eρ 4 + ... (A.33)

166
The virial coefficients k-th can be generally defined as,
1 ∂ k −1 Z
lim (A.34)
(k − 1)! ρ →0 ∂ρ k −1

Second virial coefficient:

∂   ∂α r  
 = lim 1  ∂α r ∂α r 
B (T ) = lim 1 + δ  


 ∂δ + δ (A.35)
ρ →0 ∂ρ   ∂δ τ
 δ →0 ρ c
 ∂δ 
 

 ∂α r 
B(T ) ρ c = lim  
 (A.36)
δ →0  ∂δ τ
Third virial coefficient:

 
∂2 1 + δ  ∂α 
r
C (T ) =
1
lim 2   (A.37)
2 ρ →0 ∂ρ   ∂δ  
  τ 

 1  ∂ 2α r   ∂ 2α r   ∂ 3α r  
C (T ) =
1
lim
1    + 1   + δ    (A.38)
2 ρ →0 ρ c  ρc  ∂δ 2   ∂δ 2   ∂δ 3  
  τ ρ c  τ ρ c  τ 

 ∂ 2α r 
C (T ) ρ c2 = lim  
 (A.39)
δ →0  ∂δ
2
τ
Forth virial coefficient:

∂3   ∂α r  
lim 3 1 + δ  
1 
D(T ) =  (A.40)

6 ρ →0 ∂ρ
  ∂δ τ 

1  1  ∂ 3α r   3 r
 + 1 ∂ α
  3 r
 + 1 ∂ α
  4 r
 + δ ∂ α
 
lim 2  
1  
D(T ) =  ∂δ 3        (A.41)
6 ρ →0 ρ c ρ c 
τ ρ c  ∂δ τ ρ c  ∂δ τ ρ c  ∂δ
3 3 4
  τ 

1 ∂ 3α r
D(T ) ρ c 3 = lim (A.42)
2 ρ →0 ∂δ 3

A.1.9 Vapor-Pressure Relations at Saturation

Correlation of Helmholtz free energy at saturation is derived. Four properties consisting
of saturated-liquid density, saturated-vapor density, vapor pressure, and saturation
temperature are focused. As vapor pressure can be indicated with saturated-liquid
density or saturated-vapor density at the same saturation temperature, it can be
represented from pressure relation as,

167
   ∂α r  
p s = ρ ' RTs 1 + δ '  

 (A.43)
 
 ∂δ
'
 τ 

  ∂α r  
p s = ρ '' RTs 1 + δ ''  '' 

 (A.44)
 
  ∂δ τ 

where the properties with superscript of single dash is saturated-liquid state, superscript
of double dash is saturated-vapor state, and Ts is saturation temperature. Other rule
should be required is Maxwell criterion for the saturation state. Correlation of
Helmholtz free energy and Maxwell criterion can be obtained from p-v diagram as
shown in Fig. A.1.

critical point

v''
v' m
v
p

Fig. A.1 Maxwell criterion on p-v diagram.

The criterion will be required when the same energy scale of valley and mount in the
saturation dome. The volume with superscript m is the middle of saturated-liquid and
saturated-vapor state (in two-phase).
vm v ''
p s (v m − v ' ) − ∫ p(Ts , v)dv = ∫ p(Ts , v)dv − p s (v '' − v m ) (A.45)
v' vm

Since v m − v ' = v '' − v m → v m = (v '' − v ' ) / 2 , thus

v '' v ''  ∂α r 
p s (v '' − v ' ) = ∫ p(Ts , v)dv = ∫ ρRTs (1 + δ  dv
 (A.46)
v' v'  ∂δ 
 1 1  δ  1 ∂α r 
'
ps
 dδ
RTs  ρ '' ρ '  ∫'' δ + ∂δ
− =  (A.47)
  δ 
It is finally obtained as,

168
 Ps  1 1  δ' 
 − − ln   = α r (δ ' ,τ s ) − α r (δ '' ,τ s ) (A.48)
RTs  ρ '' ρ '   δ '' 
   

A.1.10 Specific Heat Along Saturation

Saturated specific heat is a property indicates the latent heat that may be released or
absorbed to phase change at saturation. In case of the specific heat represented in
independent variables of density and temperature, the saturation state for density has
two possibilities. Consequently, this property is divided into saturated-liquid specific
heat and saturated-vapor specific heat. The saturated-liquid specific heat is not included
as isobaric specific heat or isochoric specific heat. Detailed explanation of derivations
for each type is given in Section A.2.1. The correlation derived from the basic
thermodynamic relation gives as,
(dp s / dT )
c s = (∂h / ∂T ) p + Ts (∂p / ∂T ) v (A.49)
(∂p / ∂v) T

(dp s / dT )
c s = c p + Ts (∂p / ∂T ) v (A.50)
(∂p / ∂v) T

(∂p / ∂T ) 2 (dp s / dT )
c s = c v − Ts + Ts (∂p / ∂T ) v (A.51)
(∂p / ∂v) T (∂p / ∂v) T

(∂p / ∂T ) v
c s = c v + Ts (dp s / dT − (∂p / ∂T ) v ) (A.52)
(∂p / ∂v) T

∂α r ∂ 2α r
1+ δ − δτ  dp s 
1 + δ ∂α − δτ ∂ α − 1
r 2 r
c s = cv + R ∂δ ∂δ∂τ  (A.53)
∂α r
∂ 2α r  ∂δ ∂δ∂τ ρ c Rδ dT 
1 + 2δ +δ 2 
∂δ ∂δ 2

dp s
The can be evaluated from Maxwell criterion or Clausius-Clapeyron law as given
dT
in the previous formulation. Since there are two possible densities at saturation, the
specific heat should be divided into two types, i.e. one is evaluated at saturated-liquid
state and the other is evaluated at saturated-vapor state.
Saturated-liquid specific heat:

∂α r ' ∂ 2α r '
1+ δ' − δ 'τ s
c s ' (δ ' , δ " ,τ s ) c v (δ ' ,τ s ) ∂δ∂τ 1 + δ ' ∂α ' − δ 'τ ∂ α ' − dp s 
r 2 r
= + ∂δ  (A.54)
2 r  ∂δ
s
∂δ∂τ ρ c Rδ ' dT 
R R ∂α r ' ∂ α ' 
1 + 2δ ' + δ '2
∂δ ∂δ 2

Saturated-vapor specific heat:

169
 ∂α r " ∂ 2α r "
1+ δ" − δ "τ s
c s " (δ ' , δ " ,τ s ) c v (δ " ,τ s ) ∂δ∂τ 1 + δ " ∂α " − δ "τ ∂ α " − 
r 2 r
= + ∂δ dp s

2 r  δ δ τ ρ 
c δ " dT
s
R R ∂α r " ∂ α " ∂ ∂ ∂ R 
1 + 2δ " + δ "2
∂δ ∂δ 2

(A.55)
Since both δ ' and δ " are function of temperatures, thus saturated specific heat is
function of saturation temperatures. Derivative of vapor pressure by temperature can be
derived from relation of Maxwell criterion. It is more interesting to explicitly represent
both the specific heats in Helmholtz free energy function. Therefore the above relations
can be substituted with,

dp s (δ ' , δ " ,τ s ) ρ c Rδ ' δ "   δ '   ∂α r (δ ' ,τ s ) ∂α r (δ " ,τ s )  

= ln   + α r (δ ' ,τ s ) − α r (δ " ,τ s ) − τ s  −   (A.56)
dT δ '−δ "   δ "   ∂τ ∂ τ 
 

A.1.11 Joule-Thomson Coefficient

This coefficient represents behavior of fluids at iso-enthalpy process. It can be
calculated from the basic definition usually given as,
 ∂T 
µ =   (A.57)
 ∂p  h
 ∂T  ∂p ∂h
Since h-p-T system has relation of       = −1 , it can be obtained,
 ∂p  h  ∂h  T  ∂T  p

1  ∂h  1  T  ∂p   ∂p  
µ=−   =      − 1 (A.58)
cp  ∂p T ρc p  ρ  ∂T  ρ  ∂ρ T 
 

 ∂α r ∂ 2α r ∂α r 
−  δ +δ 2 + δτ
 ∂δ ∂δ 2 ∂δ∂τ 
µRρ = 2
(A.59)
   2 0  r 
1 + δ ∂α − δτ ∂ α  −τ 2 ∂ α + ∂ α 1 + 2δ ∂α + δ 2 ∂α 
r 2 r 2 r r

 ∂δ ∂δ∂τ   ∂τ 2 ∂τ 2  ∂δ ∂δ 2 
   

A.1.12 Fugacity
Fugacity is the property extensively used for calculation of mixture under conditions of
specified composition, temperature, and pressure from fundamental equation of state. In
case of pure fluids, its relation can be derived from basic definition similar to the
mixture as,
 f 
RT ln  = RT ln (ϕ ) = − g d (A.60)
 p
Considering concept of departure function for thermodynamic properties as,

170
g d = g id − g , a r = a − a o = −a d + RT ln z , and g d = −a d − RT (1 − Z ) , thus
f 
RT ln  = a r + RT ( Z − 1) − RT ln Z (A.61)
 p
f ∂α r  ∂α r 
ln = ln ϕ = α r + δ − ln1 + δ (A.62)
p ∂δ  ∂δ 

Fugacity:

 ∂α r  ∂α r 
f = p expα r + δ − ln1 + δ  (A.63)
 ∂δ  ∂δ 

Fugacity coefficient:

 ∂α r  ∂α r 
ϕ = expα r + δ − ln1 + δ  (A.64)
 ∂δ  ∂δ 

A.1.13 Isothermal Compressibility

Basic definition:
1  ∂v 
κ = −   (A.65)
v  ∂p T

−1
  ∂p  
It is converted into κ =  ρ    , then Helmholtz free energy function is substituted
  ∂ρ T 

into partial differentiation of pressure by density, it becomes,

−1
  ∂α r   2 r  
κ=
1
1 + 2δ   +δ 2 ∂ α   (A.66)
ρRT   ∂δ 2 
  ∂δ τ  τ 

A.1.14 Volume Expansivity

Basic definition:
1  ∂v 
β=   (A.67)
v  ∂T  p

Transformation relationship for systems with two independent variables of the

following is used.
 ∂v   ∂T   ∂p 
      = −1 (A.68)
 ∂T  p  ∂p  v  ∂v T

It is then obtained as,

171
  ∂p   ∂p 
   
1 ∂T  v 1  ∂T  v
β =− = (A.69)
v  ∂p  ρ  ∂p 
   
 ∂v T  ∂ρ T

Substituting Helmholtz free energy function into each term, it is then obtained,

  ∂α r   2 r  
1 + δ   − τ δ  ∂ α  
  ∂τ ∂δ 
  ∂δ τ  τ 
β=  (A.70)
  ∂α r   2 r 
T 1 + 2δ   + δ 2  ∂ α  
  ∂δ 2 
  ∂δ τ  τ 

A.2 Basic Relations of Specific Heat

Three types of specific heat verified are specific heat at constant pressure, specific heat
at constant volume, and specific heat along saturation. Correlation of specific heats can
be obtained from the basic definition.
Basic definition of specific heats:
 δq   ds 
c=  = T  (A.71)
 δT  X  dT  X

where X is a specified process path.

Specific heat at constant pressure (isobaric process):
 δq   ds   ∂h 
cp =   = T  =  (A.72)
δ
 p
T  dT  p  ∂T  p

Specific heat at constant volume (isochoric process):

 δq   ds   ∂u 
cv =   = T  =  (A.73)
 δT v  dT  v  ∂T  v

Specific heat at phase-change (isothermal process):

 δq   ds 
cT =   = T  (A.74)
 δT  T  dT  T

Therefore, specific heat along saturation can be derived as,

 ds 
c s = Ts   (A.75)
 dT  T

By considering s as a function of T and v, thus,

 ∂s   ∂s 
ds =   dT +   dv (A.76)
 ∂T  v  ∂v T

Division by dT and restriction to changes along saturation curve, it yields,

 ds   ∂s   ∂s   dv 
  =  +    (A.77)
 dT  T  ∂T  v  ∂v  T  dT  T

172
From Maxwell criterion and thermodynamic laws, it is obtained relation as,
 ∂s  ∂s ∂p
and   =  
c
  = v (A.78)
 ∂T v T  ∂v T  ∂T  v

Hence:
 ds  c  ∂p   dv 
  = v +    (A.79)
 dT  T Ts  ∂T  v  dT  T

 ∂p   dv 
c s = c v + Ts     (A.80)
 ∂T  v  dT  T

Analogy for v as function of p and T, it is then obtained as,

 dv   ∂v   ∂v   dp 
  =  +     (A.81)
 dT  T  ∂T  v  ∂p  T  dT  T
Substituting this result to cs, it gives
 ∂p   dp 
   
 ∂p   ∂v   ∂T  v  dT  T
c s = cv + T     +T (A.82)
 ∂T  v  ∂T  p  ∂p 
 
 ∂v  T

 ∂p   dp 
   
 ∂h   ∂T  v  dT  T
cs =   +T (A.83)
 ∂T  p  ∂p 
 
 ∂v  T

Since independent variables for fundamental equation of state are density and
temperature, there are two possibilities for density at saturation. It may be
saturated-liquid density or saturated-vapor density. The specific heat may be evaluated
at these both densities. If it is applied to saturated-liquid state, then it is called as
saturated-liquid specific heat and it becomes,
 ds  ds '
c s ' = Ts   = Ts (A.84)
 dT  T evaluate at sat −liq
dT

 ∂p s   dp s 
   
 ∂h'   ∂T  v  dT 
cs ' =   + T s (A.85)
 ∂T  p  ∂p s 
 
 ∂v  T

 ∂p(δ ' ,τ s ) 
 
 ∂T  v dp s (δ ' , δ " ,τ s )
c s ' = c p (δ ' ,τ s ) + Ts (A.86)
 ∂p (δ ' ,τ s )  dT
 
 ∂v '
T

Similarly, saturated-vapor specific heat is obtained as,

173
  ∂p(δ " ,τ s ) 
 
 ∂T  v dp s (δ ' , δ " ,τ s )
c s " = c p (δ " ,τ s ) + Ts (A.87)
 ∂p(δ " ,τ s )  dT
 
 ∂v T

Correlations of specific heats for constant pressure and constant volume processes are
verified. Giving s as function of T and v, its expression in partial derivatives can be
written as,
 ∂s   ∂s 
ds =   dT +   dv (A.88)
 ∂T v  ∂v T

Analogy for s as function of T and p, it is obtained as,

 ∂s   ∂s 
ds =   dT +   dp (A.89)
 ∂T  p  ∂p T
Combining both the above relation and eliminate ds, it is then obtained as,
cv  ∂s  cp  ∂s 
dT +   dv = dT +   dp (A.90)
T ∂
 T
v T  ∂p T
Substituting cv and cp, the relation becomes,
 ∂s   ∂s 
c p dT − cv dT = T   dv − T   dp (A.91)
 ∂v T  ∂p T
From Maxwell relations:
 ∂s   ∂v  ∂s ∂p
  = −  and   =   (A.92)
∂
 T
p  ∂T  p  ∂v T  ∂T T

It can be then written as,

 ∂p   ∂v 
c p dT − cv dT = T   dv + T   dp (A.93)
 ∂T T  ∂T T

Division by dT and restriction by either constant pressure or constant volume processes,

it provides a relation as,
 ∂p   ∂v 
c p = cv + T     (A.94)
 ∂T  v  ∂T  p

174
 Appendix B

Fundamental Equation of State and Its Derivatives

This part describes mathematical tasks relating to the general form of thermodynamic
fundamental equations of state for pure fluids. The fundamental equations of state are
represented in Helmholtz free energy. It is represented in two independent variables of
density and temperature.

B.1 Dimensionless Helmholtz Free Energy for Fundamental Equation of State

Helmholtz free energy function is divided into an ideal-gas part and a residual part as,
A = Ao + Ar (B.1)
Intensive property expression may be in molar basis or mass basis. The Helmholtz free
energy can be written as,
a = ao + ar (B.2)
It is reduced by dividing with RT and then finally becomes,
A( ρ , T ) a( ρ , T )
= = α (δ ,τ ) = α o (δ ,τ ) + α r (δ ,τ ) (B.3)
MRT RT

B.2 Mathematical Operation for Ideal-Gas Part

The ideal-gas part plays significant role for ideal-gas behavior and real-fluid behavior.
Accuracy of the ideal-gas part influences in reliability of the final model because the
ideal-gas part is included on development of the residual part. It absolutely affects on
caloric properties. Based on correlations of a = u − Ts , h = u + pv , and ideal-gas law
of pv = RT , the ideal-gas part can be obtained as,
a o ( ρ , T ) = h o (T ) − RT − T s o ( ρ , T ) (B.4)
T
h o (T ) = ∫ c op (T )dT + h0o (B.5)
To

From mathematical thermodynamics, relation of ideal-gas state for entropy obtained as,
∂s ∂s
ds = du + dv (B.6)
∂u ∂v

∂s ∂p ∂u ∂s
Maxwell correlation is   =   and   = T   , thus the relation becomes,
 ∂v T  ∂T  v  ∂T  v  ∂T  v
cv  ∂p 
ds = dT +   dv (B.7)
T  ∂T  v

Alternatively, it can be obtained from the first law of thermodynamics,

175
 δq rev du p
ds = → du = Tds − pdv → ds = + dv (B.8)
T T T
The relation is finally written as,
T c op − R v
R
s o ( ρ , T ) − s 0o = ∫ dT + ∫ dv (B.9)
T0 T v0 v

T c op (T ) T   v 
s o ( ρ , T ) = s 0o + ∫ dT − R ln  + R ln
 v

 (B.10)
To T  T0   0 
T T c op (T ) T   v 
a o ( ρ , T ) = h0o + ∫ c op (T )dT − RT − T s 0o − T ∫ dT + RT ln  − RT ln
 v

 (B.11)
T0 T0 T  T0   0 
o o
a o ( ρ , T ) τ h0o s 0o τ τ c p 1 τ cp τ δ 
α o (δ ,τ ) = = − − ∫ 2 dτ − 1 + ∫ dτ + ln 0 
 (B.12)
RT RTc R R τ0τ R τ0 τ  τδ 0 
where h0o and s 0o are values convened at τ 0 and δ 0 for ideal-gas state of the fluid.

B.2.1 Integration Framework

The ideal-gas part is established by integration of ideal-gas isobaric specific heat
equation with general form as,

c op m 2 exp(−η ioτ ) n tj
= N 2o + ∑ N ioη io τ 2 + ∑ N oj τ (B.13)
R i =3 {1 − exp(−η τ )} o
i
2
j = m +1

First integration part of equation (B.12) becomes,

τ τ c op N 2o m  1 1  n τ
− ∫ dτ = N 2o − τ − ∑ N ioη ioτ  −  − τ ∑ N oj ∫ τ t j − 2 dτ (B.14)
R τ0τ 2 τ0  o
1 − exp(−η ioτ ) 
i =3  1 − exp(−η i τ 0 ) j = m +1 τ0

Second integration part of equation (B.12) becomes,

1 τ cp
o
τ
dτ = N 2o ln
 m o
{ m
(
 + ∑ N i ln 1 − exp − η ioτ − ∑ N ioη ioτ )}
exp − η ioτ ( )
∫
R τ0 τ τ 0
 i =3
 i =3 1 − exp − η ioτ ( ) (B.15)
+
m
∑ N i 
o (
 exp − η ioτ 0 ) η ioτ 0 {
− ln 1 − exp(−η ioτ 0 ) }  n τ
 + ∑ N oj ∫ τ t j −1dτ
i =3
o
(
 1 − exp − η i τ 0 )  j = m +1 τ
 0

These two parts are substituted into the ideal-gas part (equation (B.12)), it can be
obtained as,

{ ( )}
m n τ t j −1 n τ t j −2
α o (δ , τ ) = ln δ + N 0o + N 1oτ + ( N 2o − 1) ln τ + ∑ N io ln 1 − exp − η ioτ + ∑ N oj ∫ τ dτ − τ ∑ N oj ∫ τ dτ
i =3 j = m +1 τ0 j = m +1 τ0

where
τ
N 0o = ln 0
 s 0o
− − + o
( − τ ) +
m  o
o  exp − η i τ 0 ( )
η ioτ 0 − ln 1 − exp(−η ioτ 0 ) { }
 δ0
 R

1 N 2 1 ln 0
i
∑
= 3
N i  o
 1 − exp − η i τ 0 ( )  and

176
  ho No m  exp(−η ioτ 0 )  
N1o =  0 − 2 − ∑ N ioη io    , tj ≠ 0 (B.16)
 RTc τ 0 i =3  η o
τ
 1 − exp( − i 0 
) 

These both coefficients can be evaluated directly if a reference for enthalpy and entropy
is determined at the ideal-gas state.

Referring the previous relation, if the ideal-gas isobaric specific heat equation is given
as,

c op m n 2 exp(−η ioτ )
= 1 + N 2o + ∑ N ioτ ti + ∑ N ioη io τ 2 (B.17)
R i =3 i = m +1 {1 − exp(−η τ )} o
i
2

The ideal-gas part becomes,

{ ( )}
m τ m τ n
α o (δ , τ ) = ln δ + N 0o + N 1oτ + N 2o ln τ + ∑ N io ∫ τ ti −1 dτ − τ ∑ N io ∫ τ ti −2 dτ + ∑ N io ln 1 − exp − η ioτ
i =3 0 i =3 0 i = m +1

where
τ
N 0o = ln 0
 s 0o
 −
m  exp − η ioτ 0
− 1 + (1 + N 2o )(1 − ln τ 0 ) + ∑ N io 
( )
η ioτ 0 − ln 1 − exp(−η ioτ 0 ) { }
 δ0  R i = 3  1 − exp − η oτ
 i 0 ( ) 
m τ0 m τ0
− ∑ N io ∫ τ ti −1dτ + τ ∑ N io ∫ τ ti − 2 dτ
i =3 0 i =3 0

 ho (1 + N 2o ) m o o  exp(−η ioτ 0 )  
N1o =  0 − − ∑ N i ηi   (B.18)
 RTc τ0 i =3 1 − exp(−η ioτ 0 )  


B.2.2 Derivatives of Ideal-Gas Part

General form for the ideal-gas part may be given as,
m n
α o (δ , τ ) = ln δ + N 0o + N 1oτ + N 2o ln τ + ∑ N ioτ ti + ∑ N io ln 1 − exp − η ioτ
i =3 i = m +1
{ ( )} (B.19)

Its derivatives have correlations as,

∂α o 1
= (B.20)
∂δ δ
∂ 2α o 1
=− (B.21)
∂δ 2 δ2

∂ 2α o
=0 (B.22)
∂τ∂δ
∂α o N 2o m o ti −1 n N oη o exp(−η oτ )
= + ∑ N i t iτ + ∑ i i i
(B.23)
∂τ τ i =3
o
i = m +1 1 − exp( −η i τ )

177
 ∂ 2α o
=−
N 2o
+ ∑ N i t i (t i − 1)τ ti − 2 − ∑  i i
m
2
i { i }
n  N oη o exp(−η oτ ) 1 − exp( −η oτ ) + N oη o exp( −2η oτ ) 
i i
2
i 
∂τ 2
τ 2
i =3 i = m +1
 ( o
1 − exp(−η i τ ))2 


(B.24)

B.2.3 Planck-Einstein Term

Structural form of the Planck-Einstein terms can be understood from correlation of

RRHO model for ideal gas. The sum of all contributions to c op can be expressed as

c op = 4 R + C HO + δC . The CHO has relation as,

 hcvi 
2 d i vi exp − 
 hc   kT 
C HO = R  ∑ 2
(B.25)
 kT  i   hcvi 
1 − exp − 
  kT 

Based on the above structural form, Planck-Einstein relation may be expressed as,
2 exp(−η ioτ )
N ioτ 2η io (B.26)
{1 − exp(−η τ )}
o
i
2

where N io are fitting parameters, and η io is calculated from correlation as,

hcvi
= η ioτ (B.27)
kT

hcvi Tc
Thus η io = and τ = (both parameters are represented in dimensionless)
kT T
where
c : Speed of light in vacuum=299792458 m·s-1
k : Boltzmann constant =1.3806503 x 10-23 J·K-1
h : Planck constant=6.62606876 x 10-34 J·s
vi : Fundamental frequency in meter.
Other universal constants are Vm gas (273.15 K and 101.325 kPa) = 22.413996 x 10-3
m3·mol-1, and gas constant R=8.314472 J·mol-1·K-1

B.3 Mathematical Operation for Residual Part

The residual part is developed with multi-property regression. Considering the
flexibility of the object for terms composing the residual part in a programming work,
structural form for terms of the residual part is defined as,
Ψ (δ ,τ ) = δ dl τ tl exp{− md (δ − s d ) te − mt (τ − st ) te } (B.28)
Other than δ and τ contained in the term are constants that may be optimized from the

178
predicted parameters. Several terms can be derived from the above general form. In case
of md=0 and mt=0, the term becomes δ dl τ tl ; in case of md=1, sd=0 and mt=0, the term
becomes δ dl τ tl exp(−δ de ) ; in case of md=1, sd=1 and mt=0, the term becomes
δ dl τ tl exp(−(δ − 1) de ) ; in case of md=1, sd=1, mt=1, and sd=1, the term becomes
δ dl τ tl exp(−(δ − 1) de − (τ − 1) te ) ; etc.
Therefore, the residual part may be generally written in a simple equation as,
n
α r (δ ,τ ) = ∑ N i × Ψi (δ ,τ ) (B.29)
i =1

B.3.1 Separation of Residual Part

Precisely differential works are very important to calculate each type of terms relating
to every thermodynamic property. The general form of terms for the residual part can be
separated into several functions to make simple mathematical operations from the
programming aspect.
By taking Ψ consists of four composing functions with relation as,
Ψ (δ ,τ ) = ∆ (δ ) × Γ(τ ) × Φ(δ ) × Θ(τ ) (B.30)
where
∆(δ ) = δ di
Γ(τ ) = τ ti
{ }
Φ (δ ) = exp − m d (δ − s d ) d e
Θ(τ ) = exp{− m (τ − s ) }
t τ
te

B.3.2 Derivatives of Residual Part

In accordance with the previous manipulation, the residual part may be written as,
n
α r (δ ,τ ) = ∑ N i Ψi (δ ,τ ) (B.31)
i =1

∂α r n ∂Ψi (δ ,τ )
= ∑ Ni (B.32)
∂δ i =1 ∂δ

∂ 2α r n ∂ 2 Ψi (δ ,τ )
= ∑ Ni (B.33)
∂δ 2 i =1 ∂δ 2

∂ 2α r n ∂ 2 Ψi (δ ,τ )
= ∑ Ni (B.34)
∂δ∂τ i =1 ∂δ∂τ

∂α r n ∂Ψi (δ ,τ )
= ∑ Ni (B.35)
∂τ i =1 ∂τ

∂ 2α r n ∂ 2 Ψi (δ ,τ )
= ∑ Ni (B.36)
∂τ 2 i =1 ∂τ 2

179
 ∂ 2α r ∂ 2α r
= (B.37)
∂δ∂τ ∂τ∂δ

Derivatives for each function composing the residual part as,

∂Ψ (δ ,τ )  ∂Φ (δ ) ∂∆(δ ) 
= Γ(τ )Θ(τ )∆ (δ ) + Φ (δ ) (B.38)
∂δ  ∂δ ∂δ 

∂ 2 Ψ (δ ,τ )  ∂ 2 Φ (δ ) ∂∆(δ ) ∂Φ (δ ) ∂ 2 ∆(δ ) 
= Γ(τ )Θ(τ )∆ (δ ) + 2 + Φ (δ )  (B.39)
∂δ 2  ∂δ 2 ∂δ ∂δ ∂δ 2 

∂Ψ (δ ,τ )  ∂Θ(τ ) ∂Γ(τ ) 
= ∆(δ )Φ (δ )Γ(τ ) + Θ(τ ) (B.40)
∂τ  ∂τ ∂τ 

∂ 2 Ψ (δ ,τ )  ∂ 2 Θ(τ ) ∂Γ(τ ) ∂Θ(τ ) ∂ 2 Γ(τ ) 

= ∆(δ )Φ (δ )Γ(τ ) + 2 + Θ (τ )  (B.41)
∂τ 2  ∂τ 2 ∂τ ∂τ ∂τ 2 

∂Ψ (δ ,τ )  ∂Θ(τ ) ∂Γ(τ )  ∂Φ (δ ) ∂∆ (δ ) 
= Γ(τ ) + Θ(τ )∆ (δ ) + Φ (δ ) (B.42)
∂δ∂τ  ∂τ ∂τ  ∂δ ∂ δ 

∂Ψ (δ ,τ ) ∂Ψ (δ ,τ )
= (B.43)
∂τ∂δ ∂δ∂τ

∂∆(δ )
= d i δ di −1 (B.44)
∂δ

∂ 2 ∆(δ )
2
= d i (d i − 1)δ di −2 (B.45)
∂δ

∂Γ(τ )
= t iτ ti −1 (B.46)
∂τ

∂ 2 Γ(τ )
2
= t i (t i − 1)τ ti −1 (B.47)
∂τ

∂Φ (δ )
∂δ
{
= −md (d e − 1)(δ − s d ) d e −1 exp − md (δ − s d ) d e } (B.48)

∂ 2 Φ(δ )
∂δ 2
{ } {
= − md d e (d e − 1)(δ − s d ) d e − 2 + md md d e d e )(δ − s d ) 2( d e −1) exp − md (δ − s d ) d e } (B.49)

∂Γ(τ )
∂τ
{
= − mt (t e − 1)(τ − st ) te −1 exp − mt (δ − st ) t e } (B.50)

∂ 2 Γ(τ )
∂τ 2
{ } {
= − mt t e (t e − 1)(δ − st ) te −2 + mt mt t e t e )(δ − st ) 2(te −1) exp − mt (δ − st ) t e } (B.51)

B.4 Convention of Reference State for Fundamental Equation of State

Revealing the existing conventions for reference state of thermodynamic properties,
most standards determining a point for the reference state are not based on ideal-gas

180
state, but it is based on a specified point of fluid as shown in Table B.1. Therefore,
integral constants for the ideal-gas part are difficult to be determined from the
correlations given in the above relation. Alternatively, it can be solved from the reduced
Helmholtz free energy function consisting of the ideal-gas part and residual part.
Therefore, for a specified density and temperature of a reference state, and using the
convened values for enthalpy and entropy at the reference state that are adopted for the
fundamental equation of state, it can be simultaneously fitted to enthalpy and entropy
relations so that two integral constants are determined without using correlations refer
to ideal-gas state.

Table B.1 Convention for Reference State of Thermodynamic Properties

s0 h0
No. Convention Reference Point a -1 -1
kJ·kg ·K kJ·kg-1
1 IIR b 273.15 K at saturated liquid 1 200
2 ASHRAE c 233.15 K at saturated liquid 0 0
3 Normal Boiling Point Normal boiling point for 0 0
saturated liquid
a
It will determine the reduced temperature and density at a reference state
b
International Institute of Refrigeration
c
American Society of Heating, Refrigeration and Air-Conditioning Engineers

As listed in Table B.1, some convection may be used on development for

thermodynamic equation of state. The values for caloric properties such as enthalpy,
internal energy, and entropy may be significantly different values due to different
convention used on establishing equation of state, but their changing values from a state
to other state is not significant difference.

181
 Appendix C
Statistical Analysis Result

Statistical definitions used for evaluation in assessments of the ancillary equations for
saturation properties and the new fundamental equations of state are as follows: AAD -
average of absolute value of deviations; BIAS - average of deviations; and STD - standard
deviation. On the other hand Max is maximum of absolute value of deviations. All these
quantities are given in percent. Detailed definitions are written in Eqs. (C.1) to (C.3).
1 N
AAD = ∑ ∆i (C.1)
N i =0

1 N
STD = ∑ (∆ i − BIAS)
2
(C.2)
N − 1 i =0

1 N
BIAS = ∑ ∆i (C.3)
N i =0
where ∆ = 100(Y − Ycal ) / Ycal , and the Y is the property being analyzed.

Table C.1 Deviations of Measurements from the Vapor-Pressure Equations

AAD BIAS STD Max
Reference Fluid N T, K Nout
% % % %
Baroncini et al. (1993) R-32 56 238 - 351 0.072 0.018 0.089 0.23 -
Bouchot and Richon (1994) R-32 8 253 - 333 0.19 0.14 0.22 0.48 -
Defibaugh et al. (1994) R-32 18 219 - 343 0.022 -0.012 0.024 0.049 -
de Vries (1997) R-32 139 223 - 351 0.015 -0.0057 0.015 0.028 -
Fu et al. (1995) R-32 60 233 - 351 0.044 -0.030 0.039 0.22 -
Fukushima et al. (1995) R-32 57 278 - 351 0.10 -0.094 0.12 0.55 -
Holcomb et al. (1995) R-32 25 295 - 349 0.11 0.091 0.15 0.58 -
Kanungo et al. (1987) R-32 11 149 - 245 0.38 0.38 0.18 0.60 -
Magee (1996) R-32 7 270 - 330 0.11 -0.11 0.032 0.14 -
Malbrunot et al. (1968) R-32 30 191 - 348 0.24 0.044 0.28 0.59 1
Matsuda et al. (2002) R-32 22 276 - 303 0.022 0.014 0.021 0.051 -
Nagel and Bier (1995) R-32 27 204 - 351 0.050 0.048 0.059 0.19 1

182
 Table C.1 Deviations of Measurements from the Vapor-Pressure Equations (continued)
AAD BIAS STD Max
Reference Fluid N T, K Nout
% % % %
Qian et al. (1993) R-32 9 280 - 350 0.033 -0.027 0.014 0.066 -
Sato et al. (1994) R-32 21 320 - 351 0.019 0.0058 0.026 0.060 1
Weber and Goodwin (1993) R-32 27 208 - 237 0.029 0.015 0.030 0.081 -
Weber and Silva (1994) R-32 17 236 - 266 0.030 0.030 0.048 0.21 -
Widiatmo et al. (1994) R-32 25 220 - 325 0.36 -0.25 0.50 1.2 -
Zhu et al. (1993) R-32 31 273 - 347 0.24 0.030 0.28 0.55 -
Baroncini et al. (1993) R-125 58 236 - 333 0.079 -0.044 0.11 0.48 -
Boyes and Weber (1995) R-125 29 273 - 335 0.024 -0.022 0.016 0.046 -
de Vries (1997) R-125 104 222 - 339 0.043 -0.023 0.043 0.11 -
Duarte-Garza et al. (1997) R-125 15 220 - 338 0.15 0.15 0.18 0.73 -
Fukushima et al. (1995) R-125 45 252 - 339 0.14 -0.058 0.17 0.55 -
Magee (1996) R-125 33 190 - 335 0.14 -0.10 0.14 0.63 1
Matsuda et al. (2003) R-125 22 278 - 303 0.019 -0.0081 0.021 0.033 -
Oguchi et al (1996) R-125 61 223 - 338 0.060 0.034 0.10 0.60 -
Sagawa et al. (1994) R-125 26 308 -339 0.036 -0.019 0.040 0.083 -
Tillner Roth (1996) R-125 30 178 - 230 0.23 -0.089 0.36 1.7 -
Tsvetkov et al. (1996) R-125 34 263 - 338 0.038 -0.0005 0.052 0.15 -
Weber and Silva (1994) R-125 103 219 - 285 0.059 -0.051 0.061 0.19 -
Widiatmo et al. (1994) R-125 20 220 - 335 0.34 -0.29 0.47 1.9 -
Ye et al. (1995) R-125 12 290 - 339 0.087 -0.036 0.045 0.15 -
Arita et al. (1991) R-134a 3 273 - 323 0.053 -0.053 0.029 0.085 -
Baehr and Tillner-Roth (1991) R-134a 37 303 - 373 0.017 -0.016 0.013 0.037 2
Baroncini et al. (1990) R-134a 64 243 - 359 0.25 0.15 0.24 0.80 -
Basu and Wilson (1989) R-134a 32 211 - 369 0.35 0.25 0.58 2.4 -
Blanke et al. (1995) R-134a 53 175 - 374 0.11 -0.031 0.13 0.49 2
Fukushima et al. (1997) R-134a 41 262 - 372 0.13 -0.030 0.19 0.78 -
Goodwin et al. (1992) R-134a 57 214 - 313 0.031 0.021 0.036 0.13 -
Hou et al. (1992) R-134a 10 180 - 360 0.43 0.40 0.76 2.2 2
Kubota et al. (1989) R-134a 25 253 - 373 0.23 0.013 0.29 0.60 -
Maezawa et al. (1990) R-134a 13 280 - 350 0.38 -0.0004 0.61 1.7 -
Magee and Bowley (1992) R-134a 19 180 - 350 0.26 -0.060 0.42 1.4 -
Morrison and Ward (1991) R-134a 12 268 - 368 0.088 0.034 0.14 0.32 1
Niesen et al. (1994) R-134a 12 316 - 370 0.072 0.056 0.054 0.13 -

183
 Table C.1 Deviations of Measurements from the Vapor-Pressure Equations (continued)
AAD BIAS STD Max
Reference Fluid N T, K Nout
% % % %
Piao et al. (1990) R-134a 51 308 - 368 0.12 -0.11 0.081 0.30 5
Tillner-Roth (1996) R-134a 33 170 - 232 0.53 0.53 0.72 2.3 -
Weber (1989) R-134a 22 313 - 373 0.020 -0.020 0.010 0.031 -
Wilson and Basu (1988) R-134a 32 211 - 369 0.37 0.24 0.61 2.5 -
Zhu et al. (1992) R-134a 43 279 - 363 0.094 -0.030 0.11 0.28 -
Baehr and Tillner-Roth (1991) R-152a 55 301 - 386 0.15 0.15 0.034 0.22 -
Blanke and Weiss (1992) R-152a 36 155 - 260 0.21 -0.0088 0.53 1.9 2
Duarte-Garza and Magee (1999) R-152a 21 155 - 250 0.018 -0.0037 0.025 0.088 -
Fransson et al. (1992) R-152a 5 303 - 385 0.65 0.65 0.29 1.0 -
Higashi et al. (1999) R-152a 44 273 - 386 0.042 0.022 0.048 0.12 -
Holcomb et al. (1993) R-152a 33 312 - 383 0.16 0.14 0.21 1.0 -
Kamei et al. (1989) R-152a 24 250 - 365 0.056 0.051 0.11 0.46 -
McLinden (1989) R-152a 21 273 - 373 0.021 0.011 0.025 0.050 -
Mears et al. (1955) R-152a 27 204 - 377 1.7 1.3 2.6 9.7 8
Sato et al. (1987) R-152a 41 273 - 373 0.018 0.0096 0.021 0.042 -
Silva and Weber (1993) R-152a 38 220 - 273 0.059 -0.059 0.024 0.12 1
Tillner-Roth (1996) R-152a 40 155 - 232 0.51 0.45 0.80 2.8 -
Yada et al. (1988) R-152a 13 305 - 385 0.13 0.078 0.19 0.54 -
Zhao et al. (1992) R-152a 167 237 - 368 0.20 -0.071 0.25 0.80 1

Table C.2 Deviations of Measurements from the Saturated-Liquid Density Equations

AAD BIAS STD Max
Reference Fluid N T, K Nout
% % % %
Defibaugh et al. (1994) R-32 21 243 - 338 0.089 0.089 0.11 0.39 -
Fukushima et al. (1995) R-32 21 232 - 351 0.56 -0.027 0.82 2.4 1
Higashi (1994) R-32 8 336 - 351 0.90 0.67 1.1 1.8 -
Holcomb et al. (1993) R-32 25 295 - 349 0.36 0.36 0.28 1.1 -
Kuwabara et al. (1995) R-32 17 330 - 351 0.31 -0.31 0.36 1.3 -
Magee (1996) R-32 13 139 - 305 0.057 0.012 0.071 0.14 1
Malbrunot et al. (1968) R-32 15 248 - 349 0.29 0.080 0.38 0.95 -
Sato et al. (1994) R-32 2 321 - 344 0.24 0.24 0.20 0.38 -
Shinsaka et al. (1985) R-32 19 150 - 222 1.0 -1.0 0.079 1.1 -

184
 Table C.2 Deviations of Measurements from the Saturated-Liquid Equations (continued)
AAD BIAS STD Max
Reference Fluid N T, K Nout
% % % %
Widiatmo et al. (1994) R-32 22 220 - 330 0.12 -0.11 0.087 0.23 -
Fukushima et al. (1995) R-125 47 227 - 339 0.15 0.035 0.29 1.0 8
Higashi (1994) R-125 6 324 - 339 0.37 0.37 0.48 1.3 -
Kuwabara et al. (1995) R-125 9 331 - 339 0.35 0.14 0.47 1.0 -
Magee (1996) R-125 7 173 - 308 0.090 0.090 0.054 0.16 -
Takahashi et al. (1994) R-125 22 241 - 338 0.17 0.084 0.19 0.46 -
Widiatmo et al. (1994) R-125 25 220 - 335 0.086 -0.044 0.11 1 -
Aoyama et al. (1996) R-134a 4 372 - 374 0.49 0.061 0.61 0.73 -
Basu and Wilson (1989) R-134a 9 239 - 372 0.25 -0.13 0.26 0.51 -
Defibaugh and Moldover (1997) R-134a 26 245 - 370 0.14 0.036 0.16 0.32 -
Fukushima et al. (1990b) R-134a 3 325 - 357 0.099 -0.073 0.099 0.15 -
Fukushima (1991) R-134a 7 244 - 292 0.13 0.10 0.11 0.23 -
Fukushima et al. (1990a) R-134a 8 370 - 374 0.82 0.79 0.61 1.4 2
Higashi (1994) R-134a 8 359 - 374 0.84 0.61 1.7 4.8 -
Hou et al. (1992) R-134a 12 180 - 360 0.16 -0.16 0.26 0.84 -
Kabata et al. (1989) R-134a 12 343 - 374 1.0 0.95 1.7 4.9 3
Maezawa et al. (1990) R-134a 25 200 - 370 0.18 -0.11 0.25 0.79 -
Morrison and Ward (1991) R-134a 12 268 - 368 0.22 -0.21 0.22 0.52 1
Niesen et al. (1994) R-134a 12 316 - 370 0.18 -0.18 0.12 0.36 -
Takahashi et al. (1994) R-134a 7 258 - 306 0.18 0.18 0.038 0.21 -
Yokoyama and Takahashi (1991) R-134a 21 252 - 367 0.16 -0.021 0.20 0.51 -
Blanke and Weiss (1992) R-152a 19 160 - 308 0.038 -0.036 0.053 0.13 -
Geller (1979) R-152a 19 160 - 340 0.070 0.024 0.090 0.22 -
Higashi et al. (1987) R-152a 10 371 - 386 0.23 0.12 0.41 1.0 3
Holcomb et al. (1993) R-152a 33 312 - 384 0.057 -0.018 0.066 0.13 -
Kabata (1988) R-152a 10 371 - 386 0.23 0.12 0.41 1.0 3
Kamei et al. (1989) R-152a 24 250 - 365 0.10 -0.058 0.11 0.22 -
McLinden (1989) R-152a 29 233 - 373 0.10 -0.096 0.097 0.27 -
Magee (1998) R-152a 9 157 - 318 0.051 0.031 0.055 0.11 -
Mears et al. (1955) R-152a 11 232 - 353 0.25 -0.17 0.27 0.70 -
Wang et al. (1992) R-152a 11 376 - 386 1.4 -0.99 1.9 3.7 4

185
 Table C.3 Deviations of Measurements from the Saturated-Vapor Density Equations
AAD BIAS STD Max
Reference Fluid N T, K Nout
% % % %
Defibaugh et al. (1994) R-32 28 219 - 343 0.27 0.021 0.30 0.46 -
Fukushima et al. (1995) R-32 13 298 - 351 1.4 0.066 2.0 4.4 1
Higashi (1994) R-32 9 340 - 352 2.8 2.8 1.4 5.3 -
Holcomb et al. (1993) R-32 25 245 - 349 0.60 0.56 0.71 2.9 -
Kuwabara et al. (1995) R-32 13 346 - 351 0.92 0.31 1.0 1.5 3
Sato et al. (1994) R-32 4 327 - 346 1.3 -1.3 0.61 1.9 -
Boyes and Weber (1995) R-125 10 275 - 320 0.15 0.11 0.17 0.40 -
Fukushima et al. (1995) R-125 20 315 - 339 0.46 -0.18 0.54 1.3 6
Higashi (1994) R-125 7 327 - 339 0.97 0.97 0.45 1.6 -
Kuwabara et al. (1995) R-125 8 336 - 339 0.24 0.20 0.28 0.57 2
Aoyama et al. (1996) R-134a 7 373 - 374 1.3 -0.17 1.6 3.0 -
Fukushima (1990a) R-134a 6 294 - 371 0.51 -0.34 0.59 1.2 -
Fukushima et al. (1990b) R-134a 9 371 - 374 1.6 -1.5 2.6 7.3 2
Higashi (1994) R-134a 6 362 - 374 2.0 -0.74 2.8 5.7 -
Kabata et al. (1989) R-134a 15 361 - 374 2.9 -0.069 5.0 12 3
Morrison and Ward (1991) R-134a 8 298 - 368 5.6 4.7 6.9 18 -
Niesen et al. (1994) R-134a 12 316 - 370 0.32 0.30 0.37 1.3 -
Weber (1989) R-134a 5 320 - 368 0.24 -0.14 0.38 0.79 -
Higashi et al. (1987) R-152a 11 374 - 386 0.82 0.31 1.6 2.4 -
Holcomb et al. (1993) R-152a 33 312 - 384 0.87 0.76 0.99 4.0 -
Kabata (1988) R-152a 11 374 - 386 0.82 0.31 1.2 2.4 -
Kamei et al. (1989) R-152a 24 250 - 364 0.29 -0.29 0.26 0.53 -
Wang et al. (1992) R-152a 11 377 - 386 0.85 -0.17 1.2 2.7 3

186
 Table C.4 Deviations of Measurements from the New Fundamental Equation of State for R-32
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Baroncini et al. (1993) ps 56 - 238 - 351 - 0.086 0.015 0.10 0.25 -
Bouchot and Richon (1994) ps 8 - 253 - 333 - 0.21 0.14 0.25 0.51 -
Defibaugh et al. (1994) ps 18 - 219 - 343 - 0.051 -0.017 0.057 0.11 -
de Vries (1997) ps 139 - 223 - 351 - 0.038 -0.025 0.038 0.08 -
Fu et al. (1995) ps 60 - 233 - 351 - 0.058 -0.033 0.070 0.23 -
Fukushima et al. (1995) ps 57 - 278 - 351 - 0.13 -0.13 0.11 0.51 -
Holcomb et al. (1995) ps 25 - 295 - 349 - 0.089 0.057 0.16 0.59 -
Kanungo et al. (1987) ps 11 - 149 - 245 - 0.38 0.38 0.17 0.62 -
Magee (1996) ps 7 - 270 - 330 - 0.12 -0.12 0.055 0.16 -
Malbrunot et al. (1968) ps 30 - 191 - 348 - 0.24 0.032 0.27 0.58 -
Matsuda et al. (2002) ps 22 - 276 - 303 - 0.042 0.033 0.036 0.086 -

187
Nagel and Bier (1995) ps 27 - 204 - 351 - 0.057 0.035 0.063 0.16 -
Qian et al. (1993) ps 9 - 280 - 350 - 0.046 -0.043 0.032 0.095 -
Sato et al. (1994) ps 21 - 320 - 351 - 0.054 -0.035 0.047 0.087 -
Weber and Goodwin (1993) ps 27 - 208 - 237 - 0.049 -0.012 0.039 0.12 -
Weber and Silva (1994) ps 17 - 236 - 266 - 0.032 0.049 0.055 0.25 -
Widiatmo et al. (1994) ps 25 - 220 - 325 - 0.35 -0.25 0.50 1.3 -
Zhu et al. (1993) ρ' 31 - 273 - 347 - 0.23 0.027 0.28 0.56 -
Defibaugh et al. (1994) ρ' 21 - 243 - 338 - 0.096 0.079 0.14 0.47 -
Fukushima et al. (1995) ρ' 21 - 232 - 351 - 0.42 0.070 0.60 1.4 1
Higashi (1994) ρ' 8 - 336 - 351 - 1.0 0.60 1.3 2.5 -
Holcomb et al. (1993) ρ' 25 - 295 - 349 - 0.34 0.33 0.22 0.65 -
Kuwabara et al. (1995) ρ' 17 - 330 - 351 - 0.42 -0.42 0.33 0.98 -
Magee (1996) ρ' 13 - 139 - 305 - 0.042 0.027 0.042 0.080 -
Malbrunot et al. (1968) ρ' 15 - 248 - 349 - 0.32 0.024 0.38 0.68 1
 Table C.4 Deviations of Measurements from the New Fundamental Equation of State for R-32 (continued)
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Sato et al. (1994) ρ' 2 - 321 - 344 - 0.21 0.21 0.19 0.35 -
Shinsaka et al. (1985) ρ' 19 - 150 - 222 - 0.96 -0.96 0.11 1.1 -
Widiatmo et al. (1994) ρ' 22 - 220 - 330 - 0.14 -0.12 0.1 0.23 -
Defibaugh et al. (1994) ρ" 28 - 219 - 343 - 0.32 -0.031 0.36 0.61 -
Fukushima et al. (1995) ρ" 13 - 298 - 351 - 1.3 -0.17 1.7 3.3 1
Higashi (1994) ρ" 9 - 340 - 352 - 2.6 2.3 2.0 5.1 -
Holcomb et al. (1993) ρ" 25 - 245 - 349 - 0.62 0.45 0.80 3.0 -
Kuwabara et al. (1995) ρ" 13 - 346 - 351 - 0.83 -0.13 0.92 1.3 3
Sato et al. (1994) ρ" 4 - 327 - 346 - 1.4 -1.4 0.51 1.9 -
Lüddecke and Magee (1996) cs' 101 - 141 - 342 - 0.37 -0.042 0.46 1.2
Grebenkov et al. (1994) w' 16 - 250 - 325 - 0.30 0.27 0.19 1.1 -
Baroncini et al. (1993) ρ 93 0.74 - 2.7 273 - 360 G 0.19 0.075 0.25 0.86 -
Bouchot and Richon (1994) ρ 36 1.03 - 9.5 253 - 333 L 0.081 -0.042 0.090 0.18 -

188
Bouchot and Richon (1997) ρ 1433 4.1 - 85 253 - 333 L 0.061 -0.024 0.070 0.21 2
Bouchot and Richon (1997) ρ 622 0.97 - 39 253 - 333 G 0.27 -0.057 0.43 3.2 105
Defibaugh et al. (1994) ρ 167 0.30 - 9.8 268 - 373 G 0.17 -0.059 0.27 1.7 -
Defibaugh et al. (1994) ρ 219 2.0 - 6.5 243 - 349 L 0.13 0.094 0.25 1.8 7
de Vries (1997) ρ 565 0.02 - 21 263 - 433 G 0.053 -0.011 0.12 0.72 5
de Vries (1997) ρ 490 1.5 - 18 243 - 352 L 0.12 -0.033 0.25 1.5 8
de Vries (1997) ρ 94 0.11 - 1.4 224 - 334 G 0.074 -0.068 0.077 0.24 -
Fu et al. (1995) ρ 121 0.07 - 5.7 243 - 373 G 0.19 -0.02 0.25 1.3 -
Fukushima et al. (1995) ρ 102 1.9 - 9.6 314 - 424 G 0.86 -0.77 0.86 5.3 6
Fukushima et al. (1995) ρ 56 3.8 - 10 329 - 368 L 1.3 1.3 1.3 4.1 -
Holste et al. (1993) ρ 126 1.5 - 72 150 - 375 L 0.16 -0.14 0.22 0.96 -
Magee (1996) ρ 137 3.8 - 36 142 - 396 L 0.033 0.0090 0.037 0.068 -
 Table C.4 Deviations of Measurements from the New Fundamental Equation of State for R-32 (continued)
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Malbrunot et al. (1968) ρ 86 0.89 - 20 298 - 473 G 1.3 -0.96 1.4 4.7 6
Malbrunot et al. (1968) ρ 64 0.34 - 20 248 - 343 L 0.18 -0.17 0.14 0.83 2
Matsuda et al. (2002) ρ 144 0.20 - 1.9 276 - 303 G 0.032 -0.011 0.042 0.12 -
Matsuda et al. (2002) ρ 75 1.0 - 5.0 278 - 303 L 0.018 -0.018 0.0060 0.029 -
Qian et al. (1993) ρ 95 0.13 - 6.5 290 - 370 G 0.095 0.059 0.12 0.72 1
Sato et al. (1994) ρ 69 3.5 - 9.8 330 - 420 G 0.28 -0.20 0.26 1.2 5
Sato et al. (1994) ρ 11 3.3 - 6.5 322 - 350 L 0.11 0.11 0.065 0.24 -
Takahashi et al. (1995) ρ 113 0.10 - 10 298 - 423 G 0.37 -0.19 0.55 2.3 3
Zhang et al. (1996) ρ 81 0.13 - 6.5 290 - 370 G 0.063 -0.043 0.11 0.81 -
Kunimoto et al. (1994) cp 20 0.00 - 0.50 303 - 343 G 14 -14 0.95 16 -
Kubota et al. (1995) cp 18 0.00 - 0.50 303 - 343 G 14 -14 1.5 17 -

189
Yomo et al. (1994) cp 19 2.1 - 3.0 276 - 315 L 1.1 -1.1 0.20 1.6 -
Lüddecke and Magee (1996) cv 74 5.2 - 32 253 - 341 L 0.26 -0.061 0.32 0.85 -
Grebenkov et al. (1994) w 30 1.5 - 10 287 - 341 L 0.55 0.55 0.19 1.1 -
Grebenkov et al. (1995) w 93 1.3 - 17 227 - 341 L 1.1 1.1 0.67 0.67 2
Hozumi et al. (1994) w 67 0.02 - 0.39 273 - 343 G 0.0064 -0.0059 0.0070 0.024 -
Pires and Guedes (1999) w 305 1.7 - 65 248 - 343 L 0.36 0.27 0.36 2.6 2
Sun et al. (1997) w 38 0.05 - 0.39 231 - 254 G 0.037 0.016 0.046 0.16 -
Takagi (1993) w 120 0.38 - 33 243 - 373 L 0.36 0.30 0.53 3.6 5
 Table C.5 Deviations of Measurements from the New Fundamental Equation of State for R-125
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Baroncini et al. (1993) ps 58 - 236 - 333 - 0.084 -0.039 0.12 0.47 -
Boyes and Weber (1995) ps 29 - 273 - 335 - 0.035 -0.016 0.037 0.068 -
de Vries (1997) ps 104 - 222 - 339 - 0.046 -0.0095 0.055 0.090 -
Duarte-Garza et al. (1997) ps 15 - 220 - 338 - 0.17 0.17 0.17 0.71 -
Fukushima et al. (1995) ps 45 - 252 - 339 - 0.15 -0.045 0.19 0.57 -
Magee (1996) ps 33 - 190 - 335 - 0.14 -0.11 0.14 0.58 1
Matsuda et al. (2003) ps 22 - 278 - 303 - 0.032 -0.023 0.028 0.059 -
Lüddecke and Magee (1996) ps 9 - 170 - 250 - 0.12 -0.0048 0.25 0.48 -
Oguchi et al (1996) ps 61 - 223 - 338 - 0.074 0.047 0.10 0.59 -
Sagawa et al. (1994) ps 26 - 308 -339 - 0.052 0.0039 0.061 0.11 -

190
Takagi (1996b) ps 9 - 241 - 333 - 0.052 0.32 0.16 0.73 1
Tillner Roth (1996) ps 30 - 178 - 230 - 0.27 -0.12 0.39 1.77 -
Tsvetkov et al. (1996) ps 34 - 263 - 338 - 0.037 0.0018 0.053 0.17 -
Weber and Silva (1994) ps 103 - 219 - 285 - 0.056 0.040 0.073 0.21 -
Widiatmo et al. (1994) ps 20 - 220 - 335 - 0.34 -0.29 0.47 1.9 -
Ye et al. (1995) ps 12 - 290 - 339 - 0.59 -0.0076 0.053 0.13 -
Fukushima et al. (1995) ρ' 47 - 227 - 339 - 0.24 -0.034 0.41 1.1 8
Higashi (1994) ρ' 6 - 324 - 339 - 0.30 0.24 0.36 0.88 -
Kuwabara et al. (1995) ρ' 9 - 331 - 339 - 0.31 -0.13 0.36 0.76 -
Magee (1996) ρ' 7 - 173 - 308 - 0.11 0.098 0.063 0.16 -
Takahashi et al. (1994) ρ' 22 - 241 - 338 - 0.12 0.075 0.14 0.39 -
Widiatmo et al. (1994) ρ' 25 - 220 - 335 - 0.060 -0.049 0.065 0.22 -
 Table C.5 Deviations of Measurements from the New Fundamental Equation of State for R-125 (continued)
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Boyes and Weber (1995) ρ" 10 - 275 - 320 - 0.057 0.049 0.059 0.14 -
Fukushima et al. (1995) ρ" 20 - 315 - 339 - 0.56 -0.40 0.61 1.6 6
Higashi (1994) ρ" 7 - 327 - 339 - 0.82 0.27 0.99 1.5 -
Kuwabara et al. (1995) ρ" 8 - 336 - 339 - 0.47 -0.47 0.44 1.1 2
Lüddecke and Magee (1996) cs' 93 - 176 - 278 - 0.43 0.11 0.48 1.0 -
Kraft and Leipertz (1994) w' 13 - 293 - 339 - 4.5 -4.5 2.7 7.9 -
Kraft and Leipertz (1994) w" 9 - 307 - 338 - 2.1 -0.57 2.1 4.0 -
Baroncini et al. (1993) ρ 58 0.96 - 2.5 298 - 338 G 0.47 -0.20 0.51 1.1 -
Boyes and Weber (1995) ρ 80 0.33 - 4.6 273 - 363 G 0.20 -0.15 0.39 1.6 -
Defibaugh and Morrison ρ 232 1.59 - 6.3 275 - 369 L 0.69 -0.54 0.73 2.0 14

191
de Vries (1997) ρ 367 0.02 - 20 263 - 393 G 0.10 -0.083 0.23 1.9 4
de Vries (1997) ρ 595 1.4 - 18 243 - 341 L 0.30 -0.11 0.54 2.0 14
Duarte-Garza et al. (1997) ρ 105 1.2 - 68 180 - 350 L 0.27 -0.11 0.085 0.29 2
Duarte-Garza et al. (1997) ρ 213 1.2 - 68 180 - 350 G 0.23 -0.21 0.33 1.8 8
Fujimine et al. (1999) ρ 14 1.5 - 3.0 280 - 320 L 0.023 -0.0054 0.034 0.091 -
Fukushima et al. (1992) ρ 151 0.82 - 8.1 282 - 424 G 0.77 -0.42 0.85 2.7 9
Magee (1996) ρ 77 3.6 - 35 178 - 398 L 0.045 0.0021 0.057 0.13 -
Matsuda et al. (2003) ρ 63 0.29 - 1.6 278 - 303 G 0.061 -0.061 0.043 0.14 -
 Table C.5 Deviations of Measurements from the New Fundamental Equation of State for R-125 (continued)
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Matsuda et al. (2003) ρ 20 0.91- 5.0 283 - 303 L 0.021 0.012 0.021 0.043 -
Oguchi et al. (1996) ρ 167 0.81 - 17 280 - 473 G 0.53 0.25 0.59 1.7 -
Sagawa et al. (1994) ρ 92 0.69 - 12 308 - 423 G 0.46 0.18 0.54 2.1 -
Tsvetkov et al. (1995) ρ 44 0.60 - 6.1 273 - 443 G 0.63 -0.61 0.61 2.3 4
Ye et al. (1995) ρ 93 0.12 - 3.6 290 - 390 G 0.10 0.041 0.13 0.61 -
Zhang et al. (1995) ρ 93 0.12 - 3.6 290 - 390 G 0.071 -0.018 0.12 0.64 -
Wilson et al. (1991) cp 5 0.13 233 - 333 G 1.1 -0.28 1.7 3.1 -
Wilson et al. (1991) cp 5 3.45 216 - 323 L 1.0 0.0018 1.5 2.3 -
Zhao et al. (2001) cp 24 1.5 - 3.0 285 - 325 L 0.21 -0.053 0.27 0.64 -
Lüddecke and Magee (1996) cv 99 3.8 - 33 200 - 342 L 0.59 0.34 0.62 1.7 -
Gillis (1997) w 149 0.04 - 1.0 240 - 380 G 0.015 -0.0085 0.017 0.057 -

192
Grebenkov et al. (1995) w 29 4.8 - 16 303 - 333 L 0.31 -0.25 0.38 1.1 4
Grigiante et al. (2000) w 69 0.05 - 0.5 260 - 360 G 0.011 0.0017 0.014 0.040 -
Hozumi et al. (1996) w 72 0.01 - 0.24 273 - 343 G 0.035 0.035 0.011 0.062 -
Kojima and Sato (2002) w 55 0.01 - 0.50 303 - 343 G 0.0067 -0.0050 0.0057 0.017 -
Takagi (1996b) w 119 0.46 - 32 240 - 333 L 0.81 0.80 0.45 1.4 -
 Table C.6 Deviations of Measurements from the New Fundamental Equation of State for R-134a
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Arita et al. (1991) ps 3 - 273 - 323 - 0.057 -0.057 0.025 0.082 -
Baehr and Tillner-Roth (1991) ps 37 - 303 - 373 - 0.019 -0.015 0.015 0.035 2
Baroncini et al. (1990) ps 64 - 243 - 359 - 0.25 0.15 0.24 0.80 -
Basu and Wilson (1989) ps 32 - 211 - 369 - 0.35 0.25 0.58 2.4 -
Blanke et al. (1995) ps 53 - 175 - 374 - 0.11 -0.03 0.13 0.45 -
Fukushima et al. (1997) ps 41 - 262 - 372 - 0.13 -0.029 0.19 0.78 -
Goodwin et al. (1992) ps 57 - 214 - 313 - 0.033 0.020 0.038 0.13 -
Hou et al. (1992) ps 10 - 180 - 360 - 0.42 0.40 0.76 2.2 2
Kubota et al. (1989) ps 25 - 253 - 373 - 0.23 0.013 0.29 0.6 -

193
Maezawa et al. (1990) ps 13 - 280 - 350 - 0.38 -0.0077 0.61 1.7 -
Magee and Howley (1992) ps 19 - 180 - 350 - 0.26 -0.057 0.42 1.4 -
Morrison and Ward (1991) ps 12 - 268 - 368 - 0.084 0.034 0.13 0.31 1
Niesen et al. (1994) ps 12 - 316 - 370 - 0.074 0.057 0.054 0.13 -
Piao et al. (1990) ps 51 - 308 - 368 - 0.012 -0.11 0.080 0.31 5
Tillner-Roth (1996) ps 33 - 170 - 232 - 0.55 0.54 0.74 2.3 -
Weber (1989) ps 22 - 313 - 373 - 0.021 -0.020 0.012 0.040 -
Wilson and Basu (1988) ps 32 - 211 - 369 - 0.36 0.24 0.61 2.5 -
Zhu et al. (1992) ps 43 - 279 - 363 - 0.093 -0.032 0.11 0.27 -
Aoyama et al. (1996) ρ' 4 - 372 - 374 - 1.1 -1.1 0.65 1.7 -
Basu and Wilson (1989) ρ' 9 - 239 - 372 - 0.29 -0.19 0.45 1.2 -
Defibaugh and Moldover ρ' 26 - 245 - 370 - 0.12 0.035 0.18 0.66 -
Fukushima et al. (1990b) ρ' 3 - 325 - 357 - 0.37 -0.31 0.38 0.73 -
 Table C.6 Deviations of Measurements from the New Fundamental Equation of State for R-134a (continued)
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Fukushima (1991) ρ' 7 - 244 - 292 - 0.10 0.071 0.10 0.23 -
Fukushima et al. (1990a) ρ' 8 - 370 - 374 - 0.057 0.057 0.017 0.071 2
Higashi (1994) ρ' 8 - 359 - 374 - 0.98 0.22 1.9 4.6 -
Hou et al. (1992) ρ' 12 - 180 - 360 - 0.14 -0.14 0.20 0.69 -
Kabata et al. (1989) ρ' 12 - 343 - 374 - 0.65 0.48 1.4 4 -
Maezawa et al. (1990) ρ' 25 - 200 - 370 - 0.13 -0.11 0.13 0.46 -
Morrison and Ward (1991) ρ' 12 - 268 - 368 - 0.21 -0.21 0.24 0.77 1
Niesen et al. (1994) ρ' 12 - 316 - 370 - 0.18 -0.18 0.18 0.62 -
Takahashi et al. (1994) ρ' 7 - 258 - 306 - 0.14 0.14 0.039 0.18 -
Yokoyama and Takahashi ρ' 21 - 252 - 367 - 0.10 -0.011 0.18 0.66 -
Aoyama et al. (1996) ρ" 7 - 373 - 374 - 1.2 0.13 1.6 2.4 -
Fukushima et al. (1990b) ρ" 9 - 371 - 374 - 1.4 -0.30 2.5 6.3 2

194
Fukushima (1991) ρ" 6 - 294 - 371 - 0.24 -0.24 0.34 0.64 -
Higashi (1994) ρ" 6 - 362 – 374 - 1.9 -0.31 2.9 5.5 -
Kabata et al. (1989) ρ" 15 - 361 - 374 - 2.8 0.48 4.8 13 3
Morrison and Ward (1991) ρ" 8 - 298 - 368 - 5.6 4.8 7.0 18 -
Niesen et al. (1994) ρ" 12 - 316 - 370 - 0.43 0.35 0.57 1.9 -
Weber (1989) ρ" 5 - 320 - 368 - 0.17 -0.16 0.25 0.59 -
Magee (1992) cs' 160 - 172 - 290 0.26 0.13 0.29 0.85 -
Baroncini et al. (1990) ρ 46 0.18 - 1.9 263 - 359 G 0.52 0.40 0.46 1.2 -
Basu and Wilson (1989) ρ 28 1.1 - 6.7 317 - 448 G 0.43 -0.15 0.61 1.8 2
Blanke et al. (1995) ρ 44 0.86 - 4.6 180 - 233 L 0.045 0.045 0.036 0.10 -
Defibaugh and Moldover ρ 924 0.80 – 6.5 242 - 372 L 0.083 0.083 0.092 1.2 -
 Table C.6 Deviations of Measurements from the New Fundamental Equation of State for R-134a (continued)
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Dresner and Bier (1992) ρ 121 0.28 - 58 333 - 423 G 0.058 -0.0061 0.090 0.28 -
Fukushima et al. (1990) ρ 63 0.55 - 5.7 294 - 424 L, G 0.39 0.098 0.56 2.7 2
Hou et al. (1992) ρ 429 0.75 - 71 180 - 380 L 0.16 -0.15 0.18 1.8 -
Klomfar et al. (1993) ρ 89 1.0 - 56 204 - 310 L 0.13 0.13 0.032 0.18 -
Maezawa et al. (1990) ρ 10 0.51- 2.0 280 - 340 L 0.15 -0.15 0.038 0.20 -
Magee (1992) ρ 150 2.6 - 35 187 - 343 L 0.059 -0.038 0.062 0.18 -
Morrison and Ward (1991) ρ 121 0.70 - 5.8 279 - 367 L 0.085 -0.039 0.14 0.50 -
Qian et al. (1992) ρ 21 0.13 - 1.9 320 - 340 G 0.025 0.022 0.027 0.065 -
Tillner-Roth and Baehr (1993) ρ 432 0.73 - 16 243 - 413 L 0.097 0.0075 0.17 1.3 1
Tillner-Roth and Baehr (1992) ρ 411 0.09 - 16 293 - 453 G 0.058 -0.0001 0.12 1.1 -
Weber (1989) ρ 56 1.2 - 5.3 321 - 423 G 0.067 -0.052 0.087 0.32 -
Yamamoto et al. (1993) ρ 38 0.10 - 3.1 273 - 363 G 0.19 -0.19 0.29 1.3 -

195
Zhang et al. (1995) ρ 21 0.13 - 1.9 320 - 340 G 0.026 0.017 0.029 0.066 -
Zhu et al. (1992) ρ 42 0.14 - 1.3 283 - 353 G 0.38 0.091 0.44 0.90 -
Saitoh et al. (1990) cp 31 1.0 - 3.0 276 - 356 L 0.72 0.72 0.32 1.5 -
Sato and Watanabe (1994) cp 31 1.0 - 3.0 276 - 350 L 0.73 0.72 0.32 1.5 2
Magee (1992) cv 150 2.6 - 36 387 - 343 L 0.34 0.050 0.44 1.4 -
Beckermann and Kohler w 245 0.025 - 0.50 260 - 420 G 0.021 -0.010 0.029 0.11 -
Goodwin and Moldover w 94 0.006 - 0.6 233 - 340 G 0.023 -0.023 0.014 0.052 -
Grebenkov et al. (1995) w 48 0.1 -16 296 - 351 L 0.36 0.26 0.30 1.1 -
Guedes and Zollweg (1992) w 206 0.1- 69.5 179 - 380 L 0.47 0.11 0.78 4.2 -
Hozumi et al. (1997) w 23 0.01- 0.25 273 - 298 G 0.028 -0.028 0.012 0.048 -
Takagi (1996a) w 237 0.08 - 31 243 - 333 L 0.29 0.29 0.18 0.58 11
 Table C.7 Deviations of Measurements from the New Fundamental Equation of State for R-152a

Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out

Baehr and Tillner-Roth (1991) ps 55 - 301 - 386 - 0.18 0.18 0.030 0.23 -
Blanke and Weiss (1992) ps 36 - 155 - 260 - 0.082 0.0074 0.15 0.83 5
Duarte-Garza and Magee ps 21 - 155 - 250 - 0.092 0.0019 0.13 0.30 -
Fransson et al. (1992) ps 5 - 303 - 385 - 0.67 0.67 0.28 0.89 -
Higashi et al. (1999) ps 44 - 273 - 386 - 0.059 0.037 0.057 0.17 -
Holcomb et al. (1993) ps 33 - 312 - 383 - 0.18 0.17 0.20 0.10 -
Kamei et al. (1989) ps 24 - 250 - 365 - 0.069 0.044 0.10 0.41 -
McLinden (1989) ps 21 - 273 - 373 - 0.051 0.019 0.054 0.083 -
Mears et al. (1955) ps 27 - 204 - 377 - 2.1 1.7 3.0 9.7 7
Sato et al. (1987) ps 41 - 273 - 373 - 0.047 0.018 0.047 0.074 -
Silva and Weber (1993) ps 38 - 220 - 273 - 0.097 -0.097 0.026 0.14 1
Tillner-Roth (1996) ps 40 - 155 - 232 - 0.24 0.22 0.36 1.2 6

196
Yada et al. (1988) ps 13 - 305 - 385 - 0.13 0.11 0.17 0.53 -
Zhao et al. (1992) ps 167 - 237 - 368 - 0.20 -0.092 0.24 0.75 1
Blanke and Weiss (1992) ρ' 19 - 160 - 308 - 0.061 -0.061 0.029 0.12 -
Geller (1979) ρ' 19 - 160 - 340 - 0.11 0.0050 0.13 0.22 -
Higashi (1987) ρ' 10 - 371 - 386 - 0.35 0.047 0.46 0.90 3
Holcomb et al. (1993) ρ' 33 - 312 - 384 - 0.15 -0.15 0.077 0.27 -
Kabata (1988) ρ' 10 - 371 - 386 - 0.35 0.047 0.46 0.90 3
Kamei et al. (1989) ρ' 24 - 250 - 365 - 0.14 -0.12 0.13 0.29 -
Magee (1998) ρ' 9 - 157 - 318 - 0.027 0.026 0.032 0.095 -
McLinden (1989) ρ' 29 - 233 - 373 - 0.15 -0.15 0.11 0.34 -
Mears et al. (1955) ρ' 11 - 232 - 353 - 0.27 -0.22 0.26 0.77 -
 Table C.7 Deviations of Measurements from the New Fundamental Equation of State for R-152a (continued)
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Wang et al. (1992) ρ' 11 - 376 - 386 - 1.1 -0.77 1.4 2.9 4
Higashi et al. (1987) ρ" 11 - 374 - 386 - 1.0 0.40 1.6 2 2
Holcomb et al. (1993) ρ" 33 - 312 - 384 - 1.0 1.0 0.75 2.8 1
Kabata (1988) ρ" 11 - 374 - 386 - 1.0 0.40 1.2 1.9 2
Kamei et al. (1989) ρ" 24 - 250 - 364 - 0.50 -0.50 0.55 1.0 -
Wang et al. (1992) ρ" 11 - 377 - 386 - 1.1 0.16 1.4 1.9 3
Magee (1998) cs' 66 - 163 - 315 - 0.58 0.58 0.27 1.0 -
Gunther and Steimle (1997) cp' 8 - 203 - 323 - 0.42 -0.065 0.55 1.1 -
Beliajeva et al. (1994) w' 17 - 250 - 330 - 0.4 0.17 0.42 0.8 -
Blanke and Weiss (1992) ρ 209 0.59 - 31 160 - 453 L 0.073 -0.059 0.074 0.72 1
Defibaugh and Moldover ρ 1011 0.70 - 6.5 243 - 371 L 0.035 -0.028 0.025 0.061 -
Dressner and Bier (1993) ρ 150 0.23 - 58 333 - 423 G 0.11 -0.035 0.20 0.92 2

197
Geller et al. (1979) ρ 97 0.65 - 58 159 - 470 L 0.14 -0.069 0.17 0.57 2
Iso and Uematsu (1989) ρ 241 1.25 - 10 313 - 400 L 0.74 0.40 1.1 3.9 12
Magee (1998) ρ 134 2.05 - 35 158 - 400 L 0.045 -0.002 0.061 0.16 -
Mears et al. (1955) ρ 22 1.81 - 4.6 345 - 397 G 1.3 -1.0 0.96 0.94 6
Strom and Gren (1993) ρ 21 0.47 - 1.6 261 - 323 L 0.82 0.70 0.59 1.3 -
Tillner-Roth and Baehr (1992) ρ 335 0.10 - 16 293 - 433 G 0.068 0.005 0.19 1.7 2
Tillner-Roth and Baehr (1993) ρ 413 0.73 - 16 243 - 413 L 0.14 -0.029 0.30 1.9 4
Yada et al. (1992) ρ 16 4.2 - 8.8 383 - 423 G 12 12 7.8 30 -
Zhao et al. (1992) ρ 257 0.09 - 6.1 253 - 404 G 1.1 -0.64 1.3 5.0 82
Kubota (1987) cp 20 0.50 - 1.4 313 - 354 L 4.6 4.6 1.4 7.4 -
 Table C.7 Deviations of Measurements from the New Fundamental Equation of State for R-152a (continued)
Reference Y N p , MPa T, K Phase AAD, % BIAS, % STD, % Max, % N out
Nakagawa et al. (1993) cp 70 1.0 - 3.2 210 - 360 L 1.2 -1.2 0.20 1.6 1
Porrichanski et al. (1982) cp 304 2.0 - 20 181 - 424 L 1.2 0.38 1.6 6.6 27
Sato and Watanabe (1994) cp 36 1.0 - 3.2 276 - 360 L 1.2 -1.2 0.21 1.6 -
Magee (1998) cv 85 3.0 - 34 163 - 342 L 0.48 0.44 0.35 1.2 -
Beliajeva et al. (1999) w 52 0.54 - 18 233 - 303 L 1.2 1.2 0.34 2.0 1
Gillis (1997) w 148 0.03 - 1.0 243 - 400 G 0.016 0.01 0.021 0.082
Grebenkov et al. (1995) w 54 0.44 - 17 287 - 367 L 0.56 0.28 0.59 1.4 2
He et al. (2002) w 24 0.20 - 0.42 294 - 323 G 0.63 0.63 0.20 1.0 1
Hozumi et al. (1993) w 92 0.19 - 0.22 273 - 348 G 0.014 -0.014 0.0091 0.071 2
Takagi (1999) w 146 0.12 - 31 243 - 333 L 0.68 0.66 0.46 1.9 2

198
 Appendix D
Thermodynamic Properties for Computer Program Verification
Following thermodynamic properties were calculated from the fundamental equations of state of
this study. These values are very useful to verify the computer program. The thermodynamic
properties cover the single phase and saturation.

Table D.1 Thermodynamic Properties of R-32 in Single Phase

p T ρ cv cp w h s
3
MPa K kg/m kJ/(kg·K) kJ/(kg·K) m/s kJ/kg kJ/(kg·K)
0.01 150 1396.04 1.0230 1.5784 1358.90 3.07 0.0486
0.01 200 0.3147 0.5575 0.7242 202.54 486.67 2.6970
0.01 300 0.2088 0.6691 0.8299 243.58 563.52 3.0071
0.01 350 0.1789 0.7461 0.9065 260.52 606.90 3.1407
1 200 1271.27 0.9531 1.5582 1084.59 81.45 0.4955
1 250 1131.76 0.9386 1.6428 823.89 160.99 0.8501
1 350 19.2376 0.7908 1.0196 249.21 592.74 2.3752
1 400 16.3408 0.8491 1.0489 268.92 644.26 2.5127
5 250 1141.67 0.9388 1.6207 851.08 162.21 0.8409
5 300 974.01 0.9611 1.8666 583.57 248.09 1.1534
5 400 101.46 0.9506 1.4535 238.55 598.07 2.1683
5 450 80.68 0.9716 1.3046 267.62 665.96 2.3284
10 150 1404.21 1.0362 1.5711 1376.86 8.36 0.0363
10 300 999.35 0.9563 1.7721 638.74 247.37 1.1341
10 375 546.55 1.1194 4.0182 228.32 423.77 1.6516
10 450 191.02 1.0307 1.6680 251.82 622.30 2.1427
30 225 1249.69 0.9514 1.5218 1086.71 133.42 0.6319
30 300 1067.79 0.9548 1.6069 791.52 250.06 1.0787
30 450 623.18 1.0689 1.9103 392.87 517.90 1.7980
50 150 1433.83 1.0760 1.5486 1445.67 29.89 -0.0081
50 225 1274.76 0.9590 1.4971 1158.12 143.00 0.6041
50 375 946.33 1.0084 1.6334 684.00 375.32 1.3923
50 450 778.17 1.0738 1.7087 542.05 501.04 1.6977
70 150 1447.17 1.0902 1.5396 1478.49 40.79 -0.0279
70 225 1296.61 0.9662 1.4800 1220.32 152.96 0.5792
70 300 1149.33 0.9683 1.5033 983.06 264.52 1.0070
70 450 864.92 1.0841 1.6278 652.20 499.33 1.6399
5.110E-04 160 0.02 0.5163 0.6775 183.00 459.86 3.0215
3.931E+00 160 1375 1.0057 1.5592 1306.27 20.80 0.1448
2.363E-02 200 0.75 0.5743 0.7509 201.51 485.48 2.5549
9.257E-02 200 1270 0.9526 1.5600 1080.81 81.03 0.4969
7.842E-02 250 2 0.6222 0.8002 222.41 521.20 2.5255
8.611E+00 250 1150 0.9394 1.6038 873.86 163.40 0.8330
1.513E+00 300 40 0.8685 1.3646 211.40 525.73 2.1096
2.703E+00 300 960 0.9650 1.9313 553.26 248.82 1.1637
3.824E+00 350 100 0.9730 1.7912 209.37 539.93 2.0442
2.398E+01 350 900 0.9865 1.8028 563.66 334.39 1.3570
4.946E+00 400 100 0.9491 1.4453 238.97 598.78 2.1715
3.300E+01 400 800 1.0260 1.8141 504.71 420.92 1.5597

199
 Table D.2 Thermodynamic Properties of R-32 at Saturation
ps Ts ρ' ρ" cv' cv" cp' cp" w' w" h' h" s' s"
3 3
MPa K kg/m kg/m kJ/(kg·K) kJ/(kg·K) kJ/(kg·K) kJ/(kg·K) m/s m/s kJ/kg kJ/kg kJ/(kg·K) kJ/(kg·K)
8.440E-05 140 1420.54 0.0038 1.0509 0.5022 1.6062 0.6625 1428.67 171.76 -12.85 446.60 -0.0612 3.2206
1.048E-03 160 1371.39 0.0411 1.0014 0.5193 1.5628 0.6818 1297.16 182.88 18.76 459.72 0.1499 2.9059
6.907E-03 180 1321.37 0.2417 0.9712 0.5453 1.5528 0.7138 1185.67 192.74 49.88 472.64 0.3332 2.6818
2.961E-02 200 1269.91 0.9431 0.9526 0.5819 1.5601 0.7632 1080.55 201.04 81.00 484.94 0.4970 2.5167
9.388E-02 220 1216.38 2.7763 0.9420 0.6295 1.5814 0.8342 976.81 207.40 112.42 496.14 0.6465 2.3907
2.396E-01 240 1159.87 6.7206 0.9382 0.6876 1.6194 0.9323 872.29 211.34 144.45 505.67 0.7853 2.2904
5.213E-01 260 1099.01 14.203 0.9405 0.7557 1.6815 1.0675 765.68 212.40 177.49 512.91 0.9165 2.2066
1.007E+00 280 1031.61 27.385 0.9495 0.8338 1.7835 1.2634 655.65 210.07 212.10 517.02 1.0430 2.1320
1.775E+00 300 953.74 49.968 0.9671 0.9240 1.9640 1.5819 539.72 203.72 249.21 516.69 1.1683 2.0599
2.921E+00 320 856.86 89.821 1.0010 1.0326 2.3537 2.2424 412.44 192.35 290.66 509.27 1.2979 1.9810
4.563E+00 340 713.35 173.05 1.0855 1.1826 3.9197 4.9395 263.65 173.71 341.91 486.50 1.4467 1.8719
0.001 159.57 1372.47 0.0393 1.0022 0.5189 1.5633 0.6812 1299.73 182.66 18.08 459.44 0.1457 2.9117
0.01 184.64 1309.59 0.3418 0.9660 0.5528 1.5532 0.7235 1160.99 194.82 57.08 475.56 0.3727 2.6392

200
0.1 221.23 1213.00 2.9460 0.9416 0.6328 1.5832 0.8394 970.42 207.71 114.37 496.78 0.6553 2.3839
0.25 241.01 1156.92 6.9983 0.9381 0.6908 1.6219 0.9381 866.98 211.47 146.08 506.10 0.7920 2.2858
0.5 258.83 1102.71 13.636 0.9402 0.7514 1.6770 1.0583 771.98 212.43 175.52 512.57 0.9090 2.2112
1 279.78 1032.39 27.199 0.9494 0.8329 1.7821 1.2608 656.87 210.12 211.71 517.00 1.0416 2.1328
1.5 293.76 979.50 41.580 0.9604 0.8943 1.8949 1.4622 576.74 206.18 237.29 517.38 1.1291 2.0826
2 304.57 933.79 57.099 0.9729 0.9468 2.0254 1.6872 511.97 201.62 258.19 515.76 1.1972 2.0429
2.5 313.48 891.47 74.077 0.9871 0.9944 2.1848 1.9574 455.72 196.70 276.48 512.75 1.2546 2.0083
3 321.13 850.47 92.928 1.0039 1.0396 2.3896 2.3036 404.75 191.52 293.21 508.53 1.3055 1.9760
3.5 327.87 809.26 114.24 1.0240 1.0839 2.6690 2.7788 357.32 186.12 309.01 503.12 1.3524 1.9444
4 333.89 766.36 138.96 1.0489 1.1289 3.0837 3.4921 312.38 180.48 324.39 496.34 1.3969 1.9119
5 344.32 665.60 207.39 1.1235 1.2291 5.3465 7.4051 226.59 167.94 356.36 476.29 1.4870 1.8353
5.5 348.89 589.93 268.45 1.1929 1.2954 12.6129 19.2339 181.85 159.74 376.65 457.88 1.5432 1.7761
 Table D.3 Thermodynamic Properties of R-125 in Single Phase
p T ρ cv cp w h s
3
MPa K kg/m kJ/(kg·K) kJ/(kg·K) m/s kJ/kg kJ/(kg·K)
0.01 200 0.725942 0.5502 0.6214 124.37 293.42 1.5944
0.01 225 0.644068 0.5949 0.6655 131.53 309.51 1.6701
0.01 250 0.578991 0.6383 0.7085 138.27 326.69 1.7425
0.01 300 0.481883 0.7207 0.7905 150.76 364.18 1.8790
0.01 350 0.412788 0.7969 0.8665 162.23 405.64 2.0066
1 200 1602.3 0.7082 1.0684 785.23 116.61 0.6451
1 250 1422.1 0.7600 1.1657 563.58 172.14 0.8926
1 300 57.9631 0.7632 0.9288 129.87 351.78 1.5294
1 400 38.1186 0.8774 0.9705 165.63 445.07 1.7974
5 250 1440.27 0.7613 1.1467 593.19 173.07 0.8851
5 300 1226.11 0.8282 1.3183 381.64 234.10 1.1072
5 350 749.976 0.9501 2.9774 122.33 317.50 1.3621
5 400 256.096 0.9316 1.2743 135.57 416.50 1.6301
5 450 192.121 0.9661 1.1517 161.48 476.13 1.7706
10 175 1701.65 0.6869 1.0379 935.65 93.77 0.4941
10 225 1543.5 0.7357 1.0871 723.97 146.72 0.7600
10 300 1271.79 0.8244 1.2498 441.87 233.64 1.0923
10 375 851.402 0.9344 1.5867 199.59 338.72 1.4035
10 450 445.186 0.9933 1.3487 163.63 452.14 1.6801
30 200 1662.03 0.7166 1.0420 911.12 127.73 0.6120
30 300 1379.17 0.8183 1.1580 598.22 237.01 1.0535
30 375 1156.19 0.9106 1.2519 433.91 327.60 1.3226
30 450 943.636 0.9926 1.2971 340.48 423.52 1.5556
50 175 1760.19 0.6962 1.0304 1085.81 110.04 0.4551
50 225 1630.67 0.7416 1.0478 900.38 161.83 0.7154
50 300 1445.03 0.8180 1.1267 702.69 243.17 1.0268
50 375 1265.33 0.9078 1.2064 561.16 330.80 1.2872
50 450 1101.1 0.9899 1.2590 471.56 423.43 1.5123
70 175 1783.83 0.7044 1.0318 1146.02 118.37 0.4383
70 225 1662.86 0.7465 1.0397 966.26 169.98 0.6976
70 300 1494.39 0.8195 1.1104 784.69 250.37 1.0055
70 375 1336.53 0.9078 1.1873 655.71 336.63 1.2618
70 450 1194.46 0.9897 1.2440 570.04 427.94 1.4836
1.495E+01 175 1710 0.6882 1.0358 955.35 95.74 0.4888
1.212E-04 175 0.01 0.5017 0.5710 117.45 278.81 1.8216
3.171E+01 200 1665 0.7170 1.0411 917.87 128.41 0.6102
1.384E-03 200 0.1 0.5490 0.6186 124.84 293.65 1.7321
1.385E-04 200 0.01 0.5489 0.6182 124.91 293.68 1.8917
6.122E+00 250 1445 0.7614 1.1422 601.18 173.35 0.8831
1.724E-02 250 1 0.6387 0.7095 138.05 326.57 1.7044
3.021E+01 300 1380 0.8183 1.1575 599.50 237.07 1.0531
2.015E-01 300 10 0.7265 0.8069 147.28 362.11 1.6659
3.816E+00 350 300 0.9213 2.0248 101.87 359.50 1.4892
4.375E+00 350 500 0.9761 4.6824 93.54 337.61 1.4223
7.783E+00 400 500 0.9602 1.6128 136.44 392.28 1.5502

201
 Table D.4 Thermodynamic Properties of R-125 at Saturation
ps Ts ρ' ρ" cv' cv" cp' cp" w' w" h' h" s' s"
3 3
MPa K kg/m kg/m kJ/(kg·K) kJ/(kg·K) kJ/(kg·K) kJ/(kg·K) m/s m/s kJ/kg kJ/kg kJ/(kg·K) kJ/(kg·K)
5.677E-03 180 1666.92 0.4574 0.6868 0.5130 1.0484 0.5839 875.87 118.58 95.08 281.50 0.5349 1.5706
2.473E-02 200 1599.89 1.811 0.7074 0.5555 1.0699 0.6306 781.18 123.52 116.26 293.00 0.6464 1.5301
7.866E-02 220 1530.78 5.347 0.7268 0.6013 1.1006 0.6854 691.01 127.05 137.97 304.60 0.7497 1.5071
2.001E-01 240 1457.56 12.911 0.7479 0.6494 1.1430 0.7500 602.56 128.66 160.42 315.95 0.8470 1.4950
2.993E-01 250 1418.68 18.997 0.7595 0.6739 1.1697 0.7871 558.25 128.60 172.00 321.41 0.8940 1.4916
4.324E-01 260 1377.76 27.182 0.7719 0.6989 1.2010 0.8290 513.53 127.86 183.87 326.66 0.9401 1.4893
6.062E-01 270 1334.29 38.033 0.7851 0.7244 1.2385 0.8777 468.11 126.38 196.08 331.62 0.9857 1.4877
8.280E-01 280 1287.52 52.300 0.7992 0.7508 1.2847 0.9368 421.69 124.04 208.68 336.20 1.0309 1.4864
1.10545 290 1236.38 71.044 0.8144 0.7785 1.3445 1.0128 373.88 120.75 221.76 340.29 1.0761 1.4848
1.44703 300 1179.18 95.880 0.8312 0.8081 1.4273 1.1184 324.19 116.34 235.45 343.71 1.1215 1.4824
2.36093 320 1032.00 177.539 0.8745 0.8780 1.7947 1.5998 216.22 103.09 265.66 347.03 1.2163 1.4705
0.001 161.716 1728.11 0.0894 0.6601 0.4760 1.0313 0.5457 968.51 113.19 76.06 271.36 0.4236 1.6312

202
0.01 187.105 1643.22 0.7772 0.6946 0.5277 1.0553 0.5997 841.53 120.47 102.56 285.55 0.5757 1.5537
0.1 224.765 1513.79 6.699 0.7316 0.6126 1.1095 0.6998 669.87 127.63 143.24 307.34 0.7733 1.5034
0.25 245.416 1436.73 15.975 0.7541 0.6626 1.1569 0.7696 578.60 128.71 166.65 318.93 0.8725 1.4930
0.5 264.197 1359.86 31.374 0.7773 0.7095 1.2159 0.8485 494.56 127.33 188.95 328.78 0.9593 1.4886
1 286.448 1255.13 63.788 0.8088 0.7685 1.3213 0.9833 391.04 122.04 217.05 338.90 1.0600 1.4854
1.5 301.388 1170.64 99.940 0.8337 0.8124 1.4415 1.1367 317.13 115.62 237.40 344.11 1.1279 1.4819
2 312.953 1091.07 141.847 0.8569 0.8510 1.6086 1.3534 255.98 108.56 254.43 346.61 1.1818 1.4764
2.5 322.494 1008.20 193.042 0.8818 0.8886 1.8938 1.7338 201.64 100.89 269.86 346.87 1.2289 1.4677
2.7 325.896 972.04 217.681 0.8932 0.9043 2.0855 1.9946 181.01 97.64 275.87 346.28 1.2468 1.4629
2.8 327.522 952.84 231.318 0.8994 0.9124 2.2147 2.1716 170.83 95.97 278.88 345.80 1.2557 1.4600
3.1 332.126 887.94 280.241 0.9215 0.9394 2.8972 3.1111 140.45 90.76 288.16 343.35 1.2829 1.4490
3.2 333.575 862.37 300.690 0.9306 0.9497 3.3470 3.7280 130.22 88.97 291.46 342.04 1.2924 1.4441
3.5 337.673 751.66 396.135 0.9721 0.9912 9.8985 12.2763 97.98 83.46 303.55 334.54 1.3273 1.4191
 Table D.5 Thermodynamic Properties of R-134a in Single Phase
p T ρ cv cp w h s
3
MPa K kg/m kJ/(kg·K) kJ/(kg·K) m/s kJ/kg kJ/(kg·K)
0.01 170 1589.10 0.7891 1.1871 1108.91 72.15 0.4158
0.01 200 1508.19 0.8003 1.2022 963.09 107.93 0.6096
0.01 250 0.49273 0.6635 0.7467 150.84 388.14 1.9545
0.01 300 0.40984 0.7538 0.8361 164.35 427.72 2.0985
0.01 350 0.35100 0.8398 0.9217 176.74 471.69 2.2339
1 200 1509.79 0.8008 1.2012 967.15 108.35 0.6084
1 250 1368.80 0.8497 1.2756 731.31 170.02 0.8833
1 350 39.9145 0.8823 1.0361 159.70 459.08 1.8320
1 400 33.0285 0.9389 1.0578 177.71 511.20 1.9711
5 250 1379.60 0.8508 1.2645 756.07 171.38 0.8771
5 300 1223.47 0.9144 1.3872 542.72 237.39 1.1175
5 350 1007.11 0.9874 1.6853 321.17 312.60 1.3488
5 400 285.605 1.0717 2.1540 124.34 456.91 1.7303
5 450 178.839 1.0509 1.3506 164.62 535.14 1.9154
10 150 1653.31 0.7935 1.1837 1227.28 52.99 0.2573
10 300 1246.96 0.9137 1.3544 586.89 238.19 1.1067
10 375 954.249 1.0153 1.6603 309.99 349.22 1.4358
10 450 475.033 1.0948 1.8283 166.89 491.31 1.7804
30 175 1610.08 0.7983 1.1756 1190.75 91.65 0.4202
30 300 1314.38 0.9148 1.2906 720.03 243.82 1.0735
30 375 1125.89 1.0077 1.3985 520.45 344.69 1.3731
30 450 928.050 1.0902 1.4814 386.34 452.98 1.6362
50 150 1688.36 0.7850 1.1821 1382.49 71.45 0.2209
50 225 1519.94 0.8365 1.1855 1050.62 159.60 0.6974
50 300 1361.43 0.9189 1.2633 817.65 251.24 1.0484
50 375 1206.00 1.0118 1.3476 643.46 349.22 1.3396
50 450 1056.37 1.0956 1.4130 523.21 452.89 1.5914
70 175 1647.40 0.8136 1.1707 1306.33 110.20 0.3859
70 225 1543.95 0.8421 1.1780 1114.65 168.69 0.6798
70 300 1398.49 0.9241 1.2485 896.91 259.49 1.0276
70 375 1261.28 1.0178 1.3256 736.58 356.08 1.3147
70 450 1133.80 1.1026 1.3878 624.28 457.94 1.5621
7.465E-01 170 1590 0.7890 1.1867 1112.04 72.48 0.4150
1.385E-04 170 0.1 0.5097 0.5913 126.74 335.00 2.0483
1.133E+00 200 1510 0.8009 1.2010 967.69 108.40 0.6082
1.627E-03 200 1 0.5684 0.6506 136.38 353.52 1.9479
3.190E+01 225 1495 0.8318 1.1960 984.52 151.59 0.7152
1.814E-02 225 5 0.6207 0.7073 143.03 369.69 1.8285
5.155E+00 250 1380 0.8508 1.2642 757.00 171.43 0.8769
4.012E-02 250 10 0.6693 0.7582 149.54 387.25 1.8386
5.175E+01 300 1365 0.9193 1.2617 825.20 251.94 1.0464
4.432E-01 300 20 0.7957 0.9312 152.58 419.46 1.7692
9.824E+00 400 800 1.0518 1.8965 225.46 393.80 1.5513
6.797E+00 400 600 1.0880 2.9066 140.61 413.60 1.6113

203
 Table D.6 Thermodynamic Properties of R-134a at Saturation
ps Ts ρ' ρ" cv' cv" cp' cp" w' w" h' h" s' s"
3 3
MPa K kg/m kg/m kJ/(kg·K) kJ/(kg·K) kJ/(kg·K) kJ/(kg·K) m/s m/s kJ/kg kJ/kg kJ/(kg·K) kJ/(kg·K)
3.998E-04 170 1589.09 0.028879 0.7891 0.5100 1.1871 0.5919 1108.87 126.71 72.14 334.97 0.4158 1.9618
6.332E-03 200 1508.19 0.390754 0.8003 0.5711 1.2022 0.6556 963.07 135.98 107.92 353.24 0.6096 1.8361
2.445E-02 220 1453.09 1.3853 0.8171 0.6153 1.2255 0.7048 865.85 141.06 132.19 365.89 0.7251 1.7874
7.246E-02 240 1395.96 3.83329 0.8379 0.6647 1.2584 0.7645 771.78 144.81 157.03 378.56 0.8330 1.7561
1.155E-01 250 1366.29 5.94794 0.8495 0.6917 1.2784 0.7996 725.62 146.06 169.73 384.81 0.8847 1.7451
2.607E-01 270 1303.93 12.8951 0.8745 0.7500 1.3267 0.8821 634.34 147.06 195.80 396.90 0.9846 1.7294
5.179E-01 290 1235.97 25.1729 0.9015 0.8130 1.3899 0.9864 543.54 145.67 222.98 408.13 1.0810 1.7194
7.027E-01 300 1199.07 34.1851 0.9158 0.8460 1.4301 1.0507 497.97 143.92 237.09 413.28 1.1283 1.7156
1.21680 320 1116.70 60.7354 0.9458 0.9159 1.5417 1.2243 405.62 137.90 266.64 422.17 1.2222 1.7082
1.97219 340 1015.73 105.795 0.9794 0.9945 1.7499 1.5452 309.52 127.60 298.75 428.03 1.3173 1.6976
3.04125 360 872.15 193.637 1.0279 1.0972 2.4258 2.6419 201.76 111.11 335.93 426.95 1.4203 1.6731

204
0.001 178.712 1565.67 0.068764 0.7895 0.5275 1.1890 0.6097 1068.48 129.56 82.49 340.15 0.4751 1.9169
0.01 206.263 1491.09 0.599771 0.8050 0.5845 1.2084 0.6701 932.27 137.69 115.47 357.18 0.6467 1.8186
0.1 246.799 1375.88 5.1901 0.8457 0.6829 1.2717 0.7879 740.34 145.71 165.64 382.82 0.8683 1.7483
0.25 268.883 1307.54 12.388 0.8730 0.7466 1.3236 0.8770 639.42 147.06 194.32 396.24 0.9791 1.7301
0.5 288.896 1239.91 24.3124 0.9000 0.8094 1.3859 0.9799 548.55 145.82 221.45 407.54 1.0757 1.7199
1 312.536 1149.04 49.228 0.9343 0.8891 1.4936 1.1500 440.34 140.58 255.37 419.11 1.1872 1.7111
1.5 328.373 1077.43 76.6117 0.9592 0.9473 1.6102 1.3295 366.07 134.19 279.69 425.11 1.2617 1.7045
2 340.616 1012.17 107.642 0.9805 0.9972 1.7596 1.5603 306.46 127.20 299.79 428.14 1.3203 1.6971
2.5 350.712 947.52 144.086 1.0014 1.0442 1.9803 1.9098 254.42 119.74 317.68 428.86 1.3706 1.6876
2.7 354.314 920.73 160.835 1.0104 1.0632 2.1069 2.1148 234.75 116.62 324.49 428.51 1.3893 1.6829
2.8 356.037 906.95 169.824 1.0152 1.0729 2.1837 2.2405 225.04 115.03 327.86 428.19 1.3985 1.6803
3 359.341 878.29 189.32 1.0256 1.0929 2.3771 2.5606 205.74 111.79 334.55 427.21 1.4166 1.6744
3.5 366.872 794.81 252.065 1.0620 1.1500 3.4340 4.3450 156.48 103.17 351.73 422.12 1.4622 1.6541
4 373.495 633.28 394.011 1.1742 1.2468 29.3945 42.9854 97.57 91.92 375.78 404.80 1.5252 1.6029
 Table D.7 Thermodynamic Properties of R-152a in Single Phase
p T ρ cv cp w h s
3
MPa K kg/m kJ/(kg·K) kJ/(kg·K) m/s kJ/kg kJ/(kg·K)
0.01 160 1181.75 1.0245 1.5048 1385.30 22.39 0.1657
0.01 200 1108.30 1.0256 1.5323 1144.46 83.07 0.5040
0.01 250 0.31912 0.7905 0.9200 190.56 495.62 2.4810
0.01 300 0.26534 0.9027 1.0300 207.16 544.32 2.6583
0.01 350 0.22723 1.0180 1.1446 222.31 598.69 2.8257
1 200 1109.42 1.0264 1.5308 1148.88 83.66 0.5025
1 300 896.488 1.1415 1.8014 640.11 246.99 1.1612
1 350 26.0044 1.1016 1.3431 200.86 578.68 2.2039
1 400 21.4875 1.1647 1.3475 222.71 645.45 2.3822
5 175 1159.01 1.0281 1.5082 1304.08 48.17 0.2942
5 300 910.768 1.1410 1.7599 681.44 247.75 1.1490
5 350 768.650 1.2362 2.1000 440.60 342.77 1.4413
5 400 207.339 1.4366 4.0341 147.42 552.64 1.9909
5 450 119.364 1.3409 1.8119 202.73 667.19 2.2625
15 170 1175.88 1.0367 1.4973 1365.95 47.02 0.2378
15 300 939.645 1.1452 1.6950 763.02 250.89 1.1235
15 375 766.401 1.2677 1.9787 490.25 387.65 1.5292
15 450 524.349 1.4012 2.3893 283.93 552.06 1.9277
30 160 1203.09 1.0488 1.4817 1466.62 41.84 0.1300
30 225 1097.87 1.0660 1.5214 1146.38 139.24 0.6405
30 300 972.841 1.1586 1.6423 852.50 257.40 1.0929
30 375 836.924 1.2729 1.8153 630.67 386.88 1.4775
30 450 688.681 1.3877 1.9818 470.97 529.58 1.8239
40 175 1186.07 1.0476 1.4835 1425.63 70.59 0.2518
40 225 1108.32 1.0759 1.5125 1181.55 145.37 0.6275
40 300 991.086 1.1694 1.6204 899.71 262.43 1.0758
40 375 868.789 1.2828 1.7686 695.01 389.43 1.4530
40 450 742.220 1.3960 1.9066 549.73 527.47 1.7882
50 160 1215.59 1.0540 1.4706 1530.71 54.97 0.1087
50 225 1118.05 1.0864 1.5055 1213.07 151.60 0.6153
50 300 1007.28 1.1807 1.6042 940.57 267.83 1.0604
50 375 894.918 1.2938 1.7386 747.79 393.12 1.4326
50 450 782.051 1.4065 1.8640 612.07 528.38 1.7611
4.291E+00 160 1185 1.0325 1.5011 1391.25 25.14 0.1602
2.014E-05 160 0.001 0.5841 0.7101 156.46 423.38 2.9065
1.510E+00 200 1110 1.0268 1.5301 1151.16 83.97 0.5018
2.510E-03 200 0.1 0.6785 0.8078 172.59 453.12 2.4652
1.154E+01 250 1030 1.0754 1.5867 951.35 167.40 0.8321
3.105E-02 250 1 0.8038 0.9413 189.41 494.38 2.3346
3.377E+01 300 980 1.1626 1.6332 871.18 259.24 1.0862
1.816E-01 300 5 0.9365 1.0921 201.73 539.15 2.2806
7.328E+01 350 975 1.2815 1.6574 896.19 362.22 1.2788
7.941E-01 350 20 1.0815 1.2898 205.74 583.28 2.2428
6.507E+00 400 500 1.4090 4.3097 183.41 471.80 1.7776
3.272E+01 400 800 1.3141 1.8565 590.89 433.14 1.5883

205
 Table D.8 Thermodynamic Properties of R-152a at Saturation
ps Ts ρ' ρ" cv' cv" cp' cp" w' w" h' h" s' s"
3 3
MPa K kg/m kg/m kJ/(kg·K) kJ/(kg·K) kJ/(kg·K) kJ/(kg·K) m/s m/s kJ/kg kJ/kg kJ/(kg·K) kJ/(kg·K)
1.299E-04 160 1181.75 0.00645 1.0245 0.5855 1.5048 0.7121 1385.29 156.42 22.39 423.32 0.1657 2.6715
1.141E-03 180 1145.47 0.05046 1.0225 0.6341 1.5161 0.7629 1258.17 164.70 52.60 437.70 0.3436 2.4830
6.084E-03 200 1108.3 0.24350 1.0256 0.6887 1.5323 0.8229 1144.44 172.13 83.07 452.53 0.5041 2.3514
6.648E-02 240 1030.28 2.2805 1.0563 0.8191 1.5971 0.9800 932.39 183.31 145.53 482.43 0.7883 2.1921
1.053E-01 250 1009.67 3.5124 1.0678 0.8555 1.6211 1.0279 881.91 185.07 161.64 489.75 0.8539 2.1664
1.603E-01 260 988.452 5.2175 1.0804 0.8930 1.6486 1.0797 832.13 186.31 178.01 496.93 0.9179 2.1445
2.355E-01 270 966.526 7.5159 1.0941 0.9315 1.6802 1.1360 782.87 186.98 194.68 503.92 0.9805 2.1259
3.354E-01 280 943.761 10.5488 1.1089 0.9707 1.7162 1.1975 733.90 187.02 211.69 510.68 1.0420 2.1098
6.296E-01 300 895.05 19.5363 1.1416 1.0517 1.8061 1.3429 635.93 184.96 246.94 523.19 1.1625 2.0833
2.19001 350 739.063 75.1484 1.2496 1.2794 2.3058 2.1218 376.87 163.62 346.01 542.67 1.4612 2.0231
3.99866 380 560.608 191.097 1.3778 1.4805 5.4546 7.2007 183.40 132.56 426.18 525.75 1.6727 1.9348
0.001 178.623 1148 0.04458 1.0226 0.6305 1.5153 0.7592 1266.28 164.16 50.51 436.69 0.3319 2.4939
0.01 206.996 1095.05 0.38782 1.0289 0.7095 1.5402 0.8466 1106.00 174.48 93.82 457.77 0.5569 2.3151

206
0.1 248.82 1012.13 3.3446 1.0663 0.8512 1.6181 1.0221 887.83 184.89 159.73 488.90 0.8463 2.1692
0.25 271.636 962.864 7.9576 1.0965 0.9378 1.6857 1.1457 774.84 187.03 197.44 505.05 0.9906 2.1231
0.5 292.328 914.306 15.5539 1.1286 1.0202 1.7683 1.2828 673.60 186.11 233.21 518.57 1.1166 2.0928
1 316.778 849.771 31.2786 1.1727 1.1226 1.9088 1.5030 552.45 180.69 278.06 532.21 1.2620 2.0643
1.5 333.155 799.838 48.3301 1.2073 1.1961 2.0537 1.7266 468.29 173.89 310.24 539.01 1.3592 2.0458
2 345.817 755.354 67.2945 1.2382 1.2577 2.2269 1.9969 400.17 166.52 336.81 542.19 1.4355 2.0294
2.5 356.266 712.567 88.9241 1.2681 1.3137 2.4598 2.3668 340.84 158.78 360.31 542.68 1.5005 2.0124
2.7 359.998 695.322 98.5698 1.2802 1.3353 2.5821 2.5633 318.68 155.58 369.18 542.19 1.5245 2.0051
2.8 361.784 686.586 103.656 1.2863 1.3460 2.6526 2.6769 307.86 153.96 373.54 541.79 1.5362 2.0012
3.5 373.051 620.528 146.307 1.3333 1.4219 3.4990 4.0480 234.90 142.06 403.53 535.66 1.6148 1.9690
4 380.018 560.166 191.250 1.3781 1.4807 5.4857 7.2174 183.01 132.54 426.30 525.72 1.6730 1.9347
4.5 386.226 416.03 319.177 1.4949 1.5447 189.01 241.77 122.92 119.54 466.16 490.51 1.7743 1.8374
 Appendix E
Pressure-Enthalpy Diagram

The pressure-enthalpy diagram plays important role to draw the thermodynamic

process besides it can also be used to predict thermodynamic properties with worse

accuracy. In principle, all thermodynamic properties can be calculated from the

fundamental equations of state. However, uncertainties of the thermodynamic properties

relate to the uncertainty of the fundamental equations of state and their validity range.

The pressure-enthalpy diagrams for R-32, R-125, R-134a, and R-152a are

enclosed in the following pages. These diagrams are obtained from the fundamental

equations of state and using implementation of the calculation strategies that were

described in the Section 2.4. Especially for two-phase, quantity of vapor is introduced to

represent mass ratio of vapor and mixture of the liquid and vapor in the two-phase. This

property is usually used in thermodynamic analyses.

207
 100
5
.1 0 60 1. 0 1. 2 .3 .4 1.5 1.6 R-32

5
0.9 1 s =1

75
.3 0.

1
0 s=

0.4
s= s= s=

1.
s= 0 s= 1.7

0.
s= s= s=

s=
s=
s=
10 25 . 0 015 s=
11 0 1 3 0

8
0 0 .0 3 80 v=
0.0

1.
0.0 v=0 360 70 K K 02

s=
9
v= v= v=0.0

9
0

1.
310 K
35 .0

.00
K 3
v=0.00

s=
330
10 0 s=
2
2.1

0
34 K 5
v=0.00

300 K
s=

v=

320 K
85
0 2.2

K
290 K
K s=
v=0.01

v=0.000

280 K
.3

v=0.00080
s =2

270 K
v=0.02
v=0.0007 m3/kg

.4
s=2

250 K
v=0.00075
.5
1 v=0.05 s=2

230 K
220 K
.6
s =2
v=0.15 .7
s =2
200 K

v=0.30 .8
s=2
190 K

.9
0.1 s =2
170 K

v=0.85 .0
s =3
160 K
p , MPa
208

.1
s=3
150 K

v=2
.2
s= 3
140 K

v=5 .3
0.01 s= 3
.4

.7
.1

.3

.4

.5

.8
.6
0.2

s= 3

x=0.9
v=12

x=0
x=0

x=0

x=0

x=0

x=0
x=0
x=

.5
s =3
v=30
.6
s =3
0.001 .7
s =3

T=450 K
v=100
.8

400 K
s =3

350 K
300 K
250 K
.9

200 K
v=300 s= 3
.0 K)
s= 4 kg·
0.0001 . 1 kJ/(
s =4

0.00001
-50 50 150 250 350 450 550 650 750
h , kJ/kg
Fig. E.1 Pressure-enthalpy diagram for R-32 calculated from the new fundamental equation of state
 100
1 .2
1.3
R-125
1.4

0
5 9 s=

0.6
1.0 s=

1
0.7 0. s= 1.5

1.
s= s= s=

s=

s=
s=
5 7 5 009 1.6
0008 v=0.0

/kg
06 00 07 5
.000
8 1
v=0.00 =0.0011 s=
0.0
0
0.0 .00 v=0 v=0.

m3
v
v= v= v=0 4 125
38 390 00 K v=0.00 015
06
0 0 K 0 .0 7
37 v= 1.
0.0
K

310 K
10 36 0
v=0.002 s=
0K K
v=

35 3
v=0.00

290 K
0K
34 v=0.005

270 K
0 1.8
K s=

260 K
v=0.01

240 K v=0.05 1.9

s=
1
230 K

v=0.1
210 K
p , MPa

2.0
s=
200 K
209

v=0.15
190 K

0.1 v=0.3
2 .1
s=
v=0.8
.1

.9
.8
.6

.7
.4
0. 2

.3

.5
x= 0

x=0
x=0
x= 0

x=0
x=0
x= 0

x= 0
x=

.2
v=2 s =2

T=450 K
400 K
0.01

350 K
)

300 K
g· K

250 K
v=5 k
200 K k J/(
2.3
s=

0.001
75 125 175 225 275 325 375 425 475 525

h , kJ/kg
Fig. E.2 Pressure-enthalpy diagram for R-125 calculated from the new fundamental equation of state
 100
1.1 1.2 1.3 .4 1 .5
75 0.9 .0 s= s=1 s=

5
0. s= 1 s= s= 1.6 R-134a

.4
s= s= 008 .0009 s=

0
0085

s=
0 .0 =0.0 v =0

.30
v= v 01
v=0.0

/k g
0

5
011 7
1.

07
s=
v .0

m3
=0 s=

00
00 125 5

0.
.

70
1

38
0 0

07

360

37
v= v=0.0

310 K
v=

350 K
10

000

300 K

8
.00

0
2
v=0.00

1.
K
290 K

K
K

s=
3
v=0.00
v =0

.
v=0

280 K
.9
v=0.005 s=
1

270 K
v=0.01

260 K
.0
s =2
v=0.02

240 K
1 2.1
s=
v=0.05
220 K
210 K

.2
v=0.15 s =2
190 K
210

p , MPa

180 K

v=0.3
0.1 .3
s =2
v=0.8
2 .4
s=
v=2
.4
0.1

0. 5

.9
.7

.8
.6
0.2

.3

x=0

x=0
x=0

x= 0
x=0
0
x=

x=
0.01
x=

x=

v=5 .5
s =2

T=460 K
v=12

350 K

400 K
300 K
2.6

250 K
200 K
s=
)
v=30 ·K
0.001 ( kg
k J/
2.7
s=

0.0001
50 150 250 350 450 550

h , kJ/kg
Fig. E.3 Pressure-enthalpy diagram for R-134a calculated from the new fundamental equation of state
 100
.0 1.1 1. 2 1.3 .5 1.6 R-152a

30
15
0 75 1 .4 1 1.7

45
0.6 0. s= s= s =1

0.
0.
s= s= s= s= s= .8

0.
s=
s=
s= s= =1

s=
s
25 015
11 .0 0 1 v=0.0
v=0

9
0

330 K

1.
0.0

320 K
40 2
0 K v=0.00

s=
10 v= 39 2. 0
s= 1

37
0

90
2.

360
310 K
01
380 K 5 s=

0
v=0.0

v=0.000

300 K
K

v=0.0

K
K
.2
g

s=2

290 K
v=0.00085 m3/k

280 K
.3
v=0.02 s =2

270 K
.4
s= 2
1 v=0.05

250 K
.5
s=2
.6

230 K
220 K v=0.15 s=2
.7
s =2
210 K

v=0.30
0.1 s =2
.8
190 K

v=0.85
p , MPa

180 K

.9
s=2
211

v=2
170 K

.0
s=3
160 K

0.01 v=5
.1
s=3
v=12

.6
.5

.7

x=0.9
.1

.4
.3

.8
0.2

.2

x=0
x= 0

x=0
x=0

s= 3
x= 0
x= 0

x=0
x=

v=30
.3
s= 3
0.001 .4
v=100 s=3

T=450 K
.5

400 K
s= 3

350 K
300 K
250 K
200 K
v=300 .6 )
s= 3 g·K
0.0001 J/(k
.7 k
s=3

0.00001
0 100 200 300 400 500 600 700

h , kJ/kg
Fig. E.4 Pressure-enthalpy diagram for R-152a calculated from the new fundamental equation of state