Louisiana State UniversityLSU Digital Commons

LSU Historical Dissertations and Theses Graduate School

1994

The Studies of Rigid PVC Compounds:Morphology, Rheology, and Fusion Mechanism.Cheng-ho ChenLouisiana State University and Agricultural & Mechanical College

Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses

This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion inLSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contactgradetd@lsu.edu.

Recommended CitationChen, Cheng-ho, "The Studies of Rigid PVC Compounds: Morphology, Rheology, and Fusion Mechanism." (1994). LSU HistoricalDissertations and Theses. 5784.https://digitalcommons.lsu.edu/gradschool_disstheses/5784

INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

University Microfilms International A Bell & Howell Information Company

300 North Zeeb Road. Ann Arbor. Ml 48106-1346 USA 313/761-4700 800/521-0600

Order Number 9508557

The studies of rigid PV C compounds: Morphology, rheology, and fusion mechanism

Chen, Cheng-Ho, Ph.D.The Louisiana State University and Agricultural and Mechanical Col., 1994

UMI300 N. ZeebRd.Ann Arbor, MI 48106

THE STUDIES OF RIGID PVC COMPOUNDS: MORPHOLOGY, RHEOLOGY, AND FUSION MECHANISM

A Dissertation

Submitted to the Graduate Faculty of the Louisiana State University and

Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

in

The Department of Chemical Engineering

byCheng-Ho Chen

B.S. National Cheng-Kung University, Taiwan, R.O.C., 1986 M.S. National Cheng-Kung University, Taiwan, R.O.C., 1988

August, 1994

To my wife, Yu-Wen Lo and

my son, I-Ying Chen

ii

ACKNOWLEDGEMENTS

The author would like to express his deep appreciation to Dr. Rosemarie D. Wesson, Assistant Professor, and Dr. John R. Collier, Professor, of Chemical Engineering at Louisiana State University for their able direction and assistance on this research. Also, appreciation is extended to the members of the advisory committee; Dr. Frank R. Groves, Dr. William H. Daly, and Dr. Paul S. Russo.

The financial assistance granted to the author by both the Department of Chemical Engineering at Louisiana State University and the DOW Chemical Company is also appreciated. The help from my student worker, Erick J. Comeaux, is also greatly appreciated.

My appreciation also goes to Dr. Lawrence J. Effler, Dr. Deepak R. Parikh, Dr. Mark T. Berard, Suzanne Wiliams, and Rhonda Neal, who work for the Dow Chemical Company, for their assistance.

The author also wishes to thank Cindy Henk, Dr. Sharon W. Matthews, and Dr. loan I. Negulescu at Louisiana State University for their strong and useful recommendations.

TABLE OF CONTENTS

DEDICATION ................................................. iiACKNOWLEDGEMENTS ...................................... iii

LIST OF TABLES ......................................... viiLIST OF FIGURES ....................................... ixABBREVI AT IONS........................................... xv i i iNOMENCLATURE ........................................... xixABSTRACT .................................................... xxCHAPTER1. INTRODUCTION ....................................... 12. MORPHOLOGICAL ANALYSES OF PVC, PVC/CPE, AND

PVC/CPE/OPE COMPOUNDS ........................... 102.1 Basic Principles of Electron

Microscopy .................................... 102.2 Compatibility between PVC and CPE .......... 132.3 Morphology of PVC Compounds ................. 132.4 Experimental Methods......................... 15

2.4.1 Preparation of PVC Compounds ....... 152.4.2 DSC Thermal Analysis ................. 152.4.3 Morphological Analysis................ 16

2. 4. 3.1 SEM ........................... 162. 4. 3. 2 TEM ............................ 192. 4.3.3 S-TEM ......................... 19

2.5 Results and Discussion ...................... 212.5.1 DSC Analysis of CPE and PVC Compounds 212.5.2 SEM Analysis .......................... 212.5.3 TEM Analysis .......................... 3 02.5.4 S-TEM Analysis ........................ 38

3 FUSION PROPERTIES OF PVC IN A HAAKE TORQUERHEOMETER .......................................... 6 53.1 Fusion of PVC ................................ 653.2 Methods to Assess the Degree of Fusion of

PVC ............................................ 723.2.1 Introduction to Capillary

Rheological Analysis ................. 723.2.2 Introduction to DSC Thermal Analysis 81

iv

3.3 Experimental Method ......................... 843.3.1 Preparation of PVC Compounds ....... 843.3.2 Capillary Rheological Analysis ..... 853.3.3 DSC Thermal Analysis ................. 873.3.4 SEM Analysis .......................... 87

3.4 Results and Discussion ..................... 873.4.1 Haake Torque Rheometer .............. 873.4.2 Capillary Rheological Analysis ..... 913.4.3 DSC Thermal Analysis ................. 953.4.4 Rheological Analysis vs. Thermal

Analysis ............................... 1003.4.5 SEM Analysis .......................... 103

4 CORRELATION OF COMPOUNDING ENERGY WITH FUSIONLEVEL BY MATHEMATICAL MODELING ................. 1124.1 Literature Review ........................... 1124.2 Experimental Method ......................... 115

4.2.1 Preparation of PVC Compounds ....... 1154.2.2 DSC Thermal Analysis ................. 1154.2.3 Capillary Rheological Analysis ..... 117

4.3 Theoretical Method, Results and Discussion 1175 APPLICATION OF FACTORIAL EXPERIMENTAL DESIGN

TO DEMONSTRATE THE INFLUENCE OF TEMPERATURE,ROTOR SPEED, AND TOTALIZED TORQUE ON THEFUSION^OF PVC COMPOUNDS ......................... 1225.1 Introduction ................................. 1225.2 Experimental Method ......................... 125

5.2.1 Preparation of PVC Compounds ....... 1255.2.2 DSC Thermal Analysis ................. 12 65.2.3 Capillary Rheological Analysis ..... 126

5.3 Results and Discussion ..................... 1275.3.1 DSC Thermal Analysis ................. 1275.3.2 Capillary Rheological Analysis ..... 146

6 FUSION CHARACTERISTICS AND MORPHOLOGY ANALYSIS 1516.1 Introduction to CPE, OPE, and Calcium

Stearate ...................................... 1516.2 Introduction to Rabinovitch's Theory

[Nass, 1992] 1546.3 Experimental Method ......................... 158

6.3.1 Preparation of PVC Compounds ....... 1586.3.2 SEM Analysis .......................... 159

6.4 Results and Discussion ..................... 1606.4.1 Fusion Characteristics and

Lubrication Mechanisms .............. 1606.4.2 SEM Analysis .......................... 176

v

7 CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK 193REFERENCES ............................................. 200

V I T A ..................................................... 2 04

vi

1722

59

86

116

120

128

129130

140

147

161

161

162

162

LIST OF TABLES

PVC/CPE blend samples* analyzed by SEM, TEM, and S-TEM ........................................Tg of CPE and PVC/CPE blends .................Melt index v s . domain size and IZOD impact ...........................................

Test conditions for rheological measurements ....................................Digital data from the Haake Torque Rheometer at a present bowl temperature of 180°C .....Enthalpy of the PVC sample at time t, Ht, for various starting temperatures (where W = H0 + Wt) ..........................................Main effect of temperature (xx) on PVC compounds ........................................Main effect of rotor speed (x2) on PVC compounds ........................................Main effect of TTQ (x3) on PVC compounds . . .Summary of main, two - factor interaction, and three - factor interaction effects of PVC, PVC/CPE, and PVC/CPE/OPE compounds measured by DSC thermal analysis ............Summary of main, two - factor interaction, and three - factor interaction effects of PVC, PVC/CPE, and PVC/CPE/OPE compounds measured by capillary rheometer .............

Fusion characteristics of PVC blends at temperature = 190°C, rotor speed = 100 rpm .Fusion characteristics of PVC blends at temperature = 190°C, rotor speed = 60 rpm . .Fusion characteristics of PVC blends at temperature = 175°C, rotor speed = 100 rpm .Fusion characteristics of PVC blends at temperature = 175°C, rotor speed = 60 rpm . .

vii

6.5 Fusion characteristics of PVC/CPE compounds at temperature = 190°C, rotor speed = 60r p m ............................................... 168

6.6 Fusion characteristics of PVC/CPE/0.3phrOPE compounds at temperature = 19 0°C, rotorspeed = 60 rpm .................................. 168

6.7 Fusion characteristics of PVC/OPE compounds at temperature = 190°C, rotor speed = 60r p m ............................................... 174

6.8 Fusion characteristics of PVC/5phrCPE/0PE compounds at temperature = 190°C, rotorspeed = 60 rpm .................................. 174

2

11

1218

2023

24

25

26

27

28

29

31

LIST OF FIGURES

A schematic model of a suspension PVC particle [Butters, 1982] .....................

Basic difference between Transmission (TEM) and Scanning (SEM) Electron Microscopes [Bozzola et a l ., 1992] ........................Log scale (middle) of the range of resolving power of various magnifying tools (left), and the structures that they are capable of resolving (right) [Bozzola et a l ., 1992] ...Three different directions of samples ......Staining reaction sequence[Fleischner et a l . , 1977] ....................DSC traces of CPE and PVC/CPE blends .......0 phr 3615P CPE in PVC/CPE blend coated with 200 A of gold-palladium and examined by SEM (perpendicular) ................................10 phr 3615P CPE in PVC/CPE blend coated with 200 A of gold-palladium and examined by SEM (perpendicular) ............................20 phr 3615P CPE in PVC/CPE blend coated with 200 A of gold-palladium and examined by SEM (perpendicular) ............................0 phr 3 615P CPE in PVC/CPE blend coated with 200 A of gold-palladium and examined by SEM (parallel) ......................................10 phr 3615P CPE in PVC/CPE blend coated with 200 A of gold-palladium and examined by SEM (parallel) .................................20 phr 3615P CPE in PVC/CPE blend coated with 200 A of gold-palladium and examined by SEM (parallel) .................................0 phr 3615P CPE in PVC/CPE blend treated by chemical reaction and examined by TEM (perpendicular) ................................

ix

32

33

34

35

36

39

40

41

42

43

44

45

10 phr 3615P CPE in PVC/CPE blend treated by chemical reaction and examined by TEM (perpendicular) ................................

20 phr 3615P CPE in PVC/CPE blend treated by chemical reaction and examined by TEM (perpendicular) ................................0 phr 3615P CPE in PVC/CPE blend treated by chemical reaction and examined by TEM (parallel) ......................................10 phr 3615P CPE in PVC/CPE blend treated by chemical reaction and examined by TEM (parallel) ......................................20 phr 3615P CPE in PVC/CPE blend treated by chemical reaction and examined by TEM (parallel) ......................................10 phr 3615P CPE in PVC/CPE blend without chemical treatment examined by TEM (parallel) ......................................10 phr CPE (MI = 0.4) in PVC/CPE/OPE blend treated by chemical reaction and examined by TEM (perpendicular) .............0 phr 3 615P CPE in PVC/CPE blend without chemical treatment examined by S-TEM (perpendicular) ................................10 phr 3615P CPE in PVC/CPE blend without chemical treatment examined by S-TEM (perpendicular) ................................20 phr 3615P CPE in PVC/CPE blend without chemical treatment examined by S-TEM (perpendicular) ................................

0 phr 3615P CPE in PVC/CPE blend without chemical treatment examined by S-TEM (parallel) ......................................10 phr 3615P CPE in PVC/CPE blend without chemical treatment examined by S-TEM (parallel) ......................................

x

2.25 20 phr 3615P CPE in PVC/CPE blend withoutchemical treatment examined by S-TEM (parallel) ...................................... 46

2.2 6 10 phr CPE (MI = 0.4) in PVC/CPE/OPEblend treated by chemical reaction andexamined by S-TEM (perpendicular) ........... 48

2.27 10 phr CPE (MI = 1.0) in PVC/CPE blendwithout chemical treatment examined by S-TEM (perpendicular) ................................ 49

2.2 8 10 phr CPE (MI = 0.2) in PVC/CPE blendwithout chemical treatment examined by S-TEM (perpendicular) ................................ 50

2.29 10 phr CPE (MI = 1.0) in PVC/CPE blendwithout chemical treatment examined by S-TEM (parallel) ...................................... 51

2.3 0 10 phr CPE (MI = 0.2) in PVC/CPE blendwithout chemical treatment examined by S-TEM (parallel) ...................................... 52

2.31 10 phr CPE (MI =1.0) in PVC/CPE blend without chemical treatment examined by S-TEM(flat cut) ...................................... 54

2.32 10 phr CPE (MI = 0.2) in PVC/CPE blendwithout chemical treatment examined by S-TEM(flat cut) ...................................... 55

2.33 Three dimensional visions of low and highIZOD impact samples ............................ 5 6

2.34 CPE particle size distribution curves of low IZOD impact sample (MI = 1.0) and high IZODimpact sample (MI = 0.2) (perpendicular) ... 57

2.35 Melt index vs. domain size and IZOD impact . 603.1 A schematic illustration of the three

patterns of fusion [Krzewki et a l ., 1981] .. 673.2 Relationship between the degree of fusion

and physical properties [Benjamin, 1978] ... 683.3 PVC fusion mechanisms in a extrusion

equipment [Gilbert, 1983] .................... 70

xi

3.4 Normal fusion mechanism of the rigid PVC ina torque rheometer [Pedersen, 1991] 71

3.5 A schematic diagram of a capillary rheometer[Nielsen, 1977] 7 4

3.6 Bagley correction for capillary rheometers[Nielsen, 1977] 77

3.7 Standard fusion curves for two PVC compounds[Krzewki et a l . , 1981] 79

3.8 Normalized fusion curves for two PVCcompounds [Krzewki et a l . , 1981] 80

3.9 DuPont DSC thermal analysis of unplasticised PVC compounds compression molded atdifferent temperatures [Gilbert, 1981] 82

3.10 Fusion curve measured by DSC thermal analysis - heat of fusion vs. differentmolded temperatures [Gilbert, 1981] 83

3.11 A typical temperature/torque/totalized torque curves of PVC melted in a HaakeTorque Rheometer ............................... 8 8

3.12 Effect of temperature on fusion time ........ 903.13 Effect of rotor speed on fusion time ........ 923.14 Entrance pressure drop measured by a zero -

length capillary die at a constant shearrate of 23.6 sec'1 and 180°C ................. 93

3.15 Effect of temperature and totalized torqueon the degree of fusion of PVC compounds measured by capillary rheological analysis . 94

3.16 DSC thermal analysis curves of PVC compoundsprocessed in a Haake Torque Rheometer at rotor speed = 60 rpm, TTQ = 10 kg-m-min, and various temperature ............................ 96

3.17 Effect of temperature and totalized torque on the degree of fusion of PVC compoundsmeasured by DSC thermal analysis ............ 98

xii

3.18 Effect of temperature on the degree of fusion of PVC compounds processed in a Haake Torque Rheometer ................................... 9 9

3.19 Effect of rotor speed on the degree of fusion of PVC compounds processed in a Haake Torque Rheometer (blending Time = 6 minutes) ................................ 101

3.20 Morphological changes of PVC compounds ((a):PVC powder, (b): low, (c): medium, and (d): high fusion levels), extruded from thecapillary rheometer ............................... 102

3.21 Surface morphology of a suspension PVCpowder ............................................... 104

3.22 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature =190°C, rotor speed = 100 rpm, and TTQ = 1kg-m-min ............................................ 105

3.23 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature =160°C, rotor speed = 60 rpm, and TTQ = 10kg-m-min ............................................ 107

3.24 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature =180°C, rotor speed = 60 rpm, and TTQ = 10kg-m-min ......................................... 108

3.25 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature =200°C, rotor speed = 60 rpm, and TTQ = 10kg-m-min ......................................... 109

3.26 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature =200°C, rotor speed = 60 rpm, and TTQ = 15kg-m-min ......................................... Ill

4.1 Simple total energy balance for a torque rheometer [Pedersen, 1991] 114

4.2 Experimental results vs. mathematical result based on a simple total energy balance in atorque rheometer ............................... 121

xiii

124

131

132

133

134

135

136

137

141

142

143

144

Diagrammatic representation of the standard ordering of a 23 factorial experimental design ..........................................

Diagrammatic representation of the observed yields (heat of fusion: mJ/mg) and the standard ordering of experiments of PVC compounds ......................................Determination of the main effect of temperature (xj on PVC compounds ...........Determination of the main effect of rotor speed (x2) on PVC compounds ..................Determination of the main effect of TTQ (x3) on PVC compounds ...............................Determination of the temperature vs. rotor speed interaction effect (x3 vs. x2) on PVC compounds ...................................... .Determination of the temperature vs. TTQ interaction effect (x3 vs. x3) on PVC compounds ...................................... .Determination of the rotor speed vs. TTQ interaction effect (x2 vs. x3) on PVC compounds .......................................Determination of the three - factor interaction effect of temperature vs. rotor speed vs. TTQ (x3 vs. x 2 vs. x3) on PVC compounds ((+) tetrahedron) ..................Determination of the three - factor interaction effect of temperature vs. rotor speed vs. TTQ (x3 vs. x 2 vs. x3) on PVC compounds (-) tetrahedron) ...................Diagrammatic representation of the observed yields (heat of fusion: mJ/mg) and the standard ordering of experiments of PVC/5phrCPE compounds .........................Diagrammatic representation of the observed yields (heat of fusion: mJ/mg) and the standard ordering of experiments of PVC/5phrCPE/0.3phrOPE compounds .............

xiv

148

149

150

153

155

163

165

166

169

170

172

Diagrammatic representation of the observed yields (entrance pressure drop: MPa) and the standard ordering of experiments of PVC compounds .......................................

Diagrammatic representation of the observed yields (entrance pressure drop: MPa) and the standard ordering of experiments of PVC/ 5phrCPE compounds ..............................Diagrammatic representation of the observed yields (entrance pressure drop) and the standard ordering of experiments of PVC/ 5phrCPE/0.3phrOPE compounds ..................Chemical structures of PVC, CPE, OPE, and calcium stearate ............. ..................A model of PVC lubrication showing metal lubrication and lubrication between PVC microparticle flow units [Rabinovitch et al., 1984] ......................................A postulated lubrication mechanism, showing metal lubrication and lubrication between PVC microparticle flow units, of PVC/OPE compounds ........................................A postulated lubrication mechanism, showing metal lubrication and lubrication between PVC microparticle flow units, of PVC/CPE compounds ........................................A postulated lubrication mechanism, showing metal lubrication and lubrication between PVC microparticle flow units, of PVC/CPE/OPE compounds ........................................A postulated lubrication mechanism, showing metal lubrication and lubrication between PVC microparticle flow units, of PVC compounds ........................................Fusion time curves, varying with the phr of CPE, of PVC/CPE and PVC/CPE/0.3phrOPE compounds .......................................Changes of fusion temperature (+) and fusion torque (□) of, varying with the phr of CPE, PVC/CPE/0.3phrOPE compounds ..................

xv

175

177

178

179

180

181

182

183

185

186

Fusion time curves, varying with the phr of OPE, of PVC/OPE and PVC/5phrCPE/OPE compounds .......................................

Changes of fusion temperature (+) and fusion torque (□) of, varying with the phr of OPE, PVC/5phrCPE/0PE compounds ....................Surface morphology of PVC prepared in a Haake Torque Rheometer at temperature =190°C, rotor speed = 60 rpm, and TTQ = 1 kg-m-min .........................................Surface morphology of PVC/OPE prepared in a Haake Torque Rheometer at temperature = 190°C, rotor speed = 60 rpm, and TTQ = 1 kg-m-min .........................................Surface morphology of PVC/CPE prepared in a Haake Torque Rheometer at temperature = 190°C, rotor speed = 60 rpm, and TTQ = 1 kg-m-min .........................................Surface morphology of PVC/CPE/OPE prepared in a Haake Torque Rheometer at temperature = 190°C, rotor speed = 60 rpm, and TTQ = 1 kg-m-min .........................................Surface morphology of PVC/OPE prepared in a Haake Torque Rheometer at temperature = 175°C, rotor speed = 60 rpm, and blending time = 15 minutes ..............................Surface morphology of PVC/OPE prepared in a Haake Torque Rheometer at temperature =17 5°C, rotor speed = 100 rpm, and blending time = 15 minutes ..............................Surface morphology of PVC prepared in a Haake Torque Rheometer at temperature =175°C, rotor speed = 60 rpm, and TTQ = 1 kg-m-min .........................................Surface morphology of PVC/CPE prepared in a Haake Torque Rheometer at temperature = 175°C, rotor speed = 60 rpm, and TTQ = 1 kg-m-min .........................................

xvi

6.19 Surface morphology of PVC/CPE/OPE prepared in a Haake Torque Rheometer at temperature =175°C, rotor speed = 60 rpm, and TTQ = 1kg-m-min ......................................... 187

6.20 Surface morphology of PVC prepared in a Haake Torque Rheometer at temperature =160°C, rotor speed = 20 rpm, and TTQ = 15kg-m-min ......................................... 188

6.21 Surface morphology of PVC/OPE prepared ina Haake Torque Rheometer at temperature =160°C, rotor speed = 20 rpm, and TTQ = 15kg-m-min ......................................... 189

6.22 Surface morphology of PVC/CPE prepared ina Haake Torque Rheometer at temperature =160°C, rotor speed = 20 rpm, and TTQ = 15k g - m - m i n ......................................... 190

6.23 Surface morphology of PVC/CPE/OPE prepared in a Haake Torque Rheometer at temperature =160°C, rotor speed = 20 rpm, and TTQ = 15kg-m-min ......................................... 191

xvii

ABBREVIATIONS

ACER Advanced Capillary Extrusion RheometerCCD Central Composite DesignCPE Chlorinated PolyethyleneDBU 1,8-Diazabicyclo [5,4,0] undecene-7DSC Differential Scanning CalorimetryHDPE High Density PolyethyleneMI Melt IndexNF No FusionOPE Oxidized Polyethylenephr per hundred resinPVC Polyvinyl Chloriderpm revolution per minute (min'1)SEM Scanning Electron MicroscopyS-TEM Scanning-Transmission Electron MicroscopyTEM Transmission Electron MicroscopyT e m p . Temperature (°C)Tg Glass Transition Temperature (°C)TTQ Totalized Torque (kg-m-min)Wu Unit work (J)

xviii

NOMENCLATURE

A : Bowl surface area, (m2) .Cp : Heat capacity of material, (J/kg°C) .D : Diameter of a Capillary (mm)Et : Amount of heat transferred by time t, (J).G : Torque, (kg-m).H0 : Initial enthalpy of material, (J).Ht : Enthalpy of material at time t, (J).AHa : The endothermy energy of peak A (mJ/mg)IZOD : Impact strength (ft-lbs/in)L : Length of a capillary (mm)m : Charge size, (kg).N : Viscous lossP : The extrusion pressure for processed PVC sample

(MPa)P* : The extrusion pressure for unprocessed PVC sample

(MPa)P0 : Pressure upon charging, (N/m2) .P220 : The extrusion pressure for PVC sample, processed at

a constant heating rate of 10°C/min, rotor speed= 60rpm, and discharged at 220°C (Considered as 100% fusion level (kg/cm2) ) [Krzewki et a l . , 1981] .

P20o, 15: The extrusion pressure for PVC sample, processedat temperature=2 0 0°C, rotor speed=60rpm, and TTQ=15 kg-m-min (Considered as 100% fusion level (MPa)) Elastic entrance pressure drop (MPa)Total entrance pressure drop (MPa)Viscous entrance pressure drop (MPa)Rate of heat transfer, (J/min).Shear strain resulted from elasticity Time, (min).Temperature before charging, (°C) .Temperature of PVC Melt, (°C) .Temperature of Bowl, (°C) .Overall heat transfer coefficient, (J/°Cm2min) .Bowl volume, (m3) .Amount of work generated by time t, (J).Represent the temperature effect in a 23 factorialexperimental design

X2 : Represent the rotor speed effect in a 23 factorialexperimental design

X3 : Represent the TTQ effect in a 23 factorialexperimental design

r : Wall shear stress

alrelaA Pent APvioQsRtT0TmeltTbowlUVwtX,

xix

ABSTRACT

Morphological characteristics of PVC, PVC/CPE, and PVC/CPE/OPE compounds were examined by scanning electron microscopy, transmission electron microscopy, and scanning- transmission electron microscopy. A correlation has also been established between the morphological analyses and the IZOD impact strength of PVC compounds. PVC compounds were prepared in a Haake Torque Rheometer using various blending conditions. The fusion level of processed compounds was evaluated by DSC thermal analysis and capillary rheological analysis, based on the heat of fusion and entrance pressure drop, respectively. S-shaped fusion curves were obtained. The effects of temperature, rotor speed, and totalized torque on the fusion level of PVC compounds were discussed.

Also, the influence of temperature and rotor speed on the fusion time was studied. The morphological changes of the various fusion processes were characterized by SEM.

The compounding energy was calculated by using an energy balance in an internal mixer [Pedersen, 1991]. The compounding energy was then correlated with the fusion level, and the differences between experimental and theoretical results were compared and discussed.

A 23 factorial experimental design was applied to study the main, two-factor interaction, and three-factor interaction effects of temperature, rotor speed, and

xx

totalized torque (TTQ) on the fusion of PVC compounds. The fusion characteristics of these PVC compounds (fusion time, fusion torque, and fusion temperature) were studied, and the surface morphological changes of these PVC compounds were also examined by SEM. The fusion mechanisms of these PVC compounds have also been postulated in this study.

xx i

CHAPTER 1INTRODUCTION

Poly(vinyl chloride), PVC, was first synthesized and characterized more than 120 years ago but, due to the fact that its poor thermal stability makes processing difficult, it was not until about 1930 that commercial production of PVC be g a n .

The suspension, mass, emulsion, and microsuspension polymerization processes are the main commercial routes to producing PVC in industrial today. In this research, PVC(supplied by the DOW Chemical Company) was produced via the suspension polymerization process. After the suspensionpolymerization of vinyl chloride monomers, various levels of morphology exist in PVC powder [Faulkner, 1975 and Hori, 1969]. These are illustrated for suspension PVC in Figure1.1 [Butters, 1982]. Powder particles, which are stage III particles and are visible to the naked eye, are known as grains. These are irregular in shape and are about 100-150 ^m in diameter. Each grain consists of many primary particles, which are stage II particles and are about 1-2 fim in diameter. The primary particles are loosely packed together, giving the grain its porous nature. Each primary particle is made up of still smaller structures, known as

1

2

100-150 -urn

Stage III Powder Particle

Grain

Stage II

Microparticle

100 -3 0 0 A

/ Stage I \

Submicroparticle'

40 A

Crystallite Structures

(Approx 10% Crystallinity)

1 0 0 -3 0 0 A

Figure 1.1 A schematic model of a suspension PVC particle [Butters, 1982].

stage I particles (domain), which are approximately 100-300 A in diameter and contain about 10 % crystallinity.

In order to overcome its poor thermal stability and photochemical degradation, researchers have developed suitable stabilizer systems [Titow, 1990], heat stabilizers (e.g. lead compounds, organotin compounds, and other metal compounds) and light stabilizers (e.g. phenyl salicylate, oxalic anilide, and phenyl formamidine) to solve these problems, and PVC is now one of the world's leading synthetic polymers.

Currently, PVC products are widely used in automobile parts, in components for houses and buildings, and in packaging for everything from food to electronic parts. To perform such diverse functions, various additives for PVC have been found and used.

Plasticizers are materials which have low or negligible volatility and are suitably compatible with the polymer in which they are incorporated. The compatibility between PVC and plasticizers is very important in predicting and understanding the behavior of a PVC compound during processing of the plasticizers [Patel, 1986, and Luis, 1990] . While processing, the principle role of the plasticizer is to be compatible with the ultimate product in varying degrees, depending on the plasticizers' nature and proportion. Diesters of phthalic anhydride (with C4-Ca

alcohols), triaryl phosphates, alkyl esters of dibasic alkyl acids and alkyl trimellitate esters are commonly used as plasticisers in PVC processing [Titow, 1990].

Process aids are the materials which have the specific function of promoting the fusion of rigid PVC compounds. These materials are often polymeric, and their mode of action is complex. Generally, they melt much easier than PVC at lower temperatures so that a viscous mix of process aid melt and PVC particles is formed. During processing, this flow is not just particle slippage, but it is also a melt flow system. Therefore an effective degree of fusion can occur earlier than would be the case in the absence of a process aid [Butters, 1982].

Fillers, which are broadly defined as solid, particulate or fibrous materials that are chemically inert, are incorporated into the PVC composition in order to modify

the properties or reduce the material cost. Titanium dioxide (Ti02) and calcium carbonate (CaC03) are the most commonly

used fillers in PVC processing.The main function of lubricants is to improve the flow

characteristics of PVC compositions, under the heat and shear existing in the processing machinery, so that processing can be made easier. Depending on the lubricating action and effects, lubricants can be divided into, two categories: internal lubricants and external lubricants.

Internal lubricants lower intermolecular friction of PVC when the PVC composition is being hot-sheared, or fused into a melt. External lubricants reduce the friction and sticking between the hot PVC composition and the working surfaces of the processing machinery. Depending upon the proportion added and the processing conditions, the lubricants themselves cannot be classified with absolute accuracy because many lubricants can perform both functions in varying degrees. Lubrication is the most important factor in influencing the fusion of PVC during processing, which may

determine the physical properties of the final PVC products. Amides, esters, metallic stearates, waxes and acids are commonly used as lubricants in PVC processing. In this research, the influence of two lubricants, calcium stearate and oxidized polyethylene (OPE), on the fusion mechanism of PVC has been studied.

Impact modifiers are additives that are usually elastomeric and are added into the PVC composition to improve the impact resistance of the finished product to stress. The most commonly used impact modifiers in PVC include methacrylate-butadiene-styrene, methacrylate- acrylonitrile-butadiene-styrene, acrylonitrile-butadiene- styrene, chlorinated polyethylene (CPE), ethylene-vinyl acetate, and acrylic [Titow, 1990]. The chlorinated polyethylenes, which are used as impact modifiers in PVC,

are produced by the chlorination of high density polyethylene (HDPE). According to the results reported by Chang, et a l . [Chang, 1988], chlorination in solution gives the greatest uniformity of distribution of chlorine atoms in the HDPE polymer chains. The chlorine distribution and chlorine content of CPE are major factors in the mixing of CPE with PVC, which may influence the mechanical properties of the final product. CPE with 36 percent chlorine is usually used as an impact modifier for PVC, and is the optimum composition for impact, processing, and strength

[Siegmann, 1984] . In this work, the effects of CPE with 36 percent chlorine, produced by the solution chlorination of HDPE, being used as impact modifier in PVC compounds will be

studied.In chapter two, the morphological analyses of PVC,

PVC/CPE, and PVC/CPE/OPE compounds were examined by scanning electron microscopy (SEM), transmission electron microscopy (TEM) , and scanning-transmission electron microscopy (S-TEM) are discussed. Morphological analysis is the optimum way to characterize the particle distribution and size of CPE in PVC/CPE and PVC/CPE/OPE compounds. Rheological and physical properties may be improved through blend morphological analysis. A correlation has been established between the morphological analyses and the IZOD impact strength of PVC

compounds.

In chapter three, PVC compounds prepared in a Haake Torque Rheometer using various blending conditions are discussed. The fusion level of processed compounds was evaluated by differential scanning calorimetry (DSC) and a capillary rheometer, based on the heat of fusion and entrance pressure drop, respectively. S-shaped fusion curves were obtained. The effects of temperature, rotor speed, and totalized torque on the fusion level of PVC compounds were discussed. Also, the influence oftemperature and rotor speed on the fusion time was studied. The morphological changes of the various fusion processes were characterized by SEM.

The major contributions to the compounding energy of the PVC sample in the Haake Torque Rheometer are attributed to: starting temperature, rotor speed, and totalized torque. In chapter four, the compounding energy was calculated by using an energy balance in the internal mixer [Pedersen,1991] . The fusion levels of the PVC samples, which wereprepared in the Haake Torque Rheometer at various blending conditions (starting temperature = variable, rotor speed = 60 revolution per minute (rpm), and totalized torque = 10 kg-m-min) , were determined by DSC and a capillary rheometer. The compounding energy was then correlated with the fusion level, and the differences between experimental and theoretical results were compared and discussed.

In chapter five, PVC, PVC/CPE, and PVC/CPE/OPE compounds were prepared in the Haake Torque Rheometer at various temperatures, rotor speeds, and totalized torques. A 23 factorial experimental design was applied to study the main, two-factor interaction, and three-factor interaction effects of temperature, rotor speed, and totalized torque (TTQ) on the heat of fusion of PVC compounds, which were examined by using both DSC and a capillary rheometer. The sequence of the main effects on the heat of fusion or entrance pressure drop of PVC compounds, in ascending order, is temperature < rotor speed < totalized torque. The sequence of the two-factor interaction effects on the heat of fusion entrance pressure drop of PVC compounds, in ascending order, is temperature vs. rotor speed < temperature vs. totalized torque < rotor speed vs. totalized torque. The three-factor interaction effect is not significantly related to the heat of fusion or entrance pressure drop of PVC compounds.

In chapter six, PVC, PVC/CPE, PVC/OPE, and PVC/CPE/OPE compounds were prepared in a Haake Torque Rheometer at various temperatures, rotor speeds, and totalized torques (TTQ) . The fusion characteristics of these PVC compounds (fusion time, fusion torque, and fusion temperature) were studied, and the surface morphological changes of these PVC

compounds were also examined by SEM. The fusion mechanisms of these PVC compounds have also been postulated.

Through the morphological analyses and postulated fusion mechanisms more detailed information can be obtained, which may be related to the mechanical properties of PVC compounds. By studying the fusion characteristics of PVC compounds, the formulations which are proper for the processing of PVC compounds can be determined. From the results of a 23 experimental design the knowledge of the important sequence of three major compounding parameter for PVC compounds is obtained. This knowledge may be applied to the extrusion process in order to optimize the mechanical properties of PVC compounds in different demands.

CHAPTER 2MORPHOLOGICAL ANALYSES OF PVC, PVC/CPE, AND PVC/CPE/OPE

COMPOUNDS

2.1 Basic Principles of Electron MicroscopyIn this chapter three kinds of electron microscope

techniques, TEM, SEM, and S-TEM, have been applied to examined the morphology of PVC compounds. Figure 2.1 shows two basic types of electron microscopes [Bozzola et a l . ,1992] . Figure 2.2 illustrates the log scale (middle) of the range of resolving power of various magnifying tools (left) and the structures that they are capable of resolving (right) [Bozzola et a l ., 1992] .

The TEM projects electrons through a very thin slice of specimen to produce a two-dimensional image on a phosphorescent screen. The brightness of a particular area of the image is proportional to the number of electrons that are transmitted through the specimen. The SEM produces a three dimensional image. This microscope applies a 2 to 3 nm spot of electrons that scans the surface of the specimen to generate secondary electrons from the specimen that are then detected by a sensor. The S-TEM has features of both TEM and SEM. The S-TEM uses a scanning beam of electrons to penetrate thin specimens. The sensor of S-TEM detects both primary and secondary electrons. This technique is more

10

11

Specim en

ViewingScreen

TEM SEM

Monitor

ElectronBeam

Specim en

★Signal

From Specim en Transduced

Two Basic Types of Microscopy

Figure 2.1 Basic differences between Transmission (TEM) and Scanning (SEM) Electron Microscopes [Bozzola et a l ., 1992] .

12

UnaidedHuman

Eye

0.2 mm *

LightMicroscopy

0.2 pm - - i

1 meter

100 mm -

10 mm -

1 mm

“ 100 pm -

10 pm -

1 pm -

TOlThrfr--

TransmissionElectron 10 nm "

Microscopy1 nm -

0.2 nm —i ------T------Scanning^”! 0-1 nmTunneling

Microscopy

• Sewing Needle Diameter• Razor Blade Edge Thickness

•Human Hair Diameter

^Most Cells

3 f- Bacteria

- Viruses

— Macromolecules

~ j- Atoms

Figure 2.2 Log scale (middle) of the range of resolving power of various magnifying tools (left), and the structures that they are capable of resolving (right) [Bozzola et a l ., 1992],

13sensitive to the specimens with two components of different densities.

2.2 Compatibility between PVC and CPEDue to different crystallinity and chlorine content of

CPE and PVC, it is believed that CPE's (with various Cl content) are incompatible or less compatible with PVC in spite of their limited difference in chlorine composition [Doube et a l ., 1979]. DSC thermal analysis is a powerful technique used to discriminate the compatibility between two components of a polymer blend. For a polymer blend containing two compatible components, A and B, only one glass transition temperature (Tg) will exist between A's Tg and B's Tg in DSC analysis. In this chapter, DSC thermal analysis has been used to determine the Tg of the PVC

compound.

2.3 Morphology of PVC CompoundsThe morphology of a polymer blend has a significant

effect upon the impact behavior of an alloy. Based on compounding conditions, Siegmann [1984] reported PVC/CPE blends have shown that at concentrations less than 13 weight percent CPE, PVC is in the continuous phase, while CPE is in the noncontinuous phase. The CPE particles range between 100 and 400 nm in size. Increasing the CPE concentration

14above 13 weight percent reverses the phase distribution. A significant increase in impact properties is also observed in CPE levels above thirteen weight percent.

Several variables may effect the observed morphology and thus the effectiveness of PVC/CPE blends. The resulting blend morphology depends upon the mixing conditions used during the blending process. The molecular weight, crystallinity, chlorine content, and chlorine distribution of the CPE, as well as the ultimate CPE loading in the PVC,

have a significant effect on the mixing requirements and the resulting blend morphology. To facilitate studying each of these effects individually, a method should be available

which illustrates the morphology of the blend.Earlier researchers [Fleischner et a l . , 1977] used a

staining procedure and a TEM technique to examine the

boundary morphology of PVC/CPE blends. This method requires very time consuming chemical staining reactions. A less time consuming method has been developed for PVC/CPE blends [Blackson et a l . , 1988] . This newer method does not require sample staining and also provides contrast through the use of S-TEM [Blackson et a l . , 1988] . The fractured surfacemorphology of PVC compounds was also examined on SEM. In this chapter, the results of SEM, TEM, and S-TEM of PVC compounds will be discussed.

152.4 Experimental Methods

2.4.1 Preparation of PVC CompoundsThe materials used in this study are PVC (suspension

PVC, containing 5% additives and stabilizers, with or without fillers, Ti02 and CaC03, and 0.2 phr OPE) and various CP E s . All samples were supplied by the Dow Chemical Company.

The various blends of PVC/CPE, 0, 5, 10, 15, and 2 0parts per hundred resin (phr), were prepared by dry mixing the components and then melt blending them at 180°C until a 10 kg-m-min totalized torque was reached in a Haake Rheomix 600 batch mixer with a Haake Rheocord 90 drive. The rotor speed was 60 rpm. The samples discharged from the Haake

Rheometer were compression molded under 2 0 tons of pressure for 2 minutes at 200°C and subsequently cooled for 5 minutes at room temperature at the same pressure. These compression molded samples were used for morphological examination.

2.4.2 DSC Thermal AnalysisSamples, approximately 10 mg, were tested by a SEIKO

220C Automatic Cooling Differential Scanning Calorimeter. The thermal analysis processes were: first, cool the sample down to -50°C; second, heat the sample from -50°C to 150°C and hold the temperature constant for 5 minutes; third, cool

16the sample down again from 150°C to -50°C and hold the temperature constant for 10 minutes; and finally, heat the sample from -50°C to 270°C. The data from the final step was saved and plotted. Zero, 5, 10, 15, and 20 phr CPE PVC/CPE blends (containing 10% Ti02, 5% CaC03, 5% additives andstabilizers) and CPE were examined.

2.4.3 Morphological AnalysisSamples (see Table 2.1) were supplied by the Dow

Chemical Company. Zero, 5, 10, 15, and 2 0 phr PVC/CPEblends (with or without fillers and 2 phr OPE) were analyzed. Two different directions, perpendicular and parallel to the primary axis of the bar, were examined for each sample. In order to establish the relationship between the IZOD impact value and the shape of the CPE particles in PVC/CPE blends, the third direction was also examined, i.e. a flat cut across the surface of the bar for both the highest and lowest IZOD impact value samples on S-TEM (Figure 2.3).

2.4.3.1 SEMSamples, immersed into liquid nitrogen for 45 seconds,

fractured and coated with 200 A gold-palladium film, were examined on Cambridge S-260 SEM at the center of the fractured surfaces.

17

Table 2.1 PVC/CPE blend samples* analyzed by SEM, TEM, and S-TEM.

With fillers No fillers No fillers, but with 0.3phr OPE

Ophr 3615P CPE lOphr CM0836 CPE (MI**=. 1)

5 phr CM674 (25% Cl) CPE

5phr 3615P CPE lOphr XU63004 CPE(MI=.4)

lOphr CM674 (25% Cl) CPE

lOphr 3615P CPE lOphr 3623A CPE(MI=1.0)

lOphr 9305 (52% Cl) CPE

15phr 3615P CPE lOphr 3611 CPE(MI=.6)

5phr XU63 004 CPE(MI=.4)

2Ophr 3615P CPE lOphr 3615P CPE(MI=.2)

lOphr XU63004 CPE(MI=.4)

lOphr 3615P CPE(MI=.2)

5phr 3615P CPE(MI=.2)

lOphr 3615P CPE(MI=.2)

lOphr 3615P CPE(MI=.2)

lOphr CM674 (25% Cl) CPElOphr CM552 (36% Cl) CPElOphr CM631 (42% Cl) CPElOphr 9305 (52% Cl) CPE

lOphr 3615P #1 (Largest) CPElOphr 3615P #2 (Smallest) CPElOphr 3615P #3 (Middle) CPE

*: Supplied by the Dow Chemical Company.**: MI=Melt Index.

18

Perpendicular

<Parallel

Flat Cut

Figure 2.3 Three different directions of samples.

192.4.3.2 TEM

Specimens for the TEM were prepared by the procedure of Fleischer, et a l . [1977], Selective staining of the CPEphase is accomplished by a two-stage procedure involving selective dehydrochlorination of the CPE to form double bonds, followed by a reaction with 0s04 (see Figure 2.4)

Small rectangular blocks, perpendicular and parallel directions with respect to the external surface of the molded sample, approximately 5 by 8 by 12 mm were cut from the moldings, trimmed and immersed in 1,8-diazabicyclo- (5,4, 0)-undecene-7 (DBU) for 2 days at 0°C, then rinsed with 2N HC1 and stained for 14 days in 1% 0s04 aqueous solution. Ultrathin sections (approx 80 nm) were cut from the stained blocks by ultramicrotome, mounted on grids with carbon coating, and examined under 80 KV accelerating voltage in the JEOL 100CXII TEMSCAN.

2.4.3.3 S-TEMSpecimens for the S-TEM imaging needed no further

chemical treatments. Small rectangular blocks,perpendicular and parallel directions with respect to the external surface of the molded sample, about 5 by 8 by 12 mm were cut from the moldings. Ultrathin sections (approx 80 nm) were cut from the rectangular blocks by ultramicrotome, mounted on grids with carbon coating, and examined under 100

20

(1)wvCH„-CH-CH -CH o tCl 2 Days ^

CPE DBU

v w C H :C H -C H ,-C H ,v w + DBUHCI

14 Days 1 % 0 s 0 4M /

VW CH -C H -C H .- C H - v w/ \ o o\ Q s /

o o

(2) CH„- CH - CH - CH VW ̂ I C. IClPVC

Cl+ CfsjX:N

DBU

o-c^ 2 Days '

DBU : 1,8 - diazabicyclo [5,4,0] undecene - 7

Figure 2.4 Staining reaction sequence [Fleischner et a l ., 1977].

21KV accelerating voltage in the JEOL 100CXII TEMSCAN. In

order to avoid charging of the ultrathin sections in the S- TEM, it is necessary to focus on one area of samples then move to other area of samples to take a picture quickly

[Blackson, 1988].

2.5 Results and Discussion

2.5.1 DSC Analysis of CPE and PVC CompoundsThe Tg and DSC curves of CPE and PVC/CPE blends are

listed in Table 2.2 and Figure 2.5/ respectively. The Tg of CPE and PVC ranges from -18°C to -20°C and from 82° to 83°C, respectively. Only the Tg of PVC (approximately 82°C-83°C)

was noted in PVC/CPE blend samples.From Figure 2.5, PVC's Tg was easily observed, but

CPE's Tg was not observed until a slight signal appeared in the DSC curve of 20 phr PVC/CPE blend. The reason for this may be due to low CPE loading in the PVC compound.

2.5.2 SEM AnalysisFigures 2.6, 2.7, and 2.8 show the SEM imagings of the

perpendicular direction of 0, 10, and 2 0 phr CPE PVC/CPEblends, respectively. Figures 2.9, 2.10, and 2.11 show the SEM imagings of the parallel direction of 0, 10, and 20 phr CPE PVC/CPE blends, respectively.

22

Table 2.2 Tg of CPE and PVC/CPE blends.

CPE and PVC/CPE blendsGlass transition temperature (Tg)

3615P CPE -18,5°C0 phr 3615P CPE 82.1°C5 phr 3615P CPE 82.1°C

10 phr 3615P CPE 82 .1°C15 phr 3615P CPE 82.6°C20 phr 3615P CPE 83.1°C

23

0 phr CPE

5 phr CPE

exo 10 phr CPE

15 phr CPEendo

20 phr CPE

Pure CPE

196 228 260164-28 4 68 100 132-60 36

Temperature °C

Figure 2.5 DSC traces of CPE and PVC/CPE blends.

24

Figure 2.6 0 phr 3615P CPE in PVC/CPE blend coatedwith 200 A of gold-palladium and examinedby SEM (perpendicular).

25

Figure 2.7 10 phr 3615P CPE in PVC/CPE blend coatedwith 200 A of gold-palladium and examinedby SEM (perpendicular).

26

Figure 2.8 20 phr 3615P CPE in PVC/CPE blend coatedwith 200 A of gold-palladium and examinedby SEM (perpendicular).

27

Figure 2.9 0 phr 3615P CPE in PVC/CPE blend coatedwith 200 A of gold-palladium and examinedby SEM (parallel).

28

H u ? ? ?

. / : ' . - < * - i

* ’ :.*».*(■ *

Figure 2.10 10 phr 3615P CPE in PVC/CPE blend coatedwith 200 A of gold-palladium and examinedby SEM (parallel).

29

V/ffi

Figure 2.11 20 phr 3615P CPE in PVC/CPE blend coatedwith 200 A of gold-palladium and examinedby SEM (parallel).

30No CPE particles are found in the SEM micrographs

because the SEM equipment is designed for surface investigations of materials. Due to this fact, one cannot observe the phase boundaries between PVC and CPE using SEM. The SEM micrographs of the perpendicular and parallel directions show no difference in PVC/CPE blends with fillers. The SEM images of 0, 10, and 20 phr PVC/CPE blends display similar results in spite of their different CPE content. The white particles observed in SEM micrographs

are additives, stabilizers, or broken residues.

2.5.3 TEM AnalysisFigures 2.12, 2.13, and 2.14 show the TEM imagings of

the perpendicular direction of 0, 10, and 20 phr PVC/CPEblends with staining treatment, respectively. Figures 2.15, 2.16, and 2.17 show the TEM imagings of the parallel direction of 0, 10, and 2 0 phr PVC/CPE blends with staining treatment, respectively.

No CPE particles are found in Figures 2.12 and 2.15 because this sample is a 0 phr PVC/CPE blend. The dark particles are additives and stabilizers. The gray particles in Figures 2.13, 2.14, 2.16, and 2.17 are CPE particles,which are seen because of the staining effect. More CPE particles are found in Figures 2.14 and 2.17 (20 phr PVC/CPE

31

i

i

*

9

%

• •% t1.0 pm

Figure 2 .12 0 phr 3615P CPE in PVC/CPE blend treated bychemical reaction and examined by TEM(perpendicular).

32

A* '

S \tr

&

w

* M

* CPE

«Ti*W

%

®

*«*. Jr 4

* ^ o Hm

Figure 2.13 10 phr 3615P CPE in PVC/CPE blend treated by chemical reaction and examined by TEM (perpendicular).

2.14 20 phr 3615P CPE in PVC/CPE blend treated by chemical reaction and examined by TEM (perpendicular).

34

Figure 2 .15 0 phr 3615P CPE in PVC/CPE blend treated bychemical reaction and examined by TEM(parallel).

35

f& ’/ '

ijiV

%

a

' ̂

*#

*\<J .€*

n

0

M>vm

A'~££ }

0*

Figure 2.16 10 phr 3615P CPE in PVC/CPE blend treatedby chemical reaction and examined by TEM(parallel).

36

Figure 2.17 20 phr 3615P CPE in PVC/CPE blend treatedby chemical reaction and examined by TEM(parallel).

37blend) than those in Figures 2.13 and 2.16 (10 phr PVC/CPE blend).

Figure 2.18 shows the parallel direction TEM analysis of a 10 phr PVC/CPE blend without staining treatment. No CPE particles are found in Figure 2.18 (same area in Figure 2.24). Figure 2.19 shows the perpendicular direction TEM analysis of 10 phr CPE (MI = 0.4) in the PVC compound (containing 2 phr OPE) with staining treatment (same area in

Figure 2.26) .Without the impact modifier, CPE3615, the TEM imagings

of PVC, treated by the selective staining procedure, (perpendicular and parallel directions) are shown in Figures 2.12 and 2.15, respectively. No staining effects are observed. The dark particles are additives and stabilizers,

and the dark lines are knife marks from the microtome.With the selective staining procedure, the

perpendicular and parallel directions of 10 phr CPE and 20 phr CPE blends exhibit gray particles of stained CPE dispersed in an unstained PVC matrix (Figures 2.13, 2.14,2.16 and 2.17). For a 10 phr CPE blend, the TEM imaging of the perpendicular direction (Figure 2.13) is similar to that of the parallel direction (Figure 2.16). Most of the CPE particles are individual. The size of CPE particles ranges between 3 0 and 3 00 nm.

For a 2 0 phr CPE blend, the TEM imaging of theperpendicular direction (Figure 2.14) is also similar to the parallel direction (Figure 2.17). Many CPE clusters exist in a 20 phr CPE blend. The size of individual CPE particles and CPE clusters range between 45 and 400 nm. In some areas of the 20 phr CPE blend, (Figures 2.14 and 2.17), theappearance of the CPE particles is similar to that in the 10 phr CPE blend (Figures 2.13 and 2.16), except that the particles are more numerous and larger in size. In otherareas, the CPE is observed in elongated forms that are

sometimes aggregates of discrete particles and sometimes continuous structures probably formed by the fusion of particles during processing [Siegmann et a l ., 1984] .Without the two stage chemical staining procedure to improve the contrast between PVC and CPE, no CPE particle is observed in the TEM imaging (Figure 2.18).

2.5.4 S-TEM AnalysisFigures 2.20, 2.21, and 2.22 show the S-TEM imagings of

the perpendicular direction of 0, 10, and 20 phr CPE PVC/CPE

blends without staining treatment, respectively. Figures2.23, 2.24, and 2.25 show the S-TEM imagings of the parallel direction of 0, 10, and 20 phr CPE PVC/CPE blends without staining treatment, respectively.

39

%

%*

% %

P

Figure 2.18 10 phr 3615P CPE in PVC/CPE blend withoutchemical treatment examined by TEM(parallel).

40

1 ^ -V.

.Jk *

V t,-: •% 1.0pm

Figure 2.19 10 phr CPE (MI=0.4) in PVC/CPE/OPE blendtreated by chemical reaction and examinedby TEM (perpendicular).

41

ft&&'*■

%

■

% V%

*<k

^ 1.0pm

Figure 2.2 0 0 phr 3 615P CPE in PVC/CPE blend withoutchemical treatment examined by S-TEM(perpendicular).

42

Figure 2.21 10 phr 3615P CPE in PVC/CPE blend withoutchemical treatment examined by S-TEM(perpendicular).

43

Figure 2.22 20 phr 3615P CPE in PVC/CPE blend withoutchemical treatment examined by S-TEM(perpendicular).

Figure 2.23 0 phr 3615P CPE in PVC/CPE blend withoutchemical treatment examined by S-TEM(parallel).

Figure 2.24 10 phr 3615P CPE in PVC/CPE blend without chemical treatment examined by S-TEM (parallel).

Figure 2.25 20 phr 3615P CPE in PVC/CPE blend withoutchemical treatment examined by S-TEM(parallel).

47No CPE particles are found in Figures 2.20 and 2.23,

because this sample is a 0 phr PVC/CPE blend. The dark particles are additives and stabilizers. The gray particles in Figures 2.21, 2.22, 2.24, and 2.25 are CPE particles.

More CPE particles are found in Figures 2.22 and 2.25 (20phr CPE PVC/CPE blend) than those in Figures 2.21 and 2.24 (10 phr CPE PVC/CPE blend).

Figure 2.24 has the same area as in Figure 2.18. More CPE particles are found in Figure 2.24 than in Figure 2.18. Figure 2.26 shows the perpendicular direction S-TEM analysis of 10 phr CPE (MI=0.4) in the PVC compound (containing 0.2 phr OPE) with staining treatment.

Figure 2.26 has the same area as in Figure 2.19. More CPE particles are found in Figure 2.26 than those in Figure 2.19. This may be due to the fact that the staining effect is not completely achieved. This may also be the major reason that more CPE particles can be seen in S-TEM than in

TEM.Figures 2.27 and 2.28 show the perpendicular direction

of S-TEM imagings of 10 phr CPE (MI=1.0, IZ0D=3.7) and 10phr CPE (MI = 0 .2, IZOD=18.6) in PVC compounds withoutstaining treatment, respectively. Figures 2.2 9 and 2.30 show the parallel direction of S-TEM imagings of 10 phr CPE (MI = 1 . 0, IZOD=3 . 7) and 10 phr CPE (MI = 0.2, IZOD=18.6) in PVC compounds without staining treatment, respectively.

Figure 2.26 10 phr CPE (MI=0.4) in PVC/CPE/OPE blendtreated by chemical reaction and examinedby S-TEM (perpendicular).

49

Figure 2.27 10 phr CPE (MI=1.0) in PVC/CPE blendwithout chemical treatment examined byS-TEM (perpendicular).

50

Figure 2.28 10 phr CPE (MI=0.2) in PVC/CPE blendwithout chemical treatment examined byS-TEM (perpendicular).

51

Figure 2.29 10 phr CPE (MI=1.0) in PVC/CPE blendwithout chemical treatment examined byS-TEM (parallel).

52

Figure 2.30 10 phr CPE (MI=0.2) in PVC/CPE blendwithout chemical treatment examined byS-TEM (parallel).

Figures 2.31 and 2.32 show the flat cut direction of S-TEM imagings of 10 phr CPE (MI=1.0, IZOD=3.7) and 10 phr CPE (MI=0.2, IZOD=18.6) in PVC compounds without stainingtreatment, respectively. Samples with lower IZOD impact values (i.e. brittle sample, IZOD=3.7), Figures 2.27, 2.29

and 2.31 display no significant differences among these three directions (i.e. CPE particles are not elongated very much in the parallel direction). On the other hand, samples with higher IZOD impact values, as seen in Figures 2.28, 2.30 and 2.32, display very significant differences (i.e. CPE particles are elongated significantly in the parallel direction). Figure 2.33 shows the three dimensional views of low and high IZOD impact samples.

In order to obtain the CPE particle size distributions in PVC compounds, copies of perpendicular imagings of the samples were made and a ruler was used to measure the CPE particle size. Figure 2.34 shows the CPE particle size distributions of low IZOD (3.7) and high IZOD samples (18.6) . Comparing these two curves, we have found that the CPE particle size distribution curve of low IZOD sample

shifts slightly to lower diameter particles, resulting in a smaller mean CPE particle diameter. The mean particle size of low IZOD sample is (52 + 5 nm) while the mean particle size of high IZOD sample is (67 ± 5 n m ) .

Figure 2.31 10 phr CPE (MI=1.0) in PVC/CPE blendwithout chemical treatment examined by S-TEM (flat cut).

55

Figure 2.32 10 phr CPE (MI=0.2) in PVC/CPE blendwithout chemical treatment examined by S-TEM (flat cut).

56

High Izod Low Izod

(Low Melt Index) (High Melt Index)

ParallelDirection

O

PerpendicularDirection

Flat Cut Direction

Figure 2.33 Three dimensional visions of low and high IZOD impact samples.

Freq

uenc

y (%

)57

C P E PARTICLE S IZE D ISTR IB U TIO N

40 -■ : IZOD = 18.6 x : IZOD = 3.7

20 -

1 0 -

0 20 40 60 80 100 120 140 160180 200 220 240Diameter (nm)

Figure 2.34 CPE particle size distribution curves of low IZOD impact sample (MI=1.0) and high IZOD impact sample (MI=0.2) (perpendicular).

Figure 2.35 and Table 2.3 illustrate the relationship among melt index, domain size and IZOD impacts value of five different PVC/CPE blends. The melt index is an index for the molecular weight of polymers. High molecular weight polymers have low melt indices, while low molecular weight

polymers have high melt indices. From Figure 2.35 and Table 2.3, it has been observed that the domain size of CPE particles in PVC/CPE blends increases with the IZOD impact value, but decreases as the melt index increases. This means that when the CPE, with a higher molecular weight, was added into the PVC compound, the result was a higher IZOD impact value product and a larger domain size.

Figures 2.20, 2.23, 2.21, 2.24, 2.22, and 2.25 show the S-TEM micrographs of perpendicular and parallel directions of 0, 10, and 20 phr CPE PVC/CPE blends without further

chemical treatments in the S-TEM, respectively. No CPE particles are observed in Figures 2.20 and 2.23. The dark particles are additives and stabilizers, similar to those in Figures 2.12 and 2.15.

The areas in Figure 2.18 are the same as in Figure2.24, but many gray particles, which are CPE particles, are observed in Figure 2.24 even though no chemical staining procedure was required. The reduced beam flux per area, present in S-TEM coupled with the ability to electronically

59

Table 2.3 Melt index vs. domain size and IZOD impact.

CPE (phr) MI IZOD (ft-lbs/in) Diameter (nm)10 0.1 17 . 8 67 ± 510 0 . 2 18 .6 67 ± 510 0.4 7 .1 62 ± 510 1. 0 3.7 52 ± 510 6.0 3 . 9 52 ± 5

10 ph

r Iz

od

60

* — 10 phr Izod Domain Size

19

17

15

13

65 =11 _>_L1 __9 ---\i i I i i i

• W - ----— i-

7

5 • h - M - i I i .L.Li.3

10.1

I

Figure 2.35 Melt index vs. domain size and Izod impact.

61increase contrast, are largely responsible for the success of this technique in contrasting the two phases [Blackson et a l ., 1988].

For a 10 phr CPE blend, the S-TEM imaging of theperpendicular direction (Figure 2.21) is similar to that of the parallel direction (Figure 2.24). Most of the CPE particles are individual, and the size of these particles ranges between 3 0 and 3 00 nm.

For a 20 phr CPE blend, the S-TEM imaging of the

perpendicular direction (Figure 2.13) is also similar to the parallel direction (Figure 2.16). Many CPE clusters exist in a 20 phr CPE blend. The size of individual CPE particles

and CPE clusters ranges between 45 and 400 nm.In some regions, the micrographs of the CPE particles

in Figures 2.22 and 2.25 (20 phr) are similar to those inFigures 2.21 and 2.24 (10 phr), except that the particlesare more numerous and larger in size. In most of theregions, however, the CPE domains are interconnected,elongated, and almost form a network between the PVC particles. These phenomena are more obvious than those in TEM imaging.

Comparing the S-TEM imagings of perpendicular and parallel directions of 10 phr (Figures 2.21 and 2.24) and 20 phr CPE blends (Figures 2.22 and 2.25) with those in TEM imagings (Figures 2.13, 2.16, 2.14, and 2.17), there are two

62important observations. One observation is that a large amount of CPE particles may not have been stained during the chemical staining reaction. In order to verify this phenomenon, the same area micrographs on both TEM (Figure 2.19) and S-TEM (Figure 2.26) of 10 phr CPE (MI=0.4) in a PVC/CPE blend with staining treatment were taken.

Comparing Figure 2.19 (TEM) with Figure 2.26 (S-TEM), it is noticed that many CPE particles are not stained completely and are not observed on the TEM micrograph. Three procedures: increasing the reaction temperature ofdehydrochlorination, increasing the 0s04 concentration or extending the staining time more than two weeks, may be able to improve the staining effect.

The other observation is that the phase boundaries between PVC and CPE are clearer in TEM analysis than those in S-TEM analysis. The reason for this is that 0s04 provides a very considerable contrast effect on the phase boundaries between PVC and CPE.

Samples with lower IZOD impact values (i.e. brittle sample, IZOD=3.7), Figures 2.27, 2.29, and 2.31, have nosignificant differences between the perpendicular and parallel directions (i.e. CPE particles are not elongated very much in the parallel direction) , on the other hand, samples with higher IZOD impact values, Figures 2.28, 2.30,

63and 2.32 have very significant differences (i.e. CPE particles are elongated very much in the parallel

direction).The reason for this is that the CPE (36 % chlorine)

used in this blend with a lower IZOD impact value was made from a lower molecular weight Polyethylene (PE) which has a faster relaxation than the CPE with a higher molecular weight. CPE particles with a faster relaxation time are easily elongated when the PVC/CPE blend is compression molded at 200°C, but recover to orignal shape quickly after compression molded. Therefore it results in a brittle blend (lower IZOD impact value). That is the reason for the IZOD impact value decreasing as the melt index increased in Figure 2.35 and Table 2.3.

Figure 2.34 shows the CPE particle size distributions

of low IZOD (3.7) and high IZOD samples (18.6). When comparing these two curves, it is noticed that the CPE particle size distribution curve of low IZOD samples shifts to lower diameter particles, resulting in the mean CPE particle diameter of a low IZOD sample (52 ± 5 nm) being lower than that of a high IZOD sample (67 ± 5 n m ) .

This may be due to the possibility that the faster relaxation time CPE particles are not subject to agglomerate during heating and fusing to a melt in the Haake Torque Rheometer as well as being compression molded at 2 00°C.

64This factor results in the domain size decreasing as the melt index increases as shown in Figure 2.35 and Table 2.3.

CHAPTER 3FUSION PROPERTIES OF PVC IN A HAAKE TORQUE RHEOMETER

3.1 Fusion of PVCThe suspension polymerization process is the principal

commercial route to producing PVC resins in industry today. In this study, PVC grain particles, supplied by the DOW Chemical Company, are produced via the suspension polymerization process. Various levels of morphology exist in suspension PVC powder [Faulkner, 1975 and Hori, 1969], and these are illustrated in Figure 1.1 [Butters, 1982] .

Powder particles, which are stage III particles and are visible to the naked eye, are known as grains. These are irregular in shape and about 100-150 ixm in diameter. Each grain consists of many microparticles, which are stage II particles, and are about 1-2 fim in diameter. The microparticles are loosely packed together causing the grain to be porous. Each microparticle is made up of still smaller structures, known as stage I particles (submicroparticles), approximately 100-300 A in diameter. These stage I particles are reported to be about 5-10 % crystallinity. In order to achieve good mechanical properties, grain boundaries must be eliminated, and the microparticles must be altered and compacted together. After significant interdiffusion, the boundaries of

65

66submicroparticles disappear, and a three-dimensional network of polymer chains is formed. This is referred to as the fusion, or gelation, of PVC [Gilbert, 1985 and Krzewki et al., 1981]. Figure 3.1 [Krzewki et a l ., 1981] illustrates three main patterns of fusion. Normally the fusion mechanism of PVC particles, processed either in an extruder or a batch mixer, is the combination of these three

patterns.Figure 3.2 [Benjamin, 1978] shows the relationship

between physical properties (such as tensile strength, elongation, impact strength, and creep-rupture) and the degree of fusion for an unplasticised PVC compound processed by extrusion. Optimum values of impact ductility and of modulus occur prior to 100 % fusion. Benjamin [1978]suggested that while the strength of the material increases monotonically with increasing degree of fusion, the material reaches optimum ductility and then becomes increasingly brittle due to higher entanglement of the 3-D network.

Gilbert et a l . [1983] worked on the fusion of PVCcompounds in twin screw extrusion. They confirmed that the main route for fusion was through the deformation of grains by a low shear force, which leads to the network formation.

67

PVC Resin Particles

NoBreakdown

SubmicroParticles

Micro-Particles

V//////U

ResinParticles

Compacted Submicro- Particles

Compacted CompactedInternally Fusing Internally Fusing Microparticles Resin particles

Particles Interdiffusion

Fused Material

Figure 3.1 A schematic illustration of the threepatterns of fusion [Krzewki et a l ., 1981]

68

TensileStrength

ModulusA

ImpactStrength

Ductility

CreepStrength

FatigueStrength

Physical Properties of PVC vs. % Fusion

M odulus

Im pact

Ductility

I I I I 100 ►% Fusion (Gelation)

Figure 3.2 Relationship between the degree offusion and Physical properties [Benjamin, 1978] .

69The fusion mechanisms of PVC are shown in Figure 3.3 [Gilbert, 1983] . The fusion of the PVC is extremely dependent upon both the shear and thermal history of the PVC compound in extrusion.

Batch mixers have been used to process PVC compounds for a long time due to their superior heat transfer, slightly better dispersion, relatively short mixing cycles, capabilities with large capacities, and flexibility when frequent product and formulation changes are encountered [Rapetski, 1993]. Krzewki et a l . [1981] reported, indetail, the influence of temperature and additives on the fusion level of PVC compounds in a temperature-programmed Brabender Plasticorder. The fusion of PVC compounds is highly dependent upon the additives and the rotor speed, as well as its thermal history in a batch mixer.

Pedersen [1991] reported an improved method for presenting torque rheometer data as it applies to rigid PVC extrusion compounds. He applied the three-dimensional torque rheometry mapping in order to provide the rigid PVC formulator with greater insights into the extrusion

equipment. Figure 3.4 [Pedersen, 1991] shows the standard fusion mechanism of rigid PVC in a torque rheometer. The intention of this study is to understand how temperature, rotor speed, and totalized torque affect the degree of

Grains

Heat +High Shear

Heat+Low Shear

Primaries

0 (>

Final Network

Figure 3.3 PVC fusion mechanisms in a extrusion equipment [Gilbert, 1983].

71

"Fused" melt of submicroparticies

Fusion of agglomerates and submicroparticies

Microparticles and aggolomerates slide over one another at minimum torque

Resin grains compacted

Temperature increases, torque decreases

Torque reaches maximum

Torque increases as interparticle fusion initiates and propagates

High torques cause some pulverization

Cold material compacted in bowl

Figure 3.4 Normal fusion mechanism of the rigid PVC in a torque rheometer [Pedersen, 1991] .

72fusion of PVC compounds in a Haake Torque Rheometer. Additionally, this study has revealed the relationship between the morphology and the degree of fusion.

3.2 Methods to Assess the Degree of Fusion of PVCIn order to obtain optimum mechanical properties, an

appropriate degree of fusion is needed. Several different fusion assessment methods have been well developed [Gilbert,

1985] .Tensile property measurement [Marshall et a l . , 1981],

solvent immersion tests [Benjamin, 1978], optical methods [Summers et a l . , 1981], capillary rheometry [Gonze, 1971,Lamberty, 1974, and Summers et a l . , 1986], and thermalanalysis [Gilbert et a l . , 1981, Teh et a l ., 1989, andObande, 1991] are commonly used. In this chapter, we have applied both a capillary rheometry, based on the entrance pressure drop through a zero-length die, and DSC thermal analysis, based on the heat of fusion, to determine the degree of fusion of PVC compounds.

3.2.1 Introduction to Capillary Rheological AnalysisChen et a l . [1992] applied capillary rheological

analysis to characterize the wall slip behavior of linear low density polyethylene. Han [1974] used both slit and capillary-die rheometry method to evaluate exit pressures of

73polymer melts. The procedure for determining the entrance pressure drop in a capillary rheometry, developed by Gonze [1971] and Lamberty [1974], has been widely reported upon in literature.

In this method, the capillary entrance pressure drop is measured using a zero length, or low length to diameter ratio, die and a relatively low extrusion temperature. The simple schematic diagram of a capillary rheometer is shown in Figure 3.5 [Nielsen, 1977].

The total entrance pressure drop (APent) necessary to extrude a material through a capillary with diameter D and length L can be expressed as:

A P ent = t* (2N+Sr) + x*4L/D (3.2-1)

The first term is the capillary entrance pressure drop,

and the second term originates from the viscous dissipation during flow through a capillary [Philippoff et a l . , 1958 and Bagley, 1961] .

The product of the wall shear stress (r) and a sum of two terms is the entrance pressure drop. The first term, N, is the viscous loss, and the second term, SR, is the elastic term, also called shear strain [Krzewki et a l ., 1981 andGray, 1978]. For viscoelastic fluids, the total entrance pressure drop (APent) can be contributed from two parts: the viscous entrance pressure drop (APvis) and the elastic entrance pressure drop (APela) [Han, 1976] .

Capillary Rheometer

P

*w/wMW/mmm

A

W«-D

Figure 3.5 A schematic diagram of a capillary rheometer [Nielsen, 1977].

75

(3.2-2)

The viscous entrance pressure drop (APvis) , is related to the development of the velocity profile near the entrance of the capillary and to the viscous dissipation due to the converging flow prior to entering the capillary. The elastic entrance pressure drop (APela) is the pressure drop which may be converted into elastic energy, some of which is recoverable due to the elastic nature of the melt [Krzewki et a l . , 1981] . The value of APvis is approximately 10 % of the entire pressure drop [Han, 1976 and 1971]. Bagley et a l . [1969] reported that the viscous term (N) has been shown

to be constant for many polymers. Therefore the entrance pressure drop may be considered to be due to the changes in the elasticity of the material under test. If so, then, for practical purposes, we can assume that

There are two methods commonly applied to determine the entrance pressure d r o p : first, Bagley plots, with decreasing length/diameter (L/D) ratio, the contribution of the capillary pressure drop to the total pressure drop decreases and becomes zero for a zero-length capillary. Plots of total pressure differences vs. L/D at a constant shear rate give straight lines. Extrapolation to zero L/D provides the

^ ^ e n t ela (3.2-3)

76pressure drop at the capillary entrance (see Figure 3.6 [Nielsen, 1977]).

The second method, zero-length capillary method, with a very short capillary (very small L/D ratio) is used to determine the pressure drop which is nearly identical to the entrance pressure drop. Therefore we can obtain the entrance pressure drop from a single measurement.

Krzewki et a l . [1981] prepared samples in atemperature-programmed Brabender Electronic Plasticorder, equipped with an electrically heated mixing head. PVC compounds were charged into the mixing head at 120°C. The stock temperature was increased at a constant rate of 10°C/min, and the rotor speed was 60 rpm. A charge weight of 60 g was used with all compounds. All runs were carried out under identical conditions, and the entire sample was removed when it reached 160, 180, 190, 200, 210, and 220°C. The melt elasticity associated with progressively increasing fusion were measured by the capillary entrance pressure drop at a temperature of 140°C and a constant shear rate of 6 sec'1 on the Sieglaff-McKelvey Rheometer equipped with a flat entry zero-length capillary. The weight of sample was always 1.7 g, and the preheating time was 10 minutes.

The standard fusion curves established by plotting the capillary extrusion pressure versus the processing

77

Bagley Correction

High Shear RateCL

Low Shear Rate

Pn at Low Shear Rate

403020100

L /D

Figure 3.6 Bagley correction for capillary rheometers [Nielsen, 1977] .

temperatures are shown in Figures 3.7 and 3.8 [Krzewki et a l ., 1981] represents the corresponding normalized fusioncurves showing the relative degree of fusion obtained by- processing to different temperatures. Krzewki et a l . [1981] calculated the relative degree of fusion for each compound, the extrusion pressure for the unprocessed compound (P*) and the compound which was processed to 220°C (P220) were taken as pressure corresponding to 0 and 100% of fusion, respectively. The percentage of fusion corresponding to extrusion pressure P was then performed from the relation.

Percent of fusion = (P-P*) / (P220-P*) X 100 (3.2-4)

An anomalous initial decrease in extrusion pressure has not been completely explained, but Krzewki and Collins [1981] suggested that this may be due to the breakdown of grains and the very beginning of the formation of a molecule

network across the primary particle and domain boundaries.Capillary rheometry has been widely used because of

using a sample large enough to eliminate local inhomogeneity in fusion level. This is the main advantage of capillary rheometry. The main disadvantage of this method is the necessity of getting a standard fusion curve for each compound processed in certain processing equipment.

Extru

sion

Pr

essu

re

, kg

/ cm

160

140

120

100

80

60

oUnprocessed Processing Temperature , C

Compound

Figure 3.7 Standard fusion curves for two PVC compounds [Krzewki et a l ., 1981] .

Rela

tive

Degr

ee

of Fu

sion

, %

100

80

60

40

20

0

-20

[ 160 180 200 220o

Unprocessed Processing Tem perature, CCompound

Figure 3.8 Normalized fusion curves for two compounds [Krzewki et a l ., 1981].

813.2.2 Introduction to DSC Thermal Analysis

Thermal analysis is completed by using a thermalanalyzer fitted with a DSC cell. Figure 3.9, establishedand interpreted by Gilbert et a l . [1981], shows the DSCthermograms of compression molded PVC compounds at various temperatures. For the DSC thermal analysis experiments, 10 mg samples were heated from room temperature to 240°C at 20°C/min using a DuPont 990 Thermal Analyzer, fitted with a DSC cell. They identified two endothermic peaks A and B as

shown in Figure 3.9. The melting of the crystallites is shown on peak B . When the molding temperature wasincreased, this peak decreases in size and shifts to ahigher temperature. They suggested that the size decrease was due to the melting of less perfect, or smaller, crystallites, and the temperature shift was caused by annealing of unmelted crystallites.

Endothermy A appears and increases in size when the molding temperature is increased. The energy of this endothermy (AHa) increases with processing temperature, and when plotted against temperature, produces an S-shaped curve (see Figure 3.10) similar to the fusion curve obtained by capillary rheometry. Therefore, Gilbert, et a l . [1981]suggested that the area of A was related to the degree of fusion. DSC thermal analysis has attracted attention because it provide a very convenient quantitative measure

82

170* C

OXuI<

180 'C

Ooz192

200*C

22Q*C

200t50( T em p era tu re *C )

1O0

Figure 3.9 DuPont DSC thermal analysis of unplasticised PVC compounds compression molded at different temperatures [Gilbert, 1981] .

83

87

6

5

4

3

2

1

0

160 180 200 220

Molding Temperature (°C)

Figure 3.10 Fusion curve measured by DSC thermalanalysis - heat of fusion vs. different molded temperatures [Gilbert, 1981].

84of fusion and uses a very small sample (approximately 10 mg) . Since the sample size analyzed is small, some additives may interfere with the final results. This is the main disadvantage of thermal analysis.

3.3 Experimental Method

3.3.1 Preparation of PVC CompoundsThe material used in this study is suspension PVC

masterbatch powder, containing 100 parts of PVC grain particle (M.W. <= 150,000), 1.5 parts of process aid (K120N) , 1.0 part wax (XL165) , 1.0 part Calcium Stearate, and 1.5parts heat stabilizer (T-137) . All samples were supplied by the Dow Chemical Company.

In order to determine the influence of temperature and totalized torque on the degree of fusion, PVC samples with various fusion levels were prepared in a Haake Rheomix 600 batch mixer with a Haake Rheocord 90 drive, equipped with an electrically heated mixing head. The noninterchangeable roller-type rotors were used. The rotor speed was set at 60 rpm, and the sample weight was always 65 g. Samples were charged into the mixer at 160, 170, 180, 190, and 200°C and removed when the various totalized torques (kg-m-min) were reached in the Haake Torque Rheometer. In order to determine the influence of rotor speed on the degree of

85fusion, samples were charged into the mixer at 180°C and removed after 6 minutes processing time had passed in the Haake Torque Rheometer. The rotor speeds were changed from 10 to 100 rpm by increasing 10 rpm each time.

3.3.2 Capillary Rheological AnalysisThe viscoelastic property of all processed PVC samples

was characterized by the zero-length capillary entrance pressure drop at 180°C. In order to put samples into the barrel with a diameter of 20 mm, all PVC samples were cracked into small pieces by a hammer. The test conditions were summarized in Table 3.1. The rheological measurements were performed on the advanced capillary extrusion rheometer system (ACER 2000, Polymer Laboratory Limited, now Rheometrics, Inc.) equipped with a pressure transducer and a measuring range of up to 700 MPa. The entrance pressure drop was read by this pressure transducer every 4 seconds. One average value was obtained after approximately 150 entrance pressure drop data points were measured. A "zero- length" capillary die was used. In order to minimize the possibility of slip and maximize the amount of internal deformation, a flat entry was selected for the zero-length capillary [Krzewki, 1981] . In order to melt rigid PVC compounds, PVC sample was put in the barrel at an isothermal temperature (180°C) for 25 minutes before the measurement.

86

Table 3.1 Test conditions for rheological measurements.

Rheometer ACER 2000Temperature 180°CCapillary Diameter: 1.1 mm

(Flat entry) Length: 0.33 mm L/D: 0.3

Barrel Diameter: 2 0 mmShear rate (0.5 mm/min)

23.6 sec"1, constant

Sample • Small pieces with a diameter less than 20 mm • Always the same amount,

25 g• Preheating 25 minutes

Measured variable Entrance pressure drop for extrusion at a constant shear rate, 23.6 sec'1

873.3.3 DSC Thermal Analysis

PVC compounds prepared by the Haake Torque Rheometer were cut randomly into small pieces, weighted approximately 10 mg and then characterized by a SEIKO 220C automatic cooling differential scanning calorimeter. For further thermal analysis, samples were heated from room temperature to 270°C, at 20°C per minute. Three DSC measurements from each Haake preparation were conducted for each PVC compound

in order to obtain an average value for the heat of fusion. This average value was used to determine the fusion level of the PVC compound.

3.3.4 SEM AnalysisPVC samples, were immersed into liquid nitrogen for 45

seconds, fractured and coated with 200 A gold-palladium film, and then was examined, using a Cambridge S-260 SEM.

3.4 Results and Discussion

3.4.1 Haake Torque RheometerFigure 3.11 shows the typical fusion curves when the

PVC melted in the Haake Torque Rheometer. These curves illustrate the changes of viscosity related to torque, temperature, and totalized torque versus time. The viscosity related torque curve shows three different peaks.

88

Temperature

Torque

otalized Torque

L O D SFTime ( min)

Figure 3.11 A typical temperature/torque/totalized torque curves of PVC melted in a Haake Torque Rheometer.

The first peak, L, in the torque curve is due to loading, and the second peak, F, is due to compaction and onset of melting. When the PVC sample is loaded into the system, the initial peak is generated, then the torque begins to decrease sharply because of free material flow before it begins to compact; the torque then begins to increase and generates the second peak. At this peak, F, the material reaches a void-free state and starts to melt at the interface between the compacted material and the hot metal surface [Chung]. If the material is melted and fused for a long time in the mixer, the third peak, D, is generated. Usually, the time between the loading point L and the fusion

point F is defined as fusion time. PVC compounds begin to degrade seriously at point 0, and point D is defined as the degradation point. The processing time is defined as the time between the loading point L and the stopping point S. In normal processing, mixing is terminated between point F and point 0 in order to avoid serious degrading.

Figure 3.12 shows the influence of temperature on the fusion time of PVC compounds in the Haake Torque Rheometer, at a rotor speed of 60 rpm. It is illustrated that, at the same rotor speed, the fusion time decreases as the starting temperature increases. This is because higher thermal

Fusi

on

Tim

e (m

in)

1 1 1Rotor Speed - 60 rpmIQ-

160 170 180 190 200Temperature ( C)

Figure 3.12 Effect of temperature on fusion time.

91energy input speeds the melting of PVC compound and increases the rate of interfusion of PVC submicroparticles in order to form a 3-D network. Similarly, at the same starting temperature of 180°C, the fusion time decreases as the rotor speed increases. This is due to a higher shear force input, which also increases the rate of interfusion of PVC submicroparticles to form a 3-D network (Figure 3.13) .

3.4.2 Capillary Rheological AnalysisFigure 3.14 shows the capillary entrance pressure drop

for the PVC sample processed at 160°C, 100 rpm anddischarged at a totalized torque of 1 kg-m-min. At the very beginning of the measurement, due to packing of the PVC sample, the extrusion pressure increased slowly. After packing, the steady state extrusion pressure, with random variation, was achieved. Around 120-150 data points were measured, and an average value was taken. • This average value is the elastic response, as measured by the zero- length capillary.

In order to calculate the degree of fusion for each compound, the extrusion pressure for the unprocessed compound (P* = 5.39 MPa) was taken as 0 % fusion level and

the processed compound ( P 20o , i s = 10.83 MPa), which wasprocessed at a starting temperature of 200°C, a rotor speed of 60 rpm, and stopped at a totalized torque of 15 kg-m-min,

Fusio

n Ti

me

(min

)92

6.5- Temperature -1 0 0 C

5.5-

4.5-$

3.5-

2.5-

0.5-

30 35 40 45 50 55 60 65 70 75 80 85 90 100Rotor Speed ( rpm)

Figure 3.13 Effect of rotor speed on fusion time.

Dlaplacaacnt

93

8-7-

6-

9-

4-

J, J * L ,C's' *

i [ ! »(i • »

k r ~ ‘ "

I I*JiMJfi

2- C 1- 0- -1-

RESULT___________! lAVEflASE: 7 ■ 122E+031

7600

7400

•7200

7000

■6800

-6600

■6400

■6200

-6000

■9800

-9600

■9400

■9200"1--------1------- 1------- 1-------1------- 1------2 4 6 6 10 12 14TIM (aln)

Figure 3.14 Entrance pressure drop measured by azero-length capillary die at a constant shear rate of 23.6 sec'1 and 180°C.

Praa

aura

(K

Pa)

[■

Fusi

on

Leve

l (%

)

94

r 10.83'“l200oC/“90- / _ 10 CO. 180 C80- 170 C70-

° *-60-

50-

40

30-

20-

1 0 -

L 5.395 10 15 20 25 30 35 40 45 50 55 60 650

T T Q (kg-m-min)

Figure 3.15 Effect of temperature and totalizedtorque on the degree of fusion of PVC compounds measured by capillary rheological analysis.

95were taken as 100 % fusion level. The percentage of fusion corresponding to the extrusion pressure (P) was then calculated from equation 3.2-4. The degree of fusion of the PVC compound increased with totalized torque, and when plotted against totalized torque, produced an S-shaped curve (see Figure 3.15). It was noticed that, as the starting temperature increased, the slope of the fusion level curve also increased. This is due to the fact that more thermal energy input occurs at higher starting temperatures.

3.4.3 DSC Thermal AnalysisFigure 3.16, similar to Figure 3.9 [Gilbert, 1981],

displays the DSC thermal analysis curves of PVC compounds blended in the Haake Torque Rheometer at various temperatures. A DSC trace of PVC powder has an endothermic baseline shift at the glass transition temperature (Tg) (approximately 80°C) . There are two peaks, A and B, in the DSC traces. Peak B may result from the endothermic energy of the PVC crystals that are not melted in the Haake Torque Rheometer. When the processing temperature increases, this peak decreases in size and shifts to a higher temperature. Gilbert et a l . [1981 and 1985] suggested that the sizedecrease was due to the melting of less perfect, or smaller, crystallites, and the temperature shift was caused by annealing of unmelted crystallites. Peak A is related to

96

- 2 5 0 0

PVC POWDER- 4 0 0 0

160 C

exo170 C

.80C,endo oo □

190 C.- 7 0 0 0

200 C

- 8 5 0 0 120 8 0 200 260140

TEMP C ( H e a t i n g )

Figure 3.16 DSC thermal analysis curves of PVCcompounds processed in a Haake Torque Rheometer at rotor speed=60 rpm, TTQ=10 kg-m-min, and various temperature.

97the endothermic energy of the PVC crystals that are melted in the Haake Torque Rheometer and recrystallized after cooling at room temperature. Peak A was measured to determine the heat of fusion of the PVC compound.

The endothermic peak A appears and increases in size when the totalized torque is increased at the same starting temperature and a rotor speed of 60 rpm. The energy of this endothermy, AHa, increases with totalized torque, and, when plotted against totalized torque, produces an S-shaped curve

(see Figure 3.17) . It was noticed that as the starting temperature increased the slope of the fusion level curve also increased. This is due to the fact that more thermal energy input occurs at higher starting temperatures. To calculate the relative fusion level for each PVC compound, the heat of fusion for the unprocessed compound (0 mJ/mg) was taken as 0 % fusion level, and the processed compound (13.72 mJ/mg) (processed at a starting temperature of 200°C, a rotor speed of 60 rpm, and stopped at a totalized torque of 15 kg-m-min) was taken as 100 % fusion level.

Meanwhile, endotherm A appears and increases in size when the processing temperature is increased. The energy of this endotherm, AHa, increases with processing temperature, and when plotted against temperature, produces an S-shaped curve (see Figure 3.18). Similarly, the entrance pressure drop of the PVC compound increased as the processing

Fusion Lev

el (%

)

98

r 13.72

60- O)7 0 1

60-

50-

40-

30-

20-

1 0 -

5 10 15 20 25 30 35 40 45 50 55 60 650TTQ (kg-m-min)

Figure 3.17 Effect of temperature and totalizedtorque on the degree of fusion of PVC compounds measured by DSC thermal analysis .

Fusio

n Le

vel

(%)99

90-,TTQ - 1 0 kg-m-min , Rotor S peed - 60 rpm

80-

70-

60-

50-

40-

+ : DSC Thermal Analysis

□ : Capillary Rheological Analysis

30-

20-

10 -

180 190 200170160

Temperature ( C)

Figure 3.18 Effect of temperature on the degree of fusion of PVC compounds processed in a Haake Torque Rheometer.

100temperature increased, and when plotted against temperature, produced an S-shaped curve.

Figure 3.19 illustrates the effects of the rotor speeds on the fusion level of PVC compounds at a temperature of 180°C and a processing time of 6 minutes. It shows that the rotor speed has a linear effect on the fusion level. This is because higher shear force input results in not only increasing interfusion of PVC submicroparticles to form a 3- D network, but also increasing the fusion level.

3.4.4 Rheological Analysis vs. Thermal AnalysisAlthough similar fusion curves are obtained comparing

Figure 3.15 (rheological analysis) with Figure 3.17 (thermal analysis), one important different exists. The degree of fusion of PVC, processed at low processing temperature or low totalized torque, and determined by the rheological analysis is much higher than the result indicated by thermal

analysis.This may be due to the fact that the preheating

process, 25 minutes at 180°C, provides enough thermal energy to improve the fusion of non well-fused PVC particles which were pre-processed in a Haake Torque Rheometer. Figure 3.20 shows the morphological changes of PVC powder, low fusion level, medium fusion level, and high fusion level PVC

Fusio

n Le

vel

(%)

101

100 1Temperature - 1800 C90 -

80 -

70-

60 -

50 -

40 -

30 -

20 -

10 -

0 10 20 30 40 50 60 70 90 10080Rotor S p eed ( r p m )

Figure 3.19 Effect of rotor speed on the degree of fusion of PVC compounds processed in a Haake Torque Rheometer (blending time=6 minutes).

102

P V CPowder

MediumFusion

HighFusion

LowFusion

Figure 3.2 0 Morphological changes of PVC compounds((a) :PVC powder, (b) .-low, (c) :medium, and (d):high fusion levels), extruded from the capillary rheometer.

103samples before and after the rheological analysis. It illustrates that the preheating process does improve the fusion of non-fused PVC particles. For well-fused PVC compounds, this preheating process provides very little effect on the degree of fusion.

The capillary rheological analysis has a main advantage in that this method uses a sample large enough to eliminate local inhomogeneity in the degree of fusion. The two main disadvantages of this method are that the preheating process may interfere with the degree of fusion of PVC and the necessity of obtaining a standard fusion curve for each compound processed in specific processing equipment. Also, the elastic response as measured by the rheological analysis, is strongly composition-dependent [Krzewki, 1981] .

DSC thermal analysis has attracted attention recently

because it provides a convenient, quantitative measure of the fusion level and uses a very small sample (approximately 10 m g ) . The main disadvantage of thermal analysis is that some additives may interfere with the final results because the sample size is very small.

3.4.5 SEM AnalysisThe surface morphology of a suspension PVC grain

particle, approx 100-150 /xm in diameter, is shown in Figure 3.21. Figure 3.22 shows the surface morphology of the very

104

Figure 3.21 Surface morphology of a suspension PVC powder.

105

Figure 3.22 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature = 190°C, rotor speed = 100 rpm, and TTQ = 1 kg-m-min.

106beginning of fusion of PVC particles. It illustrates that some PVC grain particles were broken down to PVC microparticles, and a few PVC grain particles still remained intact. Both the PVC grain particles and the microparticles were compacted together. After this stage, if more energy was put into the PVC compound, the boundaries of the PVC microparticles were broken down into submicroparticles and the submicroparticles began to fuse together in order to form a 3-D network.

Figures 3.23, 24, and 25 show the 3-D network surface morphological changes of PVC compounds which were blended in the Haake Torque Rheometer at a rotor speed of 60 rpm, 10 kg-m-min totalized torque, and starting temperatures of 160, 180, and 200°C, respectively. At the lowest starting temperature, 160°C, the surface morphology of the PVC compound is not very smooth because the interfusion between the PVC particles was not good enough to form a smooth, 3-D network. At the medium starting temperature, 180°C, the surface morphology of the PVC compound is much smoother than that for 160°C, because of the better interfusion between the PVC particles. At the highest starting temperature, 200°C, the surface morphology of the PVC compound is smoother and less porous than that for 180°C, because of the much better interfusion between the PVC particles.

107

Figure 3.23 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature = 160°C, rotor speed = 60 rpm, and TTQ = 10 kg-m-min.

108

Figure 3.24 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature = 180°C, rotor speed = 60 rpm, and TTQ = 10 kg-m-min.

109

Figure 3.25 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature =2 00°C, rotor speed = 60 rpm, and TTQ = 10 kg-m-min.

Figure 3.26 shows the surface morphology of the well- fused PVC compound with embedded small particles of additives. The blending condition for this sample as set at a starting temperature of 200°C, rotor speed of 60 rpm, and stopped at a totalized torque of 15 kg-m-min in the Haake Torque Rheometer. The color of this PVC sample was black because some PVC particles were overfused and burned when processed in the Haake Torque Rheometer. Because the distribution of energy in the Haake Torque Rheometer is not very uniform, few PVC submicroparticles are still observed

in Figure 3.26.

Ill

Figure 3.26 Surface morphology of PVC blended in a Haake Torque Rheometer at temperature = 200°C, rotor speed = 60 rpm, and TTQ = 15 kg-m-min.

CHAPTER 4CORRELATION OF COMPOUNDING ENERGY WITH FUSION LEVEL BY

MATHEMATICAL MODELING

4.1 Literature ReviewInternal mixers and various continuous extruders have

been commonly used to mix polymer compounds and blends in industry. The importance of the compounding and blending of polymers is increasing and becoming more attractive because of the fact that the mechanical properties of polymers can

be improved through these processes. Internal mixers have been used to process PVC compounds for many years due to their flexibility, particularly, because frequent product and formulation changes are common.

Buskirk et a l . [1975] used the automatic integration of Rheocord torque-time curves to study the independent variable "unit work" (Wu) for quantifying mixing in a torque rheometer with a Rheomix 750 (Banbury type mixer/sensor) and a Rheomix 600 (cam type rotors in mixer/sensor) . Dealy [1982] reviewed the reports that were related to the energy conservation in plastics processing, especially in the areas of extrusion, injection-molding, and heat recovery. He also applied the second law of thermodynamics for energy analysis. Menges et a l . [1984] studied the flow process in an internal mixer by using colored mixtures at various

112

113mixing conditions. They also developed a model which can be used to calculate the variation of the internal wall temperatures. White et a l . [1988] applied a hydrodynamiclubrication theory to analyze the flow in internal mixers and twin-screw extruders. Pressure fields and mean fluid patterns were also computed in their report.

The compounding energy consumed in an internal mixer can be calculated using the concept that the energy transfer in the mixing process can be obtained from the total energy balance [Menges, 1984]. Figure 4.1 [Pedersen, 1991] shows a simple energy balance for a torque rheometer. The compounding procedure involves compaction of the PVC resin particles into a fixed volume during mixing and heating. The amount of work (mechanical energy) going into the PVC sample is contributed by the torque necessary to turn the non-interchangeable, counter-rotating, and nonintermeshing rotors at the preset RPM. The amount of heat transferred into the PVC sample is mainly a function of the PVC sample and bowl temperature. The melt temperature at any time (t) is dependent upon the work and heat transferred into, or out of, the PVC sample. In this chapter, the compounding energy has been correlated with the fusion level of the PVC sample by basing the determination of the compounding energy on the total energy balance, which is illustrated in Figure 4.1.

114

HoW =2irNJGdt E t = JQdt

Hc — H0 + Et + W t , Traelt t — Hc / mCpH0 = mCpT0 + P0V , Q = UA (Tbowl - Tmelt,t)

where: H0 = Initial enthalpy of material, (J).Ht = Enthalpy of material at time t, (J).Et = Amount of heat transferred by time t, (J).Wt = Amount of work generated by time t, (J) .m = Charge size, (kg).Cp = Heat capacity of material, (J/kg°C) .Pc = Pressure upon charging, (N/m2) .V = Bowl volume, (m2) .Tc = Temperature before charging, (°C) .

t = Time, (mins).Q = Rate of heat transfer, (J/min).U = Overall heat transfer coefficient,

(J/m2oCmin) .A = Bowl surface area, (m2) .T = Temperature, (°C) .G = Torque, (kg-m).

Figure 4.1 Simple total energy balance for a torque rheometer [Pedersen, 1991].

1154.2 Experimental Method

4.2.1 Preparation of PVC CompoundsThe material used in this study is suspension PVC

masterbatch powder, containing 100 parts of PVC grain particles, 1.5 parts of process aid (K120N), 1.0 part wax(XL165) , 1.0 part calcium stearate, and 1.5 parts heatstabilizer (T-137). All masterbatch samples without compounding were supplied by the Dow Chemical Company.

In order to determine the influence of temperature onthe fusion level, PVC samples with various fusion levelswere prepared in the Haake Torque Rheometer, equipped withan electrically heated mixing head. The rotor speed was set at 60 rpm, and the sample weight was always 65 g. Samples were charged into the mixer at 160, 170, 180, 190, and 200°C and removed when 10 kg-m-min totalized torque was reached in the Haake Torque Rheometer. Table 4.1 illustrates the digital data from the Haake Torque Rheometer at a preset bowl temperature of 180°C. The T2 is the melt temperature of the PVC sample at time t .

4.2.2 DSC Thermal AnalysisThe heat of fusion of the PVC compounds was measured.

The detailed experimental method has been explained in

Chapter 3.3.3.

116

Table 4.1 Digital data from the Haake Torque Rheometer at a preset bowl temperature of 180°C.

Time TORQUE TTQ R1 T1 T2 T3 T4 T5 PI P2 P3 P4 P5min sec mg kg-m-min RPM OC oC oC oC oC psi psi psi psi psi

Or 0 11 0.0 60 26 175 26 26 26 0 0 0 0 00: 6 11 0 . 0 60 26 175 26 26 26 0 0 0 0 00 12 11 0 . 0 60 26 175 26 26 26 0 0 0 0 00 :18 11 0 . 0 60 26 175 26 26 26 0 0 0 0 00:24 11 0 . 0 60 26 175 26 26 26 0 0 0 0 00:30 11 0 . 0 60 26 174 26 26 26 0 0 0 0 00:36 11 0.0 60 26 165 26 26 26 0 0 0 0 00:42 11 0 . 0 60 26 157 26 26 26 0 0 0 0 00:46 37 0 . 0 60 26 155 26 26 26 0 0 0 0 00 54 3454 0 . 4 60 26 137 26 26 26 0 0 0 0 01: 0 2180 0.7 60 26 142 26 26 26 0 0 0 0 01 6 1722 0 . 9 60 26 148 26 26 26 0 0 0 0 01 12 1439 1 .0 60 26 153 26 26 26 0 0 0 0 01 IB 1253 1.2 60 26 157 26 26 26 0 0 0 0 01:24 1157 1.3 60 26 159 26 26 26 0 0 0 0 01:30 1077 1.4 60 26 162 26 26 26 0 0 0 0 01:36 1018 1.5 60 26 164 26 26 26 0 0 0 0 01:42 981 1 . 6 60 26 166 26 26 26 0 0 0 0 01:46 965 1.7 60 26 168 26 26 26 0 0 0 0 01:54 938 1 . 8 60 26 169 26 26 26 0 0 0 0 02 0 917 1 . 9 60 26 170 26 26 26 0 0 0 0 02 6 906 2 . 0 60 26 171 26 26 26 0 0 0 0 02 12 901 2.1 60 26 172 26 26 26 0 0 0 0 02 18 912 2.1 60 26 173 26 26 26 0 0 0 0 02 : 24 901 2.2 60 26 174 26 26 26 0 0 0 0 02:30 938 2.3 60 26 175 26 26 26 0 0 0 0 02:36 965 2.4 60 26 176 26 26 26 0 0 0 0 02 42 1002 2 .5 60 26 177 26 26 26 0 0 0 0 02:46 1093 2 .6 60 26 178 26 26 26 0 0 0 0 02:54 1221 2 . 7 60 26 179 26 26 26 0 0 0 0 03 0 1370 2 . 9 60 26 180 26 26 26 0 0 0 0 03 6 1727 3 . 0 60 26 181 26 26 26 0 0 0 0 03 12 2143 3.2 60 26 182 26 26 26 0 0 0 0 03 :18 2378 3 . 5 60 26 184 26 26 26 0 0 0 0 03 :24 2420 3.7 60 26 186 26 26 26 0 0 0 0 03 : 30 2452 3 . 9 60 26 187 26 26 26 0 0 0 0 03 : 3 6 2468 4 .2 60 26 188 26 26 26 0 0 0 0 03 :42 2490 4 .4 60 26 189 26 26 26 0 0 0 0 03:48 2452 4 .7 60 26 190 26 26 26 0 0 0 0 03 54 2452 4 . 9 60 26 191 26 26 26 0 0 0 0 04: 0 2490 5.2 60 26 191 26 26 26 0 0 0 0 04 6 2522 5 . 4 60 26 193 26 26 26 0 0 0 0 04:12 2527 5 . 7 60 26 193 26 26 26 0 0 0 0 04 16 2490 5.9 60 26 194 26 26 26 0 0 0 0 04:24 2458 6.2 60 26 194 26 26 26 0 0 0 0 04:30 2410 6.4 60 26 195 26 26 26 0 0 0 0 04:36 2372 6.7 60 26 196 26 26 26 0 0 0 0 04:42 2330 6.9 60 26 196 26 26 26 0 0 0 0 04:48 2287 7.1 60 26 197 26 26 26 0 0 0 0 04 54 2255 7.4 60 26 197 26 26 26 0 0 0 0 05: 0 2228 7 . 6 60 26 198 26 26 26 0 0 0 0 05: 6 2207 7 . 8 60 26 198 26 26 26 0 0 0 0 05:12 2186 8 . 0 60 26 198 26 26 26 0 0 0 0 05:18 2170 8.2 60 26 198 26 26 26 0 0 0 0 05:24 2148 8 . 5 60 26 199 26 26 26 0 0 0 0 05:30 2127 8 . 7 60 26 199 26 26 26 0 0 0 0 05:36 2111 8 . 9 60 26 199 26 26 26 0 0 0 0 05:42 2106 9.1 60 26 199 26 26 26 0 0 0 0 05:48 2090 9.3 60 26 200 26 26 26 0 0 0 0 05:54 2079 9.5 60 26 200 26 26 26 0 0 0 0 06: 0 2063 9.7 60 26 200 26 26 26 0 0 0 0 06 6 2058 9.9 60 26 200 26 26 26 0 0 0 0 0

R2%01001110000001011110011101110101111111011110001001000000111001

1174.2.3 Capillary Rheological Analysis

The entrance pressure drop of the PVC compounds was

explained in Chapter 3.3.2.Figure 3.18 illustrates the curves of the fusion level,

measured by DSC thermal analysis and capillary rheological analysis, of PVC compounds, which were blended in the Haake Torque Rheometer at various bowl temperatures, rotor speed = 60 rpm , and removed when a 10 kg-m-min totalized torque

was reached.

4.3 Theoretical Method, Results and DiscussionBased on the total energy balance,

which is shown in Figure 4.1, the compounding energy is correlated with the fusion curves illustrated in Figure 3.18. The initial enthalpy of the material, H0, is mainly a function of the material. Here, Hc is considered as the same for each compounding condition, since a unique PVC sample was used. The amount of work generated at time t, Wt, is equal to the integration of torque (G) by time.

measured. The detailed experimental method has been

(4.3-1)

(4.3-2)

where N = Rotor speed (rpm)G = Torque

118The torque (G) is an instantaneous force required to

cause the material to flow [Haake Buchler Technical Bulletin, TB-826]], and it is expressed as:

Torque = 2 n R 2Lx (4.3-3)

where R = Rotor radius (m)L = Rotor length (m)r = Instantaneous shear stress (N/m2) t = Time (min)

If Equation 4.3-3 is substituted into Equation 4.3-2, then

Wt = 2n N J 2 n R 2Lxdt (4.3-4)

Since the PVC sample was removed from the Haake Torque Rheometer when a 10 kg-m-min totalized torque was reached, the amount of work transferred into the system can also be considered as the same for each case. Therefore, the onlydifferent term for each case is the amount of heattransferred at time t, Ec, into the system.

E t = f Qdt (4.3-5)

The rate of heat transfer, Q, is expressed as:

Q = UA(Tbowl-Tmeltit) (4.3-6)

Where U = Overall heat transfer coefficient (J/m2 °C min)A = Bowl surface area (m2)

119Tbo«i = Preset bowl temperature (°C)

Tmeit,t = Melt temperature of PVC sample at time t (°C) Substitute Equation 4.3-6 into Equation 4.3-5, then

E t = f UA(Tbowl-Tmelt/t) (4.3-7)

Equation 4.3-7 can be expressed as the following formula:

E c = fuA(Tbowl-TmelC'C)dt=J2UA(Tbowl-Tmeltit)bt (4.3-8)

If we substitute Tbowl = 180°C and Tmelt t = various with time (see Table 1) into Equation 4.3-8, then

H t = H a + Wt + 32UAA t (4.3-9)

Using a similar procedure, we can obtain the H t for thedifferent bowl temperatures: 160, 170, 190, and 200°C (see Table 4.2).

When the compounding energy, Hc, (in Y-axis) is plotted versus various bowl temperatures (in X-axis), we can obtain one S-shaped fusion curve similar to experimental results (Figure 4.2). It is apparent that the simple total energy balance based on Figure 4.1 (Pedersen 1991] can be appliedto correlate the compounding energy with the fusion level ofthe PVC compound prepared in the torque rheometer.

120

Table 4.2 Enthalpy of the PVC sample at time t, H t, forvarious starting temperatures (where W = H0 + W t) .

Starting temperature160°C W-372UAAt170°C W-241UAAt180°C W+32UAAt190°C W+67UAAt200°C W+92UAAt

Com

poun

ding

En

ergy

(J

)121

TTQ - 10 kg-m-min, Rotor Speed - 60 rpmi- 90

- 80

W+ OUAit - - 70

- 60

W-100UAAt - 50

- 40— : DSC Thermal Analysis : Capillary Rheological Analysis— : Mathematical Result

- 10

W-400UAAH200190160 170 180

P r e s e t Bowl T e m p e r a tu re ( 0 C )

Figure 4.2 Experimental results vs. the mathematical result based on a simple total energy balance in a torque rheometer.

CHAPTER 5APPLICATION OF FACTORIAL EXPERIMENTAL DESIGN TO

DEMONSTRATE THE INFLUENCE OF TEMPERATURE, ROTOR SPEED, AND TOTALIZED TORQUE ON THE FUSION OF PVC COMPOUNDS

5.1 IntroductionExperimental designs and their statistical analyses

have been well developed and applied widely in many research areas, such as basic science, engineering, sociology, ... etc. The main advantage of the experimental design is that it can cover a larger area of researchers' experimental interest, obtain unambiguous results at a minimum cost [Hahn, 1975], and shorten product development and production time [Schmidt et a l . , 1992] . Because this technique ispowerful and easy to handle, the factorial experimental design is one of the most commonly used methods to characterize the effects of independent variables which significantly affect the final experimental results.

The fusion of a PVC compound is highly dependent upon the additives and the rotor speed, as well as its thermal history in a batch mixer. Bambrick et a l . [1993] studiedthe fusion characteristics, which are the dependent variables, of PVC compounds (fusion time, fusion temperature, and fusion torque) by using a Rheocord System 40 torque rheometer, equipped with a three piece Rheomix 600

122

123bowl and roller mixing blades. They applied a central composite design (CCD) of the experiment to find the optimal formulation of additives for PVC compounds by changing the following six independent formulation variables: amounts of: impact modifier, paraffin wax, calcium stearate, ester wax, and processing aid.

In chapter three, it has been reported that the starting temperature, rotor speed, and TTQ are the three major factors that affect the heat of fusion of a PVC compound prepared by the Haake Torque Rheometer, equipped with a three-sectioned mixing chamber and two non- interchangeable roller mixing rotors. In order to realize the main, two-factor interaction, and three-factor

interaction effects of these three independent blending variables on the heat of fusion of PVC compounds, a 23 factorial experimental design (three independent variables with high (+), and low (-) levels) was applied.

Figure 5.1 illustrates the standard figure of a 23 factorial experimental design where three independent variables exist (starting temperature, rotor speed, and totalized torque) with high (+) and low (-) levels. Thus, a 23 factorial experimental design will have eight runs, thefirst in standard order being (---) , and the last instandard order being (+++). Several different fusion assessment methods have been well developed (Gilbert, 1985] .

124

( - + + )100

Rotor SpeedTTQ

X 2160 Temperature 190

X 3

Figure 5.1 Diagrammatic representation of the standardordering of a 23 factorial experimentaldesign.

125Both the heat of fusion, measured by the DSC thermal analysis, and the entrance pressure drop, determined by the capillary rheological analysis, of a PVC compound were studied as the dependent variables here. Three different PVC compounds, PVC, PVC/CPE, and PVC/CPE/OPE, were prepared

and evaluated.

5.2 Experimental Method

5.2.1 Preparation of PVC CompoundsThe materials used in this study are suspension PVC

masterbatch powder, containing 100 parts PVC resin particle,

1.5 parts process aid (K120N) , 1.0 part wax (XL165) , 1.0part Calcium Stearate, and 1.5 parts heat stabilizer (T- 137) . The other two samples are PVC masterbatch powder with 5 parts CPE (impact modifier) and PVC masterbatch powder with 5 parts CPE and 0.3 parts OPE (external lubricant) . All samples were supplied by the Dow Chemical Company.

All PVC samples were prepared by the Haake Torque Rheometer, equipped with a three-sectioned mixing chamber and two non-interchangeable roller mixing rotors. The starting temperature, rotor speed, and totalized torque were chosen as the independent variables. Two levels, high (+) and low (-) , were also defined for each independent variable. For the starting temperature, 190 and 160°C were

126chosen as high and low levels, respectively. For the rotor speed, 100 and 20 RPM were chosen as high and low levels, respectively. For the totalized torque, 15 and 1 kg-m-min were chosen as high and low levels, respectively. The sample weight was 65 grams for all runs. PVC samples were charged into the mixer at various starting temperatures and rotor speeds and removed when the set totalized torque (kg- m-min) was reached in the Haake Torque Rheometer.

5.2.2 DSC Thermal AnalysisThe detailed experimental method has been shown in

Chapter 3.3.3. The heat of fusion measured by this method was chosen as the dependent variable of the 23 factorial experimental design for the PVC compound.

5.2.3 Capillary Rheological AnalysisThe detailed experimental method has been shown in

Chapter 3.3.2. The entrance pressure drop measured by this method was chosen as the dependent variable of the 23 factorial experimental design for the PVC compound.

1275.3 Results and Discussion

5.3.1 DSC Thermal AnalysisFigure 5.2 shows the observed yields (heat of fusion;

mj/mg) and the standard ordering of experiments for PVC compounds. Figures 5.3, 5.4, and 5.5 represent thedeterminations of the main effects of temperature, rotor speed, and TTQ, respectively. According to the definition, the main effect of the controlled independent variable is the average of the difference between the values at the high level (+) and the values at low level (-).

Tables 5.1, 5.2, and 5.3 illustrate the results of the main effects of temperature, rotor speed, and TTQ, respectively. Figures 5.6, 5.7, and 5.8 illustrate thedeterminations of temperature vs. rotor speed, temperature vs. TTQ, and rotor speed vs. TTQ interaction effects, respectively. According to the definition, the two-factor interaction effect of temperature vs. rotor speed (Xi vs. X2) is equal to half the difference ( (2.70-2.70)/2 = 0 .00) between the average temperature effect with rotor speed = 100 rpm,((5.20+0.20)/2=2.70), and the average temperature effect with rotor speed = 20 rpm, ( (5 .40 + 0 . 00)/2=2 .70) . Thetemperature vs. TTQ interaction effect (X3 vs. X 3) is equal to half the difference ( (5 .30-0.10)/2=2.60) between the average temperature effect with TTQ = 15 kg-m-min,

128

Table 5.1 Main effect of temperature (XJ on PVC compounds.

Effect of temperature(XJ

individual comparisons

Conditions at which comparisons are made

Rotor Speed (X2) TTQ(X3)

(13.60-8.40)=5.20 100 15(5.63-0.23)=5.40 20 15(0.20-0 . 00)=0.20 100 1(0 . 00-0.00)=0.00 20 1

Average (main effect for temperature) (5.2 0+5.40+0.2 0+0.00)/4=2.7

129

Table 5.2 Main effect of rotor speed (X2) on PVC compounds.

Effect of rotor speed (X2) individual comparisons

Conditions at which comparisons are made

Temperature (X2) TTQ(X3)

(13.60-5.63)= 7.97 190 15(8.40-0.23)=8.17 160 15

o CO 0 1 o o o II o to o 190 1(0.00-0.00)=0 . 00 160 1

Average (main effect for rotor speed) (7.97+8.17+0.20+0.00)/4=4.09

130

Table 5.3 Main effect of TTQ (X3) on PVC compounds.

Effect of TTQ (X3) individual comparisons

Conditions at which comparisons are made

Temperature (Xx) Rotor Speed (X2)(13.60-0.20)=13.40 190 100(8.40-0. 00)=8.40 160 100(5.63-0.00)=5.63 190 20(0 . 23-0.00)=0.23 160 20

Average (main effect for TTQ) (13 .40 + 8 .40 + 5.63 + 0.23)/4 = 6 .92

131

0.00±0.00 0.20±0.01

8.40+0:64100

( - + + ) 0.00±0.000.00±0.00Rotor Speed

TTQt).23±0.01 5.63±0.20

(+■+)/'

160 Temperature 190

Figure 5.2 Diagrammatic representation of the observed yields (heat of fusion:mJ/mg) and the standard ordering of experiments of PVC compounds.

132

0.00 0.20

8.4013.60100

Rotor Speed 0.00 0.00TTQ

200.23 5.63

160 Temperature 190

Figure 5.3 Determination of the main effect oftemperature (xj on PVC compounds.

133

0.00 0.20

/N8.40

13.60100

0.00Rotor Speed 0.00TTQ

0.23 5.63

160 Temperature 190

> X

Figure 5.4 Determination of the main effect ofrotor speed (x2) on PVC compounds.

134

0.00 0.20

8.40100 13.60

Rotor Speed / 0 . 0 0 0.00TTQ

0.23 / 5 .63

160 Temperature 190

Figure 5.5 Determination of the main effect ofTTQ (x3) on PVC compounds.

135

0.00

1008.40

Rotor Speed

20 -----0.23

160 Temperature 190

0.20

/TN<4^60\/o5> / 0.00

/ X5.63

TTQ

15

XA

X

Figure 5.6 Determination of the temperature vs.rotor speed interaction effect (Xx vs. X2)on PVC compounds.

136

0.00 0.20

8.4013 .60100

Rotor Speed 0.00TTQ

200.23 7 5.63

160 Temperature 190

X 3

Figure 5.7 Determination of the temperature vs. TTQinteraction effect (X3 vs. X3) on PVCcompounds.

137

0.00 0.20

13.608.40100

Rotor Speed /0.00 0.00TTQ

200.23 5.63

160 Temperature -jgo

> X

Figure 5.8 Determination of the rotor speed vs. TTQinteraction effect (X2 vs. X3) on PVCcompounds.

138( (5.20 + 5.40)/2=5.30) , and the average temperature effect with TTQ = 1 kg-m-min, ( (0 . 20 + 0 . 00)/2 = 0 .10) . Similarly, the rotor speed vs. TTQ interaction effect (X2 vs. X3) is equal

to half the difference ((8.07-0.10)/2=3.99) between the average rotor speed effect with TTQ = 15 kg-m-min,( (7.97 + 8.17)/2 = 8 .07), and the average rotor speed effect with TTQ = 1 kg-m-min, ((0.20+0.00)/2=0.10)

Consider the individual comparisons of the effect of temperature (XJ . There are two available measurements from the experiment to estimate the three-factor interaction effect, temperature vs. rotor speed vs. TTQ (X3 vs. X2 vs. X3), one for each TTQ, TTQ = 15 kg-m-min: ((5.20-5.40))/2 =-0.10, TTQ = 1 kg-m-min: ( (0.20-0.00))/2 = 0 .10. T h edifference between these two estimates is a measure of consistency for each rotor speed, rotor speed = 100 rpm:( (5.20-0.20))/2=2.50, and rotor speed = 2 0 rpm: ( (5.40-0.00))/2=2.70. Half this difference, (-0.10-0.10)/2 = -0 .10 or (2.50-2.70)/2 = -0.10, is defined as the three-factor interaction effect of temperature vs. rotor speed vs. TTQ (X2 v s . X2 v s . X3) .

The same three-factor interaction effect will be obtained from either the effect of rotor speed (X2) individual comparisons or the effect of rotor speed (X2) individual comparisons. As in the case of the main effects and the two factor interactions, the estimate of the three

139factor interaction can be obtained from the difference between the average of vertices of the (+) tetrahedron (Figure 5.9) and the average of vertices of the (-) tetrahedron (Figure 5.10), i.e., ((13.60+0.23+0.00+0.00)/4 -

(8 .40+5.63 + 0.20 + 0.00)/4)=-0.10.Figures 5.11 and 5.12 illustrate the diagrammatic

representations of the observed yields and the standard ordering of experiments of PVC/CPE compounds and PVC/CPE/OPE compounds, respectively. Using the same calculation procedure, the three effects of PVC/CPE and PVC/CPE/OPE

compounds were also obtained.Table 5.4 illustrates the summary of the main, two-

factor interaction, and three-factor interaction effects of PVC, PVC/CPE, and PVC/CPE/OPE compounds. It shows that the sequence of the main effects on the heat of fusion of PVC compounds in ascending order is temperature < rotor speed < TTQ. This is because at the low TTQ (1 kg-m-min) the processing time of the PVC resin particles is very short and the PVC resin particles cannot be fused together well, even at the high levels of temperature and rotor speed (see Figures 5.2, 5.11, and 5.12). Similarly, at the same TTQ, the rotor speed affects the heat of fusion of PVC compounds more significantly than temperature does. This is because the rotor speed dominates the mechanical energy (torque) and the uniformity of the heat transfer in the mixer.

140

Table 5.4 Summary of main, two-factor interaction, and three-factor interaction effects of PVC, PVC/CPE, and PVC/CPE/OPE compounds measured by DSC thermal analysis.

PVC PVC/CPE PVC/CPE/OPEMain Effect

Temperature 2.70 2 . 75 2.62rotor speed 4.09 3 .95 4.13

TTQ 6.92 6.92 6.68Two-Factor Interaction

Temperature v s .

Rotor Speed 0.00 -0.38 -0 .02

Temperature v s . TTQ 2 .60 2 . 55 2.50

Rotor Speed v s . TTQ 3 .99 3 .75 4 .02

Three-Factor InteractionTemperature

v s .Rotor Speed

v s . TTQ-0.10 -0.58 -0.13

141

( + ) Tetrahedron

0.00( - + - ) 0.20 ( ++_)

8 -40 100 -----( - + + )

Rotor Speed

<13.60

(++t)0.00 0.00

TTQ

/ 0.23 . / 5.63 ( + - + )

160 Temperature 190

X

Figure 5.9 Determination of the three-factorinteraction effect of temperature vs. rotor speed vs. TTQ (x3 vs. x2 vs. x3) on PVC compounds ((+) tetrahedron).

142

( - ) Tetrahedron

0.00( - + - ) 0.20 ( ++_)

8.40100 -----

( - + + )

Rotor Speed

20

. • 'I y/ //1 3 .6 0

( + / + )jo.oo 0.00

/NjX/ \

( +-

0.23(--+)

160 Temperature 190

5.63

(+■+)

TTQ

15

X

Figure 5.10 Determination of the three-factorinteraction effect of temperature vs. rotor speed vs. TTQ (xx vs. x2 vs. x3) on PVC compounds ((-) tetrahedron).

143

0.00±0.00 0.40±0.01

1000.00±0.00

TTQ

0.00±0.00Rotor Speed

6.30±0.08Z0.03i0.01HO

160 Temperature 190

Figure 5.11 Diagrammatic representation of the observed yields (heat of fusion: mJ/mg) and the standard ordering of experiments of PVC/5phrCPE compounds.

144

0 .0 0 ±0.00 0 .23±0.02

( + + - )

100( - + + )

0.00+0.00 0.00±0.00Rotor Speed

TTQ5.30 ±0.67

160 Temperature 190

Figure 5.12 Diagrammatic representation of the observed yields (heat of fusion: mJ/mg) and the standard ordering of experiments of PVC/5phrCPE/0.3phrOPE compounds.

Therefore, TTQ and rotor speed are the first and second most important factors, respectively. The sequence of the two-

factor interaction effects on the heat of fusion of PVC compounds, in ascending order, is temperature vs. rotor speed < temperature vs. TTQ < rotor speed vs. TTQ. The TTQ, which is the most important individual factor, is equal to the integral of torque by time. Furthermore, the torque inside the mixer is mostly contributed by the rotor speed. Therefore, the interaction effect between rotor speed and TTQ is the most important factor in determining the heat of fusion of PVC compounds. Some of the torque in the mixer is also affected by the temperature of mixer because the melt viscosity of the PVC compound changes with respect to the mixer temperature. Therefore, the interaction effect between temperature and TTQ is the second most important factor in determining the heat of fusion of PVC compounds.

Because there is no significant interaction between rotor speed and temperature in the mixer, the interaction effect between rotor speed and temperature is not significantly related to the heat of fusion of PVC compounds. Since the interaction effect between rotor speed and temperature is not significant, the three-factor interaction effect is not significantly related to the heat of fusion of PVC compounds.

1465.3.2 Capillary Rheological Analysis

Figures 5.13, 5.14, and 5.15 illustrate the observedyields (entrance pressure drop; MPa) and the standard ordering of experiments for PVC, PVC/5phrCPE, and PVC/5phrCPE/0.3phrOPE compounds, respectively. Using the same calculation procedures that were demonstrated in Chapter 5.3.1, Table 5.5 is obtained.

Table 5.5 illustrates the summary of the main, two- factor interaction, and three-factor interaction effects of

PVC, PVC/CPE, and PVC/CPE/OPE compounds, measured by the capillary rheological analysis. It shows that the sequence of the main effects on the heat of fusion of PVC compounds in ascending order is temperature < rotor speed < TTQ. The sequence of the two-factor interaction effects on the heat

of fusion of PVC compounds, in ascending order, is temperature v s . rotor speed < temperature v s . TTQ < rotor speed vs. TTQ. The three-factor interaction effect is not significantly related to the fusion of PVC compounds. These results are similar to those from the DSC thermal analysis.

147

Table 5.5 Summary of main, two-factor interaction, and three-factor interaction effects of PVC, PVC/CPE, and PVC/CPE/OPE compounds measured by capillary rheometer.

PVC PVC/CPE PVC/CPE/OPEMain Effect

Temperature 0.892 1. 061 1. 077

Rotor Speed 1.308 1.556 1.412

TTQ 2 .208 2 .711 2 .733

Two-Factor Interaction EffectTemperature

v s .Rotor Speed

0 .204 0 . 000 -0.002

Temperature v s .TTQ

0.814 0.973 0.997

Rotor Speed v s .TTQ

1.110 1.039 1.047

Three-Factor Interaction EffectTemperature v s . Rotor

Speed v s . TTQ0.150 -0.055 -0.044

148

7.122±0.115 7 .257±0.107( + + - )

8.745 ±0:203100

6 .447± 0.1206.427±0.11Rotor Speed

TTQ7 .2 3 2 + 0 .2 1 2

160 Temperature 190

Figure 5.13 Diagrammatic representation of the observed yields (entrance pressure drop: MPa) and the standard ordering of experiments of PVC compounds.

149

5.033±0.108

7.865± 0.200100

Rotor Sp eed

20

( - + + )

5 .175±0.085

( + + - )

9.843^0:211

\ ( + + + )

4.570±0.116_ _

'5 .21510.123

(~+)

4 .605± 0.123

(+-)TTQ

7.304±0.112 — 15(+-+)

160 Temperature 190x

- > X ,

X

Figure 5.14 Diagrammatic representation of the observed yields (entrance pressure drop: MPa) and the standard ordering of experiments of PVC/5phrCPE compounds.

150

5.070±0.140

100 -

Rotor Sp eed

20 -----

7 .897*0 .212

(-++)

5.171 ±0.112

(++■)9.903*0:301

|4.726±0.10|2

"5.371 ±0.132

(-+)

160 Temperature 190

( + + + )4 .786*0.101

(+-)TTQ

7 .5 1 2 * 0 .2 2 0— 15

(+■+)XA

X

Figure 5.15 Diagrammatic representation of the observed yields (entrance pressure drop: MPa) and the standard ordering of experiments of PVC/5phrCPE/0.3phrOPE compounds.

CHAPTER 6FUSION CHARACTERISTICS AND MORPHOLOGY ANALYSIS

6.1 Introduction to CPE, OPE, and Calcium StearateCPE is commonly used commercially as an impact modifier

of P V C . The CPE used as impact modifiers in PVC areproduced by the solution chlorination of H D P E . According to the results reported by Chang et a l . [1988], chlorination in

solution gives the greatest uniformity of distribution of chlorine atoms in the HDPE polymer chains. The chlorine distribution and chlorine content of CPE are major factors in the mixing of CPE with PVC, which may influence the mechanical properties of the final product. The melting temperature of CPE ranges from 110°C to 13 0°C. Due to better elasticity and less compatibility with PVC, the CPE with 36% chlorine is usually used as an impact modifier for PVC, and is the optimum composition for impact, processing, and strength [Siegmann et a l . , 1984] . In this study the CPE(M.W. <= 160,000) with 36% chlorine was used.

OPE and calcium stearate are two important lubricants used in PVC processing. OPE can be made by oxidizing polyethylenes at the end of the chain, at sites of unsaturation or at points of branching. This oxidization creates slight polarity and compatibility at the point of oxidation. A typical OPE lubricant for PVC processing might

151

152be a medium-density (molecular weight 2000 to 4000) and slightly oxidized (acid number 6 to 20). In this study the OPE with (acid number 6 to 8) was used. Due to carboxylic groups, this kind of OPE is polar. The melting temperature of OPE ranges from 100 to 13 0°C. Calcium stearate is a salt containing a highly polar CaOOC-group, and a long, non-polar hydrocarbon chain - (CH2)16CH3. Calcium stearate with C18 in the R-C linkage is the most popular lubrication material for PVC processing and has a melting temperature between 145 and

160°C. Figure 6.1 shows the chemical structures of PVC, CPE, OPE, and calcium stearate.

Depending on the lubricating action and effects, lubricants may be divided into two categories: internallubricants and external lubricants. Internal lubricants lower the intermolecular friction of PVC when the PVC composition is being hot-sheared, or fused into a melt. External lubricants reduce the friction and sticking between the hot PVC composition and the working surfaces of processing machinery. Generally speaking, externallubricants (e.g. OPE) delay the fusion time of a PVC compound, and internal lubricants (e.g. esters) speed the fusion time or have very little effect on it. In actual applications, however, lubricants do not always abide by the rules, and some exceptions have been found and studied.

153

POLY (VINYL CHLORIDE) (PVC):

- CH - C H - CH - C H -2 I 2 I

Cl ClCHLORINATED POLYETHYLENE (CPE):

Cl- CH - C H - CH - C - C H - C H - C H - CH= CH

2 I 2 I 2 I I Cl Cl Cl Cl

OXIDIZED POLYETHYLENE (OPE):

c h 2 x c h \ c h 3 ---------------> CH2 \ C H ^ CI 3 ^ I \

c h 2 c h 2p c I OPE X

CH,O OH

CALCIUM STEARATE:

\C - o - C a - O - c.

O O

OH

Figure 6.1 Chemical structures of PVC, CPE, OPE, and calcium stearate.

154Bower [1986] reported that an internal/external classification is inadequate, particularly for a rigid PVC lubrication mechanism.

Several reports discussing the functions of lubricants for the processing of PVC blends are available [Chung TB- 842, Cook 1974, Hartitz 1974, Bower 1979, Rabinovitch 1982, Ditto 1982, McMurrer 1982, Riley 1983, and Yu et a l ., 1984] Lindner et a l . [1989] reported that the interaction betweenOPE and calcium stearate can be used to speed the fusion time of PVC blends significantly. Lubrication is the most important factor in influencing the fusion of PVC during processing which may determine the physical properties of

final PVC products.

6.2 Introduction to Rabinovitch's Theory [Nass, 1992]Rabinovitch et a l . [1984] studied the lubrication

mechanism by considering haze, glass transition temperature, metal release, morphology, fusion and extrusion characteristics of rigid PVC compounds. The effect of calcium stearate, paraffin wax and their synergistic combined lubricity was evolved, as shown in Figure 6.2. Figure 6.2 illustrates a model of PVC lubrication showing metal lubrication and lubrication between PVC microparticle flow units. The following discussion was demonstrated by Rabinovitch et a l . also.

CALCIUM STEARATE

£LU2

PVCPVC

PARAFFIN WAX

_i£ PVCLU

CALCIUM STEARATE

PARAFFIN WAX

Figure 6.2 A model of PVC lubrication showing metal lubrication and lubrication between PVC microparticle flow units [Rabinovitch et al., 1984] .

156Due to C-Cl groups in PVC polymer chains, PVC polymer

chains are somewhat polar molecules. Because of this polarity, PVC is also attracted and sticks to the polar

metal surface of processing equipment (Figure 6.2-a).As mentioned before, calcium stearate is a salt

containing a highly polar CaOOC-group and a long, non-polar hydrocarbon chain - (CH2)1SCH3. If calcium stearate is added to PVC, the polar ends of its molecules adhere not only to the polar PVC, but also to the polar metal surface, leaving the non-polar tails to do the lubrication, or, in other words, displacing the PVC melt from the metal surface, preventing it from sticking and providing limited slip

between the metal and PVC melt surfaces. The polar groups in calcium stearate are also strongly attracted to each other and form a viscous layered structure which is very similar to a liquid crystal (Figure 6.2-b). The OPE with acid number 6 to 8, which was used in this study, is polar also. Therefore in the presence of only OPE, it will form a phenomenon similar to that of a calcium stearate.

Because paraffin wax is a non-polar, straight chain hydrocarbon, it does not wet the metal surface and does not displace the more polar PVC microparticles. Therefore in the presence of only paraffin wax, PVC sticks to the metal

surface (Figure 6.2-c).

157If calcium stearate and paraffin wax are used together,

interaction between the tails of adjacent calcium stearate molecules is hindered by the intervening presence of mobile non-polar alkane paraffin wax molecules. These wax molecules are not attracted to the polar groups of calcium stearate, to PVC, nor to the metal surface, but are weakly attracted to themselves and the non-polar hydrocarbon tail of calcium stearate molecules (Figure 6.2-d). This mechanism demonstrates the formation of a mobile layer which increases the separation between the thin calcium stearate

films on the metal surface and on the PVC melt surface. It also provides a more mobile interface between the calcium stearate molecules adhered to adjacent PVC microparticles due to the synergistic lubrication effect of the calcium stearate/paraffin wax system [Nass 1992].

The fusion characteristics of PVC compounds are highly dependent upon their composition. In this study, the

influences of CPE and two lubricants, calcium stearate and OPE, on the fusion characteristics of PVC blends have been studied. Also the lubrication mechanisms for different PVC blends have been postulated. Moreover, the surface morphological changes of these PVC blends have been illustrated here.

1586.3 Experimental Method

6.3.1 Preparation of PVC CompoundsThe materials used in this study are suspension PVC

masterbatch powder, containing 100 parts PVC resin particle (M.W. «= 150,000), 1.5 parts process aid (acryloid, K120N) ,

1.0 part wax (intermediate lubricant, XL165), 1.0 partCalcium Stearate, and 1.5 parts heat stabilizer (tin thioglycolate, T-137). The other four samples are PVC masterbatch powder with various parts CPE (0, 1, 3, 5, 10, and 15), PVC masterbatch powder with various parts CPE (0, 1, 3, 5, 10, and 15) and 0.3 parts OPE (external lubricant) , PVC masterbatch powder with various parts OPE (0, 0.1, 0.3, 0.5, 1.0, and 1.5), PVC masterbatch powder with variousparts OPE (0, 0.1, 0.3, 0.5, 1.0, and 1.5) and 5 parts CPE. All samples were supplied by the Dow Chemical Company.

All PVC samples were prepared in a Haake TorqueRheometer, equipped with a three-sectioned mixing chamber and two non-interchangeable roller mixing rotors. The starting temperature, rotor speed, and totalized torque were chosen as the blending variables. Three levels, high, medium, and low, were also defined for each blending variable. For the starting temperature, 190, 175, and 160°C

were chosen as the high, medium, and low levels,respectively. For the rotor speed, 100, 60, and 20 rpm were

159chosen as the high, medium, and low levels, respectively. For the totalized torque, 15, 8, and 1 kg-m-min were chosen as the high, medium, and low levels, respectively. The sample weight was 65 grams for all runs. PVC samples were charged into the mixer at various starting temperatures and rotor speeds and removed when the set totalized torque (kg- m-min) was reached in the Haake Torque Rheometer.

The fusion characteristics, fusion time, fusion temperature, and fusion torque, have been summarized and compared for different PVC blends prepared at high and medium level blending conditions. Also, the surface morphological differences were examined and compared for different PVC blends, especially those which were prepared at low totalized torque (1 kg-m-min) and low temperature

(160°C).

6.3.2 SEM AnalysisPVC samples which were not well-fused in the Haake

Torque Rheometer were cut randomly, coated with 200 A gold- palladium film and examined on the Cambridge S-260 SEM. Well-fused PVC samples were immersed into liquid nitrogen for 45 seconds, fractured, coated with 200 A gold-palladium film and examined on the Cambridge S-2 60 SEM only at the

center of the fractured surfaces.

1606.4 Results and Discussion

6.4.1 Fusion Characteristics and Lubrication MechanismsThe pictures, similar to Rabinovitch's theory, of

lubrication mechanisms of these four PVC blends have been postulated. The fusion characteristics of PVC blends which were prepared at a low starting temperature or a low rotor speed are not discussed in this report, because of no apparent fusion peaks. Tables 6.1, 6.2, 6.3, and 6.4 show the fusion characteristics of PVC blends which were prepared at high or medium starting temperatures and high or medium

rotor speeds.If the fusion time of a PVC compound without OPE and

CPE is considered as the normal fusion time, 0.3 phr OPE and 5 phr CPE delay the fusion time of PVC/OPE and PVC/CPE

compounds, respectively. These tables illustrate that the fusion time of the PVC/OPE compound is the longest among these four PVC blends. Moreover, no fusion occurred when the medium starting temperature and high or low rotor speed were used. This may be due to a mobile layer, resulting from the synergistic combined lubricity between OPE and calcium stearate, which increases the separation and slipping among PVC particles (Figure 6.3-e). At the high starting temperature (190°C) , this mobile layer can be destroyed, then PVC particles fuse together, but the fusion

161

Table 6.1 Fusion characteristics of PVC blends attemperature = 190°C, rotor speed = 100 rpm.

PVC PVC/OPE PVC/CPE PVC/CPE/OPE

Fusion T e m p . (°C)

189 198 193 .5 189

FusionTime

(minute)0.8 1.3 1. 05 0.7

FusionTorque(g*m)

2946 2468 2645 2724

Table 6.2 Fusion characteristics of PVC blends attemperature = 190°C, rotor speed = 60 rpm.

PVC PVC/OPE PVC/CPE PVC/CPE/OPE

Fusion Te m p . (°C)

188 196 193 188

FusionTime

(minute)1.25 2.3 1.8 1.2

FusionTorque(g*m)

2701 2116 2339 2465.5

162

Table 6.3 Fusion characteristics of PVC blends attemperature = 175°C, rotor speed = 100 rpm.

PVC PVC/OPE PVC/CPE PVC/CPE/OPE

FusionTemp.(°C)

186 .5 NF* 192 .5 187 .5

FusionTime

(minute)1.15 -- 2 .25 1.1

FusionTorque(g*m)

3074.5 -- 2767.5 2916

Table 6.4 Fusion characteristics of PVC blends attemperature = 175°C, rotor speed = 60 rpm.

PVC PVC/OPE PVC/CPE PVC/CPE/OPE

Fusion T e m p . (°C)

186 .5 NF* 189 185

FusionTime

(minute)3.1 3 .45 2.05

FusionTorque(g*m)

2562.5 -- 2547 2665.5

* : No fusion.

163

^ 1 nm PVC \/ Microparticle PVC

C ^ Row Units i A

CALCIUM STEARATE

PARAFFIN W AX

C) LU2J PVC ® PVC PVC

OPE

d) f/ -

OPE and CALCIUM STEARATE^ t N *

e)° = l S U (

PARAFFIN WAX

Figure 6.3 A postulated lubrication mechanism, showing metal lubrication and lubrication between PVC microparticle flow units, of PVC/OPE compounds.

164time is delayed already. At the medium starting temperature (175°C) and high (100 rpm) or medium (60 rpm) rotor speed,

this mobile layer still exists, therefore no fusion occurs.These’ tables also illustrate that 5 phr CPE delays the

fusion time of PVC/CPE compounds. This may be due to two factors. The first factor may be the less compatibility between the PVC resin and the CPE particles, and the second may be that the CPE particles may decrease the gummy property of calcium stearate, which can hold the PVC resin particles together and allow them to fuse together easily (Figure 6.4-e) . In this case, CPE acts as an external

lubricant.However, the fusion time of PVC/CPE/OPE compounds is

the shortest among these PVC blends. This may be due to the fact that 0.3 phr OPE and 5 phr CPE interact and form a powerful, viscous material which appears to act as a glue that allows the PVC resin particles fuse together easily (Figure 6.5-f). The fusion time of the PVC/CPE/OPE compound

is especially more significantly different than those of PVC, PVC/OPE, and PVC/CPE compounds at the medium starting temperature (175°C) and the medium rotor speed (60 rpm) (Table 6.4). Because a high starting temperature or a high rotor speed improves the gummy property of calcium stearate [Lindner 1981], therefore, the mobile layer has very little influence on delaying the fusion time. The fusion times of

165

a) £1 tim PVC

Microparticle Flow Unit

CALCIUM STEARATE

b) £LU2PVCPVC

PARAFFIN WAX

CPE

d)

CALCIUM STEARATE

I \PARAFFIN WAX

CPE

Figure 6.4 A postulated lubrication mechanism, showing metal lubrication and lubrication between PVC microparticle flow units, of PVC/CPE compounds.

CALCIUM STEARATE

£2PVCPVC

PARAFFIN WAX

£1112

PVC

PVC PVC CPE

OPE\

_iLUs

PVC PVC

PARAFFIN WAXOPE and CALCIUM STEARATECPE

PVC

Figure 6.5 A postulated lubrication mechanism, showing metal lubrication and lubrication between PVC microparticle flow units, of PVC/CPE/OPE compounds.

167PVC compounds are very close to those of PVC/CPE/OPE compounds in Tables 6.1, 6.2, and 6.3. In this case,calcium stearate functions as a-process aid. However, at a medium starting temperature and rotor speed, the gummy property of calcium stearate is not very significant and a mobile layer is formed (Figure 6.6-d). This mobile layer results in a longer fusion time (Table 6.4).

Since the fusion time of PVC/OPE compounds is the longest, it can be reasoned that more energy is needed to be absorbed by PVC/OPE compounds in order for the PVC particles to fuse together, and, therefore, the fusion temperatures of PVC/OPE compounds are the highest (Tables 6.1 and 6.2) . Because a higher fusion temperature results in decreasing the melt viscosity of samples in the Haake Torque Rheometer,

the fusion torques of PVC/OPE compounds are the lowest (see Tables 6.1 and 6.2). Applying the same reasoning used above, an explanation is provided for the fact that the fusion torques and fusion temperatures of PVC/CPE compounds are the highest in Tables 6.3 and 6.4.

Tables 6.5 and 6.6 illustrate the influence of varying phr of CPE on the fusion characteristics of PVC/CPE and PVC/CPE/0.3phrOPE compounds prepared at a temperature of 190°C and a rotor speed of 60 rpm. Figure 6.7 illustrates the fusion time curves, varying with the phr of CPE, of PVC/CPE and PVC/CPE/0.3phrOPE compounds. It illustrates

168

Table 6.5 Fusion characteristics of PVC/CPE compounds at temperature = 190°C, rotor speed = 60 rpm.

lphr 3 phr 5phr lOphr 15phr

Fusion T e m p . (°C)

196 195 193 187 184

FusionTime

(minute)2.5 2.2 1.8 1.1 0.9

FusionTorque(g*m)

2244 2223 2339 2415 2559

Table 6.6 Fusion characteristics of PVC/CPE/0.3phrOPEcompounds at temperature = 190°C, rotor speed = 60 r p m .

lphr 3phr 5phr lOphr 15phr

Fusion T e m p . (°C)

194 193 188 179 174

FusionTime

(minute)1.8 1.6 1.2 0.7 0.5

FusionTorque(g*m)

2256 2368 2465 .5 2619 2773

169

a) £LU1 |im PVC

Microparticle Flow Unit

b)

CALCIUM STEARATE

c)

PARAFFIN WAX

£LU2

PVC

CALCIUM STEARATE

d)W(

PARAFFIN WAX

Figure 6.6 A postulated lubrication mechanism, showing metal lubrication and lubrication between PVC microparticle flow units, of PVC compounds.

Fusio

n Ti

me

(m

inu

te)

170

PVC/CPE

PVC/CPE/O PE2. 5-

V

+

0. 5-

0 1 3 5 10 15

phr of CPE

Figure 6.7 Fusion time curves, varying with the phr of CPE, of PVC/CPE and PVC/CPE/0.3phrOPE compounds.

171that the fusion times of PVC/CPE/0.3phrOPE compounds decreased as the phr of CPE was increased. It also shows that the fusion time curve of PVC/CPE compounds has a plateau (between 0 phr and 3 phr CPE) , then decreased as the phr of CPE was increased.

For PVC/CPE compounds, a lower phr of CPE (less than 10 phr) may function like an external lubricant which decrease the gummy property of calcium stearate and form a mobile layer among PVC particles, therefore, delays the fusion time. When increasing the phr of CPE in PVC/CPE compounds, sufficient CPE melt can act as a process aid, which allows the PVC particles to fuse together easily. ForPVC/CPE/0.3phrOPE compounds, a higher phr of CPE may result in more effective and powerful material, similar to a glue, due to the synergistic reaction between CPE particles and

OPE wax. This material can easily destroy the boundary of a PVC particle, increase the friction and improve the fusion

at a shorter time.Figure 6.8 illustrates the changes of fusion

temperature (+) and fusion torque (□) of, varying with the phr of CPE, PVC/CPE/0.3phrOPE compounds. It shows the fusion temperature increases with fusion time and the fusion torque decreases with fusion time (same reason as mentioned

above).

Fusio

n T

empe

ratu

re

( C

)

172

2 2 0 1

210 ^

2 0 0 -

190-

180

170-

160-

150-

140

130-

1 2 0 -

110 -

100-

Tem perature: 190 C

Rotor Speed : 60 rpm

+ + +

0 1 3 5

+□

10

phr of CPE

+

r 3000

-2900

-2800 If*

-2700 32600 <D

32500 D"i_

O2400 1 -

c2300 o

‘c/52200

-2100

-2000

3LL

15

Figure 6.8 Changes of fusion temperature (+) andfusion torque (□) of, varying with the phr of CPE, PVC/CPE/0.3phrOPE compounds.

173Tables 6.7 and 6.8 illustrate the influence of varying

phr of OPE on the fusion characteristics of PVC/OPE and PVC/5phrCPE/0PE compounds prepared at a temperature of 190°C and a rotor speed of 60 rpm. Figure 6.9 illustrates the fusion time curves, varying with the phr of OPE, of PVC/OPE and PVC/5phrCPE/OPE compounds. It illustrates that the fusion times of PVC/5phrCPE/OPE compounds decreased as the phr of OPE was increased. It also shows that the fusion time curve of the PVC/OPE compounds passes a plateau (between 0.1 phr and 0.3 phr OPE) , then decreased as the phr of OPE was increased.

For PVC/OPE compounds, a lower phr of OPE (less than1.0 phr) may function like an external lubricant and result

in a mobile layer which increases the separation and slipping among PVC particles and delay the fusion times. When increasing the phr of OPE in PVC/OPE compounds, the interaction between OPE and calcium stearate, may act as a process aid, which allows the PVC particles to fuse together easily [Lindner 1989]. For PVC/5phrCPE/OPE compounds, a higher phr of OPE may result in more effective and powerful material, similar to a glue, due to the synergistic reaction between CPE, OPE, and calcium stearate. This material can easily destroy the boundary of a PVC particle, increase the friction and improve the fusion at a shorter time (similar

result as mentioned above).

174

Table 6.7 Fusion characteristics of PVC/OPE compounds at temperature = 190°C, rotor speed = 60 rpm.

0.lphr 0.3phr 0.5phr 1.Ophr 1.5phr

Fusion T e m p . (°C)

195 196 194 188 187

FusionTime

(minute)2.2 2.3 1.8 1.2 1.0

FusionTorque(g*m)

2079 2116 2223 2351 2404

Table 6.8 Fusion characteristics of PVC/5phrCPE/OPEcompounds at temperature = 190°C, rotor speed = 60 r p m .

0.lphr 0 . 3phr 0.5phr 1.Ophr 1.5phr

FusionTemp.(°C)

194 188 185 184 174

FusionTime

(minute)1.7 1.2 0.9 0.8 0.5

FusionTorque(g*m)

2367 2465.5 2447 2527 2692

Fusio

n Ti

me

(min

ute)

175

PVC/OPEPVC/CPE/OPE2.5-

0.5-

0.5 1.00 0.1 0.3 1.5

phr of CPE

Figure 6.9 Fusion time curves, varying with the phr of OPE, of PVC/OPE and PVC/5phrCPE/OPE compounds.

176Figure 6.10 illustrates the changes of fusion

temperature (+) and fusion torque (□) of, varying with the phr of OPE, PVC/5phrCPE/0PE compounds. It shows the fusion temperature increases with fusion time and the fusion torque decreases with fusion time (same reason as mentioned above).

6.4.2 SEM AnalysisFigures 6.11, 6.12, 6.13, and 6.14 show the surface

morphology of PVC, PVC/OPE, PVC/CPE, and PVC/CPE/OPE compounds which were prepared in the Haake Torque Rheometer at temperature = 190°C, rotor speed = 60 rpm, and TTQ = 1 kg-m-min, respectively. There is not much difference between Figure 6.11 and Figure 6.14. Figures 6.12 and 6.13, however, illustrate that more PVC resin particles remain intact. This phenomenon results in delaying the fusion

times of PVC/OPE and PVC/CPE compounds (Table 6.2).Similar morphological results were obtained for PVC

blends which were prepared in the Haake Torque Rheometer at temperature = 190°C, rotor speed = 100 rpm, and TTQ = 1 kg- m-min or at temperature = 175°C, rotor speed = 100 rpm, and TTQ = 1 kg-m-min.

Figures 6.15 and 6.16 reveal the surface morphology of PVC/OPE compounds which were prepared in the Haake Torque Rheometer at temperature = 175°C and rotor speed = 60 rpm

Fusio

n T

empe

ratu

re

( C

)

177

220-

2 1 0 -

200-

190-

180-

170-

160-

150-

140-

130-

120-

110 -

100 -0 0.1

Temperature: 190 C

Rotor Speed : 60 rpm

+ +

0.3 0.5

+□

1.0

phr of OPE

2900

2800

2700 'e*■2600 3-2500

-2400

<DCJST£

2300 co2200 V)=s

LI-2100

-20001.5

Figure 6.10 Changes of fusion temperature (+) andfusion torque (□) of, varying with the phrof OPE, PVC/5phrCPE/OPE compounds.

178

Figure 6.11 Surface morphology of PVC prepared ina Haake Torque Rheometer at temperature= 190°C, rotor speed=60 rpm, and TTQ=1 kg-m- min .

Figure 6.12 Surface morphology of PVC/OPE preparedin a Haake Torque Rheometer at temperature= 190°C, rotor speed=60 rpm, and TTQ=1 kg-m- min .

180

Figure 6.13 Surface morphology of PVC/CPE preparedin a Haake Torque Rheometer at temperature= 190°C, rotor speed=60 rpm, and TTQ=1 kg-m- min .

181

Figure 6.14 Surface morphology of PVC/CPE/OPEprepared in a Haake Torque Rheometer at temperature=190°C, rotor speed=60 rpm, and TTQ=1 kg-m-min.

Figure 6.15 Surface morphology of PVC/OPE preparedin a Haake Torque Rheometer at temperature= 175°C, rotor speed=60 rpm, blending time=15 minutes.

183

Figure 6.16 Surface morphology of PVC/OPE preparedin a Haake Torque Rheometer at temperature= 175°C, rotor speed=100 rpm, and blending time=15 minutes.

and 100 rpm, respectively. Even though the blending time

was 15 minutes, no fusion occurred (Tables 6.5 and 6.6). Figures 6.17, 6.18, and 6.19 reveal the surface morphology of PVC, PVC/CPE, and PVC/CPE/OPE compounds which were prepared in the Haake Torque Rheometer at temperature = 175°C, rotor speed = 60 rpm, and TTQ = 1 kg-m-min,respectively. In Figure 6.19, more PVC resin particles were

broken into PVC microparticles than those in Figures 6.17 and 6.18. Therefore, the fusion time of the PVC/CPE/OPE compound is the shortest. The fusion time of the PVC/CPE/OPE compound is more significantly different than

those of PVC and PVC/CPE compounds at the medium starting temperature and the medium rotor speed (Table 6.4). This may be due to the interaction of CPE and OPE (as mentioned before). More PVC resin particles remain intact in Figure 6.18; this results in a longer fusion time for the PVC/CPE compound (Table 6.4) . Without the CPE particles, the fusion time of the PVC compound is between that of the PVC/CPE compound and that of the PVC/CPE/OPE compound.

Figures 6.20, 6.21, 6.22, and 6.23 reveal the surface morphology of PVC, PVC/OPE, PVC/CPE, and PVC/CPE/OPE compounds which were prepared in the Haake Torque Rheometer at temperature = 160°C, rotor speed = 20 rpm, and TTQ = 15 kg-m-min, respectively. These four figures illustrate the significantly different fusion phenomena among PVC, PVC/OPE,

185

Figure 6.17 Surface morphology of PVC prepared ina Haake Torque Rheometer at temperature= 175°C, rotor speed=60 rpm, and TTQ=1 kg-m- min .

186

.AV-j

Figure 6.18 Surface morphology of PVC/CPE preparedin a Haake Torque Rheometer at temperature= 175°C, rotor speed=60 rpm, and TTQ=1 kg-m- min .

187

500UIH

Figure 6.19 Surface morphology of PVC/CPE/OPEprepared in a Haake Torque Rheometer at temperature=175°C, rotor speed=60 rpm, and TTQ=1 kg-m-min.

188

Figure 6.20 Surface morphology of PVC prepared ina Haake Torque Rheometer at temperature= 160°C, rotor speed=20 rpm, and TTQ=15 kg-m- min .

189

Figure 6.21 Surface morphology of PVC/OPE prepared in a Haake Torque Rheometer at temperature= 160°C, rotor speed=20 rpm, and TTQ=15 kg-m- min .

V

190

Figure 6.22 Surface morphology of PVC/CPE prepared in a Haake Torque Rheometer at temperature =160°C, rotor speed=20 rpm, and TTQ=15 kg- m-min .

191

Figure 6.23 Surface morphology of PVC/CPE/OPEprepared in a Haake Torque Rheometer at temperature=160°C, rotor speed=20 rpm, and TTQ=15 kg-m-min.

PVC/CPE, and PVC/CPE/OPE compounds. In Figures 6.21 and 6.22, most of the PVC resin particles remain intact. In Figure 6.20, some of the PVC resin particles still remain intact. In Figure 6.23, however, the PVC resin particles were fused together and formed a 3-D network. This phenomena reveals that the interaction between CPE and OPE has a very significant influence upon the degree of fusion of PVC blends, especially those which were prepared at a low starting temperature and rotor speed.

CHAPTER 7CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK

In chapter two, the morphology of a series of various concentrations of CPE in PVC/CPE blends was examined in SEM, TEM, and S-TEM. Using SEM, the CPE particles in PVC/CPE blends cannot be discerned even though this technique does

reveal similar surface morphologies in all PVC/CPE blends. Although the TEM technique allows clearer determination of phase boundaries, this technique is time consuming because of the required sample staining. Furthermore, the TEM staining process does not stain all of the CPE phase and this procedure may alter the CPE phase. Three approaches may improve the staining effect: increasing reactiontemperature of dehydrochlorination, increasing 0s04 concentration, or extending staining time more than two weeks. The S-TEM technique is fast and does not require further chemical treatments. Additionally, this technique does not alter the CPE phase. One disadvantage of this

method is that it does not provide as clear a phase discrimination of the boundaries between PVC and CPE as TEM analysis.

Brittle samples (low IZOD impact value) display no significant difference between the perpendicular and parallel directions, but ductile samples (high IZOD impact

193

194value) do display significant differences (i.e. CPE particles are highly elongated in the parallel direction). The domain sizes of CPE in PVC/CPE blends increase with IZOD impact values, but decreases as the melt indexes increase.

In chapter three, it was illustrated that at the same rotor speed of 60 rpm, the fusion time decreases with increasing starting temperatures of the mixer. Similarly, at the same starting temperature, the fusion time decreases

with increasing rotor speeds. S-shaped fusion curves were obtained. Starting temperature of the mixer, rotor speed, and totalized torque are the three major factors that affect the degree of fusion of PVC compounds blended in the Haake Torque Rheometer. In order to obtain high impact strength PVC compounds, optimal blending conditions may be to set the starting temperature of the mixer at 180 or 190°C, the rotor speed at 60 or 70 rpm, and the totalized torque at 10 kg-m- min. Both capillary rheological and DSC thermal analysis can be applied to determine the degree of fusion of a PVC compound.

The capillary rheological analysis has a main advantage in that this method uses a sample large enough to eliminate local inhomogeneity in the degree of fusion. The two main disadvantages of this method are that the preheating process may interfere with the degree of fusion of PVC and the necessity of getting a standard fusion curve for each

195compound processed in certain processing equipment. Also the elastic response, as measured by the rheological analysis, is strongly composition-dependent [Krzewki, 1981].

The DSC thermal analysis has attracted attention recently because it provides a convenient, quantitative measure of the fusion level and uses a very small sample (approximately 10 mg) . The main disadvantage of thermal

analysis is that some additives may interfere with the final results because the sample size is very small.

In chapter 4, the simple total energy balance based on Figure 4.1 [Pedersen, 1991]] can be applied to correlate the compounding energy with the fusion level of the PVC compound prepared in the Haake Torque Rheometer. Both DSC thermal analysis and capillary rheological analysis can be used to evaluate the fusion level of PVC compounds.

In chapter 5, both the DSC thermal analysis and

capillary rheological analysis were used to evaluate thedependent variable of this 23 experimental design. Similarresults have been obtained. The sequence of the main effects on the heat of fusion of PVC compounds, in ascending order, is temperature < rotor speed < TTQ. The sequence of the two-factor interaction effects on the heat of fusion of PVC compounds, in ascending order, is temperature vs. rotor speed < temperature vs. TTQ < rotor speed vs. TTQ. The

196three-factor interaction effect is not significantly related to the heat of fusion of PVC compounds.

In chapter six, the pictures, similar to Rabinovitch's theory, of lubrication mechanisms of these four PVC blends have been postulated. Due to a mobile layer, resulting from the synergistic combined lubricity between OPE and calcium stearate, which increases the separation and slipping among PVC particles, the fusion time of PVC/OPE compounds is the longest among these four PVC blends.

However, the fusion time of PVC/CPE/OPE compounds is

the shortest among these PVC blends. This may be due to the fact that OPE and CPE interact and form a powerful, viscous material and appears to act as a glue, which enables the PVC resin particles to fuse together easily. The fusion time of the PVC/CPE/OPE compound is significantly different from those of PVC, PVC/OPE, and PVC/CPE compounds at the medium starting temperature and the medium rotor speed. Because a high starting temperature or a high rotor speed improves the

gummy property of calcium stearate [Lindner, 1981], the fusion times of PVC compounds are very close to those of PVC/CPE/OPE compounds in Tables 6.1, 6.2, and 6.3. In this case, calcium stearate functions as a process aid.

Since the fusion time of PVC/OPE compounds is the longest, it can be concluded that more energy is needed to be absorbed by PVC/OPE compounds in order for the PVC

197particles to fuse together, thus, the fusion temperatures of PVC/OPE compounds are the highest. Because a higher fusion temperature results in decreasing the melt viscosity of samples in the Haake Torque Rheometer, the fusion torques of PVC/OPE compounds are the lowest.

Figure 6.7 illustrates that the fusion times of these two compounds decreased as the phr of CPE increased. For PVC/CPE compounds, the lower phr of CPE may be due to less compatibility with PVC and decreasing the gummy property of calcium stearate, therefore the longer fusion time occurs. When increasing the concentration of CPE in PVC/CPE compound, sufficient CPE melt can act as a processing aid which helps PVC particles to fuse together easily. For PVC/CPE/OPE compounds, a higher phr of CPE may result in more effective and powerful material, similar to a glue, due to the synergistic reaction between CPE particles and OPE wax. This material can easily destroy the boundary of a PVC

particle, increase the friction and improve the fusion in a shorter amount of time. Meanwhile, fusion torques and fusion temperatures increased with fusion time (same as mentioned above).

SEM analyses successfully revealed the surface morphological changes of the fusion of PVC, PVC/OPE, PVC/CPE, and PVC/CPE/OPE compounds, which were prepared at

198various conditions. Based on the above conclusions, the following recommendations for future work are offered.

(1) In order to obtain a comprehensive knowledge of the influence of impact modifiers, one may prepare PVC compounds, containing two, three, or four different impact modifiers with various phr. Then correlate their morphological changes, characterized by SEM, TEM, and S-TEM, with their mechanical properties (IZOD impact and tensile property). Moreover, study the lubrication mechanism for PVC/OPE compounds, blended with various impact modifiers

instead of only CPE. Through this study, one may find a

more suitable formulation for PVC compounding.(2) In order to obtain more practical results and apply

this knowledge in order to produce PVC end products through extrusion, one may change the internal mixer system to an extrusion system and study the fusion characteristics and the morphological changes of extruded PVC compounds.

(3) In order to have a better understanding of the influence of the interaction between OPE and CPE on fusion characteristics, one may use OPE's with various acid number, CPE's with various chlorine content, and other calcium soaps to process PVC compounds and examine the morphological changes of these compounds. Through this study, one may obtain a better lubricant system for PVC processing.

(4) Develop new PVC resins or copolymers of vinyl chloride monomer (VCM) and other monomers in order to overcome the poor physical properties of pure PVC resins. Try to prepare and study the characteristics of atactic, isotactic, and syndiotactic PVC resins. The physical properties of PVC resins may vary with different crystallinity, configuration, and molecular weight. Through this study, one may obtain PVC resins with more suitable processing properties without using such a large number of

additives.

REFERENCES

Allsopp, M. W . , "Manufacture and Processing of PVC",Burgess, R. M. (Ed), Applied Science Publishers Ltd.,London, (1982).Bagley, E. B., Transaction Society of Rheology, 5, 355(1961).Bagley, E. B., Schreiber, H. P., "Rheology-Theory andApplications", Eirich F. R. (Ed), Academic Press, New York, (1969) .Bambrick, C. R., Society of Plastics Engineers ANTEC, II, 1797 (1993) .Benjamin, P. , Processing of Plastics and Rubber Institute of International Conference, PVC Processing, April, B51, Egham Hill, England, (1978).Blackson, J. H., and Liou, D. W., Institute of Physics Conference Series. No.93, 2, 457 (1988).Bower, J. D., Plastics Compounding, January/February, 64(1979).Bower, J. D., Journal of Vinyl Technology, 8, 179 (1986) .

Bozzola, J. J., and Russell, L. D., "Electron Microscopy : Principles and Techniques for Biologists", Jones and Bartlett Publishers, Boston (1992).Buskirk, P. R. Van, Turetsky, S. B., and Gunberg, P. F. , abstracted by Blake W. T., "Haake Buchler Technical Abstract", TA-56.Butters, G., "Particulate Nature of PVC-Formation, Structure, and Processing", Applied Science Publishers Ltd., London (1982).Chang, B. H. , Zeigler, R., and Hiltner, A., Polymer Engineering and Science, 28, 1167 (1988) .Chen, Y., Kalyon, D. M., and Bayramli, E., Society of Plastics Engineers ANTEC, II, 1747 (1992) .Chung, J. T., "Haake Buchler Technical Bulletin", TB-842.

200

201Cook, P. J. , Society of Plastics Engineers ANTEC, I , 526 (1974) .Ditto, P. E., Journal of Vinyl Technology, 4, 124 (1982).McMurrer, M. C., Plastics Compounding, July/August, 74(1982) .Doube, C. P., and Walsh, D. J., Polymer, 20, 1115 (1979).Faulkner, P. G., Journal of Macromolecular Science-Physics, Bll, 251 (1975).Fleischner, D., Fisher, E., and Brandrup, J., Journal of Macromolecule Science-Physics, B14, 17 (1977).Gilbert, M. , Plastics and Rubber International, 10, 16(1985) .Gilbert, M. , Hemsley, D. A., and Miadonye, A., Plastics and Rubber Processing and Applications, 3, 343 (1983) .Gilbert, M . , and Vyvoda, J. C., Polymer, 22, 1134 (1981) .Gonze, A., Plastica, 24, 49 (1971).Gray, A., Processing of Plastics and Rubber Institute of International Conference, PVC Processing, April, 10.1, Egham Hill, England, (1978).Hahn, G. J., Chemical Technology, 5, 496 (1975).Hahn, G. J . , Chemical Technology, 5, 561 (1975).Han, C. D., Transactions of The Society of Rheology, 18.1, 163 (1974) .Han, C. D., "Rheology in Polymer Processing", AcademicPress, New York, (1976) .Hartitz, J. E., Polymer Science and Engineering, 14, 392(1974) .Hori, Y., Japan Plastics, 3, >48 (1969).Kato, K . , Polymer Engineering and Science, 7, 38 (1967) .Krzewki, R. J., and Collins, E. A., Journal of Macromolecule Science-Physics, B20, 443 (1981).

202Krzewki, R. J., and Collins, E. A., Journal of Macromolecule Science-Physics, B20, 465 (1981).

Lamberty, M., Plastiques Modernes et Elastomeres, 26, 82(1974) .Lindner, R. A., Plastics Compounding, July/August, 50(1989) .Lindner, R. A., Plastics Compounding, September/October, 35(1981) .Luis, R. V., and Gilbert, M . , Plastics and Rubber Processing and Applications, 13, 151 (1990).Luis, R. V., and Gilbert, M., Plastics and Rubber Processing and Applications, 13, 157 (1990).Marshall, G. P., and Birch, M. W., "Processing of SERC PED Review Meeting", Loughborough, (1981).

Nass, L. I., " Encyclopedia of PVC, Volume 3: CompoundingProcesses, Product Design, and Specifications ", Second Edition, Marcel Dekker Inc., New York and Basel, (1992) .Nielsen, L. E., " Polymer Rheology ", Marcel Dekker Inc.,New York and Basel, (1977).Obande, 0. P., Journal of Applied Polymer Science, 42, 1433 (1991) .Patel, S. V., and Gilbert, M . , Plastics and Rubber Processing and Applications, 6, 321 (1986).Pedersen, T. C., Vinyl RETEC 1991, "Applying 90's Technology to Vinyl: Products, Processing, Methodology, and Ecology", 31 (1991).Philippoff, W., and Gaskins, F. H., Transaction Society o f , Rheology, 2, 263 (1958).Rabinovitch, E. B., Journal of Vinyl Technology, 4, 62(1982) .Rabinovitch, E. B . ; Lacatus, E., and Summers, J. W., Journal of Vinyl Technology, 6, 98 (1984).Rapetski, W. A., Vinyl RETEC 1993, "Vinyl 201: AComprehensive Processing Update", 17 (1993).

203Riley, D. W., Society of Plastics Engineers ANTEC, I, 890(1983) .Schmidt, S. R., and Launsby, R. G . , "Understanding Industrial Designed Experiments - 3rd Edition", Air Academy Press, Coloado Springs, Colorado, (1992).Siegmann, A., and Hiltner, A., Polymer Engineering and Science, 24, 869 (1984).Summers, J. W., and Rabinovitch, E. B., Journal of Macromolecular Science-Physics, B20, 219 (1981).Summers, J. W. , Rabinovitch, B. E., and Booth, P. C. Journal of Vinyl Technology, 8, 2 (1986).Teh, J. W., Cooper, A. A., Rudin, A., and Batiste, J. L. H . ,Journal of Vinyl Technology, 11, 33 (1989).Teh, J. W., Cooper, A. A., Rudin, A., and Batiste, J. L. H.,Makromolekulare Chemie, Macromolecular Symposia, 29, 123(1989) .Titow, W. V., "PVC Plastics Properties, Processing and Applications", Elsevier Science Publication Ltd., London and New York (1990).

Titow, W. V., "PVC Plastics Properties, Processing and Applications", Elsevier Science Publication Ltd. , London and New York (1990).

Titow, W. V., "PVC Plastics Properties, Processing and Applications", Elsevier Science Publication Ltd., London and New York (1990).Yu, A. J., Boulier, P., and Sandhu, A. Society of Plastics Engineers ANTEC, I, 83 8 (1984).

VITA

Cheng-Ho Chen was born on September 16, 1963, in Chu- Nan, Taiwan, the third child of his parents. He received Bachelor and Master Degrees from Chemical Engineering Department of National Cheng-Kung University in Taiwan in June, 1986 and June, 1988, respectively. He is currently pursuing a doctorate in the Chemical Engineering Department at Louisiana State University.

204

DOCTORAL EXAMINATION AND DISSERTATION REPORT

Candidate: Cheng-Ho Chen

Major F ield : Chemical Engineering

T it le o f D isserta tion : The Studies of Rigid PVC Compounds: Morphology,Rheology, and Fusion Mechanism

Approved:

Major Professor and Chairman

EXAMINING COMMITTEE:

Date o f Examination:

May 6 , 1994