Pressure drop, heat transfer and flow pattern maps inside smooth and enhanced tubes during convective boiling and condensation of r134a and r1234ze

POLITECNICO DI MILANO DEPARTMENT OF ENERGY DOCTORAL PROGRAM IN ENERGY AND NUCLEAR SCIENCE AND TECHNOLOGY Pressure drop, heat transfer and flow pattern maps inside smooth and enhanced t

POLITECNICO DI MILANO

DEPARTMENT OF ENERGY DOCTORAL PROGRAM IN ENERGY AND NUCLEAR SCIENCE

AND TECHNOLOGY

Pressure drop, heat transfer and flow pattern maps inside smooth and enhanced tubes during convective boiling and condensation

of R134a and R1234ze

Tutor: Prof LUIGI PIETRO MARIA COLOMBO

Advisor: Prof ANDREA LUCCHINI

Doctoral dissertation of:

PHAN THANH NHAN

2018 CYCLE XXX

Acknowledgement

I would like to express my gratitude to my Tutor Prof Luigi Pietro Maria Colombo, my advisor Prof Andrea Lucchini, and consultant of the project prof Muzzio They gave me the tasks, teach, support, encourage and also work with me for four years in Energy department, Politecnico di Milano, Italy.

I would like to present my thankful to Vietnam government scholarship (VIED) and thank to the project of Italy – Program PRIN 2015 (Grant number 2015M8S2PA), provide the funds to do this research

I would like to say thank you to Samanta (my PhD friend) and my friends who always stay, listen and give some advises to me without any conditions

And two maid’s ladies in energy department, even though the barrier language, they so nice to just say hello to me

Besides, I would like to say thank you to my University in Vietnam, Ho Chi Minh City University of Technology, the place I work as lecturer for 10 years

Finally, most importance I would like to thank to all my Family and my Fiancee, especially my father and my mother, they always stay behind, say nothing and keep watching on my step

Milan, Italy, October 31 th , 2018

PHAN THANH NHAN

Declaration

- What written in this thesis is original, only the chapter about theory (chapter 2) is getting from available text book Every information presented in here have been made on the PhD time of candidate

- With all experiments have been done for 4 years PhD of candidate in Energy department, Politecnico di Milano

Contents

Chapter 1 INTRODUCTION 1

1.1 Introduction 1

1.1.1 Motivation of the research 1

1.1.2 Originality; 2

1.2 State of the art of the specific research field 2

1.3 Objects: 4

1.3.1 Micro-fin tube: 4

1.3.2 New refrigerants: 4

Chapter 2 THEORY OF TWO-PHASE FLOW 8

2.1 Basic equations for two phase flow 8

2.2 Parameters of two-phase pipe flows 9

2.2.1 Phase density function 9

2.2.2 Instantaneous space-averaging operators 9

2.2.3 Local time-averaging operators 10

2.2.4 Commutativity of average operators 10

2.2.5 Qualities 11

2.2.6 Volumetric quality 11

2.3 Composite averaged equations 11

2.4 Friction pressure drop 12

2.5 Dimensionless numbers 13

2.6 Void fraction: 15

Chapter 3 CORRELATIONS FOR MICROFIN TUBES 17

3.1 Flow pattern: 17

3.1.1 Boiling model of Wojtan et al 2005 [11] 18

3.1.2 Boiling model of Rollmann and Spindler 2015 [45] 20

3.1.3 Boiling model of Zhuang et al 2016 [46] 22

3.1.4 Boiling model of Yang et al 2018 [47] 23

3.1.5 Condensation model of Breber et al 1980 [41] 24

3.1.6 Condensation model of Tandon et al 1982 [42] 24

3.1.7 Condensation: Hajal et al 2003 [9] 25

3.2 Correlations for heat transfer coefficients 28

3.2.1 Correlation for boiling model of Han et al 2017 [87] 28

3.2.2 Correlation for boiling model of Rollmann and Spindler et al 2016 [78] 28

3.2.3 Correlation for boiling model of Chamra and Mago 2006 [77] 29

3.2.4 Correlation for boiling model of Yun 2002 [76] 30

3.2.5 correlation for boiling model of Cavallini et al 1999 [17] 31

3.2.6 Correlation for boiling model of Thome et al 1997 [1] 32

3.2.7 Correlation for condensation model of Cavallini et al 2009 [30] 33

3.2.8 Correlation for condensation model of Han et al 2005 [81] 34

3.2.9 Correlation for condensation model of Kedzieski and Goncalves 1999 [80] 35

3.2.10 Correlation for condensation model of Yu and Koyama 1998 [79] 35

3.3 Correlation for pressure drop 37

3.3.1 Correlation for two phase flow model of Choi et al 2001[82] 37

3.3.2 Correlation for two phase flow model of Goto et al 2001[83] 37

3.3.3 Correlation for boiling model of Wongsa et al 2004[88] 38

3.3.4 Correlation for boiling model of Kou and Wang 1996 [89] 38

3.3.5 Correlation for condensation model of Han et al 2005[81] 38

3.3.6 Correlation for condensation model of Kedzieski and Goncalves 1999[80] 39

3.3.7 Correlation for condensation model of Harraguchi 1993[85] 39

PART A EXPERIMENTAL APPARATUS 40

Chapter 4 DESCRIPTION OF EXPERIMENTAL APPARATUS 41

4.1 Experimental apparatus 41

4.2 Refrigerant loop 42

4.2.1 Purpose 43

4.2.2 Component 43

4.3 Demineralized water loop 46

4.3.1 Purpose 46

4.3.2 Components 47

4.3.3 Operating conditions 48

4.4 Glycol and water loop 51

4.4.1 Purpose 51

4.4.2 Components 51

4.5 Test section 60

4.5.1 Purpose 60

4.5.2 Components 60

4.6 Setting up the measuring instruments 62

4.6.1 Measuring temperatures: 62

4.6.2 Measuring pressure 64

4.6.3 Visualization apparatus 67

Chapter 5 DATA REDUCTION AND UNCERTAINTY ANALYSIS 68

5.1 Data reduction 68

5.1.1 Wall temperatures 70

5.1.2 Refrigerant 70

5.1.3 Demineralized water 71

5.1.4 Heat transfer coefficient 72

5.1.5 Quality 72

5.1.6 Operating conditions 73

5.1.7 Experimental results 74

5.2 Uncertainty analysis: 74

5.2.1 Uncertainty of pressure and pressure drop of refrigerant 75

5.2.2 Uncertainty of temperatures 75

5.2.3 Uncertainty of thermophysical properties: 76

5.2.4 Uncertainty of thermal power supply: 78

5.2.5 Uncertainty of quality: 78

5.2.6 Evaluate the uncertainty of heat transfer coefficient: 79

5.2.7 Choosing the working condition from analyzing uncertainty of heat transfer coefficient 80

5.2.8 Determine the measurement uncertainty by repeated experiments: 82

Chapter 6 PRE-TEST 85

6.1 Causes and solution 85

6.1.1 Problem 1 - thermocouples 85

6.1.2 Problem 2: un-equilibrium between temperature measure directly by thermocouple and temperature as saturation pressure at the inlet of test section of refrigerant 86

6.1.3 Problem 3: the offset of pressure drop 89

6.1.4 Problem 4: the stability of refrigerant mass flow rate 90

6.1.5 Problem 5: unsteady of electric net 90

6.2 Single phase test 91

6.2.1 Single phase test method 91

6.2.2 Single phase of refrigerant R134a 92

6.2.3 Single phase of refrigerant R1234ze: 92

PART B EXPERIMENTAL RESULTS 95

Chapter 7 EXPERIMENTAL RESULTS FOR BOILING 101

7.1 Boiling in smooth tube of R134a 101

7.1.1 Effect of mass flux 101

7.1.2 Effect of heat flux 104

7.1.3 Comparison between experimental data and correlations 105

7.2 Boiling in microfin tube J60 of R134a 106

7.2.1 Flow pattern for R134a in microfin tube J60 107

7.2.2 Effect of mass flux 110

7.2.3 Effect of mean quality 111

7.2.4 Effect of heat flux 113

7.3 Boiling in microfin tube J60 of R1234ze 115

7.3.1 Flow pattern of boiling R1234ze in microfin tube J60 116

7.3.2 Effect of mass flux 121

7.3.3 Effect of mean quality 122

7.3.4 Effect of heat flux 124

7.4 Comparison between smooth tube and microfin tube J60 – R134a 125

7.4.1 Comparison between smooth tube and microfin tube J60 – R134a 125

7.4.2 Dry out in smooth tube and microfin tube J60 127

7.5 Comparison between R134A and R1234ze boiling in microfin tube J60 131

7.5.1 Effect of mass flux 132

7.5.2 Effect of mean quality 133

7.6 Comparison between experimental data and correlations 136

7.6.1 Heat transfer coefficient 136

7.6.2 Pressure drop 137

Chapter 8 EXPERIMENTAL RESULTS FOR CONDENSATION 140

8.1 Condensation in smooth tube of R134a 140

8.2 Condensation in microfin tube J60 of R134a 144

8.2.1 Flow pattern 144

8.2.2 Effect of mass flux 149

8.2.3 Effect of mean quality 149

8.3 Condensation in microfin tube J60 of R1234ze 152

8.3.1 Flow pattern 153

8.3.2 Effect of mass flux 158

8.3.3 Effects of mean quality 158

8.4 Comparison between smooth tube and microfin tube J60 – R134a 162

8.5 Comparison between R134A and R1234ze condensation in microfin tubes 163

8.5.1 Effect of mass flux 164

8.5.2 Effect of mean quality 165

8.6 Comparison between experimental and correlations 167

8.6.1 Heat transfer coefficient for microfin tube J60 167

8.6.2 Pressure drop for condensation for microfin tube J60 169

Chapter 9 CONCLUSIONS 171

9.1 Conclusions 171

9.2 Future works 171

List of figures

Figure 1.1 microfin tube 4

Figure 1.2 comparison the flow pattern map for R134a, R1234yf and R1234ze 5

Figure 1.3 Gungor and Winterton 1986 [48] 6

Figure 1.4 Liu and Winterton 1991 [49] 6

Figure 1.5 Wojtan et al 2005 [12] 6

Figure 1.6 Saitoh et al 2007 [52] 6

Figure 1.7 Gronnerud 1972 7

Figure 1.8 Fridel 1979 [57] 7

Figure 1.9 Muller and Heck 1986 [59] 7

Figure 1.10 Quiben et al 2007 [14] 7

Figure 2.1 phase in a circular tube 9

Figure 2.2 void fractions from various correlations 16

Figure 3.1 Flow regime for boiling from Wojtan et al 2005 18

Figure 3.2 Flow regime for condensation from Hajal et al 18

Figure 3.3 flow pattern map of Wojtan et al 2005 18

Figure 3.4 geometrical parameters of stratified flow in circular tube 19

Figure 3.5 flow pattern map of Rollmann and Spindle 2015 20

Figure 3.6 flow pattern map of Zhuang et al 2016 22

Figure 3.7 flow pattern map of Yang et al 2018 23

Figure 3.8 flow pattern map of Breber et al 1980 24

Figure 3.9 flow pattern map of Tandon et al 1982 25

Figure 3.10 flow pattern map of Hajal et al 2003 26

Figure 4.1 schematic diagram of experiment apparatus 42

Figure 4.2 refrigerant loop 42

Figure 4.3 condensation system on refrigerant loop 43

Figure 4.4 evaporator: electric heaters connection 45

Figure 4.5 evaporation system on refrigerant loop 45

Figure 4.6 the total power of evaporation system on difference connecting method 46

Figure 4.7 demineralized water loop 46

Figure 4.8 plate heat exchangers between Demineralized water and water glycol 48

Figure 4.9 Proportional Kp vs mw 50

Figure 4.10 Integral time Ti vs mw 50

Figure 4.11 Derivative time Td vs mw 50

Figure 4.12 glycol water loop 51

Figure 4.13 rotameters 52

Figure 4.14 the general working map of condensing unit for boiling test at 5°C 55

Figure 4.15 working map for boiling - Model S09-04 56

Figure 4.16 working map for boiling - Model S22-06 56

Figure 4.17 working map for boiling - Model S22-10 56

Figure 4.18 working map for boiling - Model S22-18 56

Figure 4.19 the general working map of condensing unit for condensation test at 35°C 57

Figure 4.20 working map for condensation test - Model S09-04 58

Figure 4.21 working map for condensation test - Model S22-06 58

Figure 4.22 working map for condensation test - Model S22-10 58

Figure 4.23 working map for condensation test - Model S22-10 58

Figure 4.24 Proportional Kp vs V_g-c 59

Figure 4.25 integral time Ti vs V_g-c 59

Figure 4.26 derivative time Td vs V_g-c 60

Figure 4.27 simply scheme of every test section 60

Figure 4.28 structure of micro-fin tube 62

Figure 4.29 the connection way to insert thermocouple on the pipe 63

Figure 4.30 the connection way to insert thermocouples on water side 63

Figure 4.31 the connection method to set thermocouple on the wall 64

Figure 4.32 the cold junction reference 64

Figure 4.33 calibration for pressure transducer 65

Figure 4.34 calibration for differential pressure transducer 66

Figure 4.35 the pressure connection method 66

Figure 4.36 visualization apparatus 67

Figure 5.1 Diagram of test section - refrigerant and water 72

Figure 5.2 derivative of cd(T) on [20÷100]( o C) 77

Figure 5.3 relative uncertainty of heat transfer coefficient 81

Figure 6.1 pre-test – heat transfer coefficient for boiling 85

Figure 6.2 repeated test at the same condition in difference days 86

Figure 6.3 supporter for thermocouples 86

Figure 6.4 diagram of measuring Thermocouple and pressure transducer 87

Figure 6.5 check the equilibrium on the condensation test 35°C 88

Figure 6.6 check the equilibrium on the boiling test 5°C 88

Figure 6.7 diagram of add one plate heat exchanger as a calm down part 89

Figure 6.8 comparison the equilibrium before and after solving problem 89

Figure 6.9 single phase - test section 91

Figure 6.10 energy balance between liquid R134a and demineralized water 92

Figure 6.11 energy balance between subcooling liquid R1234ze and demineralized water 93

Figure 6.12 heat transfer coefficient of subcooling liquid R1234ze 94

Figure 6.13 Nusselt number vs Reynolds number of subcooling liquid R1234ze 94

Figure 6.14 pressure drop of subcooling liquid R1234ze 94

Figure 7.1 flow pattern – heat transfer coefficient for boiling at q = 9[kW/m 2 ] 102

Figure 7.2 flow pattern map for boiling from Wojtan et al 2005 102

Figure 7.3 the effecting of G[kg/m 2 s] on heat transfer coefficient at q = 9[kW/m 2 ] 103

Figure 7.4 the effecting of G[kg/m 2 s] on pressure drop at q = 9[kW/m 2 ] 103

Figure 7.5 the effecting of G[kg/m 2 s] on heat transfer coefficient at q = 18[kW/m 2 ] 103

Figure 7.6 the effecting of G[kg/m 2 s] on pressure drop at q = 18[kW/m 2 ] 104

Figure 7.7 flow pattern map for boiling with q = 9[kW/m 2 ] and q = 18[kW/m 2 ] 104

Figure 7.8 the effecting of q [kW/m 2 ] on heat transfer coefficient 105

Figure 7.9 the effecting of q [kW/m 2 ] on pressure drop 105

Figure 7.10 comparison of HTC - experimental with correlations 106

Figure 7.11 comparison of pressure drop of experimental -correlations 106

Figure 7.12 flow pattern for boiling R134a at G = 67 kg/m 2 s and G = 80 kg/m 2 s 108

Figure 7.13 Flow pattern map of boiling R134a on microfin tube J60 109

Figure 7.14 Flow pattern of boiling R134a on microfin tube J60 –Model Wojtan et al 2005 109

Figure 7.15 Flow pattern of boiling R134a - microfin tube J60 – Model Rollmann and Spindler 2015 110 Figure 7.16 Flow pattern of boiling R134a on microfin tube J60 – Model Zhuang et al 2016 110

Figure 7.17 HTC of boiling R134a on J60 – Δx = 0.6 – xm =0.5 111

Figure 7.18 ΔP/L of boiling R134A on J60 –Δx = 0.6 – xm =0.5 111

Figure 7.19 Flow pattern of boiling R134a on microfin tube J60 at the same Δx = 0.2 112

Figure 7.20 HTC of boiling R134a on J60 – Δx = 0.2 112

Figure 7.21 ΔP/L of boiling R134a on J60 – Δx = 0.2 113

Figure 7.22 HTC of boiling R134a on J60 – various G at q = 9 kW/m 2 114

Figure 7.23 ΔP/L of boiling R134a on – various G at q = 9 kW/m 2 114

Figure 7.24 HTC of boiling R134a on J60 – various G at q = 18 kW/m 2 114

Figure 7.25 ΔP/L of boiling R134a on – various G at q = 18 kW/m 2 114

Figure 7.26 HTC of boiling R134a on J60 – q = 9 kW/m 2 and q = 18 kW/m 2 114

Figure 7.27 ΔP/L of boiling R134a on J60 –q = 9 kW/m 2 andq = 18 kW/m 2 115

Figure 7.28 flow pattern of boiling R1234ze on microfin tube J60 – Δx = 0.2 117

Figure 7.29 flow pattern of boiling R1234ze on microfin tube J60 – various Δx 118

Figure 7.30 Flow pattern map of boiling R1234ze on microfin tube J60 119

Figure 7.31 Flow pattern of boiling R1234ze on microfin tube J60 – Model Wojtan et al 2005 120

Figure 7.32 Flow pattern of boiling R1234ze-microfin tube J60–Rollmann and Spindler 2015 120

Figure 7.33 Flow pattern of boiling R1234ze on microfin tube J60 – Model Zhuang et al 2016 121

Figure 7.34 HTC of boiling R1234ze on J60 – Δx = 0.6 – xm =0.5 121

Figure 7.35 ΔP/L of boiling R1234ze on J60 – Δx = 0.6 – xm =0.5 122

Figure 7.36 flow pattern of all experimental results for boiling R1234ze 123

Figure 7.37 HTC of boiling R1234ze on J60 – Δx = 0.2 123

Figure 7.38 ΔP/L of boiling R1234ze on J60 – Δx = 0.2 123

Figure 7.39 HTC of boiling R1234ze on J60 – q = 8.3kW/m 2 – G = 110, 163, 222kg/m 2 s 124

Figure 7.40 ΔP/L of boiling R1234ze onJ60 – q=8.3kW/m 2 – G = 110, 163, 222kg/m 2 s 124

Figure 7.41 HTC of boiling R1234ze on J60 – G= 222kg/m 2 s q = 8.3; 16.5; 18kW/m 2 124

Figure 7.42 ΔP/L of boiling R1234ze onJ60 – G= 222kg/m 2 s, q = 8.3; 16.5; 18kW/m 2 124

Figure 7.43 HTC of boiling R1234ze on J60 – G=163kg/m 2 s, q = 8.3; 12.4kW/m 2 124

Figure 7.44 ΔP/L of boiling R1234ze onJ60 – G=163kg/m 2 s q = 8.3; 12.4kW/m 2 124

Figure 7.45 comparison of HTC – R134a –q = 9kW/m 2 – smooth tube vs microfin tube 126

Figure 7.46 comparison of ΔP/L – R134a –q = 9kW/m 2 – smooth tube vs microfin tube 126

Figure 7.47 comparison of HTC – R134a –q = 18kW/m 2 – smooth tube vs microfin tube 126

Figure 7.48 comparison of ΔP/L – R134a –q = 9kW/m 2 – smooth tube vs microfin tube 127

Figure 7.49 changing pattern of flow in smooth tube xm = 0.15 to xm = 0.90 128

Figure 7.50 distribution of T_wallat G = 333kg/m 2 s - q = 17kW/m 2 - xi = 0.77 - xo = 0.92 128

Figure 7.51 distribution of T_wall at G = 333kg/m 2 s - q = 17kW/m 2 - xi = 0.82 - xo = 0.95 129

Figure 7.52 distribution of T_wall at G = 333kg/m 2 s - q = 17kW/m 2 - xi = 0.85 - xo = 0.98 129

Figure 7.53 HTC – detailed data results at G = 333kg/m 2 s - q = 18kW/m 2 129

Figure 7.54 pressure drop – detailed data results at G = 333kg/m 2 s - q = 18kW/m 2 130

Figure 7.55 HTC – smooth – microfin– G=111kg/m 2 s, q = 9kW/m 2 130

Figure 7.56 Δp/L –smooth – microfin– G=111kg/m 2 s, q = 9kW/m 2 130

Figure 7.57 HTC – smooth – microfin– G=222kg/m 2 s, q = 9kW/m 2 130

Figure 7.58 Δp/L –smooth – microfin– G=222kg/m 2 s, q = 9kW/m 2 130

Figure 7.59 HTC – smooth – microfin - G=333kg/m 2 s, q = 9kW/m 2 131

Figure 7.60 Δp/L –smooth – microfin– G=333kg/m 2 s, q = 9kW/m 2 131

Figure 7.61 HTC – smooth – microfin– G=222kg/m 2 s, q = 18kW/m 2 131

Figure 7.62 Δp/L –smooth – microfin– G=222kg/m 2 s, q = 18kW/m 2 131

Figure 7.63 HTC – smooth – microfin– G=333kg/m 2 s, q = 18kW/m 2 131

Figure 7.64 Δp/L –smooth – microfin– G=333kg/m 2 s, q = 18kW/m 2 131

Figure 7.65 comparison HTC of boiling R1234ze – R134a – Δx = 0.6 – xm =0.5 132

Figure 7.66 comparison Δp/L of boiling R1234ze – R134a – Δx = 0.6 – xm =0.5 133

Figure 7.67 comparison HTC of boiling R1234ze – R134a – G = 80, 111, 221kg/m 2 s – Δx = 0.2 134

Figure 7.68 comparison Δp/L of boiling R1234ze – R134a – G = 80, 111, 221kg/m 2 s – Δx = 0.2 134

Figure 7.69 comparison HTC of boiling R1234ze – R134a – G = 221kg/m 2 s – Δx = 0.1; Δx = 0.2 134

Figure 7.70 comparison Δp/L of boiling R1234ze – R134a – G = 221kg/m 2 s – Δx = 0.1; Δx = 0.2 135

Figure 7.71 comparison HTC of boiling R1234ze – R134a – G at q = 9; 18kW/m 2 135

Figure 7.72 comparison Δp/L of boiling R1234ze – R134a –G at q = 9; 18kW/m 2 135

Figure 7.73 comparison experimental data and correlation prediction for heat transfer coefficient of boiling R134a on microfin tube J60 136

Figure 7.74 comparison experimental data and correlation prediction for heat transfer coefficient of boiling R1234ze on microfin tube J60 137

Figure 7.75 comparison experimental data and correlation prediction for pressure drop of boiling R134a on microfin tube J60 138

Figure 7.76 comparison experimental data and correlation prediction for pressure drop of boiling R1234ze on microfin tube J60 138

Figure 8.1 flow pattern map for condensation from Hajal et al 2003 140

Figure 8.2 flow pattern map for condensation from Breber et al 1980 141

Figure 8.3 the effect of mass flux on heat transfer coefficient at q = 7.7kW/m 2 141

Figure 8.4 the effect of mass flux on pressure drop at q = 7.7kW/m 2 141

Figure 8.5 the effect of heat flux on heat transfer coefficient 142

Figure 8.6 the effect of heat flux on pressure drop 142

Figure 8.7 comparison of HTC - experimental with correlations 143

Figure 8.8 comparison of pressure drop of experimental -correlations 143

Figure 8.9 Flow pattern of condensation R134a on microfin tube J60; Δx = – 0.2 145

Figure 8.10 Flow pattern of condensation R134a on microfin tube J60 – various Δx 146

Figure 8.11 Flow pattern map of condensation R134a on microfin tube J60 147

Figure 8.12 Flow pattern of condensation R134a on microfin tube J60 – Model Hajal et al 2003 147

Figure 8.13 Flow pattern of condensation R134a on microfin tube J60 – Model Tandon et al 1982 148

Figure 8.14 Flow pattern of condensation R134a on microfin tube J60 – Model Breber et al 1980 148

Figure 8.15 HTC of condensation R134a on J60; Δx = –0.6; xm =0.5 149

Figure 8.16 Δp/L of condensation R134a on J60; Δx = –0.6; xm =0.5 149

Figure 8.17 HTC of condensation R134a on J60; Δx = – 0.2 150

Figure 8.18 Δp/L of condensation R134a on J60; Δx = – 0.2 150

Figure 8.19 HTC of condensation R134a on J60 – various G at q = 7.7 kW/m 2 151

Figure 8.20 Δp/L of condensation R134a on J60 – various G at q = 7.7 kW/m 2 151

Figure 8.21 HTC of condensation R134a on J60 – various G at q = 15.4 kW/m 2 151

Figure 8.22 Δp/L of condensation R134a on J60 – various G at q = 15.4 kW/m 2 151

Figure 8.23 HTC of condensation R134a on J60 – q = 7.7 kW/m 2 and q = 15.4 kW/m 2 151

Figure 8.24 Δp/L of condensation R134a on J60 – q = 7.7 kW/m 2 and q = 15.4 kW/m 2 151

Figure 8.25 flow pattern for condensation R1234ze on J60; Δx = – 0.2 154

Figure 8.26 flow pattern for condensation R1234ze on J60 for other conditions 155

Figure 8.27 Flow pattern map of condensation R1234ze on microfin tube J60 156

Figure 8.28 Flow pattern of condensation R1234ze on microfin tube J60 – Model Hajal et al 2003 156

Figure 8.29 Flow pattern of condensation R1234ze-microfin tube J60 – Model Tandon et al 1982 157

Figure 8.30 Flow pattern of condensation R1234ze on microfin tube J60 – Model Breber et al 1980 157 Figure 8.31 HTC of condensation R1234ze on J60; Δx = – 0.6; xm = 0.5 158

Figure 8.32 Δp/L of condensation R1234ze on J60; Δx = – 0.6; xm = 0.5 158

Figure 8.33 HTC of condensation R1234ze on J60; Δx = – 0.2 159

Figure 8.34 Δp/L of condensation R1234ze on J60; Δx = – 0.2 159

Figure 8.35 HTC of condensation R1234ze on J60 – various G at q = 7.2kW/m 2 160

Figure 8.36 Δp/L of condensation R1234ze on J60 – various G at q = 7.2kW/m 2 160

Figure 8.37 HTC of condensation R1234ze on J60 – various G at q = 14.4kW/m 2 160

Figure 8.38 Δp/L of condensation R1234ze on J60 – various G at q = 14.4kW/m 2 160

Figure 8.39 HTC of condensation R1234ze on J60 – various G at q = 15.5kW/m 2 160

Figure 8.40 Δp/L of condensation R1234ze on J60 – various G at q = 15.5kW/m 2 160

Figure 8.41 HTC of condensation R1234ze on J60 – various G at q = 7.2kW/m 2 and q=14.4kW/m 2 161

Figure 8.42 Δp/L of condensation R1234ze on J60 – various G at q = 7.2kW/m 2 and q=14.4kW/m 2 161 Figure 8.43 comparison of HTC – condensation of R134a – q = 7.7kW/m 2 – smooth tube vs microfin tube J60 162

Figure 8.44 comparison of Δp/L – condensation of R134a – q = 7.7kW/m 2 – smooth tube vs microfin tube J60 162

Figure 8.45 comparison of HTC – condensation of R134a – q = 15.4kW/m 2 – smooth tube vs microfin tube J60 163

Figure 8.46 comparison of Δp/L – condensation of R134a – q = 15.4kW/m 2 – smooth tube vs microfin tube J60 163

Figure 8.47 comparison HTC of condensation R1234ze – R134a; Δx = – 0.6; xm =0.5 165

Figure 8.48 comparison Δp/L of condensation R1234ze – R134a; Δx = – 0.6; xm =0.5 165

Figure 8.49 comparison HTC of condensation R1234ze – R134a – various G; Δx = – 0.2 166

Figure 8.50 comparison Δp/L of condensation R1234ze – R134a – various G; Δx = – 0.2 166

Figure 8.51 comparison HTC of condensation R1234ze – R134a – various G – q =15.4kW/m 2 167

Figure 8.52 comparison Δp/L of condensation R1234ze – R134a – various G – q =15.4kW/m 2 167

Figure 8.53 comparison HTC of condensation R1234ze – R134a – various G – q = 7.7kW/m 2 167

Figure 8.54 comparison Δp/L of condensation R1234ze – R134a – various G – q = 7.7kW/m 2 167

Figure 8.55 comparison experimental data and correlation prediction for heat transfer coefficient of condensation R134a on microfin tube J60 168

Figure 8.56 comparison between experimental data and correlation prediction for heat transfer coefficient of condensation R1234ze on microfin tube J60 168

Figure 8.57 comparison experimental data and correlation prediction for pressure drop of condensation R134a on microfin tube J60 170

Figure 8.58 comparison between experimental data and correlation prediction for pressure drop of condensation R1234ze on microfin tube J60 170

List of tables

Table 1.1 Information of some Refrigerants 5

Table 1.2 at the same working conditions 5

Table 2.1 list of void fraction correlations 15

Table 4.1 the list of condensers 43

Table 4.2 the list of total power on evaporator based on various connecting methods 45

Table 4.3 the list of pumps 47

Table 4.4 PID components from autotuning for demineralized water loop 49

Table 4.5 the list of flowmeters on the loop of glycol from chiller to tank 52

Table 4.6 the list of flowmeters on the loop of glycol to condensers 53

Table 4.7 PID components from autotuning closed loop of glycol to condenser 59

Table 4.8 Geometry parameters of tubes 62

Table 4.9 voltage change vs temperature for various thermocouple 64

Table 4.10 calibration results for pressure transducer 65

Table 4.11 calibration results for differential pressure transducer 66

Table 5.1 list of measured quantities and computed quantities of all experiment 68

Table 5.2 the list of direct and indirect measurement parameters 69

Table 5.3 the uncertainties of specific heat capacity of water cd 77

Table 5.4 function values of R134a 77

Table 5.5 function values of R1234ze 78

Table 5.6 The list of instruments and errors supplied from manufactures 80

Table 5.7 the relative uncertainties of heat transfer coefficients depend on τ and θ 81

Table 5.8 general parameter with measured value and error 82

Table 5.9 the general information for data results of repeated experiments 84

Table B.1 Number of experiments performed for two phase flow 96

Table 7.1 experimental conditions for boiling of R134a in smooth tube 101

Table 7.2 experimental conditions 1: for boiling of R134a on microfin tube J60 107

Table 7.3 experimental conditions 2: for boiling of R134a on microfin tube J60 107

Table 7.4 experimental conditions 3: for boiling of R134a on microfin tube J60 107

Table 7.5 experimental conditions 1: for boiling of R1234ze on microfin tube J60 – Δx = 0.6 115

Table 7.6 experimental conditions 2: for boiling of R1234ze on microfin tube J60 – Δx = 0.2 115

Table 7.7 experimental conditions 3: for boiling of R1234ze on microfin tube J60 116

Table 7.8 mean index factor of E, P and I 127

Table 7.9 average of HR, PR, IR for all working condition 132

Table 7.10 data of HR, PR, IR on the tests Δx = 0.6 – xm =0.5 133

Table 7.11 eA% and σ% of boiling heat transfer coefficient 136

Table 7.12 eA% and σ% of boiling pressure drop 139

Table 8.1 experimental conditions for condensation of R134a on smooth tube 140

Table 8.2 experimental conditions 1: for condensation of R134a on microfin tube J60 – Δx=0.6 144

Table 8.3 experimental conditions 2: for condensation of R134a on microfin tube J60 – Δx = 0.2 144

Table 8.4 experimental condition 3: for condensation of R134a on microfin tube J60 – various Δx 144

Table 8.5 experimental conditions 1: for condensation of R1234ze on microfin tube J60; Δx= – 0.6 152

Table 8.6 experimental conditions 2: for condensation of R1234ze on microfin tube J60; Δx= – 0.2 152

Table 8.7 experimental conditions 3: for condensation of R1234ze on microfin tube J60; Δx 152

Table 8.8 some main thermophysical properties of R1234ze and R134a at 35°C 164

Table 8.9 data of HR, PR, IR on the condensation tests Δx = – 0.6; xm =0.5 164

Table 8.10 eA% and σ% of condensation heat transfer coefficient on J60 169

Table 8.11 eA% and σ% of condensation pressure drop on J60 169

List of symbols

M [kg/kmol] molecular mass

Tri = Tsat(pri) [°C] refrigerant inlet saturation temperature

Tro= Tsat(pro) [°C] refrigerant outlet saturation temperature

Twi

Two

Umd [kg/s] uncertainty of mass flow rate of water

∆Tri= Tri∗ − Tsat(pri) [°C] refrigerant inlet temperature difference

∆Tro= Tro∗ − Tsat(pro) [°C] refrigerant outlet temperature difference

Chapter 1 INTRODUCTION

a Research theme; Pressure drop, heat transfer and flow pattern maps inside smooth and enhanced tubes during convective boiling and condensation of R134a and R1234ze

b Field (of basic research/of application); heat transfer in two phase flow

c Type of study (theoretical, numeric, experimental); experimental

1.1 Introduction

1.1.1 Motivation of the research

The continuous demand for size and weight reduction of chillers and heat pumps led to the replacement of smooth tubes with enhanced tubes (microfin tubes, herringbone tubes, and so on) Much different geometries were developed, but it is still unknown, which of them has the best performances because the increase of the heat transfer comes together with the increase of the pressure drop, but their growth is not proportional Even though the analytical description of a two-phase flow is available, the equations are so complex that they can be solved only in few (and very simple) cases

Also, approximate solutions, based on computational approach and numerical simulation (except, again, for very simple cases) is not viable For instance, the shape of the liquid vapor interface usually is very complex shape and would require, to be properly discretized, such a huge number of elements that it is not possible to get, with the computers nowadays available, reliable results in reasonable times The only reliable way to study heat transfer in two-phase flows is based on the experimental approach, and empirical correlations are developed to predict heat transfer coefficient and pressure drop Their fine-tuning or their range extensions continuously require new data

Besides, the challenge is come from the environment, especially from the climate change, two phase flow (boiling and condensation) of current refrigerant in smooth and enhancement tubes are keep working and in the near future should be replaced by some new friendly refrigerants For the new friendly refrigerants, the lack of researches has been done, the insufficient data about the performance of new fluid is facing Numeral value can be used to design is not enough reliable for manufacture So that more and more data results from the real experiment is urgent required

In case of internal flow, the main difficulty, concerning to the evaluation of the heat transfer coefficient and friction factor, is related to the flow regimes (distribution of liquid and vapor in a cross-section) As it is easily understood, the main heat transfer mechanism changes if, at the wall, there is liquid, vapor, both of them, droplet impingement, bubble formation and so on Every flow regime requires a specific analytical description and appropriate criteria to state if it occurs or not As quality, liquid mass flux and vapor mass flux change along a duct, several flow regime onsets along

a duct in the presence of heat transfer The criteria are usually represented in two-dimensional diagrams called flow pattern maps The internal geometry of the enhanced tubes and the thermal properties of the fluid affect the distribution of the phases in the cross section As a consequence, every time a new enhanced tube is developed, or a different fluid is used, a new flow pattern map has

to be drawn and compared with the existing ones to check what are the best operating conditions?

1.1.2 Originality;

The experimental approach is the only way to evaluate the performances of the enhanced tubes, moreover the post processing of the measures and the comparison with the data from the literature is the only method to get hints about the changing of the geometry that could improve the performances

of the enhanced tubes Usually a single parameter is used in the correlations to account for the characteristics of the tube geometry, but frequently this is not enough Experiment and data processing are necessary to identify innovative parameters to reduce the uncertainty of the correlations Furthermore, the same geometry could result in different performances if new refrigerant blends are adopted Hence, the comparison of different fluids could be very interesting in the perspective of phasing out the ones currently used to reduce the Global Warming Potential

1.2 State of the art of the specific research field

In the past few decades, a large number of researches are published with the clearly described about the phenomenon performance and behaviors of many refrigerants on the boiling and condensation processes flow inside the horizontal smooth tubes The general flow pattern map, heat transfer and pressure drop models had been built with many various versions not only for boiling but also for condensation of the previous and current refrigerants

The typical correlation of two phase on smooth tube which were frequently used as one of the most accurate correlation created and published by Thome et al [1], [2], [3]–[14] With their huge contribution to build the general flow pattern map, heat transfer coefficient and pressure drop models based on the various flow pattern not only for boiling but also for condensation flow in smooth horizontal tube which are used as reference by almost all later researches, even for microfin tube or research with the newest refrigerant nowadays In 1998, they already published the full method to define flow pattern map, heat transfer and pressure drop for boiling in horizontal smooth tube [2], [7], [8], but their models did not easily to use, so keep working and updating with the new models were presented in the next few years [11]–[14] Besides, they also focused on building the models for condensation combine with the group of Cavallini et al [9], [10] Until now, they keep working with two phase flow for the new refrigerant [5], new shape of tube [6] or some interested topic as dryout

on the tube [4] or the comparison the performance at different saturation temperature of boiling [3] There is a huge deficiency when studying two-phase flow of fluids without mentioning the group of Cavallini et al They have been researching many different issues on the condensation phenomena in long period times started from early 1970s to now [5], [6], [15]–[31] They supplied a lot of information about the phenomenon, effect of performance of those processes on experiments and also giving their correlations on many types of tubes In the recent years the members of Cavallini group such as Davide Del Col, Luisa Rossetto … are keep continuing with these fields, concentrating to condensation on tubes (smooth, micro-fin tubes, micro channel…) with some new refrigerants and also concerning about the flow boiling

These days, the group of Bengt Sunden are also focused on two phase phenomenon on microfin tube with some difference geometries for the current refrigerants by using their experimental data as well

as collected data results from the others group to make an overview for their purposes [32]–[35]

In the group of Muzzo et al, started in 1997 with the two phase flow in the microfin tubes with refrigerant R22 [36], [37], then moved to use refrigerant R134a for the same topic which had been done by the next authors Colombo et al [38] Until now, keep researching with new method approach

not only about the experimental system but also about the accurate equipment and modern controlling method [39] Then continuing with the new refrigerants, promising for the next publishes

Beside some group researchers are always pursuing and chasing from the past to now The contribution from the others is highly appreciated for the concerning topics One of the most common approaches to describe evaporation and condensation is the flow regime characterization, since flow patterns and transport rates between the phases are mutually related To discriminate flow patterns inside horizontal smooth tubes under adiabatic and diabatic conditions, numerous flow patterns maps

or criteria for transition between flow patterns have been proposed for condensation [9], [40]–[44] and [11], [45]–[47] for evaporation Even though the coordinate systems proposed for the representation are different and there is not agreement on which is the best, a widespread knowledge about flow patterns is by now available for horizontal smooth tubes On the contrary, only little information can be found about the effect of microfins on the flow pattern behavior

More detailed, on smooth tube, analyzed about the dynamics and heat transfer of boiling and condensation which were not only in the experimental field but also in the numerical, simulation fields and presented the general correlations Some normal use correlations about heat transfer coefficient for evaporation from Gungor and Winterton [48], Liu and Winterton [49], Kandlikar [50], [51], Saitoh et al [52], Wojtan et al [12], or some new correlations have just published in the previous year from Fang et al [53], Mohseni et al [54] Also for condensation, the correlations of Dobson et al [55], Thome et al [10], Cavallini et al [31], Shad et al [56] could be used Besides, the correlation for pressure drop of two phase flow are published by Friedel [57], Jung et al [58], Muller and Heck [59] and Quiben et al [14]

It is not be doubt that the outstanding performance of heat transfer in microfin tubes for both evaporation and condensation is valuable to consider without introducing excessive penalization in the pressure drop Most of the available literature deals with experimental investigations for this topic [60]–[74], concerned with different refrigerants combined with various geometrical parameters such

as tube diameter, spiral angle, fin height and shape and number of fins Actually, although predictive and empirical models have been proposed to determine the heat transfer coefficients and the pressure drop in evaporation and condensation processes Some typical correlations for heat transfer of evaporation process on microfin tube are published by Thome et al [1], Cavallini et al [17], Yu et al [75], Yun et al [76], Chamra et al [77], Wu et al [32], Rollmann and Spindle [78] Also the heat transfer correlations for condensation are came from Yu and Koyama [79], Kedzieski and Goncalves [80], Han et al [81], Cavallini et al [30] Also for the pressure drop of two phase flow, some correlations were built for both evaporation and condensation choi et al [82], Goto et al [83], the others used for separated purposed Some particular pressure drop correlations could be mentioned for evaporation Bandarra et al [84], Rollmann and Spindler [78] or for condensation Kedzieski and Goncalves [80], Haraguchi et al [85], Cavallini et al [25] The fluid dynamics and heat transfer have not been yet completely characterized and the most reliable approach in the heat exchanger design is essentially based on empirical correlations Therefore, the collection of extensive experimental data

is further needed to improve the knowledge of the governing phenomena as well as to refine the existing models

- U [W/m2K] overall heat transfer coefficient

- ΔTm [K] appropriate mean temperature difference

- R [mK/W] overall thermal resistance per unit length

Figure 1.1 microfin tube

The performance of heat exchanger improves if the overall resistance per unit length reduces An enhanced surface geometry may be used to increase either or both the heat transfer coefficient and heat transfer surface area Here is below presented some advantages of applying microfin tube on two phase flow and also some drawbacks:

 Advantages:

- Microfin surface could create the mixture of flow go through and promote annular flow happened earlier than smooth surface

- On the microfin tube, the nuclear boiling could be represented more than smooth tube

- Maybe the dryout phenomenon will be occurred later or doesn’t affect

- Reduced volumes of heat exchangers

- Small refrigerant charge

- High heat exchanger efficiency

Table 1.1 Information of some Refrigerants

Refrigerant type safety ODP GWP Critical point MM NBP lifetime

T[°C] p [Mpa] [Kg/kmol] [°C] [day]

Simply analysis to compare 3 different refrigerants, one is widely used as current refrigerant R134a, two others are the newest refrigerants would be promising replacement refrigerants R1234yf, R1234ze for the near future The comparison has been taken by using data results from some correlations at the same working conditions analyzed for 3 different refrigerants

Table 1.2 at the same working conditions

of the phase for refrigerants R134a, R1234yf and R1234ze

As showing in Figure 1.2, flow pattern maps of 3 various refrigerants nearly the same for boiling, just only a slightly shifted the transient curves but very small Just only one particular note is that the changing of flow pattern from intermittent into annular of R1234ze is sooner than R134a and the later changing is R1234yf, with x = 0.29, x = 0.31 and x = 0.34 respectively

Figure 1.2 comparison the flow pattern map for R134a, R1234yf and R1234ze

Intermittent flow

Annular flow

Mist flow

Slug + stratified wavy flow

Slug

Dryout

black : R134a green : R1234yf

red : R1234ze

Heat transfer coefficient:

Four heat transfer coefficient correlations of boiling in smooth tube are used to calculate and compare the effect on various refrigerants The correlations of Gungor-Winterton 1986 [48], Liu-Winterton

1991 [49], Wojtan-Ursenbacher and Thome 2005 [12] and Saitoh-Daiguji and Hihara 2007 [52], are applied to compute the heat transfer coefficient As show in Figure 1.3, Figure 1.4, Figure 1.5 and Figure 1.6, the heat transfer coefficient of R1234ze is nearly equal to R134a and those are always higher than R1234yz on the whole range of quality.

Figure 1.3 Gungor and Winterton 1986 [48] Figure 1.4 Liu and Winterton 1991 [49]

Figure 1.5 Wojtan et al 2005 [12] Figure 1.6 Saitoh et al 2007 [52]

0 500 1000 1500 2000 2500 3000 3500 4000

0 500 1000 1500 2000 2500 3000 3500 4000 4500

Figure 1.7 Gronnerud 1972 Figure 1.8 Fridel 1979 [57]

Figure 1.9 Muller and Heck 1986 [59] Figure 1.10 Quiben et al 2007 [14]

Due to the finding replacement refrigerant for nearest future, the refrigerant has the same performance

of heat transfer is priority in this project, so that, the research will be done with R134a and R1234ze

The targets of project:

Experimental researches about the behavior of two-phase flow (Flow patterns, heat transfer coefficients, pressure drop) on the micro-fin tubes with some refrigerants

- Implementing the experimental system

- Build the procedure to operate the system

- Experimental test for Smooth tube of R134a

- Experimental test for micro-fin tube of R134a

- Experimental test for micro-fin tube of R1234ze

- Analyze and compare data results

- Compare experimental results with correlations

0 1 2 3 4 5 6

0 1 2 3 4 5 6

Chapter 2 THEORY OF TWO-PHASE FLOW

2.1 Basic equations for two phase flow

Local balance laws at point x are expressed in terms of partial differential equations

Called k is denoting the phases:

t k

𝑔̇ [kg/m2s] mass transfer per unit interface area and unit time;

2.2 Parameters of two-phase pipe flows

A fluctuation behavior is presented on two phase flow To describe two phase flow, simplify Navier stokes equations averaging operators, acting on space and time is needed

Figure 2.1 phase in a circular tube

2.2.1 Phase density function

The presence or absence of the phase k at a given point x and a given time t is characterized by the phase density function defined:

𝑃𝑘(𝑥, 𝑡) = {0 if point x does not pertain to phase k 1 𝑖𝑓 point x pertains to phase k

2.2.2 Instantaneous space-averaging operators

Instantaneous field variables may be averaged over a n-dimension domain:

Dn,k n-dimension domain where is present the k-th phase

In the following area average quantity will be considered, so the development of averaging operators will be for n equal to 2 consequently two different instantaneous space averaging operators are introduced:

 Instantaneous space fraction:

 Instantaneous mass flow rate:

Ṁk = ∫D ρkυk,adD = A ∈k< ρkυk,a>2,k

where: Ṁk [kg/s] mass flow rate of kth phase

2.2.3 Local time-averaging operators

The local field variables can be averaged over a time interval, whose magnitude T must be chosen:

 Large enough compared with the time scale of the turbulence fluctuations;

 Small enough compared with the time scale of the overall flow fluctuations

In a given point x of a two-phase flow, the kth phase passes this point intermittently and a field variable fk(x,t) is a piecewise continuous function Denoting by Tk(x,t) the cumulated time of the

kth phase is within the interval T, it is possible to define two different local time averaging operators:

2.2.4 Commutativity of average operators

By the definitions, the following identity holds:

Pn,k< fk>n,k=< αkfkk>n (2.18) the time-averaged volumetric

V̇k= A ∈k< υk,a >2,k= A < αkυk,ak>2 (2.19) mass flow rates

Ṁk = A ∈k< ρkυk,a >2,k= A < αkρkυk,ak>2 (2.20)

2.2.5 Qualities

The mass velocity is defined:

where: G [kg/m2s] total mass flux

The quality is defined:

where: V̇ [m3/s] total volumetric flow rate

The local volumetric flux is defined:

2.3 Composite averaged equations

The time-average over an interval T of the instantaneous area-averaged balance equations brings:

∂zAk< (Τ̃k⋅ vk) ⋅ nz >2,k − ∂

2.4 Friction pressure drop

The frictional pressure drop is not usually measured directly, most of the experimental data provides values of the total pressure drop A void fraction evaluation is needed to determine the acceleration and gravity pressure drops to be subtracted from the total pressure drop to get the frictional component

Simplified balance equations:

The averaging operators defined over a cross section area and over a time interval give averages

of products expressed as a function of the product of the averages by means of correlation coefficient:

Where: Cs space correlation coefficient

Where: Ct time correlation coefficient

Some hypothesizes are assumed:

- The space correlation coefficients are equal to 1

- The time correlation coefficients are equal to 1

- The equation of state valid for local quantities applies to averaged quantities

- Longitudinal conduction terms in each phase as well as their derivatives are negligible

- The phase viscous stress derivatives and the power of these viscous stresses are negligible

- The flow is symmetrical with respect to a straight line

 Boiling number

Bo =Ghq

Due to the complicated defining by experiment, void fraction correlations from theoretical models as well as the empirical correlations are presented in Table 2.1 which are collected from the literatures and most frequently used on numerous researches in this field

Table 2.1 list of void fraction correlations

Simply apply Working condition in condensation case at mr = 50kg/h; Tr = 35oC; heat flux q = 7.44kW/m2, average quality x = [0.1 – 0.95] of R134a to analyze and compare void fraction by using various correlations as showing in Figure 2.2, the result of void fraction from homogeneous model

is far from the others, it could be explained by the assuming of moving between two phases of flow

at the same velocity While the result from the others are quite close together, just at the low-quality

x < 0.3 the void fraction from different correlations are a little bit high error, in contract at the quality

x > = 0.3, the result of void fraction can get right values The increasing quality the increasing of agreement of all correlations

Figure 2.2 void fractions from various correlations

Zivi 1964 Chisholm 1973 Chen 1986 Rouhani and Axelsson 1970 Huq and Loth 1992

Chapter 3 CORRELATIONS FOR MICROFIN TUBES

et al 1980, Tandon et al 1982 or Hajal et al 2003 or the map for evaporation from Wojtan et al 2005, Rollmann and Spindler 2015, Zhuang et al 2016 or the newest map from Yang et al 2018

Even many different group researches, they also classified the map into some main regimes: bubbly flow, plug flow, slug flow, intermittent flow, stratified flow, stratified wavy flow, annular flow, dry-out regime, mist flow or transition regime, other differences just only the difference name they called for the same regime Depend on every single group they have every different name for their classification regimes Every regime can be described as below:

+ bubbly flow: due to buoyancy force, gas bubbles focus on the upper part of tube, and normally take place at the high mass flow rate

+ plug flow: the individual small bubbles have coalesced to create long bubble, can call the name elongated bubble flow

+ slug flow: at high velocity, the wave amplitude is so large which increasing and touching to the top of the tube

+ intermittent flow: could be defined instead of plug and slug flow

+ fully stratified flow: at low velocity, liquid and vapor are completely separated, interface between them is smooth

+ stratified wavy flow: the formation of wave in the interface between liquid and vapor of stratified flow when the velocity of gas rises up

+ annular flow: when the liquid forms around perimeter of the tube, vapor flows in the core separate and make the annular shape of liquid flow

+ dry-out regime: at higher quality of vapor, thinner of annular liquid flow will be disappeared,

at the top of tube becomes dry first, then gradually spread around tube to bottom

+ mist flow: liquid will be entrained to the core of gas phase as small droplets

+ transition regime: on the changing pattern regimes, it is not clear to define which pattern they are, so that the name transition regime is used

For micro-fin tube, those the last few years, some other detailed regimes are called for flow patterns: + helix flow: the formation of helix flow due to the helical structure of microfin where the liquid flow helically through out

+ slug + helix flow: on the regime, slug and helix flow are spontaneously occurred

Figure 3.1 Flow regime for boiling from Wojtan et al 2005

Figure 3.2 Flow regime for condensation from Hajal et al

Here are some typical and available maps were built from previous group authors for Boiling and condensation:

3.1.1 Boiling model of Wojtan et al 2005 [11]

The model of Wojtan et al 2005 describe the full procedure to build the flow pattern map for boiling with the boundary transitions between two regimes and introduced the new transitions for dry-out and mist flow And apply the method of Biberg to determine the wetted angle in two phase stratified pipe flow With this method, the iteration does not a mater to compute dimensionless geometrical variables of stratified two-phase flow

Figure 3.3 flow pattern map of Wojtan et al 2005

1 void fraction ε: Rouhani-Axelsson correlation

1

200(1 − ε)ε[1 − 2(1 − ε)][1 + 4((1 − ε)2+ ε2)]} (3.2)

2 Geometrical parameters for two phase flow in a circular tube:

Figure 3.4 geometrical parameters of stratified flow in circular tube

Slug+ stratified wavy zone: Gstrat < G < Gwavy and x < Xia Stratified wavy zone: x > xIA

4 The transition boundary curve between stratified wavy flow and fully stratified flow Gstrat

Gstrat = {(226.3)3ALD AVD2 ρ V (ρ L −ρ V )μ L g

and Gstrat = Gstrat(xIA) at x < xIA (3.10)

5 The transition boundary curve between Intermittent and annular flows: xIA

If Gstrat(xi) ≥ Gdryout(xi), then Gdryout(xi) = Gstrat(xi)

If Gwavy(xi) ≥ Gdryout(xi), then Gdryout(xi) = Gwavy(xi)

If Gdryout(xi) ≥ Gmist(xi), then Gdryout(xi) = Gmist(xi)

xi = [0:1] the considered vapor quality

3.1.2 Boiling model of Rollmann and Spindler 2015 [45]

One of the newest and rarest models of flow pattern map for microfin tube is Rollmann and Spindle

2015 This model was built from their experimental results with refrigerant R134a, they put the new name for new regime of flow, helix, that is effect of helical fin inside of the tube as they observed The model classified the pattern into six regime, annular, helix, slug/helix, stratified wavy, slug/ slug+stratified wavy and stratified flow

Figure 3.5 flow pattern map of Rollmann and Spindle 2015

+ Transition from stratified flow to stratified wavy flow:

Gstrat = {4μL g(ρL−ρV)ρ V ε(1−ϵ)

S2 = 0.02844

C5 = 22.9 kg/sm2

+ Transition from slug flow to stratified wavy flow:

G < Gstrat Stratified flow

G > Gslug and G < Gsw Stratified wavy flow

G < Gslug and G < Gsw Stratified wavy/slug flow or slug flow

G < Gslug-helix Slug/helix flow

G < Ghelix Helix flow

G > Ghelix Annular flow

3.1.3 Boiling model of Zhuang et al 2016 [46]

The model of Zhuang et al 2016 was built in the terms of dimensionless weber number We and Mattinelli parameter Xtt, based on the model of Kim et al combine with their experimental data for R170 on the test rig designed to observe flow patterns with the working range of saturation pressures from 1.5Mpa to 2.5Mpa on mass flux from 100kg/m2s to 250kg/m2s

This model classified 4 transition curves to build the map with 5 different patterns named smooth annular, wavy annular, transition, slug, plug

Figure 3.6 flow pattern map of Zhuang et al 2016

Calculate Martinelli number and dimensionless parameter Weber:

Marttinelli number: Xtt

Xtt = (1−xx )0.9(ρG

ρL)0.5(μl

Dimensionless parameter Weber: We

Based on the Reynold range of liquid flow

Wavy-annular to transition flow:

3.1.4 Boiling model of Yang et al 2018 [47]

The map of Yang et al 2018 presented two transitions line on two difference maps as a function of Martinelli parameter Xtt, the map for plug and slug regime, and another map for slug and annular for refrigerant R1234ze The transition lines based on three dimensionless numbers K1, K2, K3 which represented the inertia force, surface tension force, shear force, gravity force and evaporation momentum force

Figure 3.7 flow pattern map of Yang et al 2018

Dimensionless number K1 is a ratio of evaporation momentum force with inertia force:

K1 = evaporation momentum forceinertia force = (

q hlv)

2 1 ρv G2 ρl

= (hq

lv)2 ρl

Dimensionless number K2 is a ratio of evaporation momentum force with surface tension force:

K2 =evaporation momentum forcesurface tension force = (

q hlv)

2 1 ρv σ D

= (hq

lv)2 Dσρ

Dimensionless number K3 is a ratio of shear force with gravity force:

K3 =gravity forceshear force =

μl ρl

G D

K = KS−A = K1−0.2963K20.3620K30.1941= 0.3044Xtt0.5671 (3.37)