 

 

V /
.IIIIIIII'IIIII‘IIIII

()3 ‘ a

This is to certify that the
thesis entitled

BIOCOMPOSITES FROM ENGINEERED NATURAL FIBERS AND
UNSATURATED POLYESTER RESIN FOR HOUSING PANEL
APPLICATIONS

presented by

GEETA MEHTA

has been accepted towards fulfillment
of the requirements for the

MASTER OF degree in Chemical Engineering

 

 

 

 

SCIENCE
Ma or Professoyr’s Sbnﬂure
a0, 3009!
Date

MSU is an Affirmative Action/Equal Opportunity Institution

' -~‘-—--n-n—-o-.--o-o-a-oon-.---.-.-.-.-.

 

LIBRARY ‘
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:lCIRC/DateDue.p65op.15

 

 

BIOCOMPOSITES FROM ENGINEERED NATURAL FIBERS AND
UNSATURATED POLYESTER RESIN FOR HOUSING PANEL APPLICATIONS

By

Geeta Mehta

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemical Engineering and Material Science

2004

ABSTRACT

BIOCOMPOSITES FROM ENGINEERED NATURAL FIBERS AND
UNSATURATED POLYESTER RESIN FOR HOUSING PANEL APPLICATIONS

By

Geeta Mehta

The aim of this project is to make low cost ‘green materials' (biocomposites) for use in
various housing panel applications, for example, wall panels, roofs, doors, floors, etc, to
compete with and substitute for glass-polyester panel systems used presently. The
starting materials for these products are natural ﬁbers as reinforcements, and unsaturated
polyester resin as polymer matrix. The project consists of three interconnected parts:
modiﬁcation and engineering of natural ﬁbers to make them suitable as reinforcements in
polymer composites, alteration of the polymer composite matrix to make it more ‘green’
while maintaining its properties and performance, and the development of a sheet
molding compounding panel process (SMC) as a new way for continuous production of
biocomposites. The research on ‘designed’ ﬁber reinforcement sought to synergistically
combine up to four different ﬁbers, which have been effectively surface treated,
physically or chemically, to improve the ﬁber matrix adhesion in the resulting bio-
composite. Research was also being conducted to modify the polyester polymer matrix
through the addition of modiﬁed vegetable oils while maintaining properties and
processability. The functionalization of vegetable oils is done in a manner that it is
compatible with polyester resin. We have also been successful in developing a SMC

process for the continuous manufacturing of bio-composites using chopped natural ﬁbers.

 

C0pyn'ght by
GEETA MEHTA
2004

DEDICATION

And Life, a little bald and gray,
Languid, fastidious, and bland,
Waits, hat and gloves in hand,
Punctilious of tie and suit
(Somewhat impatient of delay)
On the doorstep of the Absolute.

T. S. Eliot January 1910

This work is dedicated to my father, mother, sisters, and niece, who have given me the
wings to ﬂy, strength to persist, inspiration to move forward, faith to be brave in
desperate situations, and love to conquer it all. A special dedication to my best ﬁ‘iend,
philosopher and guide, who has always been there for me through all thick and thin. I
thank all of you, I’m so lucky to have you all. Words can not express the warmth of love
that I feel all around me because of you all.

ACKNOWLEDGEMENTS

I would like to thank Dr. Lawrence T. Drzal, my research advisor, for giving me the
opportunity to work on the wonderful area of bio-composites. He has provided me with
much opportunity to travel, network, and present my own work in all settings, showing
his dedication to my academic and professional growth. My experience of working with
him has made me realize exactly how to handle my future graduate students or
employees. He was instrumental in writing this dissertation, and I am thankful for his
guidance.

I am indebted to Dr. Manjusri Misra, my supervisor for all her guidance, encouragement,
and support throughout my graduate study. She acted as the driving force behind this
research. She provided the opportunity for each student to do his/her own research, yet
when asked, she was always willing to provide her knowledge and expertise to assist in
understanding the problem. Her professional and personable attitude has made my stay at
Michigan State a very rewarding and memorable experience. I have enjoyed her
mentorship and ﬁiendship, and have learned about life as well as research from her. She
always has an answer for experimental challenges. Her dedication to the work of the
students and postdocs of the CMSC goes above and beyond the call of duty and her
expertise has proven priceless.

I would also like to thank Dr. Amar K. Mohanty, who not only wrote the original NSF-
PATH proposal, but also guided my research. He is one of the most hard working and

motivating individuals I have ever come across. When graduate life was bleak, it was his

 

 

tough love therapy that got me through the dark patches. I have always cherished his
honesty and forthrightness, and his wonderful habit of helping people, especially
students. I am thankful for this thoughtfulness and his dedication towards my project.
Special thanks and recognition also go to Dr. Rigoberto Burgueno, who has always been
extremely generous with his time, whether I needed technical or professional advice. I
am grateful to him for serving on my committee, for helpful discussions, and for
reviewing my research program. He has been a great counselor and I wish him all the
luck in the world for his future.

Also, I am grateful to acknowledge all of past and present CMSC group members who
provided not only scientiﬁc, but also moral support, and most of all friendship, and
camaraderie throughout my study and research. Some of the special people I would like
to acknowledge particularly are, Arief, LaKeya, Mario, Jean Pierre, Praveen, Masud,
Hwan Man, Tao, Dana, Wanjun, and, Prasad who made coming into the lab fun and
interesting and were always ready to lend a helping hand.

My utmost appreciation and thanks are given to my best friend, Pinks, for all the love and
support throughout my graduate career. My best friend has been my rock to whom I
could always turn for comfort, understanding, and motivation. He made me see the silver
lining in the clouds when all I could see was an abysmal black hole. My family has
always stood by me, through all the ups and downs, and has been very patient and
encouraging in everything I have ventured to do. Ma and Papa sacriﬁced their lives for us
kids. And I wouldn’t be myself without my lovely sisters Gudia and Pooja. My niece J iya

has ﬁlled our lives with joy. I could not have made it this far without them.

vi

I am thankful to Mr. Michael Rich and Mr. Kelby Thayer for opening my eyes to US
work culture. Kelby also helped me with the SMC line. Bob ‘Jurek and Brian Rook have
always been accommodating with all my queries.

Jean Rooney and Karen Lillis, who keep things running in the CMSC, whether this calls
for ordering materials or remembering somebody on a special day. The work they do is
invaluable, and without them, students, staff and faculty alike would be lost.

Special thanks are given to many faculty and staff members in the Department of
Chemical Engineering and Materials Science, and the Graudate School, Michigan State
University for their assistance during my graduate study.

Finally, I would like to acknowledge the ﬁnancial support provided by Dr. Drzal through
grants from the NSF -PATH (National Science F oundation-Partnership for Advanced
Technologies in Housing, Award number 0122108). I am also thankful to Kemlite Inc.
and Flax Craft Inc., for their participation

1 am exceedingly appreciative of the Department of Chemical Engineering and Materials
Science of Michigan State University for providing me with graduate assistantship,

research assistantships and fellowship.

vii

TABLE OF CONTENTS

LIST OF TABLES ................................................................................... xi

LIST OF FIGURES ................................................................................ xiii

CHAPTER ONE

1 INTRODUCTION ......................................................................... 1
References ................................................................................... 11

CHAPTER TWO

2 LITERATURE REVIEW ................................................................. 13
2.1 Natural Fibers ..................................................................... 13
2.2 Unsaturated polyester resin ..................................................... 34
2.3 Bioresins .......................................................................... 70
2.4 Processing of composites ....................................................... 77
2.5 Weathering ........................................................................ 89
References ................................................................................... 95

CHAPTER THREE

3 EXPERIMENTAL METHODS ......................................................... 110
3.1 Materials .......................................................................... 1 10
3.2 Surface Modiﬁcations of natural ﬁbers ....................................... 113
3.3 Bioresin ........................................................................... 118
3.4 Fabrication of composites ...................................................... 122
3.5 Testing ............................................................................. 151

3.5.1 Natural ﬁber ............................................................. 151
3.5.2 Bioplastics and biocomposites ....................................... 158

CHAPTER FOUR

4 RESULTS ON ENGINEERED FIBERS .............................................. 173
References ................................................................................. 231

CHAPTER FIVE

5 RESULTS ON MATRIX MODIFICATION ......................................... 233
5.1 Curing Kinetics ................................................................... 233
5.2 Bioplastics from Bioresins ...................................................... 253
References ................................................................................. 294

viii

CHAPTER SIX

6 RESULTS ON NEW PROCESSING .................................................. 296
Biocomposites from SMC ............................................................... 296
References ................................................................................. 324

CHAPTER SEVEN

7 RESULTS ON DURABILITY ......................................................... 325
7.1 Moisture Absorption ............................................................ 325
7.2 QUV Accelerated Weathering ................................................. 330
References ................................................................................. 340

CHAPTER EIGHT

8.0 CONCLUSIONS AND FUTURE WORK ............................................ 341
8.1 CONCLUSIONS ................................................................ 341
8.2 FUTURE WORK ................................................................ 347

ix

LIST OF TABLES

Table 2.1.1 Typical properties of glass ﬁber and natural ﬁbers ............................... 18

Table 2.2.] Cost ofvarious polyester resins. .35

Table 2.2.2 Examples of monomers used for UPE system cross linking .................... 39
Table 2.2.3 Examples of inhibitors used for UPE system ..................................... 42
Table 3.1.1 Reinforcements for composites made in this project ........................... 111
Table 3.1.2 Materials and suppliers ............................................................ 112
Table 3.2.1 Pre-treatment and post treatment for surface treatments of the natural
ﬁbers ................................................................................................ 1 17
Table 3.3.1 Bioresins from grafting of vegetable oils ........................................ 119
Table 3.3.2 Bioplastics obtained from graﬂed oils ............................................ 120
Table 3.4.1 Samples made using silicone molds .............................................. 124
Table 3.4.2 Samples made using compression molding ..................................... 127
Table 3.4.3 Samples made using SMC line .................................................. 142
Table 4.1.1 Thermogravimetric results for surface treated hemp ﬁbers ................... 175

Table 4.1.2 Elemental composition of surface treated hemp ﬁbers (from XPS
analysis) ............................................................................................. 17 8

Table 4.1.3 Elemental ratios of surface treated hemp ﬁbers (ﬁ'om XPS analysis) ....... 178
Table 4.2.1 Thermogravimetric results for surface treated pure hemp ﬁbers. . . . . . ....196

Table 4.2.2 Elemental composition of surface treated pure hemp ﬁbers (from XPS
analysis) ............................................................................................. 199

Table 4.2.3 Elemental ratios of surface treated pure hemp ﬁbers (from XPS
analysis) ............................................................................................. 199

Table 4.3.1 Thermogravimetric results for surface treated kenaf ﬁbers . .......217

Table 4.3.2 Elemental composition of surface treated hemp ﬁbers (from XPS analysis)
......................................................................................................... 219

Table 4.3.3 Elemental ratios of surface treated hemp ﬁbers (from XPS analysis) . ..219

Table 5.1.1 Comparison of maximum conversion at different temperatures from FTIR

experiments .......................................................................................... 252
Table 6.1 Maximum degradation temperatures for natural ﬁbers .......................... 301
Table 6.2 Atomic concentrations on the surfaces of ﬁbers used for SMC line ........... 301
Table 6.3 Ratio of atomic concentrations of ﬁbers ........................................... 302
Table 7.1.1 Maximum degradation temperature of the bioplastics ........................ 330

xi

LIST OF FIGURES

Figure 1.1 Schematic representation of the research plan for this project .................... 6
Figure 2.1.1 Examples of natural ﬁbers .......................................................... 14
Figure 2.1.2 Classiﬁcation of bioﬁbers .......................................................... 16
Figure 2.1.3 Comparison of cost and density between glass ﬁbers and bioﬁbers. . . . . .....16
Figure 1.1.4 Chemical structure of cellulose molecules: Poly-B(1,4)-D-Glucose. . . ..18
Figure 2.1.5 Comparison of moduli of glass ﬁbers and bioﬁbers ............................ 20
Figure 2.1.6 Comparison of natural ﬁber (NF) and glass ﬁber (GF)—unsaturated polyester
(UPE) composites ................................................................................... 23
Figure 2.1.7 North American major end markets of natural ﬁber composites, year
2000 ................................................................................................... 24
Figure 2.2.1 Matrix pattern of polymer composites ............................................ 35
Figure 2.2.2 A representative chemical structure of unsaturated polyester ................. 36
Figure 2.2.3 Preparation of polyester resin ...................................................... 38
Figure 2.2.4 Chemical structure of methyl ethyl ketone peroxide (MEKP) ................ 40
Figure 2.2.5 Chemical structure of benzoyl peroxides (BPO) ............................... 40
Figure 2.2.6 Reaction mechanism of polyester resin with cobalt activator . .....41
Figure 2.2.7 Crosslinking mechanism of UPE resins .......................................... 43
Figure 2.2.8 Curing reaction rate of UPE ........................................................ 45
Figure 2.2.9 A depiction of a simple IPN material ............................................. 69
Figure 2.2.10 Soybean plant ....................................................................... 72
Figure 2.2.10 Flax plant and ﬂower .............................................................. 73

Figure 2.2.1] Photograph of ESO-clay nanocomposite showing high ﬂexibility. . . . . ....77
Figure 3.1.1 Hemp mat with ﬁller (Bastrnat 115) obtained from F laxCraﬁ Inc .......... 112

Figure 3.2.1 Vacuum oven for drying bioﬁbers ............................................... 114

xii

Figure 3.4.1 Conventional oven for curing silicone molds ................................. 124
Figure 3.4.2 Carver compression Press ......................................................... 126

Figure 3.4.3 Picture of bioplastic samples made using compression molding

........................................................................................................ 129
Figure 3.4.4 Scheme of BCSMCP process .................................................... 132
Figure 3.4.5 Scheme of vibratory feeder and screw feeder ................................ 134
Figure 3.4.6 Mixing of resin components in the Ross mixer .............................. 141
Figure 3.4.7 BCSMCP process for natural ﬁber-polyester resin composites ............. 143
Figure 3.4.8 Feeding natural ﬁbers though a vibratory feeder .............................. 145
Figure 3.4.9 An E-Glass-CaCO3-UPE composite ............................................. 147
Figure 3.4.10 Composite samples made using SMC line .................................... 148
Figure 3.4.11 Composite samples made using SMC line .................................... 149

Figure 3.4.12 Composite samples made using SMC line 150
Figure 3.5.1 Hi-Res TGA 2950 (TA instruments) in CMSC lab ............................ 152
Figure 3.5.2 Perkin Elmer spectrum 2000 FT-IR in CMSC lab . .......153
Figure 3.5.3 PHI 5400 ESCA (XPS) system in CMSC lab .................................. 155
Figure 3.5.4 TA 2920 Modulated DSC (TA instruments) in CMSC lab. . . . . . . ...157
Figure 3.5.5 Sputter coater(1eﬁ) and ESEM (right) Model no. 2020 at CMSC lab. . ...158
Figure 3.5.6 TA DMA 2890 in CMSC lab 161
Figure 3.5.7 UTS in ﬂexural testing mode at CMSC lab16l

Figure 3.5.8 Impact testing machine (left) and notching machine (right) at CMSC

lab .................................................................................................... 162
Figure 3.5.9 United Corporation “SFM-20” Test System ................................... 163
Figure 3.5.10 TMA 2980 (TA instruments) in CMSC lab .................................... 164

Figure 3.5.11 (a) Working of a contact mode AFM. (b) MultiMode Scanning Probe
Microscope (SPM) from Digital Instrument ................................................... 165

xiii

Figure 3.5.12 Humidity chamber in CMSC lab ............................................... 167

Figure 3.5.13 QUV weatherometer in Dr. Kamdem’s lab ................................... 169
Figure 3.5.14 Open panel of QUV weatherometer ........................................... 169
Figure 3.5.15 CIELAB system for color measurement ...................................... 163
Figure 3.5.16 Surface proﬁlometer for roughness measurement ........................... 170
Figure 3.5.17 Weathered samples, on left, a hempmat-UPE biocomposite, on right, a
BBSG—GFC-UPE biocomposite ................................................................. 171
Figure 4.1.1 Schematic representation of the compression molding process ............. 173
Figure 4.1.2 TGA of surface treated hemp ﬁbers ............................................ 174
Figure 4.1.3 Derivative thermo-gravimetric analysis (DTGA) of surface treated hemp
ﬁbers ................................................................................................. 175
Figure 4.1.4 DSC of surface treated hemp ﬁbers .............................................. 176
Figure 4.1.5 Proposed reaction for surface treatment of hemp ﬁbers. . . . . . . . . .....179

Figure 4.1.6 Proposed reaction for curing of surface treated hemp ﬁbers and UPE. . ...180
Figure 4.1.7 Proposed reaction for curing of surface treated hemp ﬁbers and UPE. . l 81

Figure 4.1.8 Optimization of ﬁber volume fraction by evaluation of mechanical

properties ............................................................................................ 182
Figure 4.1.9 Comparison of tensile properties of surface treated composites. . . . . . ..184
Figure 4.1.10 Comparison of ﬂexural properties of surface treated composites... . . . . ...186
Figure 4.1.11 Impact strength of surface treated composites ................................ 188
Figure 4.1.12 Typical storage modulus curves of surface treated composites ........... 189
Figure 4.1.13 Typical tan delta curves of surface treated composites ..................... 190
Figure 4.1.14 Storage modulus of surface treated composites at 40 OC .................... 191
Figure 4.1.15 ESEM micrographs of surface treated hemp mat ﬁbers ..................... 192
Figure 4.1.16 Schematic representation of the ﬁber matrix interface and various factors
affecting the ﬁber-matrix adhesion .............................................................. 193
Figure 4.2.1 TGA of surface treated pure hemp ﬁbers ........................................ 195

xiv

Figure 4.2.2 Derivative therrno-gravimetric analysis (DTGA) of surface treated pure

hemp ﬁbers ........................................................................................... 196
Figure 4.2.3 DSC of surface treated pure hemp ﬁbers ....................................... 197
Figure 4.2.4 Optimization of styrene content by evaluation of ﬂexural
properties ............................................................................................ 200
Figure 4.2.5 Optimization of styrene content by evaluation of mechanical
properties ............................................................................................. 201
Figure 4.2.6 Optimization of styrene content by evaluation of mechanical
properties ............................................................................................. 202
Figure 4.2.7 Comparison of tensile properties of surface treated pure hemp mat based
composites ........................................................................................... 203
Figure 4.2.8 Comparison of ﬂexural properties of surface treated composites ........... 205
Figure 4.2.9 Impact strength of surface treated composites ................................. 207
Figure 4.2.10 Typical storage modulus curves of surface treated composites. . . . . . . . ....208
Figure 4.2.11 Typical tan delta curves of surface treated composites ...................... 210
Figure 4.2.12 Storage modulus of surface treated composites at 40 0C .................... 211
Figure 4.2.13 ESEM micrographs of surface treated pure hemp mat ﬁbers. . . . . . . . ........212
Figure 4.2.14 ESEM micrographs of tensile fractured surfaces of biocomposites from
pure hemp mat ﬁbers ............................................................................... 214
Figure 4.3.1 TGA of surface treated kenaf ﬁbers ............................................... 215
Figure 4.3.2 Derivative therrno-gravimetric analysis (DTGA) of surface treated kenaf
ﬁbers ................................................................................................... 216
Figure 4.3.3 DSC ofsurface treated kenafﬁbers................................................218
Figure 4.3.4 Comparison of ﬂexural properties of surface treated kenaf based
composites ............................................................................................ 220
Figure 4.3.6 Speciﬁc Flexural Properties of Kenaf-UPE composites ...................... 222
Figure 4.3.5 Notched Izod Impact strengths of kenaf-UPE composites ................... 224

Figure 4.3.6 Typical storage modulus curves of surface treated kenaf composites. . ....225

Figure 4.3.7 Typical tan delta curves of surface treated kenaf composites. ................. 226

XV

Figure 4.3.8 Storage modulus of Kenaf-UPE composites at 40 OC.. ............................. 227
Figure 4.8.9 SEM micrographs of surface treated kenaf ﬁbers. ................................... 229

Figure 4.3.10 SEM micrographs of surface treated kenaf ﬁbers based composites ....... 230

Figure 5.1.1 Temperature sweep for curing UPE from 25 0C to 160 0C. . ...............235
Figure 5.1.2 Temperature sweep for curing UPE from 35 0C to 300 0C..... ..........2.35
Figure 5.1.3 Isothermal DSC cure ofUPE at 50, 60,70, 80 0c. . 236
Figure 5.1.4 Isothermal DSC cure ofUPE at 90, 100, 110, 120 0C237
Figure 5.1.5 Isothermal DSC cure ofUPE at 130, 140, 150, 160 0C. 237
Figure 5.1.6 Residual heat ﬂow (temperature sweep) DSC UPE curing .................. 238

Figure 5.1.7 Comparison of isothermal heat of reaction at different temperatures. . .....239

Figure 5.1.8 Comparison of maximum conversions at different temperatures. . . . . ........239
Figure 5.1.9 FT IR spectra of pure UPE. . . .. ................................................................... 242
Figure 5.1.10 FTIR spectra of pure styrene .................................................................... 243

Figure 5.1.11 FT IR spectra of UPE system at 150 0C at t=0 minutes and

F 250 minutes ...................................................................................... 244
Figure 5.1.12 Fractional conversion at 50 0C from in situ F TIR study .................... 245
Figure 5.1.13 Fractional conversion at 60 0C from in situ F TIR study .................... 245
Figure 5.1.14 Fractional conversion at 70 0C from in situ FTIR study .................... 246
Figure 5.1.15 Fractional conversion at 80 0C from in situ FTIR study .................... 247
Figure 5.1.16 Fractional conversion at 90 0C from in situ FTIR study .................... 247
Figure 5.1.17 Fractional conversion at 100 0C from in situ FT IR study ................... 248
Figure 5.1.18 Fractional conversion at 110 0C from in situ FTIR study. . . . . . . . . . . . . . ......248
Figure 5.1.19 Fractional conversion at 120 0C from in situ FTIR study ................... 249
Figure 5.1.20 Fractional conversion at 130 0C from in situ FTIR study. . . . . . . .............250
Figure 5.1.21 Fractional conversion at 140 0C ﬁom in situ FTIR study ................... 250

xvi

Figure 5.1.22 Fractional conversion at 150 0C from in situ F TIR study .................... 251

Figure 5.1.23 Fractional conversion at 160 0C from in situ FTIR study .................... 252
Figure 5.2.1 FT IR spectra of pure acrylonitrile ................................................ 255
Figure 5.2.2 FT IR spectra of pure castor oil ................................................... 255
Figure 5.2.3 FT IR spectra of pure soybean oil ............................................... 256
Figure 5.2.4 FTIR spectra of pure methyl ester of soybean oil ............................ 256
Figure 5.2.5 FTIR spectra of grafted castor oil ............................................... 257
Figure 5.2.6 FTIR spectra of grafted soybean oil ............................................ 257
Figure 5.2.7 FTIR spectra of grafted methyl ester of soybean oil ......................... 258
Figure 5.2.8 Scheme showing probable reaction between vegetable oil and reactive
monomer, resulting in modiﬁed vegetable oil ................................................ 259
Figure 5.2.9 Impact strength of the bioplastics ............................................... 260
Figure 5.2.10 Flexural properties of the bioplastics .......................................... 261
Figure 5.2.11 Impact strength of biocomposites containing bioresins .................... 262
Figure 5.2.12 Flexural properties of biocomposites containing bioresins ................ 262

Figure 5.2.13 ESEM micrographs of impact fractures surface of bioplastic containing
MSO-PBMA-UPE, a) Magniﬁcation 2000X, Scale 25 um, b) 250 X, .................... 263

Figure 5.2.14 ESEM micrographs of impact fractures surface of bioplastic containing
MSO-UPE, a) Magniﬁcation 300 X, Scale 150 um b) Magniﬁcation 2000X, Scale 25
um ................................................................................................... 264

Figure 5.2.15 a) Impact ﬁactured surface of neat polyester resin, scale 20 pm .........265

Figure 5.2.15 b) Impact ﬁactured surface of bioplastic (MSO-UPE-PBMA), scale 20 um
......................................................................................................... 265

Figure 5.2.15 c) Impact Fractured surface of bioplastic (MSO-UPE-PBMA), scale 2 pm
. . ...................................................................................................... 265

Figure 5.2.16 AFM picture of neat unsaturated polyester resin (deﬂection image, scan
size: 5 pm X 5 pm) ............................................................................... 266

Figure 5.2.17 AFM picture of bioplastic (MSO-UPE-PBMA) (deﬂection image, scan
size: 60 um X 60 um) ............................................................................ 267

xvii

Figure 5.2.18 AFM picture of bioplastic (MSO-UPE-PBMA) (deﬂection image, 9 mm X

9 pm) ................................................................................................ 268
Figure 5.2.19 TGA of pure soybean oil phosphate ester polyol (SOPEP) ................ 271
Figure 5.2.20 DSC of pure soybean oil phosphate ester polyol (SOPEP) ................ 272
Figure 5.2.21 FTIR spectra of pure maleic anhydride (MA) ............................... 273
Figure 5.2.22 FTIR spectra of pure soybean oil ............................................. 274
Figure 5.2.23 FTIR spectra of pure SOPEP ................................................... 274
Figure 5.2.24 FTIR spectra of modiﬁed SOPEP ............................................. 274
Figure 5.2.25 Proposed reaction mechanism for. grafting of SOPEP with MA .......... 276
Figure 5.2.26 Storage modulus of bioplastics made with GFTSOPEPI ................... 278
Figure 5.2.27 Storage modulus of bioplastics made with GFTSOPEP2 .................. 278
Figure 5.2.28 Storage modulus of bioplastics made with GFTSOPEP3 .................. 279
Figure 5.2.29 Tan delta of bioplastics made with GFTSOPEPI ........................... 279
Figure 5.2.30 Tan delta of bioplastics made with GFTSOPEP2 ........................... 280
Figure 5.2.31 Tan delta of bioplastics made with GFTSOPEP3 ........................... 280

Figure 5.2.32 Glass transition temperature of bioplastics made with GFTSOPEPl ....281
Figure 5.2.33 Glass transition temperature of bioplastics made with GFTSOPEP2 ....282

Figure 5.2.34 Glass transition temperature of bioplastics made with GFTSOPEP3 ....282

Figure 5.2.35 Crosslinking density of bioplastics with GFTSOPEPl ..................... 284
Figure 5.2.36 Crosslink density of bioplastics made with GFTSOPEP2 .................. 284
Figure 5.2.37 Crosslink density of bioplastics made with GFTSOPEP3 .................. 285
Figure 5.2.38 Impact strength of bioplastics made with GFTSOPEPI ................... 285
Figure 5.3.39 Impact strength of bioplastics made with GFTSOPEP2 ................... 286
Figure 5.2.40 Impact Strength of bioplastics made with GFTSOPEP3 ................... 286

Figure 5.2.41 ESEM micrographs of impact fractures surface of bioplastic containing
a) 10%GFTSOPEP1-UPE, Magniﬁcation 3000 X, Scale bar 15 um

xviii

b) 20% GFTSOPEPl-UPE, Magniﬁcation 300 X, Scale bar 150 um
c) 30% GFTSOPEPl-UPE, Magniﬁcation 800 X, Scale bar 50 um
(I) 40% GFTSOPEPl-UPE, Magniﬁcation 285 X, Scale bar 150 um ...................... 288

Figure 5.2.42 ESEM micrographs of impact fractures surface of bioplastic containing

a) 10%GFTSOPEP3—UPE, Magniﬁcation 400 X, Scale bar 100 um
b) 20% GFTSOPEP3-UPE, Magniﬁcation 400 X, Scale bar 100 um
c) 30% GFTSOPEP3-UPE, Magniﬁcation 600 X, Scale bar 200 pm

(1) 40% GFTSOPEP3-UPE, Magniﬁcation 300 X, Scale bar 150 um ...................... 289
Figure 5.2.42 AFM picture of 20% GFTSOPEP3-UPE (deﬂection image, 50 um X 50
pm) .................................................................................................. 290
Figure 5.2.44 AFM picture of 20% GFTSOPEP3-UPE (deﬂection image, 50 um X 50
um) .................................................................................................. 291
Figure 5.2.45 AFM picture of 20% GFTSOPEP3-UPE (Z scale: 1000 nm, scan size, 16
um X 16 um) ...................................................................................... 291

Figure 5.2.46 AFM picture of 20% GFTSOPEP3-UPE (deﬂection image, 16 um X 16
um) .................................................................................................. 292
Figure 5.2.47 Picture of 20% GFTSOPEP3-UPE on TV screen ........................... 292
Figure 6.1 Flexural properties of glass composites .......................................... 297
Figure 6.2 Storage modulus of glass composites at 40 0C .................................. 298
Figure 6.3 DMA plot for glass composites ................................................... 298
Figure 6.4 DSC plots of natural ﬁbers used for SMC line .................................. 300
Figure 6.5 TGA plots of natural ﬁbers used for SMC line ................................. 300
Figure 6.6 Tensile properties of biocomposites .............................................. 304
Figure 6.7 Flexural properties of biocomposites ............................................. 306
Figure 6.8 Impact properties of biocomposites ............................................... 308
Figure 6.9 Storage modulus of biocomposites at 40 OC ..................................... 309
Figure 6.10 Tensile properties of biocomposites ............................................. 312
Figure 6.11 Flexural properties of biocomposites ........................................... 314
Figure 6.12 Impact properties of biocomposites ............................................. 316

xix

Figure 6.13 Storage modulus of biocomposites at 40 OC .................................... 317

Figure 6.14 ESEM micrographs of SMC biocomposites, at magniﬁcation of 100 X and

scale bar of 450 um ............................................................................... 318
Figure7.1.l Moisture absorption of bioresins ................................................ 325
Figure7.1.2 Moisture absorption of biocomposites of hemp mat .......................... 326
Figure7.1.3 Moisture absorption of SMC manufactured biocomposites .................. 326
Figure 7.1.4 TGA of bioplastics ................................................................ 329
Figure 7.1.5 DSC of bioplastics ................................................................ 329
Figure 7.2.1 Change in color parameter ‘L’ over time for biocomposites ................ 330
Figure 7.2.2 Change in color parameter ‘a’ over time for biocomposites ................ 331
Figure 7.2.3 Change in color parameter ‘b’ over time for biocomposites ................ 331
Figure 7.2.4 Change in color parameter ‘E’ over time for biocomposites ................ 332
Figure 7.2.5 Change in weight over time for biocomposites ............................... 333
Figure 7.2.6 Change in roughness parameter Ra over time for biocomposites ........... 335
Figure 7.2.7 Change in roughness parameter Rz over time for biocomposites ........... 335
Figure 7.2.8 Change in roughness parameter Rmax over time for biocomposites .........336
Figure 7.2.9 Change in storage modulus over time for biocomposites .................... 337
Figure 7.2.10 Change in T8 over time for biocomposites ................................... 338

Images in this thesis are presented in color

XX

CHAPTER ONE: INTRODUCTION

1. INTRODUCTION

Fiber-reinforced plastic composite materials were ﬁrst produced by combining cellulose
ﬁbers with phenolic resin in 1908. Since that time, composite materials have advanced to
applications ranging from packaging materials to aerospace components and structure,
largely based on the utilization of petroleum based constituents, i.e., glass, carbon or
ararnid ﬁbers reinforced with epoxy, unsaturated polyester resins, polyurethanes, or
phenolics. Because of growing environmental consciousness, traditional composite
structures are now becoming the subject of legislative initiatives governing their
manufacture, use and removal. The most difﬁcult aspect of the problem is the removal at
the end of lifetime, as the components are closely interconnected, relatively stable, and
therefore difﬁcult to separate and recycle [1]. There is a new and growing requirement
for composite materials which are recyclable and/or degradable.

Natural Fibers

By combining natural reinforcing ﬁbers with biopolymer matrices, new ﬁber reinforced
materials called ‘biocomposites’ are being developed [2-10]. These biocomposites consist
of bioﬁber as the reinforcing element and usually a biodegradable polymer as the matrix
material. Since both components are biodegradable, the composite is also expected to be
biodegradable. Biopolymers are generally biodegradable, but they do not possess the
necessary thermal and mechanical properties useful for engineering applications. On the
other hand, common engineering plastics are obtained ﬁom synthetic polymers, but are

non-biodegradable. Although much research has been carried out on bioﬁber reinforced

synthetic polymers, they are not completely biodegradable. However, bioﬁbers derived
from annually renewable resources, used as reinforcing ﬁbers in both thermoplastic and
thermoset matrix composites, provide environmental beneﬁts with respect to ultimate
disposability and raw material utilization [1, 5, 7].

In biocomposites, the bioﬁbers serve as reinforcement by enhancing the strength and
stiffness of the resulting composite structures. The conventional ﬁbers like glass, carbon,
aramid, etc., are produced with very speciﬁc properties having little variability, whereas
the properties of natural ﬁbers vary considerably. Bioﬁber properties depend on whether
the ﬁbers are taken from plant stem or leaf, the quality of the plants as a function of
agricultural practices, the age of the plant, the preconditioning of ﬁbers, and the
processing methods adopted for extraction of ﬁbers [5, 6]. Properties such as density,
ultimate tensile strength, initial modulus, and toughness are related to the internal
structure and chemical composition of ﬁbers [5]. However, pristine natural ﬁbers have
the mechanical properties that would allow them to be used as glass ﬁber replacements in
certain composite applications.

In addition to constituent selection, to obtain a good reinforcing effect for plastics
composites, it is necessary to increase the adhesion between ﬁbers and resins by using
surface treatments [3, 9]. Surface chemical modiﬁcations of natural ﬁbers, including
dewaxing, alkali treatment, cyanoethylation, vinyl grafting, and treatment with various
coupling agents, have been shown to improve ﬁber matrix adhesion in biocomposites [9].
Thermoset composites are more desirable than therrnoplastics because of their superior
mechanical properties. Petroleum based matrix resins in thermoset biocomposites are in

general, non-biodegradable. Compared to thermoplastics, thermoset polymers have a

lower viscosity, complex formulations, better damage tolerance, easier ﬁber
impregnation, longer processing cycles, and higher fabrication costs. Thermoplastics are
also recyclable, easy to handle, and have unlimited shelf life. However, they are prone to
creep, while thermoset are not.

Unsaturated polyester resin

Among various thermoset resins, unsaturated polyesters (UPEs) are the most widely used
matrix materials in polymeric composites [11, 12]. They are processed over a wide
temperature range involving hand lay-up at low temperatures, resin-transfer molding at
medium temperatures, and sheet-molding compound (SMC) compression molding, bulk-
molding compound (BMC) injection molding, and pultrusion at high temperatures [19].
UPE resins have many applications in automotive, aircraft, electrical, and appliance
components [1 1-20].

Unsaturated polyester resins (UPE) refer to a large class of polymers containing reactive
double bonds in the polymeric chain. The reaction of a UPE resin is a free-radical chain-
growth crosslinking copolyrnerization between the styrene monomer (25-45 wt %) and
the UPE molecules. A source of free radicals, usually organic peroxides, is needed to
initiate the reaction. The initiator is disassociated with help of a promoter. Optimum
performance can be achieved by controlling the type and concentration of peroxide and
promoter. Typical initiator concentrations run from 1 to 3% by weight based on clear
resin and, for accelerators, (mostly cobalt) from 0.25 to 4% based on 1% metal content
solution [1 1-14]. Under or overdosing can result in deterioration of cure and mechanical
properties.

Project Motivation

The ambitious goals set by the US. government for the creation of a biobased economy
as an alternative to the existing petroleum-based products are challenging the industry,
academia, and agriculture [21]. This project seeks to replace conventional glass ﬁber-
polyester composites by a novel low-cost natural/bioﬁber composite for housing panel
applications. Through use of bioﬁbers, our goal is to make composites, which can
outperform current housing panels, while maintaining competitive economic structure.
Objective

The overall objective of this project is to develop environmentally friendly biocomposites
as alternatives to the present glass based composites. Although polyester resin is a
petroleum based product and not eco-friendly, we have chosen it as the matrix resin as it
is required to produce a material which can be integrated easily into the current housing
material industries. In the future, replacement or modiﬁcation of the petroleum based
resin by a suitable biobased resin is a logical and desirable direction for this research.

In order to achieve the above objective, several interrelated components are required: i)
development of a Biocomposite Sheet Molding Compound Panel (BCSMCP) high
volume manufacturing process; ii) fundamental studies for determining the relationship
between structures and properties of the materials;, and iii) utilization of ‘engineered’
natural ﬁbers having good ﬁber-matrix adhesion, a balance of mechanical properties and
processability.

Design of Fibers

‘Engineered’ bioﬁbers are deﬁned as the suitable blends of surface treated bast, leaf and
link ﬁbers, which provide an optimum balance in mechanical properties of the ﬁnal

composite. The blending of differently modiﬁed ﬁbers is based on the principle that the

individual ﬁber properties vary with their location in the plant and the level of adhesion
will determine their contribution to the composite stiffness and toughness. Previous
research has shown that ﬁbers originating ﬁom the leaf have very high toughness while
those originating from bast have very high modulus [1]. Thus, by combining different
ﬁbers in optimum weight ratios, it is possible to have very tough and stiff bioﬁbers. The
bast ﬁbers used in this research are kenaf, hemp, jute, ﬂax, while the leaf ﬁbers are sisal,
henequen, and pineapple leaf ﬁber.

An optimum ﬁber matrix interface bond is critical for performance of composite
materials. Surface chemical modiﬁcations of bioﬁbers have been successful in improving
the ﬁber-matrix adhesion of biocomposites. Surface treatments also improve
compatibility between ﬁber and matrix, dimensional stability and resistance to
biodeterioration. The use of low cost, water based surface treatment and sizing for
bioﬁbers, for example, low concentrations of alkali treatment, and silane treatment is
mandatory if the balance between performance and economic viability is to be
maintained. The reinforcing effect of the ﬁber is also dependent on other parameters,
such as, ﬁber diameter, ﬁber length, chemical constituents of the ﬁber, ﬁber orientation
in the composite, number of individual ﬁbers in a ﬁber bundle, and the ﬁber volume
fraction used in the composite. All of these variables and their effect on the ﬁnal

biocomposite properties will be factored into the experimental plan.

Project: Biocomposites from Engineered Natural Fibers for Housing
Panel Applications

 

 

 

 

 

 

 

 

 

 

 

 

Innovative bioﬁber Formulation of Novel Processing
treatment & ‘Engineered polyester resin (Biocomposite sheet
bioﬁbers’ molding compound
.___I— panel processing)
Novel l-step water BCSMCP

 

Optimization

based sizing

  
  
 
 
 
  

 

of all curing
constituents Fabrication of

l biocomposite
l .
Water ‘ panels usrng

emulsion
ﬂame Curing BCSMCP process
retardant studies I

 

 

 

 

 

 

 

 

 

 

 
   

 

 

 

 

Evaluation of
Surface characterization physico-mechanical
of treated bioﬁbers propertles and
optimization of
Blending of BCSMCP
different bioﬁbers 1. .
Design of
‘Engineered’ bioﬁber

 

 

 

 
    

Research output:
Integration of
education with
R&D, possible
future
commercialization

 
   
     
   
 

Figure 1.1: Schematic representation of the research plan for this project [22]

Curing of polymer matrix

Temperature and time are the most important parameters in a cure cycle. Curing of
thermosets requires an intimate knowledge of the chemical kinetics of the polymerization
and crosslinking reactions. The parameters that must be determined in a cure cycle are
the number of stages in the cure, the rate of temperature increase, the hold temperature at
each stage, the pressure at which the cure takes place, and the time allotted for the cure
cycle [23]. Once the kinetics are understood and the actual chemistry behind the curing is
established, the cure cycle parameters can be chosen based on the desired polymer
properties.

The physical nature of chemical crosslinking is quantiﬁed and represented by crosslink
density and degree of cure. The crosslink density is a quantitative measure of the number
of crosslinks that exist in a given volume in the therrnosetting polymer. The degree of
cure represents the chemical conversion of the curing reaction [24]. The crosslinking
density is determined from Dynamic Mechanical Analysis (DMA), and degree of cure is
evaluated from Diffrential Scanning Calorimerty (DSC), and Fourier Transform Infra
Red Spectroscopy (FTIR). In addition, the kinetics of curing reaction, including, rate of
reaction, order of reaction, etc. can be estimated from DSC and FTIR experiments. It
should be noted that the optimum curing conditions will have to re-established after the
addition of any ﬁllers, ﬁbers, additives to the polymer matrix.

Matrix modification

Unsaturated polyester resins (UPE) are brittle, undergoing a 7-10% volume shrinkage
after curing, and are not very resistant to alkali exposure. Traditionally, these properties

are improved by blending with various additives; for example, fracture properties of a

cured resin improve after blending with reactive liquid rubber and the shrinkage of UPE
is prevented by introduction of a polar low-shrinkage thermoplastic.

Bioresins, or thermoset resins derived from vegetable oils like, soybean, castor, corn,
peanut, cottonseed, etc, also increase toughness and reduce volume shrinkage, with the
added advantage of lower cost, abundant availability, and lower environmental impact
[25]. Adding bioresins to the matrix will reduce the amount of fossil fuel based products
in the ﬁnal composite. It is planned to eventually phase out polyester resin from the
formulation of biocomposites, to produce a completely ‘green’ product.

Processing

Using thermoset resins, biocomposites can be fabricated by compression molding, resin
transfer molding and hand lay-up. One goal of this project is to develop a continuous
process for making biocomposites, which will be similar to the existing Sheet Molding
Compound (SMC) process, and thus could be adopted in the industries without any major
change in infrastructure. This new high volume processing technique will be named
‘biocomposite stampable sheet molding compound panel’ (BCSMCP) manufacturing
process. SMC is a continuous sheet containing chopped ﬁbers and mineral ﬁllers
embedded in a highly viscous thermoset resin [26]. In the commercial SMC process,
continuous glass ﬁbers rovings are fed to a chopper, cut to 6 mm in length, and then
distributed onto a carrier ﬁlm, forming a uniform layer of chopped glass ﬁbers. Since
natural ﬁbers cannot be obtained in a continuous from, and making a continuous yarn or
roving with these ﬁbers would be difﬁcult and expensive, chopped natural ﬁbers can only

be utilized if a new process is developed. Achieving a uniform and continuous dispersion

of a controlled amount of ﬁbers with little variability on the SMC line will be an
important goal of this work.

Testing

Mechanical properties of the composites need to be determined in order to insure that the
resulting biocomposite has structural parity with the conventional composite. After the
composite has been fabricated and conditioned, it is subjected to various tests, in
accordance with ASTM standards to evaluate its properties. Properties of interest are
tensile strength and modulus, bending strength and modulus of elasticity, impact strength,
storage modulus, loss modulus, tangent delta, coefﬁcient of thermal expansion,
compressive strength, creep, shear stress and stiffness. The experimental properties are
compared with theoretical values obtained from micro-mechanical models such as the
Halpin-Tsai, Halpin and Pagano and Piggott models [27]. Morphology and ﬁber
distribution are ascertained by electron microscopy on tensile fractures samples of
composites. The durability of biocomposites is tested by moisture absorption test,
accelerated weathering test, and ﬁre test.

The untreated and surface treated bioﬁbers and hybrid bioﬁbers are characterized by
DSC, Thermal Gravimetric Analysis (TGA), Fourier Transform Infra-Red Spectroscopy
(FT IR), Environmental Scanning Electron Microscopy (ESEM), and X-Ray
Photoelectron Spectroscopy (XPS) to evaluate degree of crystallinity, maximum
degradation temperature, chemical reactions as a result of treatment, topology, and
morphology, and surface atomic concentrations, respectively.

Summary

The desire for renewable materials from sustainable sources is increasing for a variety of
applications. Polymer matrix composites reinforced with natural plant ﬁbers are one such
example. Combining natural ﬁbers with an unsaturated polyester resin matrix, novel low-
cost biocomposites with desired properties can be made. Such biocomposites can provide
many beneﬁcial additions to the American Advanced Housing program. Fundamental
and applied research into the materials, their surface treatments and fabrication processes
is necessary to transfer this technology to industry.

The objective of this research is to investigate the properties and processing of bioﬁber
reinforced Unsaturated Polyester Resin (UPE) composites. The biggest advantage of
panels made from this type of biocomposite is their low cost, combined with their
ecological and technological advantages. If composite panels with acceptable properties
can be developed, along with a viable manufacturing method for their continuous

processing, housing panels for the future can be produced.

10

References

l. Biodegradable Polymers in North America & Europe (P081), Mar Tech, USA, July
(1998), 21

2. A. K. Mohanty, M. Misra, G. Hinrichsen, Macromolecular Materials and
Engineering, 276/277, (2000), l

3. A. K. Bledzki, J. Gassan , Progress in Polymer Science, 24, (1999), 221-274

4. A. K. Mohanty, M. Misra, L. T. Drzal, Journal of Polymers and the Environment, 10
( 1/2) (2002) 19

5. J. N. McGovern, in: Encyclopedia of Polymer Science and Engineering, Vol.7, H. F.
Mark, N. M. Bikales, C. G. Overberger, G. Menges, J. I. Kroschwitz, Eds, John Wiley &
Sons; New York, (1987), p.16.

6. L. M. Lewin, E. M. Pearce, Handbook of Fiber Science and Technology, Volume IV,
Fiber Chemistry, Marcel Dekker, New York (1985)

7. AK. Rana, B.C. Mitra, A.N. Banerjee, Journal of Applied Polymer Science, 71
(1999), 531 i

8. B. van Voorn, H. H. G. Smit, R. J. Sinke, B. de Klerk, Composites: Part A, 32,
(2001),]271-1279

9. A. K. Mohanty, M. Misra, L. T. Drzal, Composite Interfaces, 8, 5 (2001), 313-343

10. D. Ray, B. K. Sarkar, A. K. Rana, N. R. Bose, Composites: Part A, 32, (2001),1 19-
127

11. R. Bums, Polyester Molding Compound; Marcel Dekker: New York, (1982)

12. V. Bellenger, B. Mortaigne, J. Verdu, Journal of Applied Polymer Science, 44, (1992),
653

13. H.V Boeing, Unsaturated polyesters: structure and properties, Elsevier Pub. Co.,
(1964)

14. M. Malik, V. Chaudhary, I. K. Varma, Journal of Macromolecular Science, Part C -
Reviews in Macromolecular Chemistry and Physics, C40, 2&3, (2000), 139-165

15. C. H. Bamford, C. F. H. Tipper, Free-radical polymerization, Elsevier Pub. Co.,
(1976)

16. Y.J. Huang, J .8. Leu, Polymer, 34, 2, (1993), 295-304

17. H. Yang, L. J. Lee, Journal of Applied Polymer Science, 84, (2002), 211—227

11

18. D. G. Hepworth, D.M. Bruce, J. F. V. Vincent, G. Jeronimidis, Journal of Material
Science, 35, (2000), 293-298

19. R. Crawford, Journal of Plastics Engineering, 2nd ed.; Pergamon: New York, (1987)
20. Y. S.Yang, L. Lee, Journal of Polymer Process Engineering, 5, (1987—1988), 327

21. C. Eckert (Kline& Co.), “Opportunities of Natural Fibers in Plastic composites”, 3 rd
Annual Ag Fiber Technology Showcase, October 4-6, 2000, Memphis, TN, USA

22. L. T. Drzal, A. K. Mohanty, “Biocomposites from Engineered Natural Fibers for
Housing Panel Applications”, NSF-PATH proposal (2001) (Award number 0122108)

23. G. Giindiz, “Unsaturated polyester resin overview”, in Polymeric Materials
Encyclopedia, 11, (J .C. Salamone Ed.) CRC Press, Boca Raton, FL, (1996), 8469-8476

24. H. Yang, L. J. Lee, Journal of Applied Polymer Science, 84, (2002), 211—227

25. G. Mehta, A. K. Mohanty, M. Misra, L. T. Drzal, Green Chemistry, 6(5), (2004),
254-258

26. US 3615979, C. J. Davis; R. P Wood; E. R. Miller, “Glass ﬁber-reinforced sheet
molding compound” (1971)

27. J. C. Halpin, and J. L. Kardos, “The Halpin-Tsai equations: A review”, Journal of
Polymer Engineering and Science, 52, (1976), 344-352

12

CHAPTER TWO: LITERATURE REVIEW

2.0 Literature Review

In this chapter, some of the basic deﬁnitions and concepts related to natural ﬁber
composites will be introduced, followed by summaries of past research related to natural
ﬁber modiﬁcation, polyester resin, bioresins, composite fabrication techniques, and

applications of natural ﬁber composites.

2.1 Natural Fibers

The idea of using cellulose ﬁbers as reinforcement in composite materials is not a new or
recent one. Ever since the beginning of human civilization, natural ﬁbers like grass and
straw were used to reinforce mud bricks. However, with the advent of high performance
man made materials, the use of natural ﬁbers diminished. Until recently, the natural ﬁbers
were mainly used in the production of rope, string, clothing, carpets and other decorative
products. Interest in the use of natural ﬁbers has grown during the last decade due to their
various advantages.

Increasing environmental consciousness and demands of legislative authorities are
leading to the scrutiny of manufacture, use, and removal of traditional composite
structures, usually made of glass, carbon or aramid ﬁbers reinforced with epoxy,
unsaturated polyester resins, polyurethanes, or phenolics. The disadvantages of such
composite materials is the use of organic materials during their manufacture, high

processing temperature and energy use, and disposal and recycling at the end of their

13

lifetime. The composite constituents are closely interconnected, relatively stable, and

 

©Drzal. Mohanty, Misra, MSU 2001
Figure 2.1.1: Examples of natural ﬁbers [1]

therefore difﬁcult to separate and recycle [1].

Sustainability, industrial ecology, coo-efﬁciency, and green chemistry are guiding the
development of the next generation of materials, products, and processes. Biodegradable
plastics and biobased polymer products based on annually renewable agricultural and
biomass feedstock can form the basis for a portfolio of sustainable, coo-efﬁcient products
that can compete and capture markets currently dominated by products based exclusively
on petroleum feedstock [2]. To be sustainable, a biobased product derived from
renewable resources has to be recyclable, triggered biodegradable, commercially viable

and environmentally acceptable.

The Technology Road Map for Plant/Crop based Renewable Resources 2020, sponsored
by the US. Department of Energy (DOE), has targeted to achieve 10 % of basic chemical
building blocks arising from plant derived renewable sources by 2020, with development
concepts in place by then to achieve a ﬁrrther increase to 50 % by 2050. The US.
agricultural, forestry, life sciences, and chemical communities have developed a strategic
vision [3] for using crops, trees, and agricultural 2100 residues to manufacture industrial
products, and have identiﬁed major barriers [4] to its implementation.

2.1.2 Structure of bioﬁbers

Bioﬁbers are generally lignocellulosic consisting of helically wound cellulose
microﬁbrils in an amorphous matrix of lignin and hemicellulose [1, 2, 5]. A single ﬁber
of all plant based natural ﬁbers consists of several cells. These‘cells are formed out of
crystalline microﬁbrils based on cellulose, which are connected to a complete layer, by
amorphouse lignin and hemicellulose. Multiple of such cellulose-lignin/hemicellulose
layers stick together to a multilayer composite, the cell wall [9]. These cell walls differ in
their composition and in the orientation of the cellulose microﬁbrils. These ﬁbers consist
of several ﬁbrils that run along the length of the ﬁber. The potential ﬁbers are separated
from the original plant in several ways like retting, scrapping, decorticating and pulping.
Cellulose is the main component of almost all natural ﬁbers. The elementary unit of a
cellulose macromolecule is anhydro-D-glucose, which contain three hydroxyl (OH)
groups (Figure 2.4). These hydroxyl groups form hydrogen bonds inside the
macromolecules itself (intramolecular) and between other cellulose molecules

(intermolecular). Therefore all natural ﬁbers are hydrophilic in nature.

15

 

 

H OH
O
OH
OH H H CHZOH 0
g "' I]

Figure 1.1.2: Chemical structure of cellulose molecules: Poly-[3(1,4)-D-Glucose

2.1.3 Advantages of natural fibers

Natural ﬁbers offer various advantages over other kinds of reinforcements. Bioﬁbers
come ﬁ'om annually renewable resources, and are biodegradable. They are much less
expensive and less dense as compared to man made ﬁbers like E-glass ﬁbers and carbon
ﬁbers [1]. Environmental gains can be made through use of renewable bioﬁbers instead

of synthetic ﬁbers. Insulation and sound absorption properties of natural ﬁbers are much

 

better than those of ﬁberglass.
180 ‘r‘ 3
150 - i n cowl" r 2.5 A
O Densrty (gm/cm3) ME
r 2 g
39
a?
is
D

 

 

 

   

Figure 2.1.3: Comparison of cost and density between glass ﬁbers and bioﬁbers
Natural ﬁbers provide a net energy savings over man-made ﬁbers. Bioﬁbers are produced
by solar energy while production of synthetic ﬁber needs large amounts of petroleum

based energy. The processing temperature to make glass ﬁber exceeds 1200°C [1, 7, 8]. It

16

takes 6,500 BTUs of energy to produce one pound of kenaf (not including the energy to
produce fertilizer, collect and process the ﬁbers) while it takes almost four times that
much energy (23,500 BTUs) to produce one pound of glass ﬁber [7].

2.1.4 Properties of Natural Fibers

Physical properties of natural ﬁbers are strongly inﬂuenced by their chemical structure
such as cellulose content, degree of polymerization, orientation and crystallinity, which
are affected by the plant genetic makeup, conditions during growth of plants as well as
extraction methods used. As a result, there is an enormous variability in ﬁber properties
depending upon which part of the plant the ﬁbers came from, the quality of the plant and
its location [1, 5]. The properties of these ﬁbers are very difﬁcult to measure, because a
considerable number of ﬁbers need to be tested to obtain statistically signiﬁcant mean
values.

The mechanical properties of bioﬁbers depend greatly on the scale and structure. Smaller
structures usually lead to more regular composition and less defects and hence a better
mechanical properties. However, with the existing technologies the smallest ﬁber
obtained is in a form of single ﬁber or microﬁbril. Research is going on the extraction of

nano scale cellulose whiskers from bioﬁbers [10].

2.1.5 Comparison between biofibers and other fibers
The properties of bioﬁbers are comparable to those of some artiﬁcial ﬁbers (Table 2.1).
Although bioﬁbers have moduli comparable to glass ﬁbers, they are lower in strengths

and toughness. They also have a higher hydrophilicity because of the presence of OH

17

groups on their backbones. Thermal degradation in bioﬁbers begins approximately
around 200 0C, therefore they must be processed below this temperature.

The speciﬁc tensile modulus (modulus of ﬁber divided by its density) and speciﬁc
modulus of elasticity of bioﬁbers are higher than those of glass ﬁbers (Figure 2.5). This is
because of the low densities of bioﬁbers.

Table 2.1.1: Typical properties of glass ﬁber and natural ﬁbers [1 1]

 

 

Fibers
Properties E-Glass Flax Hemp Jute Ramie Sisal
Density
(g/cm3) 2.55 1.40 1.48 1.46 1.50 1.33
Tensile 800 _
Strength 2400 550 - 900 400 - 800 500 600 - 700

1500

(MPa)
E'MOdulus 73 60 - 80 7o 10 - 30 44 38
(Gpa)
Speciﬁc
Modulus 29 26 - 46 47 7 - 21 29 29
(E/density)
Elonga‘io“ at 3 1.2 - 1.6 1.6 1.8 2 24.
Failure (%)
Moisture
Absorptiori%)" 7 8 12 12-17 11

 

 

El Elastic Modulus (GPa) 0 Speciﬁc Modulus (GPa/grn/cm3)
80

 

p—a
LII
O

L
O
T
O\
O

 

l
J;
O

 

 

Elastic Modulus (GPa)
m S
o o
N
o
Specific Modulus

 

 

 

 

 

 

 

 

 

Glass ﬁber Hemp th

Figure 2.1.4: Comparison of moduli of glass ﬁbers and bioﬁbers

18

2.1.6 Types of natural ﬁbers

There is a wide variety of natural ﬁbers. These include wood ﬁbers, and a variety of agro-
based ﬁbers such as stems, stalks, bast, leaves and seed hairs. These ﬁbers are abundantly
available throughout the world and they come from renewable resources [1]. Fibers are
also obtained ﬁ'om recycled agro ﬁber-based products such as paper, waste wood, and
point source agricultural residues such as rice hulls from a rice processing plant [2].
Depending on their origin, natural ﬁbers may be grouped into: grasses and straw ﬁbers,
non wood ﬁbers, wood ﬁbers, and cellulose nano-whiskers.

l . Straws, grasses and reeds

These ﬁbers come from the stems of monocotyledonous plants such as bamboo and sugar
cane, big blue stem grass, switch grass, corn [1, 2, 5]. Most grasses produce bioﬁbers that
can be used to reinforce plastics. Higher cellulose content leads to higher mechanical
properties in the bioﬁbers, so does a lower microﬁbrillar angle. The mechanical
properties of ﬁbers decrease as the diameter increases.

2. Leaf ﬁbers

Leaf ﬁbers are ﬁbers that run lengthwise through the leaves of most monocotyledonous
plants such as henequen, banana, pineapple, sisal, screw pine, and palm [1, 2, 5]. These
ﬁbers are also referred 'to as ‘hard ﬁbers’. Leaf ﬁbers generally have a higher
microﬁbrillar angle, leading to lower values of tensile modulus compared to bast ﬁbers.

3. Bast ﬁbers

These ﬁbers (bundles) come from the inner bark (phloem or bast) of the stems of

dicotyledonous plants. Common examples are jute, ﬂax, hemp, mesta, and kenaf [1, 2, 3].

l9

Bast ﬁbers generally have a lower microﬁbrillar angle, and higher cellulose content
leading to higher values of tensile modulus compared to bast ﬁbers.

4. Seed and fruit hairs

These are ﬁbers that come from seed-hairs and ﬂosses, which are primarily represented
by cotton, coir, kapok, and oil palm [1, 2, 5].

5. Wood ﬁbers

These ﬁbers come from the xylem of angiosperrn (hardwood) and gymnosperm

(softwood) trees. Examples include maple, yellow poplar and spruce [6].

 

 

 

 

 

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

REINFORCING BIOFIBERS j
Straw/ Grass Non-wood ‘ Cellulose Wood
Bioﬁbers Bioﬁbers nano-whiskers Bioﬁbers
, I I ,

Examples: Bast I I - Sources: Wood
Com/ wheatl - Leaf Seed/Fruit and non-wood
rice straw/ big ﬁbers through
bluestem grass/ explosion
switch grass, etc. Examples: Examples:

Kenaf, jute, Cotton, coir, S
etc oft and
ﬂax, hemp, etc. ' hard woods

 

Examples:
Sisal, henequen,
pineapple leaf, etc.

 

 

 

Figure 2.1.5: Classiﬁcation of bioﬁbers [2]

2.1.6 Biocomposites

20

Biocomposites, in general are materials made naturally or produced synthetically that
include some type of natural material in their structure. In our research, biocomposites
are also known as natural ﬁber composites. Biocomposites are formed through the
combination of natural cellulose ﬁbers with other resources such as biopolymers or resins
or binders based on renewable raw materials or synthetic polymers. The objective is to
combine two or more materials in such a way that a synergism between the components
results in a new material that is much better than the individual components. Some of the
plant ﬁbers with suitable properties for making biocomposites are: industrial hemp, kenaf,
henequen, jute, ﬂax, sisal, banana, kapok, etc. [1]. The most commonly used polymer
matrices include thermoset polymers such as polyesters, epoxies and phenolics, and
therrnoplastics like polyethylene (PE), polystyrene (PS), and polypropylene (PP). Since
the polymer matrix is soft, ﬂexible and light weight in comparison to ﬁbers, their
combination provides a high strength-to-weight ratio for the resulting composite [1, 2,
11].

The properties of composites also depend on those of the individual components and on
their interfacial compatibility. The interface between the ﬁber and the matrix gives the
composite its structural integrity. The interface consists of the bond between ﬁber and
matrix and the immediate region adjacent to this bond. At least three types of bonding are
thought to exist at the interface: chemical, electrical, and mechanical. A composite with
weak ﬁber matrix interface will not be able to transfer the load from the matrix to the
reinforcing ﬁber and usually leads to poor composite strength [12]. The stress transfer at
the interface between two different phases is determined by the degree of adhesion.

Strong adhesion at the interface is needed for effective transfer of stress and load

21

distribution throughout the composite. In the case of short-ﬁber composites, various
factors inﬂuencing the properties are: (i) the ﬁber dispersion, (ii) the orientation and
geometry (aspect ratio) of the ﬁbers within the composites, (iii) the ﬁber volume fraction,
and (iv) the quality of the interface between the reinforcing ﬁber and polymeric matrix

phase.

2.1.7 Applications of biocomposites

The applications for which biocomposites have been studied include interior and exterior
paneling of automobiles, interior paneling in rail vehicles, fumiture, and replacement of
wood products in housing applications. Daimler Chrysler has used biocomposites of
green ﬂax and hemp ﬁber mats with polyester for under body panels of the EvoBus (a
touring coach) [13]. They found replacing the under body panel, previously a glass ﬁber
composite, with biocomposites required 83 percent less energy to manufacture, and the
resulting part was 40 percent cheaper. In addition, the same tools and machine used to
manufacture the part with glass-ﬁber composites were used for manufacturing with
biocomposites.

Biocomposite materials have also been used for interior paneling applications in
automobiles, including door paneling and rear shelf paneling [14]. The interior paneling
of rail vehicles in Germany has also been manufactured using biocomposites, mainly for
weight savings over glass-ﬁber composites [15]. In this application the standards for ﬁre
protection are very high in Germany. Thus, the biocomposites were treated with ﬂame

retardants to attain the high levels of ﬁre-protection required by the standards.

22

 

 

I IUP 30 M%GF EIUP 35wt°/oM=
100%«1 '

       
   

  

80%"

 

38 kJ/mm

60%“

40%”

20 kJ/mmz'

20%“

 

 

 

 

 

 

0%“
Flexural Flexural inpact Density
Strength Modulus Strength

Figure 2.1.6: Comparison of natural ﬁber (NF) and glass ﬁber (GF)—unsaturated
polyester (UPE) composites [13]

Phenix Biocomposites of Mankato, Minnesota have developed biocomposite materials
with decorative surfaces for furniture, table tops, wall panels, and other home and ofﬁce
ﬁnished surface applications [16]. These applications show that biocomposites can be
manufactured and used in products where attractive surface ﬁnishing is required. Jute
and coir based biocomposites have been developed in India as substitutes for plywood
and medium density ﬁber boards for low-cost housing needs [17]. The engine and
transmission covers of Mercedes-Benz transit buses now contain biocomposites of
polyester resin and natural ﬁbers [18].

Other products under development include the use of sisal—based biocomposites as panels
and rooﬁng sheets, which could also be used as false ceilings and partition boards. The
current market uses of bio-based composites in North America are shown in Figure 2.7

[19].

23

Others (7%)
. Automotive (8%)

  
 
  

Industrial!
Consumer (10%)

Building
Products (75%)

Figure 2.1.7: North American major end markets of natural ﬁber composites, year 2000
[19]

2.1.8 Modification of Natural Fiber

A major disadvantage of bioﬁbers is their highly polar nature, which makes them
incompatible with non-polar polymers. This incompatibility usually leads to poor
dispersion of the ﬁbers in the matrix material, poor interfacial strength and lower
mechanical performance. In addition, the poor resistance to moisture absorption makes
the use of natural ﬁbers less attractive for exterior applications or applications where they
are exposed to a moist environment. In order to improve compatibility with polymer
matrices and to minimize water absorption, the natural ﬁbers need to be modiﬁed.

This situation calls for the development of strategies for the modiﬁcation of the cellulose
ﬁber surface, thereby gaining control over the ﬁber-polyrner interface.

Fiber surface can be modiﬁed by either physical or chemical treatments. Physical
treatments include ﬁbrillation, plasma treatment, and corona treatment [20]. Physical
treatments change the chemical, structural and surface properties of the ﬁber surface and
thereby inﬂuence the mechanical bonding with the matrix polymer.

Chemical treatments introduce chemical bonds between the ﬁber and matrix, and include
dewaxing, alkali treatment, cyanoethylation, vinyl grafting, and treatment with various

coupling agents (silane, isocyanate, titanate, maleic anhydride etc.) [21]. Pretreatment of

24

ﬁbers by encapsulated coating with coupling agents also provides better dispersion by

reducing the ﬁber-ﬁber interaction with the formation of coating on the ﬁber surface.

2.1.9 Past Research on bioﬁbers and biocomposites

2.1.9.1 Studies on bioﬁbers

Han [22] reviewed the characteristics of the non wood ﬁbers, and suggested their use in
pulping. The important properties of the ﬁbers are ﬁber length, lignin content, and
cellulose content. Rowell [23] reviewed the various applications possible for the agro-
based ﬁber reinforced composites, ranging from geotextiles to ﬁlers to sorbents to
structural and nonstructural composites, to packaging and molded products.

Peijs [24] reviewed the developments in the ﬁeld of biocomposites and their applications.
From wood ﬁber composites, to natural ﬁber reinforced composites to all green
biocomposites, biocomposites have come a long way in the recent years. These ﬁbers
have found their own special place in the automotive, sports and transportation sectors,
using conventional processing techniques including sheet molding compound (SMC),
bulk molding compound (BMC), laminating and resin infusion (RTM and VARTM), etc
[11 ]. Riedel and Nickel [25] advocated the use of natural ﬁbers reinforced with
biopoylmers as construction materials in their 1999 article. They discovered a new
application of biocomposites, in covering structural elements in automobiles, railway,

furniture, and leisure industry.

2.1.9.2 Studies on surface treatments

25

Mishra et al. [26] surface treated pineapple and sisal ﬁber (alkali treatment,
cyanoethylation, and acetylation) to make composites with polyester resin. They also
made hybrid composites reinforced with a combination of glass ﬁbers and sisal, as well
as glass and pineapple ﬁbers. The water absorption tendency of the biocomposites
decreased after surface treatments, and also after glass ﬁbers were introduced in the
system.

Mwaikarnbo and Ansell [27] studied the effect of mercerization and acetylation on the
properties of hemp, sisal, jute and kapok by XRD, DSC, FT-IR, and SEM. After the
chemical treatments, the surface of the ﬁbers became rough and clean. From XRD they
observed that at low alkali concentrations, there was a slight increase in the crystallinity
of the ﬁbers, but at higher alkali concentrations, the ﬁber crystallinity index decreases.
Thermal analysis pointed at the optimization of acetyl groups at elevated temperatures.
They concluded that the structure of the ﬁbers was altered by alkalization and acetylation.
Bisanda [28] studied the effect of alkali treatment on the adhesion characteristics of sisal
ﬁber reinforced epoxy composites. He observed that the alkali treatment of sisal ﬁber
improves the wetting ability of the ﬁber with the resin, reduces the voids, and improves
strength and water resistance.

Rout et al. [29] studied the graft copolymerization of acrylonitrile onto coir ﬁbers using
CuSo4 and NaIOa, as well as inorganic salts and organic solvents as the initiators.
Morphological studies showed the evidence of grafting taking place on the ﬁbers and
even penetrating onto the ﬁber matrix. Grafting resulted in higher maximum stress at

break, and increased hydrophobicity of the coir ﬁbers.

26

Stamboulis et al. [30] studied the effect of environmental conditions on mechanical and
physical properties of ﬂax ﬁbers. They treated the ﬂax ﬁbers with Duralin treatment,
which consists of heating the ﬁbers in steam or water at 160 0C for 30 min in an
autoclave, followed by drying at 250 0C for 2 hours. It was found that Duralin treated
ﬁbers absorbed less moisture than untreated green ﬂax ﬁbers. Treated ﬁbers were smooth
and well separated, and have uniform higher strength. Zeta potential measurements
conﬁrmed higher hydrophilicity of green ﬂax, and hydrophobicity of Duralin treated
ﬁbers.

Silva et al. [31] performed mechanical and thermal characterization of alkali treated
native Brazilian coir ﬁber. Mechanical and morphological results pointed at higher
ultimate tensile strength and modulus of coir ﬁbers after treatment. Thermal stability of
ﬁbers increased up to 48 hours of mercerization.

Gassan and Bledzki [32] determined the relationship between the structure and
mechanical properties of alkali treated jute ﬁbers. Due to alkali treatment, there is
shrinkage of the jute ﬁbers which inﬂuences ﬁber structure, crystallinity ratio, degree of
polymerization and Herrnans factor, as well as mechanical properties.

Bismarck et al. [33] studied the effect of ﬁber surface treatments on the thermal and
electrokinetic properties of coir and sisal ﬁbers. The ﬁbers were dewaxed, alkali treated
and grafted with methylmethacrylate (MMA). It was seen from SEM micrographs that
coir ﬁbers were larger in diameter than sisal. After dewaxing and alkali treatment, the
ﬁbers became rougher, while, after MMA grafting they became very smooth. The
maximum degradation temperature of the ﬁbers increased after alkali treatment and

MMA grafting. All treatments led to an increase in the water absorbed by the ﬁbers.

27

Higher accessibility of the surface functional groups resulted in lower zeta potential,
improving the interaction between the ﬁbers and the matrix.

Mohanty et al. [34] alkali treated and exploded the corn stalks with ammonia and carbon
dioxide and studied their morphology and thermal properties of the ﬁbers. As a result of
these treatments, some ﬁbrils got separated from the ﬁber bundle, and the thermal
stability increased.

Samal and Ray [35] studied FTIR spectra of chemically modiﬁed pineapple leaf ﬁbers.
The treatments done on the ﬁbers were: alkali, dinitrophenylation, benzoylation, and
benzoylation-acetylation.

Pavithran et al. [36] evaluated the impact performance of unidirectional sisal-polyester
composites. Owing to the optimum microﬁbrillar angle of sisal ﬁbers, they exhibit a very
impact amongst lignocellulosic materials. Energy absorption in the sisal biocomposites
occurred by ﬁber failure and ﬁber pull cuts. The same authors also examined the impact
performance of pineapple, banana, coir and sisal ﬁbers in an unsaturated polyester resin
matrix [37].

Oksman et al. [38] evaluated the mechanical and morphological properties of
unidirectional sisal-epoxy composites made using resin transfer molding (RTM). The
sisal ﬁbers were non-uniformly distributed in the polymer matrix, and the adhesion
between ﬁbers and matrix was weak.

Belcher et al. [39] studied the ﬁber matrix adhesion of the silane treated aligned
henequen ﬁber reinforced epoxy composites. Coupling with y-glycidoxypropy
ltrimethoxy silane (GPS) led to compatibilization between henequen ﬁbers and epoxy

matrix, thereby improving the thermal and mechanical properties of the biocomposites.

28

Belcher et al. [40] also studied the effects of epoxy modiﬁcations and surface
modiﬁcations on the aligned henequen ﬁber reinforced epoxy composites. They
introduced various amounts of epoxidized soybean oil and epoxidized linseed oil into the
matrix, and also modiﬁed the ﬁbers by using alkali treatment, plasma treatments, silane
treatment and ultraviolet treatment.

Varma et al. [41] made hybrid composites with glass ﬁbers and jute ﬁbers in a polyester
resin matrix. The jute ﬁbers were modiﬁed with y-aminopropyl trimethoxy silane,
isopropyl triisostearoyl titanate, and tolylene diisocyanate (TDI). Although titanate
treatment resulted in the properties of hybrid composites, Silane and TDI treated jute
ﬁbers were not compatible with the polyester resin.

Mishra et al. [42] studied the tensile, ﬂexural, impact and hardness properties of the
unidirectional untreated and bleached jute-epoxy composites. Bleaching the ﬁbers caused
deligniﬁcation which led to improvement in toughness, hardness, interlaminar shear
strength and ﬂexural properties of the composite.

Rout et al. [43] studied the inﬂuence of surface modiﬁcations of coir ﬁbers on the coir-
polyester resin composites. The ﬁbers were either alkali treated, bleached or grafted with
acrylonitrile. The authors observed that adhesion between the coir ﬁbers and the matrix
improved after all surface treatments. The water absorption tendencies of the composites
also decreased after the treatments.

Zimmerman and Losure [44] reported the use of non-woven kenaf ﬁber mats as
reinforcements for epoxy matrix. Although they found voids in the samples led to lower

properties, they recommended pursuing this research.

29

Rong et al. [45] studied the interfacial interaction in sisal-epoxy composites and its effect
on impact properties. The ﬁbers were subjected to alkali treatment, acetylation,
cyanoethylation, y-aminopropyl triethoxy silane, heat treatment, and mixed treatments.
The modiﬁcations led to increased surface energy of sisal ﬁbers after treatments, and
chemical bonding between the ﬁber and the matrix. The microfailure mechanism of the
composites was a function of interfacial adhesion and the ﬁber length continuity. For
chopped ﬁber reinforced composites, the interfacial strength should be tailored to
enhance the energy dissipation through debonding and pull out of the ﬁber bundles.
Mishra et al. [46] analyzed the effectiveness of various types and degrees of surface
modiﬁcations of sisal ﬁbers in improving mechanical properties of sisal—polyester
composites. 5% alkali treatment, 10% acrylonitrile grafting and cyanoethylation at 60 0C
were the optimum modiﬁcation conditions with respect to mechanical properties.
Sydenstricker et al. [47] performed thermal, morphological and pull out analysis on siasl-
polyester composites. The Brazilian sisal ﬁbers were treated with alkali and N-isopropyl-
acrylamide. The ﬁbers treated with acrylamide solution performed better in tensile
properties, moisture content and pull out lengths as compared to untreated and alkali
treated ﬁber based composites.

Hill and Khalil [48] studied the effect of treatments on mechanical properties of
composites made with polyester resin and coir or oil palm as reinforcements. The ﬁbers
were treated with acetylation, y-methacrylopropyltrimethoxy silane, and
nepenty1(diallyl)oxytri(dioctyl)pyrophosphate titanate.

Khalil et al. [49] observed the effect of various anhydride treatments on the mechanical

properties and water absorption tendency of oil palm empty fruit bunches in a polyester

30

matrix. The ﬁbers were reacted with acetic, succinic, or propionic anhydride without any
catalysts at 100 0C fbr 1 hour. Modiﬁcations resulted in hydrophobic ﬁbers and improved
ﬁber-matrix bonding. Acetylated ﬁbers had better properties compared to propionylated
or succinylated ﬁbers.

Mohanty et al. [21] reviewed the surface modiﬁcations of natural ﬁbers and the
properties of resulting biocomposites. They highlighted recent studies and developments
in the area of improvement of the ﬁber-matrix adhesion for thermoplastic as well as
thermoset matrices reinforced with natural ﬁbers.

Mohanty et al. [50] also reported surface treatments of natural ﬁbers in a polyester resin
matrix which improved the interface. The modiﬁcations included: alkali treatment,
acrylonitrile grafting, methyl methacrylate grafting, and cyanoethylation.

Ray et al. [51, 52] studied the effect of alkali treatments on the ﬂexural and tensile
properties of jute-vinyl ester composites. The improvements in mechanical properties
were highest for a composite made with 4 hours alkali treated jute ﬁbers with 35% ﬁber
weight.

Ray et al. [53] analyzed the dynamic mechanical properties of alkali treated jute-vinyl
ester composites. For alkali treated ﬁber based composites, the rate of all of storage
modulus with temperature was inversely proportional to the defect concentration in the
composites.

Gassan and Bledzki [54] evaluated the effect of cyclic moisture absorption-desorption on
the mechanical properties of silane treated jute-epoxy composites. The absorption-

desorption cycle led to debonding of the resin from the ﬁbers, and introduced craks in the

31

matrix. It also changed the fracture mechanism of the composite, but did not alter the
tensile strength.

Depworth et al. [55] produced high volume fraction of hemp and ﬂax ﬁber composites
using low viscosity epoxy and phenolic resins. The ﬁbers used were untreated, retted,
mechanically decorticated or soaked in urea.

Rout et al. [56] studied the effect of ﬁber surface treatments on the mechanical properties
of coir-polyester composites. The surfaces of coir ﬁbers were changed by use of
dewaxing, alkali treatment, and graft copolymerization with methyl methacrylate onto
alkali treated ﬁbers.

Joseph et al. [57] studied the effect of hybridization in jute/cotton fabric reinforced
polyester resin composites. Giacomini et al. [58] manufactured composites containing
curaua ﬁber and unsaturated polyester resin. Pothan et al. [59] studied the vicsoelastic
properties of banana ﬁber reinforced polyester composites. Oksman et al. [60] analyzed
the morphology and mechanical properties of sisal-epoxy composites made using resin

transfer molding (RTM).

2.1.9.3 Studies on bioﬁber reinforced thermoplastics

As mentioned before, one of the major drawbacks of using cellulose ﬁbers as
reinforcement is because of their poor dispersion characteristics in many thermoplastic
melts, such as polypropylene and polystyrene, due to their hydrophilic nature. Several
methods have been suggested and described in the literature to overcome this problem.
Among them are ﬁber surface modiﬁcation, use of dispersing agents such as stearic acid,

and ﬁber pre-treatrnents such as acetylation. Fiber dispersion can also be improved with

32

increased shear force and mixing time [61]. A careful selection of initial ﬁber lengths,
processing aids, processing techniques as well as processing conditions then is necessary
in order to produce high performance composites.

Raj and Kokta [62] investigated the inﬂuence of using various dispersing aids (stearic
acid and mineral oil) and a coupling agent (maleated ethylene) in cellulose ﬁber
reinforced polypropylene composites.

Good ﬁber dispersion is generally the ultimate objective of any mixing process [61].
Various mixers are available to mix short ﬁbers in thermoplastics such as extruders,
plasticorder, injection molding machines and kneaders. Different mixing techniques,
however, do not produce composites with the same degree of ﬁber dispersion.
Therrnokinetic mixers have also been used to mix cellulose ﬁbers with thermoplastics to
effectively disperse the cellulose ﬁbers within thermoplastic matrices [63, 64]. Pereira et
al. [65] investigated the effect of several processing techniques on the properties of
polypropylene composites reinforced with short sisal ﬁbers. The best processing method
involved a twin-screw extruder.

Childress and Selke [66] investigated the effectiveness of several additives in enhancing
mechanical properties of wood ﬁber/high-density polyethylene composites. The additives
used were ionomer-modiﬁed polyethylene (ION), maleic anhydride modiﬁed
polypropylene (MAPP), and two low molecular weight polypropylenes [67]. The
mechanical properties of the composites studied increased with increasing additive
concentration. The most effective additive was MAPP, followed with ION.

Thomas et al. [68] reinforced polystyrene with benzoylated sisal ﬁbers. The results

revealed better compatibility between treated cellulose ﬁbers and the polystyrene matrix,

33

and this resulted in enhanced tensile properties of the resulting composite. These
improvements were attributed to the similarity between the phenyl-structure present in
both benzoylated sisal ﬁbers and polystyrene, which makes them thermodynamically
compatible with each other.

Ali et al. [69] studied the effects of processing conditions on the viscoelastic and
mechanical properties of biodegradable composites made with extrusion of sisal ﬁbers
and mater Bi-Y and mater Bi-Z. Tensile and creep behavior of composites was affected
by type of polymer matrix and processing conditions such as: speed of mixing, time of
mixing and temperature.

Alvarez et al. [70] evaluated the mechanical properties and water uptake of the
compression molded composites made of alkali treated sisal ﬁbers and MaterBi-Y, which
is a biodegradable polymer. They found that the ﬁber treatment produced an increase in

the equilibrium moisture content and a decrease in the diffusion coefﬁcient.

2.2 Unsaturated polyester resin

The majority of resins used in the composite industry are thermosets (see Figure 2.2.1)
[71]. About 65 % of all composites produced currently for various applications, use glass
ﬁber and polyester or vinyl ester resins. Unsaturated polyester resins (UPE) are widely
used in the composite industry because of their relatively low price, low density, ease of
handling, thermal and dimensional stability, good chemical and weather resistance, and
excellent mechanical, chemical and electrical properties. Furthermore, compared to other
thermosets, unsaturated polyester resins over can be pigmented, and can be easily ﬁlled

and ﬁber reinforced in a liquid form. The use of reinforced thermoset composites by

34

automakers has nearly doubled in the last decade, and is expected to increase 47 percent

during the next ﬁve years through 2004 [50, 72].

l 5%
other
thermosets

   
    

1 5%
. 70%
Reinforced .
. Reinforced
thermoplastics

unsaturated polyester

Figure 2.2.1: Matrix pattern of polymer composites [71]
The polyester resins, because of their versatility and low cost, are widely used throughout
the world. Polyester resins are classiﬁed as: (l) ortho resins, (2) isoresins, (3) bisphenol-

A-fumarates, (4) chlorendics, and (5) vinyl ester [73]. Ortho resins, known as general-

Table 2.2.1: Cost of various olyester resins

 

 

 

 

 

Cost (Cents/lb)
Polyester Resins
Ortho resin 60-68
Iso-resin 73-83
Bisphenol-A 123-153
Vinyl ester 147-161

 

purpose polyester resin, are based on phthalic anhydride, maleic anhydride and glycols
(Figure 2.2.2). Ortho-resin is the least expensive among all the polyester resins (Table
2.2.1). The solutions of unsaturated polyesters and styrene vinyl monomers (reactive
diluent) are known as UPE resins.

2.2.1 Making UPEs:

35

The ﬁrst unsaturated polyester resins of similar type as used today were synthesized in
the 1930s. Polyesters are subdivided into three classes: aliphatic, aromatic, crosslinked.
The third category is thermosetting polymers. Polyester resins are manufactured by a step
growth polymerization reaction of unsaturated acids (or anhydrides), saturated aromatic

acids, and difunctional alcohols glycols [73].

 

 

HO/

 

 

Figure 2.2.2: A representative chemical structure of unsaturated polyester

The basic chemistry of linear unsaturated polyesters is rather simple. A mixture of
unsaturated and saturated dicarboxylic acids is reacted with diols in a melt
polycondensation. Monofunctional alcohols and acids are also used in some formulations
to tailor the properties. The most traditional composition is maleic anhydride, o-phthalic
anhydride and 1.2-propanediol, which are cheap raw materials. Other common raw
materials are ﬁlmaric acid, isophthalic anhydride, terephthalic acid, adipic acid, ethylene
glycol, diethylene glycol, dipropylene glycol, neopentyl glycol and bisphenol A [74]. The
properties of the ﬁnal product can be varied almost endlessly by changing the
composition of the unsaturated polyester using these raw materials. Generally aromatic
groups improve the hardness and the stiffness while aliphatic chain components increase
the ﬂexibility.

The unsaturated polyester has typically a molecular mass between 1000 and 5000. The

molecular mass is regulated by the diol/dicarboxylic acid ratio [75]. Usually the diol is in

36

excess, as the used diols are liquids, while the dicarboxylic acids and anhydrides are
solids. An excess of solid reactants can cause a problem in the form of sublimation of the
reactants during polycondensation. A high molecular mass will give a higher hardness,
tensile and ﬂexural strength of the ﬁnal cured material. If the molecular mass is too low,
the mechanical properties of the cured resin will be poor. A too high molecular mass
increases the viscosity of the resin solution, which will cause problems with the
processing of the resin. Air entrapment in the laminate, poor wetting of the reinforcement,
long mould ﬁlling times and processing times are typical practical problems due to the
resin viscosity.

Almost all commercial production of unsaturated polyesters is done by the melt
polycondensation of unsaturated and saturated acids or anhydrides with glycols. No
solvents are used, and the formed water is continuously removed, in order to force the
esteriﬁcation reaction towards completion. The condensation temperature is typically
between 170 and 230 °C. At the end of the condensation, vacuum is often applied in
order to remove remaining water from the viscous melt. The total reaction time can be
from 8 h up to 25 h, and the reaction is followed by acid number titrations and viscosity
measurements [76]. Azeotropic polycondensation in the presence of organic solvents
such as xylene or toluene can also be used. The reaction takes place at lower temperatures
and it is possible to avoid losses of volatile reactants. The drawbacks are longer reaction
times and environmental problems with solvent removing and recycling.

The most common glycol used is propylene glycol. It is low cost and has good balance
of properties. A combination of bisphenol A and propylene glycol provides good

chemical resistance and high HDT to the oligomer. The most common anhydride used is

37

maleic anhydride, it provides cure sites. Fumaric acid is the best unsaturated acid used.
Amongst saturated acids/anhydrides, phthalic anhydride is low cost and hard, isophthalic
acid improves strength and chemical resistance, and adipic acid imparts ﬂexibility and
toughness. The UPE provides polymer properties, including modulus, toughness, glass

transition temperature, and durability to the resin formulation [77].

O O
.l O + O + CH3—‘CH2—CH2 -—->
OH OH
O o

Phthalic anhydride Maleic anhydride Propylene glycol

it 1? it it

-— C—O—(EH-CHz—O—C—CH=CH—C—O—CH—CH2—0——

— . _ + H20 _ n
Unsaturated polyester resrn

 

 

Figure 2.2.3: Preparation of polyester resin [74]

2.2.2 Diluent (Monomer)

The polyester resins are usually diluted by adding a low molecular weight comonomer to
adjust the viscosity of the mixture. Reactive diluent or monomer is added to the oligomer
in the weight range of 0-60 %, typically 35-45 %. It helps to control the viscosity of the
UPE, acts as a crosslinker, and improves wetting behavior [78]. Because of its lower cost,
the most common monomer used for UPE is styrene, but other monomers can also be
used. The examples of monomers used for UPE system cross linking are shown in table

2.2.2

38

Table 2.2.2: Examples of monomers used for UPE system cross linking

 

Mainly used for translucent parts. Always used in combination

 

Methylmethacrylate with styrene (50:50). Excellent weathering.
Used for some special applications where a good control of the
:égtzrenethyl reactivity and good control of shrinkage is needed. Unpleasant

odor

 

High reactivity. Provide high chemical resistance. Lower impact

Divinyl benzene .
res1stance.

 

Low volatility. Used in molded parts with good electrical

Diallyl phthalate p erformances.

 

 

Used in special applications like stratiﬁcation of PS foam.

““34 acetate Low hydrolysis resistance.

 

 

2.2.3 Curing of UPE

The curing of polyester resin occurs by free-radical chain-growth crosslinking. In
addition to the monomer and oligomer, curing agents need to be added for the reaction to
proceed. Curing agents include initiator, promoter, accelerator and inhibitor.

Initiator

To initiate the reaction a source of free radicals is needed. Organic peroxides are used as
the source of ﬁ'ee radicals [74]. They refer to a family of molecules containing at least

two oxygen atoms, single-bonded together. The general structure is:

R—‘-———0—0—R2 R‘—o—o—R—o—3 o—R2 Rl—o—o-H

The 0-0 bond can be easily broken to generate very reactive species called free radicals.
The free radicals can be generated either by the action of heat (homolitic scission), or at
room temperature in presence of an activator like metallic salts compounds or an amine.

The ﬁrst step of initiation is the decomposition of the peroxide. The second step is chain

propagation reaction via styryl radical formation capable of reacting with a double bond.

39

 

CH3\ H CH3\ /C H3
2 H2 H2
OOH— (It—OOH OOH—C— O—O— T— OOH
C H3 CH3 CH3
Monomer Dimer

Figure 2.2.4: Chemical structure of methyl ethyl ketone peroxide (MEKP)

The main families of organic peroxides used for curing unsaturated polyester are: ketone
peroxides, diacylperoxides, peresters, hydroperoxides, and perketals. Ketone peroxides
are generally used at room or mid temperature in combination with Cobalt salts.
Diacylperoxides like Benzoyl peroxide can be used at room temperature, activated by
tertiary aromatic amines, or without promoter at higher temperature (100°C). Peresters
and perketals are used at higher temperatures and are preferred in injection and
compression molding processes. Hydroperoxides are used at higher temperatures in
injection and compression molding processes [79].

Ketone peroxides are the most commonly used for curing UPE systems. They have 8-10
% active oxygen. For room temperature curing as in the case of hand-lay-up structures,
methyl ethyl ketone peroxide (MEKP) is used; for moderate temperature curing benzoyl
peroxide is used. For hot press di-t-butyl peroxide or t-butyl perbenzoate is used. A

mixture of initiators is used when a large temperature increase is expected.

Q—l‘l-o—o—il

Figure 2.2.5: Chemical structure of benzoyl peroxides (BPO)

40

The selection of the peroxide determines the kinetics of reaction and is also an important
parameter for the "pot life". Final part quality is also linked to the peroxide used (aspect,
curing efﬁciency). The selection of the right peroxide system involves several parameters
including: part produced, associated process, processing temperature, timing of the
process, nature of the pure resin used, and formulation. Optimum performance can be
achieved by controlling the type and concentration of peroxide and promoter [80].
Promoters and accelerators

Promoters and accelerators are used to help initiate cure at room temperature by
accelerate the decomposition of peroxides. Commom accelerators include metallic salts,
amines, or mercaptants. Cobalt naphthenate (CoNap) and cobalt octanoate (CoOc) are the
most widely used accelerators.

Cobalt naphthenate (CoNap) is usually added to systems containing MEKP, in weight
range of 0-0.3%. Dimethyl aniline (DMA) is usually added in the weight range of 0-
0.3% to systems containing BPO and MEKP. Gel time for a given resin depends on

initiator level, promoter level, second promoter level, and temperature.

ROOH+Co3+——-> Roo‘ + H+ + Co2+
lroorr+Co2+ ———> my + on" + (303+

Figure 2.2.6: Reaction mechanism of polyester resin with cobalt activator

Inhibitors

In order to prevent premature curing and to extend pot-life, inhibitors are also introduced
into the curing system, typically in concentrations less than 100 ppm. Inhibitors are free
radical scavengers like Hydroquinone derivatives, terbutylcatechol, cresol derivatives.

Some examples of inhibitors used for UPE system are shown in Table 2.2.3

41

Table 2.2.3: Examples of inhibitors used for UPE system

 

Hydroquinone Excellent for stabilization during storage No color issues

 

Can be used in a wide range of temperatures. Very efﬁcient

Tertlobutylhydroqumone with systems initiated by Dibenzoylperoxide

 

Provides very good stability to the resin until 60°C. Used

D'tertmbutylhyquumone when short gelling time is needed (cold and hot process)

 

Can be used in a wide range of temperature. Very efﬁcient
Toluhydroquinone with systems initiated by Dibenzoylperoxide. Provides
excellent thermal resistance

 

Excellent performances. Provides yellow color to the

 

 

 

 

Benzoquinone
system
. Active until 40°C. Rapidly absorbed by the resin.
Tertrobutylcatechol Excellent adjustment of pot-life
C rosslinking

Polymerization may occur by any of three mechanisms: free radical, cationic, anionic.
Chain polymerization is characterized by the presence of a few active sites which react
and propagate through a sea of monomers. It may be formed via a two step process,
where a polymer with unsaturations is ﬁrst formed (a thermoplastic formed via step
polymerization), and then these unsaturated sites are reacted with a crosslinking agent in
a second step to produce the ﬁnal structure. The reaction of a UPE resin is a free-radical
chain-grth crosslinking copolymerization between the styrene monomer and the UPE
molecules.

The curing reaction involves a sequence of steps: initiation, inhibition and/or retardation,
propagation, and termination. Polyester molecules are the crosslinkers while styrene
serves as the agent to link adjacent polyester molecules. The polymer chains grow and
become crosslinked in three possible reactive processes: styrene—polyester co-

polymerization, styrene homopolyrnerization, and polyester homopolymerization [81].

42

 

The curing behavior is complex due to the interaction between the resin chenristry and
the variation of the physical properties. There are four stages in the crosslinking of UPE:
Gelation: incipient formation of an inﬁnite molecular network

Vitriﬁcation: glass transition temperature of the forming polymer rises above the
temperature of cure

Full cure: highest attainable degree of cure

Devitriﬁcation: degradation

Before the ﬂee—radical polymerization proceeds, the resin is a viscous liquid. During the
curing reaction, the chain length of the resin molecules grows through the crosslinking
reaction of functional groups and the resin becomes more viscous. As the curing
advances further, the reacting system forms a highly crosslinked network, which results
in a rapid increase in the resin viscosity [73].

Gelation corresponds to the formation of the ﬁrst insoluble fraction of the polymer with
an inﬁnite molecular weight. The onset of gelation is of critical importance to processing

thermoset systems. The general crosslinking mechanism is shown in Figure 2.2.7.

 

_ Liquid Solid: 3D network
‘8 Styrene monomer : Double bond

~

Polyester chain

Figure 2.2.7: Crosslinking mechanism of UPE resins

43

The reactivity, the viscosity and the ﬁnal properties of UP resins can be adjusted by
changing either the nature of the UPE chain, chemical composition (aliphatic, aromatic,
number of double bonds), resin molecular weight, nature and concentration of the
crosslinking monomer (containing at least one double bond or ratio of the
styrene/polyester content. [76-81 ].

Autoacceleration

Autoacceleration effect can occur when the increasing viscosity limits the rate of
termination because of diffusional limitations. The free radicals can't terminate with as
easily as they can ﬁnd other monomers to propagate with. Rate decrease occurs because

of further diffusional limitations caused by vitriﬁcation [].

Effect of Additives on Cure

Surfaces, sizings, and surface treatments on ﬁbers and ﬁllers can signiﬁcantly alter not
only the ﬁnal properties of a composite, but also resin cure kinetics. Polyester resins are
generally associated with ﬁllers (silica, CaCO3, MgO, clay), and ﬁbers (glass ﬁbers, glass

ﬁbers mats or fabrics) acting mainly as a binder.

2. 2. 3.1 Curing performance measurement

Lots of methods are used for determination of curing kinetics. As an example, the curing
kinetic and performance in a cold cure system can be measured by the viscosity increase
over time and/or the temperature increase over time. The crosslinking reaction between
the unsaturated polyester and styrene is exothermic. The exothermic heat generated is

proportional to the level of unsaturation of the chain and the amount of styrene. The heat

44

generated can auto accelerate the hardening reaction but some processes cannot sustain
high temperatures. In molding processes, high temperatures can lead to excessive
shrinkage, warpage and cracking for high thicknesses. Figure 2.2.8 below shows the

principle of one possible measurement and the typical curves obtained.

   
 
     
    
   

Propagation/
Autoacceleration

[I

Diffusion
limited

[I

Reaction rate

Induction
time

B

 

 

Time

Figure 2.2.8: Curing reaction rate of UPE

2.2.4 Studies on UPE
Considerable work has been reported on the synthesis, characterization, curing behavior

of UP resins [82-155].

2.2.4.1 Studies on curing of UPE
Kama] and Sourour [82] modeled of the DSC cure reaction with empirical equation

dx/dt=kx"'(1-x)" by assuming the total reaction order m+n=2 and a zero initial cure rate

[82].

45

Salla and Martin [83] observed the dynamic, isothermal and residual heats of curing UPE
at different scanning rates, temperatures and setting times by use of DSC. At room
temperature, UPE curing follows an autocatalytic reaction system. It was difﬁcult to ﬁnd
extent of reaction for such a system because at lower temperatures, the instrument was
not sensitive enough, and at higher temperatures, some heat was lost even before the
instrument could detect it. Residual heat can help to determine the ultimate extent of
curing of any system.

Salla et al. [84] used DSC to study the therrno-dependence and thermodynamics of the
curing of UPE with different catalyst systems. For curing the UPE, they used BPO (0.03-
3.0 % wt), DMA, MEKP (1 wt "/o) and CoOc (0.2-10 wt %). In dynamic DSC
experiments, more than one exothermic peak was seen. In the absence of a promoter,
only one peak was observed. When the amount of initiator is varied, the exothermic peak
gets shifted. Without post curing, all the formulations are only partially cured.

Paci et al. [85] used pulsed nuclear magnetic resonance (NMR) at 20 MHz to monitor the
middle to end stage of curing of UPE. The gelling and network formation process was
followed by measuring spin lattice relaxation time of the system. A semi-empirical
kinetic equation was used to evaluate degree of cure as a function of time, kinetic
constants, and activation energy.

Rarnis and Salla [86] investigated the theoretical and experimental curing reaction of
UPE system with different polyester/styrene ratios by means of DSC and gel permeation
chromatography (GPC). The theoretical heat of reaction corresponding to complete
conversion was extrapolated from DSC curves. Experimentally, complete conversion

was not seen in nay of the resin systems. From the molecular weight distribution obtained

46

from GPC, it was seen that part of styrene and polyester, as well as low molecular weight
oligomer s of styrene and branch styrene on polyester molecules, do not form the
polyester network.

Rouison et al. [87] used an autocatalytic model to study the curing kinetics of a polyester
resin containing various promoters and an inhibitor. They used an isothermal DSC
method to obtain curing rate, rate constants, enthalpy of reaction, and order of equation.
The experimental plots ﬁtted well with the model and were temperature dependent.
Nzihou et al. [88] utilized DSC for studying the polymerization kinetics of a thermoset
resin under isothermal and dynamic conditions. They proposed a phenomenological
kinetic model, which took into account diffusion effects, for curing reaction. They found
that under isothermal conditions conversion was less than 100 % because of presence of
unreacted monomer in the system. Dynamic studies in DSC revealed higher conversion at
higher temperatures.

Vilas et al. [89] studied the curing kinetics of UPE using DSC and thermal scanning
rheometry (TSR). When the concentration of initiator (MEKP), or the cure temperature is
increased, the gel time decreases, and the reaction rate increases.

Huang and Leu [90] studied the effects of temperature, initiator and promoter on low
temperature curing (30-50 0C) of UPE. In the early stages of reaction, the
copolymerization is azeotropic. During diffusion controlled propagation stage, the
conversion of styrene is greater than that of vinylene groups. The cross link lengths of
styrene decrease with increasing temperature. At moderate reaction rates, the

intermicrogel crosslinking reaction is more dominant than intramicrogel crosslinking

47

.a'

“J

.lrl

‘1'

,
u.

 

reaction. At later stage of reaction, intramicrogel crosslinking is more favorable than
interrnicrogel crosslinking.

Yang and Lee [91] quantiﬁed the effects of resin type, temperature, curing agents, on the
reaction kinetics of UPE and vinyl ester cured at low temperatures by DSC and
rheometry. When MEKP is used as the initiator, inhibitor benzoquinone provides a longer
induction period and higher ﬁnal conversion for UPE compared to vinyl ester resin.
Pentanedione is a good retarder for vinyl ester but acts as an accelerator for UPE. They
also developed a gel time model to quantify the effects of temperature, inhibitor, retarder,
on both polyester and vinyl ester resin. UPE is more sensitive to temperature changes
than vinyl ester resin.

Yang and Lee [92] studied the reaction kinetics of three UPE resins having different C=C
bonds per molecule, by using a differential scanning calorimeter (DSC) Fourier transform
infrared (FT IR), and electron spin resonance (ESR) spectroscopy. They also developed a
diffusion-controlled model to simulate the reaction rates and conversion proﬁles of
polyester vinylene and styrene vinyl groups, as well as the total reaction rate and
conversion. The results from DSC and F TIR match well. The reaction rate increases with
increase in the degree of C=C unsaturation per molecule.

The diffusion-limitation effect is more signiﬁcant for the polyester resin with a higher
degree of unsaturation, leaving more unreacted C=C bonds trapped inside the matrix after
vitriﬁcation, resulting in lower ﬁnal conversions of polyester and styrene C=C bonds
The glass transition temperature measured by the DSC is a linear function of the ﬁnal
conversion within the temperature range studied, and it is used to monitor the change of

the ﬁnal conversion.

48

Elli

t’i '

..4
cl...

,4
I ‘5)

According to ESR measurements, for the polyester resin with a higher degree of
unsaturation, the polymer formed is more compacted due to more intramolecular reaction,
and more radicals are trapped in the cured resin without termination. The mechanistic
kinetic model was capable of reasonably predicting the reaction rate and conversion
proﬁles of polyester vinylene and styrene vinyl groups, as well as the total reaction rate
and conversion at different temperatures and at nonisothennal conditions.

Chen and Yu [93] studied the microgelation phenomenon of UPE curing by static and
dynamic light scattering as well as DSC. The formation of microgels due to
intramolecular crosslinking reaction inside the UPE coils, was observed by studying the
variation between UPE particle size and the depolarization ratio. There is a decrease in
the compatibility between partially cured UPE and styrene monomer as the
intermolecular crosslinking reactions amongst the microgel particles proceeds further.
Kosar and Gomzi [94] evaluated the thermal effects of curing reaction of UPE by
isothermal and dynamic DSC experiments. They also developed a numerical model
which took into account the heat transferred by conduction through the resin, as well as
the kinetics of heat, generated by cure reaction. There was good agreement between the
predicted model and the experimental data.

Yun et al. [95] examined the curing kinetics of unsaturated polyester resin system
exhibiting apparent induction periods by modeling free radical initiation and propagation
processes. After making a master curve of the reaction exotherms, a single activation
energy of the reacting system was determined for both induction time (ti) and the
maximum reaction-rate time (tm). Using the power-law equation for the initiation

efﬁciency function, two elementary model equations were proposed for the radical and

49

monomer reaction rates. The developed model equations predicted isothermal as well as
dynamic heating curing reactions very well with only ﬁve model parameters without any
modiﬁcation in different thermal conditions.

Ziaee and Palmese [96] evaluated the effects of cure temperature on proprieties of vinyl
ester resin system using NMR, FTIR, AFM and tensile properties. Post curing the
samples after isothermal curing showed the same conversion of styrene and vinyl ester
bonds regardless of the initial cure temperature. The glass transition temperature was
observed between 118-122 0C, irrespective of curing temperatures. The fracture
toughness of samples cured at lower temperature (30 0C) was three times higher than
those cured at higher temperature (90 0C). AF M revealed nodular morphology which
could be microgels.

Brill and Palrnese [97] used FTIR spectroscopy to investigate vinyl ester-styrene bulk
copolymerization curing kinetics. The depletion of vinyl ester and styrene double bonds
was monitored separately. Styrene monomer continues to react even after vinyl ester has
completely reacted. Overall conversion increases with increase in isothermal cure
temperature.

Near IR and mid IR spectroscopic techniques were used for cure characterization of UPE
by Grunden and Sung [98]. In addition to NIR analysis, they calculated the extent of
reaction of styrene and vinylene C=C, in comparison to extent of reaction values from
conventional mid—IR. The differences between the conversion of styrene and vinylene
C=C is due to reduced diffusion of styrene into tightly crosslinked microgels.

Lin and Hsu [99] studied the cure reactions of aromatic dicyanate, diepoxide, diamaine

system using in situ FTIR and DSC. They proposed six reaction paths for dicyanate

50

/diepoxide system, four reaction paths for diepoxide /diamine system, and four reaction

paths for dicyanate Idiepoxide /diarnine system.

2.2.4.2 Studies on properties of UPE

Shan et al. [100] studied the effect of vinyl ester-styrene network structure on the thermal
and mechanical properties by using DSC and DMA. The crosslink density of the vinyl
ester resin was changes by two methods: by changing molecular weight of oligomer and
by changing amount of styrene added to the curing system. Vinyl ester /styrene systems
have secondary relaxations, which are indistinguishable from each other, though the
physical network and the composition of network chains are different. Tg of the systems
increases with decrease in cross link density.

Han and Lem [101] studied the curing kinetics and the rheological properties of
unsaturated polyester resins. They found that as cure progressed, the steady shear
viscosity increased very rapidly with cure time at all shear rates investigated. The normal
forces showed negative values at low shear rates and positive values at high shear rates.
The negative normal forces may be due to material shrinkage during cure, and positive
normal forces due to deformation of large molecules, formed by crosslinking reactions
during cure. They explained the increase of viscosity with the degree of cure, at various
values of isothermal curing temperature by combining the rheological and DSC
measurements.

Han and Lem [102] investigated the effect of particulates on both the rheological

properties and the curing kinetics of unsaturated resin. In the formulation, calcium

51

carbonate and clay were used as inorganic particulates and high-density polyethylene
powder as organic particulates. They authors found that as the particulate content
increased, the resin formulation gave rise to shear-thinning behavior and the rate of cure
increased. The observed shear-thinning behavior is attributable to breakage of the
particle agglomerates as the shear rate is increased, behavior which is typical of
concentrated suspensions. The CaC03 particles also helped to control shrinkage during
cure when the material was subjected to steady shear deformation. The gel time was
shorter for mixtures of resin and particulates than for the neat resin alone. A UPE system
with particulates showed negative normal stress effects at rest, which originated from
polymerization shrinkage. However, when the ﬂuid is subjected to steady shearing
deformation, the normal stress response was found to vary greatly among the various
particulates investigated.

Han and Lem [103] explored the effect of low-proﬁle thermoplastic additives on
rheology and curing kinetics of unsaturated polyester resin. The authors used UPE with
two different types of LPAs: poly(vinyl acetate) (PVAc) and poly(methyl methacrylate)
(PMMA). They observed that during cure the UPE/PMMA system exhibits shear
thinning behavior even before the cure time reaches the critical value whereas the
UPE/PVAc system does not. Both PVAc and PMMA helped to reduce the shrinkage of
the resin during cure. The rate of cure and the ﬁnal degree of cure are decreased when the
amount of low-proﬁle additive is increased. The shrinkage control becomes effective
only when the shear rate is greater than a certain critical value. The shrinkage of the resin
during cure, judged by the normal stress response is decreased by the addition of

thermoplastic additive. The shrinkage control is found to be more effective when the

52

mixture of UPE/LPA is subjected to intensive shearing deformation than when the
material is at rest. An optimum amount of low-proﬁle additive must be used in order to
minimize the sacriﬁce of the rate of cure and yet maximize the shrinkage control.

Serre et al. [104] examined the morphology of the outermost layer of a compression
molded samples made of UPE, glass ﬁbers, calcium carbonate, low proﬁle additives and
release agents by using atomic force microscopy (AFM), dispersive x-ray spectrometry
(EDX), scanning electron microscopy SEM, and XPS. The positions of ﬁbers, ﬁller, and
mold release, were observed.

Serre et al. [105] used atomic force microscopy (AF M) in the tapping mode to investigate
the morphology of cured four blends based on miscible UPE/LPA (low proﬁle
additive)/ST (styrene) systems. AFM identiﬁed particles, nanoparticles and nanogels
(with sizes varying from 20 to 57 nm), besides aggregates and microgels previously
observed by scanning electron microscopy (SEM). All samples were miscible with each
other, and indicated that miscibility was an important parameter in the morphology of the
whole network. Higher miscibility resulted in more numerous nanogels, and smaller
microgels. They also developed a relation between the microgels sizes and the void
volumes.

Marieta et al. [106] utilized AFM for studying the morphology of thermoset resins
toughened by incorporation of core shell particles and thermoplastics. They used AFM in
tapping and contact modes to look at the topographical, force and phase images of these

systems.

2.2.4.3 Studies on ﬂammability of UPE

53

Mazrouh [107] analyzed the ﬁre retardant characteristics of polyvinyl chloride (PVC) and
antimony trioxide (Sb203) in a UPE-ﬁberglass formulation. Addition of PVC (l-6%) and
Sb203 (4-9%) enhanced the ﬁre retardancy of the composites, with very little effect on
the mechanical properties of the composites.

Hangzhou JLS Flame Retardants Chemical Co. [108] developed a ﬂame retardants for
thermosets or for intumescent coatings. These retardants are based on ammonium
polyphosphate, which is a non halogen, environmentally friendly product. When
combined with the standard ﬂame retardant ﬁller, aluminum trihydrate (ATH), it lowers
the levels of ATH needed to reach the required ﬁre retardation regulations.

M. Chaote [109] described some of the chemical approaches for halogen free ﬂame
retardants in circuit boards industry. He recommended the use of phosphate esters,
melamine cyanurate, and phosphorous compounds as ﬁre retardants. However, in these
compounds, nitrogen and phosphorus have an intrinsic afﬁnity towards moisture. They
are also greatly susceptible to oxidative attack, and are readily thermally degraded.
Another approach is to reduce the amount of resinous fuel, and increase the amount of
hydrated ﬁller content (aluminum trihydrate).

Avtec Industries has developed a polymeric additive, TSWBTM, for ﬁre retardancy and
smoke suppressing, which is ﬂee from bromine, antimony, and magnesium. It is designed
and engineered to be incorporated within composite laminate structures (UPE based) to

provide ﬁre protection and thermal insulation properties and smoke suppression [110].

2.2.4.4 Studies on toughening of UPE

54

Rosa and Felisberti [111] modiﬁed the UPE with poly(organosiloxanes) to improve its
impact resistance. Glycidyl methacrylate, 1,3-aminopropyltriethoxy silane(ATPS), and
l,l,3,3-tetramethyl-l,3—diehtoxysiloxane were used as modiﬁers for UPE resin. From
dynamic mechanical analysis, it was observed that APTS was incorporated into the resin
by reaction of its amino group with GMA. In systems with lower weight fraction of
additives (7.23%), there was improvement in impact strength, however due to phase
separation; there was no improvement in systems with higher amounts of additives (14.5
%).

Rosa et al. also studied the adhesion between siloxane modiﬁed UPE and glass ﬁber by
acoustic emissions of composites [112].

Rajulu, et al. [113] blended epoxy and unsaturated polyester resin in chloroform and

studied the viscosity, ultrasonic velocity, and refractive indices of the blends.

Gawdzik et al. [114] evaluated the effect of concentrations of toluene diisocyanate (TDI)
on the thermo-mechanical properties of unsaturated polyester resins. 1-3 % by weight of
TDI was added to the UPE, and the samples were subjected to FTIR, thermo-gravimetric
analysis, Charpy impact test, and rheometry. TDI should not be added to UPE resins in
amounts exceeding 1 wt %. At higher concentrations if TDI, there is a huge increase in
the viscosity of the materials, and thixotropy is observed in samples containing 3 % TDI.

Min et al. [115] used polyurethane (PU) was used as a modiﬁer for UPE to improve the
toughness of the resin. They studied the effect of the polyol molecular weight as a
polyurethane (PU) soft segment and the PU contents on the toughness of PU-modiﬁed

UPE resins. The PU was prepared by reaction between polymeric methyl diisocyanate

55

(MDI), a difunctional poly(ether polyol), a trifunctional poly(ether polyol) and dibutyltin
dilaurate as the catalyst. The maximum toughness was observed in the case of 2 wt % of
PU in the curing system, due to the network formation of UPE with PU. Above 2 wt %,
the toughness value deteriorated because of the presence of unreacted polyol in the
system. In the case of difunctional polyol, polyol cannot participate in the network
reaction effectively as the molecular weight increases, so the toughness decreased
because the chain mobility decreased. But in the case of trifunctional polyol, that
touglmess was not affected considerably by the molecular weight.

Cherian and Thacil [116] performed the toughening studies of UPE by introducing
maleated elastomers using physical or chemical methods. The elastomers are either
dissolved in styrene and then blended with UPE, or elastomers are modiﬁed by grafting
with maleic anhydride.

Cherian and Thacil [117] prepared blends of UPE with functional rubbers, and studied
their mechanical properties. As the compatibility of the UPE with unmodiﬁed elastomers
is poor, and a suitably distributed rubber phase of optimum size is not formed, they used
elastomers bearing ftmctional groups. The elastomers used were epoxidized natural
rubber, hydroxyl terminated natural rubber, hydroxyl terminated polybutadiene, and
maleated nitrile rubber. Toughness and tensile properties were maximum at 2.5 wt % of
rubber in the system (studied range: 0-5 wt % rubber). Maleated nitrile rubber had the
best performance compared to all other elastomers.

Thouless et al. [118] discussed the mechanics of toughening of brittle polymers. There
are tow principles for toughening of brittle polymers. The ﬁrst is by manipulating the

cohesive processes acting across a crack tip, to increase the intrinsic toughness of the

56

materials. The second is by manipulation of the microstructure of the polymer, to make it
easier to trigger non-linear dissipative processes around the crack tip.

Huang et al. [119] reviewed the developments in toughening mechanisms of thermoset
polymers. A toughened thermoset usually contains elastic or thermoplastic domains
dispersed in discrete for throughout the matrix resin to increase the resistance to crack
growth initiation. The phenomenon of viscoelasticity, shear yielding, and dilatational
deformation involving cavitations govern the fracture behavior of thermosets.

McGarry and Subramaniam [120] prepared a toughened cross linked molecular network
system with unsaturated polyester, styrene, diglycidyl ether of bisphenol A, and a
reactive liquid runner, namely, amino terminated butadiene-acrylonitrile. The network is
a two phase system,iwhich has one phase containing no rubber, and the second phase

with uniform distribution of rubber.

2.2.4.5 Studies on modiﬁcation of UPE

Synthetic modiﬁcations of unsaturated polyester resins have been done in order to
achieve lower styrene emission, better styrene solubility and lower processing viscosity.
Two concepts were used; introduction of liquid crystalline segments into the unsaturated
polyester and end-capping the unsaturated polyester with poly(ethylene glycol)s of
various molar mass[121] .

Hegemann [122] developed a comonomer free low viscosity unsaturated polyester resin
for electrical industry. Two C=C bonds were found in a system by combining standard
UPE resin (has a C=C from maleic/ﬁrmaric unsaturation), and the adduct of

Cyclopentadiene with a functional group, which allows for a condensation reaction. These

57

unsaturations were incorporated in one resin, without copolymerization during the
polycondensation of the polyester. Crosslinking was studied by using DSC.

Douglas and Pritchard [123] studied the effect of substituents in the styrene ring, on the
water absorption characteristic and thermal stability of UPE. They used styrene, 4-mehtyl
styrene, 4-ehtyl styrene, 4-n-butyl styrene, 4-isoproyl styrene, tertiary butyl styrene, 4-
chlorostyrene, and 3, 4,-dichlorostyrene to crosslink a maleic/phthalic anhydride based
unsaturated polyester. Loss of water resistance was seen in systems with chloro
substituents and most systems with alkyl groups. System with tertiary butyl group leads
to a decrease in water absorption. Thermal stability is decreased by chloro substituents
and increased by n-alkyl groups.

Guo and Zheng [124] studied the miscibility and crystallization of thermosetting polymer
blends of unsaturated polyester resin and poly(s-caprolactone) (PCL). Before curing, only
one glass transition temperature (Tg) is present in (PCL) UPE blends, showing
miscibility. But after crosslinking, the blend was partially miscible. FTIR studies revealed
that intermolecular hydrogen bonding interaction between the components is an
important driving force to the miscibility of PCL/UPE blends and the partial miscibility
of crosslinked PCL/UPE blends. The spherulitic morphology of the blends was
remarkably affected by crosslinking.

Guo and Zheng [125] also studied the crystallization kinetics of thermosetting polymer
blends of poly(e-caprolactone) and unsaturated polyester resin. The overall crystallization
rate of PCL decreases with the addition of amorphous component, UPE. The kinetic rate

constant decreases rapidly for both the PCL/UPE blends and the crosslinked PCL/UPE

58

blends with decreasing PCL concentration. However, neither of these papers presented a

study on effect of PCL addition on volume shrinkage of the unsaturated polyester resin.

2.2.4.6 Studies on low proﬁle additives for UPE

In low temperature molding processes, control of resin shrinkage and residual monomer
is an important concern. The presence of low proﬁle additives (LPAs) can reduce the
shrinkage of unsaturated polyester (UPE)/styrene resins under proper processing
conditions but may increase the residual styrene content. It is believed that the reaction-
induced phase separation and the polymerization shrinkage in both the LPA-rich and UP-
rich phases result in the formation of microvoids, which partially compensates the resin
shrinkage. The relative reaction rate in the two phases plays an important role in
shrinkage control.

Han and Lem [126] measured the bulk rheological properties of polyester resin
formulation containing UPE/ﬁllers and UPE/ﬁllers/LPAs using cone-and-plate rheometer.
The ﬁllers used were clay, calcium carbonate, and milled glass ﬁber. The low-proﬁle
additives, poly(vinyl acetate) (PVAc) and poly(methyl methacrylate) (PMMA), were
dissolved in styrene. The bulk viscosities of all blends of polyester resin and PVAc lie
between those of the individual components, whereas the bulk viscosities of some blends
of polyester resin and PMMA go through a minimum and a maximum, depending on the
composition of the mixture. On adding both the ﬁller and low-proﬁle additive together in
polyester resin, the rheological behavior became quite complex, indicating presence of

interactions between the ﬁller and the low-proﬁle additive.

59

Han and Lem [127] investigation was made of the rheological behavior of unsaturated
polyester resin during thickening in the presence of ﬁller or low-proﬁle additive alone
and, also, in the presence of both ﬁller and low-proﬁle additive. They used CaCO3 and
clay as ﬁllers and poly(methyl methacrylate) (PMMA) and poly(vinyl acetate) (PVAc) as
the low-proﬁle additives. The viscosity behavior of the resin/ﬁller/LPA system was
similar to that of the resin/ LPA. From the rheological standpoint, the different types of
low-proﬁle additive had little inﬂuence on the thickening behavior of the respective ﬁlled
systems.

Rufﬂer et al. [128] characterized the ﬁssures occurring during curing of blends of UPE
with PVAc using fractal analysis. The LPA effect arises by ﬁssuring with fractal
geometry. The fractal dimensions of ﬁssures depend on the PVAc concentration and the
cure temperature.

Huang and Su [129] studied the effects of two LPAs, PVAc and PMMA on the curing
kinetics during the cure of UPE resins by use of DSC and FTIR. There is a kinetically
controlled plateau (small shoulder) in the initial portion of the DSC rate proﬁle
depending on the molar ratio of styrene to polyester C=C bonds, in the sample containing
LPA. Adding LPA enhanced the relative conversion of polyester C=C bonds. to styrene
throughout the reaction, as shown by F TIR results. A microgel-based kinetic model was
developed to explain the effects of LPAs on reaction kinetics (intramicrogel and
intermicrogel crosslinking reactions), relative conversion of styrene and polyester C=C
bonds, and the ﬁnal conversion. The LPAs induced a globule microstructure during the

curing, which enhanced the ﬁnal conversion of polyester, but would result in an increase

60

or a decrease of the ﬁnal conversion of styrene, depending on the initial molar ratio of
styrene to polyester C=C bonds.

Huang and Su [130] investigated the effects of two LPAs, PVAc and PMMA, on the
morphological changes during the cure of UPE resins by using DSC, SEM, optical
microsc0py, and low-angle laser light-scattering (LALLS). Under SEM, the microvoids
and microcracks which are responsible for the volume shrinkage control were observed at
the later stage of reaction. The morphological changes during curing varied considerably,
depending on the types of LPA and the initial molar ratios of styrene to polyester C=C
bonds. Adding LPA in the neat UP resin reaction system enhanced the clear identiﬁcation
of microgel particles, because a layer of LPA could cover the surface of microgel
particles when the microgel particle phase separates from the matrix of unreacted resin
containing LPA. The segregating effect of LPA led to reduced merging of microgel
particles and helped to retain the identity of the individual microgel particle.

Sun and Yu [131] used gel permeation chromatography (GPC) and DSC to study the
effect of LPAs, PVAc and PMMA, on the curing reaction of UPE resins. According to
DSC results the curing reaction rate decreased as the concentration of LPA increased.
The concentration of LPA and the compatibility of LPA with UPE resins had a strong
inﬂuence on the polyester microgel formation and the curing behavior. There was
evidence of the shrinkage of UPE microgels at the early stage of curing reaction
witnessed from GPC results. The formation of microgel particles has been found to be the
key feature of the UPE-styrene copolymerization. For UPE resins mixed with LPA, the
curing reaction rate of UPE with styrene decreased as the concentration of LPA increased.

Increase of the LPA concentration facilitates the UPEs to form coils in the styrene

61

monomer and to undergo an intramolecular crosslinking reaction, which causes a delay of
gelation.

Huang and Liang [132] compared the volume shrinkage abatement caused by four LPAs,
including PVAc, PMMA, thermoplastic polyurethane (PU), and polystyrene (PS), in
polyester resin. The effectiveness of volume shrinkage agent decreased in the following
manner: PVAc > PMMA > PS. The shrinkage effect with PU depended on its
concentration. In case of relatively polar LPAs like PVAc and PMMA, the fractional
volume change after curing is controlled by microvoid formation. For nonpolar LPAs like
PS, it is controlled by intrinsic polymerization shrinkage. For polar LPAs which can react
with UPE, both factors play a role.

Siato et al. [133] studied the rheological and morphological changes in polyester resin
after incorporation of methyl diphenyl isocyanate (MDI) and magnesium oxide (MgO) as
thickeners into a UPE-LPA system. The MDI based system exhibited rapid viscosity
build up and followed by stable viscosity for long storage times, and the shrinkage
control by this system of poor due to lack of phase separation. For MgO based systems,
the viscosity increased slowly and it changed during the long storage time. There was
phase separation during curing which led to a better shrinkage control in MgO based
system.

Li and Lee [134] found out the shrinkage mechanism of UPE/LPA systems at low
temperature cure using rheometry, kinetics, morphology and dilatometry study. The
competition of the shrinkage induced by resin polymerization and the expansion induced

by microvoid formation, strongly determine the shrinkage behavior of the UPE resin. At

62

low temperature cure, the systems with higher molecular weight and lower LPA content
are more effective shrinkage controllers.

Huang and Jiang [135] explored the effects of chemical structure and composition of
UPE on the miscibility of UPE/styrene/LPA system having PU, PVAc, PMMA as LPAs.
The dipole moments of UPE and LPAs were calculated, and it was found that the
increasing order of polarity in terms of dipole moment per unit volrune was
PMMA<PVAc<UPE<PU. If the molecular weight of UPE is increased, the phase
separation rate for systems with PVAc and PMMA would decrease. The miscibility of
UPE/styrene/LPA systems and samples morphologies were correlated with polarity
differences per unit volume between UPEs and LPAs.

Huang and Chen [136] investigated the effects of three different UPE resins with
different chemicals composition on the mechanical properties of the UPE/LPA systems
with PU, PVAc, and PMMA. The UPE resin series were made of different anhydrides,
acids, alcohols in different molar ratios. The authors discussed the effects of glycol ratios,
modiﬁcation of dibasic acids, C=C unsaturation, and molecular weight on the mechanical
properties of the polymers formed.

Hsieh and Yu [137] explored the phase separation occurring during curing of UPE resin
with PVAc by low-angle laser light scattering (LALS) and scanning electron microscopy
(SEM). It was seen that the PVAc rich phase was regularly dispersed in the UPE matrix.
But as the PVAc content was increased (greater thanlO wt %), a co-continuous
morphology of PVAc and UPE was seen. According to LALS data indicated a regularly
discrete phase dispersed in a continuous matrix phase, as observed from scattering

intensity proﬁles with maximum peak intensities.

63

Huang et al. [138] studied the effects of chemical structure and molecular weight of three
series of thermoplastic polyurethane (PU) based LPA on the miscibility of UPE
resin/styrene/LPA ternary systems prior to reaction by using the DSC, SEM, and
correlated them with F lory-Huggins theory and group contribution methods. From
morphological data and DSC reaction rate proﬁle, the phase-separation characteristics of
UPE resin/styrene/LPA systems during the cure could be predicted by the calculated
upper critical solution temperature for the uncured UPE/styrene/LPA systems. Moderate
phase separation between the LPA and crosslinked UPE phases during the cure would be
indispensable for the microcrack and microvoid formation, while either too little or too
much phase separation would be unfavorable.

Zhang and Zhu [139] investigated the characteristics and mechanisms involved in the
microvoid formation of a low shrinkage UPE resin with PVAc. PVAc and styrene were
found to play vital roles in controlling the microvoid formation. The voids changed from
isolated type to interconnected type by increasing PVAc content, and their sizes were
signiﬁcantly enlarged by increasing styrene content. The microvoid formation was also
affected by the curing conditions and compositions. It was enhanced by high curing
temperature, but substantially restrained by an applied external pressure. When a ﬁller
(CaCO3) was introduced to the PVAc-containing resin or when the resin was cured with
t-butyl perbenzoate (TBPB), the number of voids increased. They suggested that the
microvoid formation did not arise from the strain-relief cracking but from the nucleation
and expansion of surplus styrene in the PVAc rich phase.

Cao and Lee [140] examined the effect of secondary monomers, divinylbenzene (DVB)

and trimethylopropane trimethacrylate (TMPTMA), and a co-promoter, 2,4-pentandione

64

(2,4-P), on the shrinkage control of UPE/styrene/ LPA cured at low temperatures.
Addition of both TMPTMA and 2,4-P results in an earlier formation of microvoids,
earlier volume expansion and better shrinkage control. a better shrinkage control can be
obtained when the volume expansion occurs at a lower conversion. There is no further
improvement ﬁnal volume shrinkage after introduction of TMPTMA and 2,4-P into UPE
formulation. The morphology of the samples with TMPTMA is quite different from that
without TMPTMA, therefore, TMPTMA affects both crosslinking and phase-separation
processes in the curing system. The samples with 2,4-P had no morphology changes.

Cao and Lee [141] carried out a systematic study to investigate the effect of the initiator
system and reaction temperature on sample morphology, ﬁnal resin conversion, and resin
shrinkage of UPE resins with LPA. It was found that the ﬁnal conversion of the resin
system could be improved by using dual initiators. The effect is more obvious at low
temperatures. Volume shrinkage measurements of the resin system initiated with dual
initiators revealed that good LPA performance was achieved at low (35 0C) and high (100
0C) temperatures but not at intermediate ones.

Cao and Lee [142] investigated the effect of a comonomer, methyl methacrylate (MMA),
on volume shrinkage and residual styrene content of UPE resin with low proﬁle additives
cured at low temperature by an integrated reaction kinetics-morphology property analysis.
MMA affects the volume shrinkage and residual styrene content differently depending on
MMA to styrene C=C bond molar ratio. At low MMA/styrene ratio, residual styrene
decreases and the volume shrinkage of the resin system remains unchanged, while at high
MMA/styrene ratio, residual styrene can be substantially reduced, but the resin system

suffers poor volume shrinkage control. MMA has a negative effect on shrinkage control

65

of UEP/styrene/LPA system because the system becomes more compatible with the
addition of MMA. Therefore, phase separation becomes more difﬁcult, leading to poor
shrinkage control. The surface quality and the ﬂexural properties of UPE depend strongly
on the volume shrinkage and ﬁnal conversion of the resin system.

Boyard et al. [143] studied the dynamic curing of a polymer blend containing unsaturated
polyester, styrene, and a saturated polyester based on adipic acid and propylene glycol as
a LPA using a homemade dilatometer that simultaneously determined conversion degree,
volume, pressure, and temperature. They also developed a model based on physical
considerations which ﬁtted well with the experimental results.

Huang et al. [144] investigated the effect of three series of LPAs on the volume shrinkage
and internal pigrnentability of the unsaturated polyester resin. Differential scanning
calorimetry, seaming electron microscopy, optical microscopy, and image analysis were
used to measure static phase characteristics of the UPE/styrene/LPA system, including
the reaction kinetics, sample morphology, and microvoid formation. Volume shrinkage
control and the internal pigrnentability are inﬂuenced by microvoid formation during the
cure of UPE/ styrene/ LPA system. A higher volume fraction of microvoids can generally
lead to better volume shrinkage control but more haziness of pigment color. A co-
continuous globule cured sample morphology leads to the generation of sufﬁcient
interfacial area between the densely crosslinked polyester phase and the LPA-rich phase
during the cure, lower stifﬁress of the UP resin, and lower Tg for the LPA in reference to
the cure temperature, which results in a higher volume fraction of microvoids and

microcracks generated during curing.

66

Huang et al. [145] examined the effect of three series of LPAs on the Tg and mechanical
properties of the crosslinked polyester resin. The LPAs were synthesized by the authors,
and included poly(vinyl acetate) (PVAc), poly(vinyl chloride-co-vinyl acetate), and
poly(vinyl chloride-co-vinyl acetate-co-maleic anhydride), with different chemical
structures and molecular weights. The mechanical properties depended on the interfacial
adhesion between the LPA-rich and major St-crosslinked polyester phases, molar ratio of
styrene/polyester C=C bonds, sample morphology, and the volume fraction in the major
continuous phase of styrene-crosslinked polyester as a result of phase separation during
cure. The Tg in the major continuous styrene-crosslinked polyester phase depends on the
relative importance of the two opposing effects, namely, the plasticization effect of the
LPA phase on the continuous phase and the crosslinking effect of the continuous phase
itself in terms of the molar ratio of styrene/polyester C=C bonds.

Wang et al. [146] examined the effect of low proﬁle additives (LPA) on the shrinkage
behavior of UPE during curing at low to medium temperatures. The shrinkage of UPE is
related not only to the temperature, but also LPA type, molecular weight of LPA,
concentration of LPA, and also the competition between the shrinkage induced by resin
polymerization and the volume compensation by LPA. Poly(vinyl acetate) (PVAc),
polystyrene and poly(methyl methacrylate) (PMMA) were used as the LPA. Volume
shrinkage was measured in an in-house dilatometer. Low concentrations of PVAc with
high molecular weight or high concentrations of PVAc at low molecular weight provide
good shrinkage control.

Xue et al. [147] developed a simulation for shrinkage mechanism in UPE-PVAc system

after characterizing the system using DSC and SEM. There are ﬁve steps in the curing of

67

UPE-PVAc system. Firstly, a homogeneous mixture of UPE, styrene and PVAc is formed
at room temperature. In the second step, reaction starts near the initiator to form
microgels. The segments of PVAc drift from microgel core to the surface because of the
decrease in the compatibility between the components. In the thirds step, microgelation
takes place. In the fourth and most important step, microvoids are formed in the LPA
phase to compensate for the shrinkage stresses. Finally, on cooling the system down,
many more microvoids appear in the LPA phase and in the interfaces between UPE and
LPA. After all these steps, the low proﬁle phase separation structure of the PVAc systems

is formed.

2.2.4.7 Studies on interpenetrating networks with UPE

An interpenetrating networks (IPN) material is a polymer network intermeshed with
another polymer network. An IPN is any material containing two polymers, each in
network form. Most of the IPNs do not form a network at molecular scale, but may form
ﬁnely divided phases of sizes in nanometers. In most IPNs polymers form phases that are
continuous on macroscopic scale. The networks either have to be made at the same time
or the second component must be polymerized inside of the ﬁrst component after its
crosslinked [148, 149]. In Figure 2.2.9, one network is represented by solid lines, while
the other network is shown by dotted lines. The circles represent points at which the

chains are connected (crosslinked)

68

,i \
Figure 2.2.9: A depiction of a simple IPN material.
IPNS are composite materials in a special way. A very general, simplistic deﬁnition of an
IPN is a material with two co-continuous phases or a material with 2 parts intertwined
throughout. The main goal of an IPN is to trap 2 polymer phases in each other thus
causing a possible synergistic behavior or an addition of properties as opposed to a
compromise of properties is the scenario which typically results in most composites [150].
Thus, there are two networks co-existing within one another.
Gunanathan et al. [151] formulated a tri-component interpenetrating network containing
castor oil, toulenediisocyanate, acrylonitrile, ethylene glycol diacrylate, and UPE with
different compositions for the toughening of UPE. The polyurethane
(PU)/polyacrylonitrile (PAN)/UPE network undergoes single decomposition showing
ideal compatibility at the composition of 10:90 (PU/PAN: UPE). The impact strengths of
the IPNS were higher than control UPE.
Chou and Lee [152] studied the IPNS based on polyurethanes and three different
polyester resins. The interaction between the two phases was studied by reaction kinetics

and rheological changes. The reaction between the end groups of UPE molecules and

isocyanate led to more reactive vinyl polymers. The presence of carboxylic groups on the

69

UPE chain increased the rate of formation of the polyurethane network. The extent of
chemical grafting governed the morphologies of the IPNS formed.

Sperling [153] reviewed the different IPNS, methods of their synthesis, applications to the
plastics and the industry, and history of IPNS.

Frischinger and Dirlikov [154] formed two phase interpenetrating epoxy networks with
homogeneous formulations of diglycidyl ether of bisphenol-A, epoxidized soybean oil,
and vemonia oil. At low oil contents, the DGEBA matrix is rigid and rubbery particles of
oils are randomly distributed, and at higher oil contents, the rubbery phase is continuous
and the DGEBA is rigid small particles. Phase inversion and particle size depends on the
miscibility of the two phases.

Frischinger et al. [155] discussed the commercial applications of interpenetrating epoxy
networks made with DGEBA and epoxidized vegetable oils. These IPNS can be used for
toughening of epoxy resin and also for preparation of crack resistant coatings. The
systems with a continuous vegetable oil rubbery phase and small DGEBA rigid particles

are especially suited for stress-resistant and crack-resistant coatings.

2.3 Bioresins

There is a growing urgency to develop novel bio-based products (based on renewable
resources) and associated technologies that reduce widespread dependence on fossil fuels.
Bio-based products can form the basis for an assortment of green materials that can
compete with and take over markets currently dominated by products exclusively from
petroleum feed stocks. Interest in the use of green materials has also grown due to their

various advantages, like ready availability and low cost. Currently, various renewable

70

resources are being used to prepare products like starches, proteins, fats, and oils for a
range of applications in building materials, structural and automotive parts, absorbents,
coatings, inks, plasticizers, lubricants, adhesives, bonding agents and agrochemicals
[156-161].

Vegetable oils such as soybean, cottonseed, linseed, castor, are available on a global basis
in large quantities at affordable costs and are possibly biodegradable. These naturally
hydrophobic plant oils are used as starting materials for a range of resin products for
example, resin precursors, produced by the transformation of vegetable oils that will
polymerize when heated in the presence of catalysts. This technology uses the reactivity
of the monomers to attack the double bonds present in the oil, forming linkages that can
subsequently be reacted into crosslinkable units. Other reactive chemicals can also be
incorporated into the formulation to facilitate curing [156].

Because of different kinds of fatty acids present in oils, on a molecular level they are
composed of many different types of triglycerides with varying levels of unsaturation
[157]. Recently, genetic engineering techniques have helped to regulate the variation of
unsaturation in plant oils such as soybean, ﬂax, and com [158]. These oils have also been
used as toughening agents [156, 159, 160]. There also has been an extensive amount of
work in their use to produce interpenetrating networks (IPNS) [161].

Vegetable oils are an abundant natural resource that has yet to be fully exploited as a
source for polymers and composites. This area of research sets a foundation from which
completely new materials can be produced with novel properties. Research continues to

try and optimize the properties of these materials and understand the fundamental issues

71

that affect them. In this manner, more renewable resources can be utilized to meet the

material demands of many industries.

2.3.1 Soybean oil

Soybean oil is a complex mixture of ﬁve fatty acids (palmitic, stearic, oleic, linoleic, and
linolenic acids) that have vastly differing melting points, oxidative stabilities, and
chemical ﬁlnctionalities [157]. Soybean oil is the world's most widely used edible oil. In

the US, more than 13 billion pounds of soybean oil are used annually in foods and food

 

Figure 2.2.10: Soybean plant

processing. Also, more than 600 million pounds of the soybean oil produced annually in
the US is used for non-edible applications, including the production of industrial
materials [162]. More than half of this soybean oil use falls in the category of fatty acids,
soaps, and feed. The remainder of the non-edible soybean market is in the manufacturing
of inks, paints, varnishes, resins, adhesives, coating, lubricants, plasticizers, plastics,
biodiesel, and specialty products. There are ﬁve primary target areas that hold great
potential for soy-based products: adhesives, coatings and printing inks, lubricants,

plastics, and specialty products.

72

2.3.2 Linseed oil (from Flax seeds)

Linseed oil is obtained from seed of the ﬂax plant. It is primarily used in industry; but

" .. ‘. "-h.
. - v. . , d.'w
Figure 2.2.10: Flax plant and ﬂower

    

some is also used for edible purposes. Flax seed is grown for oil in the US. with
production of 200 million pounds, in 2002 [163]. The global production of ﬂaxseed is
about 2750 million pounds per year; of which over 40 percent is grown in Canada.
Britain and France are the other major producers of linseed. In the US, North Dakota is
the biggest producer of linseed. Linseed oil has long been used in making linoleum,
oilcloth, certain inks, varnishes, paints, drying oils, printing inks, soaps and for some
ﬁnishing processes in the manufacturing of leather [164]. The crude linseed oil is an

organic mixture of linolenic acid, linoleic acid and oleic acid.

2.3.3 Modiﬁcations of vegetable oils

In recent times, there have been a lot of research and publication on the use of soybean
oils as epoxy, polyol, polyurethane, and polyester resins. The triglyceride molecules
contain active sites, which are prone to chemical reaction; these sites are the double bond,
the allylic carbons, the ester group, and the carbons alpha to the ester group. These active
sites can be used to introduce polymerizable groups on the triglyceride using synthetic

techniques that have been applied for the petroleum—based polymers. A higher molecular

73

weight, increased crosslink density, and incorporation of chemical functionalities that
impart stiffness, can be accomplished by several synthetic pathways [165-174].

In recent years, extensive work has been done to develop polymers for engineering
applications using vegetable oils as the main component. The following is the description
of the most notable work that has been done in this area:

Mulhaupt and coworkers [165-166] developed polyester networks based upon epoxidized
and maleinated natural oils

Wool et al. have synthesized liquid thermosetting bioresins from soybean oil which can
be reinforced with glass, carbon or kevlar ﬁbers, and can be processed by methods such
as hand lay-up, resin transfer, spray-up, ﬁlament winding, and pultrusion. These resins
have the same mechanical properties as unsaturated polyester or vinyl ester resins. They
have also synthesized and polymerized maleate half esters of fatty acid monoglycerides
derived from soybean oil. [167].

Can et al. [168-169] prepared rigid thermoset polymers from the radical
copolymerization of the soybean oil monoglyceride maleates (SOMG) with styrene.
SOMG were produced from the reaction of soybean oil with glycerol with an
optimization of the reaction to maximize the monoglyceride yield, following by reaction
with maleic anhydride to produce the SOMG maleate half esters. In another study, the
same authors reacted the copolyrners of the SOMG maleates with styrene to produce
rigid thermosetting polymers. They used mixtures of SOMG with neopentyl glycol
(NPG) and SOMG with bisphenol A (EPA), and maleinized them under the same

reaction conditions, and copolyrnerized the resulting maleates with styrene.

74

Li and Larock [170-173] were successful in direct cationic polymerization and
copolymerization of the C=C bonds of various soybean oils and several other biological
oils. They used regular soybean oil, low-saturation soybean oil (LSS) and conjugated
LSS (CLS), to develop a series of polymers ranging from rubbery to rigid by cationic
polymerization with styrene and divinyl benzene. The therrnophysical and mechanical
properties of the resulting materials were also determined. The biological oils that they
studied were tung oil, Norway ﬁsh oil ethyl esters, conjugated Norway ﬁsh oil, and
triglyceride ﬁsh oil.

Dweib et al. [174] manufactured structural panels and unit beams from soybean oil based
resin and natural ﬁbers (ﬂax, cellulose, pulp, recycled paper, chicken feathers) using
vacuum assisted resin transfer molding (V ARTM) technology.

Soucek et al. [175] synthesized norbomyl epoxidized linseed oil via Diels—Alder reaction
of cyclopentadiene with linseed oil at high pressure and high temperature, followed by an
epoxidation using hydrogen peroxide with a quaternary ammonium catalyst. In another
study, Soucek et al. [176] prepared epoxy norbomane linseed oils (ENLOs) as a ﬁmction
of the norbomene content, and carried out cationic photopolymerization of ENLOs.
Latere et al. [177-180] made polyurethanes from soybean oil derived polyols and
synthetic isocyanate, and made composites with glass ﬁbers, hemp ﬁbers and kenaf ﬁbers.
Petrovic and coworkers [181-186] have developed various polyols from different
vegetable oils. They were also successful in making polyurethanes from soybean derived
polyols. Uyama et al. [187] synthesized green nanocomposites consisting of epoxidized

soybean oil and nano clay (Figure 2.2.11).

75

The manufacturing and characterization of pultruded composites from glass ﬁbers and
epoxidized allyl soyate and epoxidized soybean oil was done by researchers from the
University of Missouri—Rolla [188].

Eren and Kusefoglu [189] have simultaneously added bromine and acrylate groups to the
double bonds of fatty acids in triglycerides soybean oil, sunﬂower oil and methyl oleate.
Eren et al. [190] also reacted soybean oil triglycerides with maleic anhydride to introduce
more than two maleate residues per triglycerides. The maleinized soybean oil (SOMA)
was then polymerized with diols to half-esters.

Neckers et al. have researched UV curing of soybean oil or epoxy soybean oil for use in
coatings. With their experiments, they developed a new series of photoinitiators that
allow the curing of epoxy soybean oil as a coating [191].

Miyagawa et al. [192-199] have developed bio-based epoxy materials containing
functionalized vegetable oils (FVO), such as epoxidized linseed oil (ELO), epoxidized
soybean oil (ESO), acrylated epoxidized soybean oil (AESO), and, octyl epoxy linseedate
(OEL) by processing with an anhydride or amine curing agent. They have also used bio-
epoxies to make nanocomposites with nano ﬁllers including clay, silica, alumina, vapor
grown carbon ﬁbers, and carbon nanotubes.

In another study, people at composite materials and structures center (CMSC) of
Michigan State University designed and engineered biocomposites from plant bioﬁbers
and a blend of unsaturated polyester resin and derivitized vegetable oil to replace existing

glass ﬁber-polyester composites for use in housing applications [156, 200].

76

 

Figure 2.2.11: Photograph of ESO—clay nanocomposite showing high ﬂexibility [187]
2.3.4 Modified matrix

In this study, we propose to make bio-based resins by the modiﬁcation of vegetable oils,
such as soybean oil, linseed oil, etc., and polymerize them with UPE. The development of
such bioresins would ensure reduced dependence on petroleum products for a variety of
industrial applications.

The vegetable oils will be modiﬁed according to the functionality desired. New methods
for adding one or more functional groups like acrylates, epoxies, ethers, hydroxyl and
esters to the natural oils will be developed. The modiﬁed vegetable oils (bioresins) will
then be crosslinked by the free radical polymerization with UPE. The amount of initiator
will be optimized by the evaluation of mechanical and thermal properties of the resulting
polymerized bioresin. A unique initiator system will be designed for each type of
modiﬁed oil.

The bioresins hence developed, will also be used for the fabrication of composites using
an optimum amount of reinforcement. The ﬁbers can be. The resulting composite will

thus be more than 50 % bio-based.

2.4 Processing of biocomposites

77

The goal of modern composite material processing is to obtain high-quality structural
components at lower cost. To this end, biocomposites are processed by injection molding,
extrusion, compression molding (CM), pultrusion, resin transfer molding (RTM), resin
injection molding (RIM), vacuum assisted resin transfer molding (VARTM), bulk
molding compound (BMC), hand lay-up, and (now) sheet molding compounding (SMC).

Discussed below are some of the processing techniques pertinent to thermoset composites.

Compression molding

Compression molding (CM) is a high-volume, hi gh-pressure method suitable for molding
complex, high-strength ﬁberglass reinforcements. Compression molding is a processing
technique in which the molding material, generally preheated, is ﬁrst placed in an open,
heated mold cavity. The mold is closed with a top force or plug member, pressure is
applied to force the material into contact with all mold areas, and heat and pressure are
maintained until the molding material has cured. Compression molding was speciﬁcally
developed for replacement of metal components with composite parts.

The molding process can be canied out with either thermosets or thermoplastics.
Compression molding is the most common method of processing thermosets. Composite
can also be compression molded with unidirectional tapes, woven fabrics, randomly
orientated ﬁber mat or chopped ﬁbers. The advantages of compression molding are the
ability to mold large, fairly intricate parts, short cycle time (typically 1-6 minutes), high
volume production, high quality surfaces [201]. Compression molding produces fewer

knit lines and less ﬁber-length degradation than injection molding. However, it involves

78

high initial capital investment, is labor intensive, and sometimes requires secondary
operations. It is very popular with automotive industry, and products like front and rear
end automobile panels, hoods, roofs, scoops, fenders, spoilers, air deﬂectors, and lift

gates are made using CM.

Resin Transfer Molding (R TM)

Resin Transfer Molding or RTM literally means the transfer of the resin mix from one
system (the RTM machine) to another system (the closed mold containing the
reinforcement) while molding a product. It bridges the gap between labor intensive hand
lay-up process and capital intensive compression molding process.

In a traditional RTM process; a dry, ﬁbrous preform is loaded into a matched-metal mold.
The mold is then closed, which compacts the preform, and a thermosetting resin is
injected into the preform. The resin and ﬁber remain in the mold until crosslinking occurs,
after which the composite can be removed. The material can be cured at room
temperature or in heated mold by pr0per choice of initiator.

There are several advantages in using RTM over more traditional composite processing
techniques. Some of these advantages are: very large and complex shapes can be made
efﬁciently and inexpensively, elimination of the labor and skill required in hand lay-ups,
better reproducibility of parts due to the high dimensional accuracy of the mold,
reduction in the exposure to volatile organic compounds (VOC). Along with the

advantages, there are the inevitable disadvantages: the mold design is critical, and it can

79

become quite expensive, resin rich areas often form in the edges and radii, reinforcements
that move during injection can alter the quality of the ﬁnished part.

Applications of RTM are in consumer goods such as helmet, bathroom ﬁxtures, boats, car
body panels, instrument panels, truck caps, building panels etc. Factors affecting RTM
processing are: resin characteristics, reinforcement characteristics, resin injection
pressure, design of the mold, mold temperature, vacuum state of the system, ﬂow rate of
resin mix, and mold ﬁlling time [202].

Some of the advantages of RTM process include: relatively low tooling cost, design
ﬂexibility, faster productivity, labor savings, energy savings, lower emissions of volatiles,
no air entrapment, smooth surface ﬁnishing on both sides, both sides can be gel-coated,
close dimensional tolerances maintained, a more consistent product, thus reducing bad
product thrown away, and more parts for the money.

O’Dell [203] made composite of jute mats and unsaturated polyester resin through resin
transfer molding (RTM). The author also evaluated the mechanical properties and
resistance of the composites to wood-degrading fungi and simulated weathering. The
suitability of resin transfer molding equipment with regard to resin ﬂow, air entrainment,
and surface properties for jute-UPE composites were determined. There were no
problems in processing, thus, jute ﬁbers could be processed just as well as glass ﬁbers in
resin transfer molding. The samples exposed in a accelerated weatherometer for 1,200
hours showed minimal surface erosion or color change, the samples also showed
negligible weight loss when exposed to wood degrading fungi.

Rowell et al. [204] manufactured composite samples with jute and UPE by RTM, jute

and epoxy resin by RTM, and jute-polyurethane by structural reaction injection molding

80

(SRIM). The samples were subjected to mechanical testing as well as weather resistance
analysis. The jute ﬁber mats used in the RTM as well as SRIM system wetted well with
the respective resins. On comparing glass ﬁber and jute ﬁber based composites, the
surface ﬁnish of both composites were nearly equal. The samples exposed to accelerated

weathering for 600 hours showed surface whitening and nearly identical ﬁber exposure.

VAR TM

The VARTM process has its roots in early composite manufacturing. However, modern
VARTM is traced to the patenting of SCRIMP — Seemann’s Composite Resin Infusion
Molding Process. This method employed the use of a vacuum bag, resin supply tubing, a
resin conduit and a distribution medium. RTM has lead to the evolution of Vacuum-
Assisted Resin Transfer Molding (V ARTM) in order to overcome some of its inherent
ﬂaws. In the VARTM process, a vacuum bag is used as one of the tool surfaces. This
eliminates the need for precise matched-metal mold making that is crucial in
conventional RTM. The exact ﬁt of the vacuum bag to the preform drastically reduces
resin rich areas, and it also allows for a more efﬁcient way of trapping and controlling
VOC emissions that are being more strictly regulated. A ﬁnal advantage of VARTM lies
in the low injection pressures (~1 atrn). These low pressures allow for little movement of
the reinforcement during processing, which produces a better quality part [205].

This lay-up begins with a dry, ﬁbrous preform laid onto an open-faced tool plate. Then, a
porous peel-ply is placed on top of the preform, but underneath a resin distribution tube.

Next, comes the resin distribution medium. This material, which is also known as the

81

high-permeable distribution medium, covers a large portion of the preform and peel-ply.
Even though VARTM processes do not technically require these high-permeable
distribution media, they can be helpful in signiﬁcantly reducing processing times by
providing a low resistance resin pathway. The ﬁnal step in the lay-up is the placement of
a vacuum bag over the entire assembly.

Once the lay-up is complete, the processing can be carried out. First, a vacuum pump is
turned on to expel any air from the preform assembly. After the system has equilibrated,
the resin is allowed to ﬂow through the resin distribution tube and across the high-
perrneable distribution medium. The resin then ﬁlls the preform in the transverse, or
through the thickness, direction by leaking downward from the high-permeable
distribution medium. The vacuum source is left on until the resin system begins to gel.
Once the vacuum is turned off, the part may either cure at room temperature, or it can be
placed in an oven to assist the curing process.

The VARTM process is currently implemented in the manufacturing of Naval structural
components (masts, hulls, and bridge decks), automobile chassis and body components,
transportation inﬁastructure, railroad cars, and turbine blades A great deal of work still
needs to be done on efforts to reduce the costly trial-and error methods of VARTM
processing that are currently in practice today. An understanding of the flow behavior of
the system prior to production is crucial. For rigid monolithic structures this provides
several advantages: reduced labor, good surface mold side, and improved ﬁber volume

fraction.

82

Shenton et al. [206] discussed a collaborative effort to manufacture a roof made from
completely natural composite material for residential construction. The raw materials
used for the composite were soy bean based resins and natural ﬁbers.

Dweib et al. [207-209] manufactured structural panels and unit beams from soybean oil
based resin (Acrylated, Epoxidized Soybean Oil, (AESO)), natural ﬁbers (ﬂax, cellulose,
pulp, hemp, old newspaper, recycled paper, chicken feathers), and woven glass ﬁber and
closed cell structural foam, using vacuum assisted resin transfer molding (VARTM)
technology. Structural parts incorporating a foam core were also manufactured and tested
to give the required strength for roof applications. Stay-in-place bridge decking form was
successfully manufactured and tested to replace the current in use metal form. Same
material can be used as a good insulation material in housing and reﬁigerating trucks for
transportation.

Hong and Wool [210] have developed a low dielectric constant material from chicken
feathers and acrylated epoxidized soybean oils by a vacuum assisted resin transfer
molding (VARTM) process. A natural, bio-based and environmentally ﬁiendly material
from hollow keratin ﬁbers and chemically modiﬁed soybean oil was thus obtained. The
dielectric constant values were lower than that of a conventional semiconductor insulator
material such as silicon dioxide, epoxies, polyimides and other dielectric materials.
Quagliata et al. [211] improved the material stiffness shortcomings of biocomposites
through efﬁcient structural conﬁgurations. They used the popular designs for advanced
composites based on this idea, namely cellular and sandwich structures which
respectively maximize material and structural performance. They investigated the

feasibility of using biocomposite materials for primary load-bearing components in civil

83

structures by utilizing high-performance structural forms. Their research work focused
on the investigation of this concept through the development, manufacturing, and
characterization, both experimental and analytical, of laboratory-scale cellular beams and
plates. These studies were used to analytically investigate the feasibility of using
biocomposite materials for full-size structural components for civil structures, and assess

the future research needs for the development of load-bearing biocomposite structures.

SMC

Biocomposites have been traditionally manufactured with techniques like extrusion,
injection molding, compression molding, resin transfer molding, and pultrusion, etc. But
they have never been produced using an SMC line. SMC is a very useful processing
technique usually used for fabricating glass-polyester resin composites. Today more and
more parts and products from cars to skis, hockey sticks to microwaves are being molded
from SMC. This is not too surprising since thermoset composite plastics have been used
for more than sixty years around the world. A recent US. report states that the demand
for thermoset composites in the automotive industry will rise by 68 per cent to 467
million pounds by the middle of this decade. The reason is easy. SMC combines high
strength and light weight. Other advantages of thermoset composites are durability and
resistance to corrosion in a wide range of temperature environments. The application of a
priming and topcoat will allow SMC components to meet automotive ’class A’ surface
ﬁnish requirements.

Ever since 1960’s there have been numerous patents on SMCs. US 3615979 granted in

1971 was a patent for glass ﬁber-reinforced sheet molding compound, by Owens-Coming

84

Fiberglas Corp [212]. US 3713927 granted in 1973, thermosetting sheet molding
compounds [213]. US 3835212 granted in 1974 was a patent for resinous sheet like
products [214]. DE 2357000 granted in 1974 was a patent for sheet molding compounds
[215]. JP 49076988 granted in 1974 was a patent for unsaturated polyester compositions
for moldings [216]. JP 49107086 granted in 1974 was a patent for unsaturated polyester
compositions [217]. JP 50005463 granted in 1975 was a patent for molding of
unsaturated polyesters [218]. JP 50008882 granted in 1975 was a patent for molding resin
compositions [219]. JP 50014792 granted in 1975 was a patent for sheet molding
compositions [220]. JP 50036588 granted in 1975 was a patent for sheet molding
compositions [221]. US 3931095 granted in 1976 was a patent for ﬁre retardant
thermosetting resin composition [222].

The traditional SMC (Sheet Molding Compound) is a combination of chopped glass
strands and ﬁlled polyester resin, in the form of a sheet. SMC is a fully formulated
system (ﬁber, resin, and additives), which can be molded without additional preparation.
Sheet molding compound can be made with random chopped ﬁber reinforcement, with
continuous ﬁbers, or with a mixture of the two. The variety of ﬁber architectures
available for sheet molding compound means a variety of properties can be achieved in
compression molded parts. SMC prepreg is made from chopped glass strands,
sandwiched between two layers of ﬁlm, onto which the resin paste has already been
applied. The prepreg passes through a compaction system that ensures complete strand
impregnation before being wound into rolls. These are stored for a few days before

molding to allow the prepreg to thicken to a moldable viscosity.

85

SMC is aged to a state near gelation. The manufacturer allows the SMC to reach an
extent of reaction close to gelation because it stabilizes the polymer. This light
crosslinking increases viscosity and improves processability of the SMC. This ready-to-
mold glass ﬁber reinforced polyester material primarily used in compression molding.
SMC is not just ﬁber mixed with resin. The other ﬁllers have an important role. SMC is
made by dispensing mixed resin, ﬁllers, maturation agent, catalyst and mold release agent
onto two moving sheets of carrier ﬁlm. SMC can be molded into complex shapes with
little scrap. It is also called the "raw material" for compression molding. It is suitable for
production runs in excess of 10,000 parts per year.

Processing of SMC by compression molding enables the production of bodywork or
structural automotive components and electrical or electronic machine housings in large
industrial volumes or body panels (hoods and deck lids). The process also penetrates
sectors such as sanitary ware (baths) and urban furniture (stadium and cinema seating),
etc. Its characteristics allow it to ﬁll a mold under the effects of the molding temperature
and pressure.

Sheet Molding Compound (SMC) produces highly cross-linked and highly ﬁlled systems,
in which polymer component is polyester. The molded product combined high modulus

with high strength. This is more expensive than metal, but has lower tooling cost.

There have been many improvements in the SMC processing since 19705, and hence

many more patents. But none of these cater to SMCs with natural ﬁbers and resins.

However, there is one patent and two papers which talk about SMC processing for bio-

86

composites. These include, US 2003088007 granted in 2003 to Wool et al, which is a
patent for sheet molding compound resins ﬁ'om plant oils [223].

van Voom et al. [224] discussed a non automated SMC process for making ﬂax ﬁber
reinforced unsaturated polyester resin composites using sheet molding compounding in
2001. A homogeneous product was obtained after development of ﬁber distributor, use of
compaction rollers, and drying of ﬁbers before compounding. The natural ﬁbers in their
process are not continuously fed to the SMC line. They do so by a batch process. Thus,
their process is not fully automated, and hence cannot be used in an industrial setting.
Goswarni et al. [225] made jute (cloth) reinforced sheets with the help of polyester resin
in “Jute reinforced sheets based on shellac ﬁlled SMC”, in 2003.

So far, no industrial process exists for using sheet molding compound panel processing to
fabricate bio-composites. No other process has been reported in literature for fabrication
of bio-composites via SMC. Our sheet molding process is different from all the above
mentioned ones, because of its uniqueness of ﬁber dispersion, ability to run the process
for any kind of natural ﬁber, automated ﬁber feeding to the production line, and
ﬂexibility with resin formulations.

Kawamura et al. [226] described warpage prediction method for compression molded
SMC products. Warpage relates to material anisotropy caused by ﬁber orientation
distribution. Charge pattern, which is one of molding conditions, has a signiﬁcant effect
on ﬁber orientation distribution. So, mechanical properties such as bending modulus and
the coefﬁcient of thermal expansion were evaluated in two different charge patterns in
order to examine the relationship with warpage. Temperature distribution of each layer

(surface layer, middle layer) in the thickness direction was also measured during curing.

87

The temperature difference between the top and bottom mold causes different heat
generation in each layer. It was found that anisotropy of the coefﬁcient of thermal
expansion caused by ﬁber orientation distribution and the inversion phenomenon of
temperature gradient in the thickness direction during the curing process were dominated
in inducing warpage. Warpage prediction method utilizing ﬁnite element method has
been developed based on these causes of warpage. Predicted warpage coincides well with
warpage of molded products quantitatively. This warpage prediction method is able to
reduce warpage of compression molded SMC products.

Odenberger et al. [227] focused on experimental visualization of the ﬂow during mold
closure in compression molding of sheet molding compound (SMC). Special attention
was given to the advancing ﬂow front, for which the full complexity is captured by
means of continuous high resolution close-up monitoring. From the experimental
visualisation of the ﬂow front, three phases were deﬁned, namely pitch, ﬂoating, and
boiling. A qualitative investigation of the surface quality of the molded plates was also
carried out. It was observed that the surface void is reduced with an increased mold
temperature and that the mould closing speed does not seem to have any inﬂuence on the
surface quality at lower mold temperatures while at higher mold temperatures, an
increased mold closing speed clearly gives a reduction of surface void.

Feuillade et al. [228] studied the inﬂuence of chemical structure of two glass ﬁber sizings
on the impregnation of the ﬁbers by an unsaturated polyester matrix before the curing
reaction. The ﬁbers were chosen to provide a low and a high surface quality respectively
for two ﬁber types for the ﬁnal SMC part. Individual characteristics of each component

and their interactions play important roles on the processing of SMC paste and on ﬁnal

88

product properties, such as surface aspect and uniformity. Two of these characteristics are
ﬁber wetting by the resin and ﬁber/matrix adhesion. Results give evidence of the various
mechanisms occurring at the surface. The main parameters which should inﬂuence ﬁber
impregnation are: the ﬁlm former quantity, the antistatic agent type, the deposition
method and the nature of the ﬁlm former.

Massardier-Nageotte et al. [229] developed a methodology based on visible light
transmission and x-ray imaging, for the characterization of glass ﬁber orientation and
distribution in large components molded using sheet molding compounds. They
observed that for planar components, the ﬁbers were uniformly distributed and

orientation was in the ﬂow direction.

2.5 Weathering

Biocomposites have always been looked at suspicion with regards to their durability. It is
known that the thermal as well as mechanical properties of the wood composites and
natural ﬁber based composites fade with time. Numerous studies in the literature have
talked about the moisture absorption or water uptake of the biocomposites. Furthermore,
even though durability of wood composites has been studied in detail [230-238], not
many people have studied the weathering of biocomposites. To counter questions about
the durability of biocomposites, it is necessary to evaluate the effect of long-term
weathering on biocomposites.

The weathering of biocomposites can be done in an accelerated weathering chamber,

where the samples can be exposed to harsh weather conditions including harmful ultra

89

violet rays, condensation, humidity, elevated temperatures, and rain simulating water
spray. At regular intervals of time, the samples can be evaluated for physical properties
like color, surface roughness, weight change, and mechanical and thermal properties like,
tensile and ﬂexural strength, modulus of elasticity, impact strength, storage modulus,
coefﬁcient of thermal expansion etc.

It is expected that with the increment in the exposure time of the samples to accelerated
weathering conditions, the surface roughness should increase, while color should turn
lighter, and thermal as well as mechanical properties should decrease, compared to their
properties at the beginning of the test. Such weathering experiments can give an insight
about the amount of damage the samples can undergo in accelerated weathering
conditions, and can also help to set guidelines about the life time of the products
performing at the acceptable levels. However, such tests cannot be compared to real time,
and no correlation exists to convert the test time in the chamber to actual time in the real
world

Ever since 19905 many researchers have actively examined the durability of wood-
ﬁber/plastic composites. Fungal attacks on wood composites have been studied by
different lab and ﬁeld tests.

Matuana et al. [230] investigated the durability of polyvinyl chloride (PVC)/ woof ﬁber
composite samples subjected to a cyclic UV light/spray weathering tests for a duration of
1200 hours. The samples were analyzed by FTIR, colorimeter, contact angle, and surface
roughness. They observed that the wood ﬁbers are sensitizers and their reinforcement in a
PVC matrix accelerated the photo-degradation of the polymer matrix. There was greater

discoloration in the composite samples compared to neat plastic.

90

Kalnins and F eist [231] exposed the western redcedar panels to outdoor weathering. They
measured the wettability of the panels with water using contact angle analysis. They
found that the contact angle decreased as the exposure time increased, leading to greater
wettability. Increased wettability might be a factor contributing to the degradation of
wood structures and articles.

Feist [232] performed outdoor study of various ﬁnish systems on aspen siding. He found

that acrylic latex paints are best protection for weathering.

Ultraviolet weathering performance of rigid PVC ﬁlled with different concentrations of
wood ﬂour was studied by Matuana and Karndem [233]. The extruded and compression
molded samples were subjected to a cyclic UV lamps, condensation, and exposed for 300
hours. The samples were regularly assessed over a wide range to time during exposure.
The samples were analyzed by FTIR, colorimeter, contact angle, tensile test and notched
Izod test. It was found that incorporation of wood ﬁbers in the PVC matrix led to a
increased degradation of the matrix material. The exposed PVC plastic samples also had
lower impact strength and lower elongation at break compared to original plastics.
However, all of the composites retained their impact strength, tensile strength, tensile
modulus as well as elongation at break.

Podgorski et al. [234] developed an artiﬁcial weathering method to assess the durability
of exterior wood coatings and wood in a short period of time. The optimized cycle they
developed contained a condensation period of 24 hours, UVA at 340 nm for 2.5 hours

and water spray for 0.5 hours. The total duration of exposure was 2016 hours. This

91

optimized cycle was tested on different wood substrates, wood coatings, and also checked
for reproducibility.

P. Chow and D. S. Bajwa studied the weathering effects on the decay resistance of
creosote treated oak [235]. They used a standard soil block culture method to test decay
resistance of naturally and artiﬁcially aged oaks infected with brown and white rot fungi.
They found that naturally aged samples were more vulnerable to fungal attack as
compared to artiﬁcially aged ones.

Nzokou and Karndem studied the weathering of African paduak and red maple, both of
which are hardwoods. The wood samples were artiﬁcially weathered for 1000 hours, and
changes in color, roughness and contact angle were measured at regular intervals of
exposure [236]. They observed that the color change for both species was very fast
during the ﬁrst 24 hours of exposure. With increment in exposure time, the roughness of
both wood types increased, and the contact angle decreased due to a decrease in lignin
and extractive content of the wood surface.

Feist [237] and team evaluated the long term effect of solvent borne and waterborne
water repellent preservative treatments on the performance of various paint systems
applied to different wood substrates. Six substrates were subjected to nine years of
outdoor exposure and evaluated for degradation, ﬁnish and decay. They found that the
performance of substrates treated with waterborne preservatives was slightly better than
those treated with solvent—bome preservatives.

Stark and Matuana [238] studied the performance of photostabilized HDPE/wood ﬂour
composites after accelerated weathering exposure of 2000 hours. They found that both

colorants and UVA are better photostabilizers than hindered amine light stabilizers for

92

HDPE/wood composites. The ﬂexural properties of neat HDPE dropped signiﬁcantly
after weathering, but those of photostabilized HDPE/wood composites remained same
until 2000 hours of exposure.

Gassan and Bledzki [239] determined the thermal degradation of untreated and alkali-
treated jute and ﬂax ﬁbers. The ﬁbers were exposed to temperatures between 170 and
210 0C for a maximum of 120 min. The exposed ﬁbers were mechanically tested in a
yarn tensile test to measure the inﬂuence on tenacity. In all cases both ﬁbers showed none
or only a slight decrease in tenacity and degree of polymerization at temperatures below
to 170°C. For temperatures above 170°C the tenacity as well as the degree of
polymerization decrease rapidly, and depended on both exposure time and temperature.
In general, a higher temperature or longer time of exposure led to an increase in level of
depolyrnerization and to a drop in tenacity. Because of chain scission due to this thermal
treatment, an increase in degree of crystallinity with a maximum of 20% was found.
Degradation consists of loss of strength and/or stiffness. The relationship between the
decay during accelerated experimental testing and long-terrn testing is expressed through
an acceleration factor k. The acceleration factor is the ratio between time exposures after
natural aging and after accelerated aging to obtain a similar decay. For instance, an
acceleration factor k = 400 means that every hour under accelerated laboratory aging

testing corresponds to 400 hours of aging under normal conditions.

Flammability
Schartel et al. [240] studied thermal and ﬁre behavior of ﬂax-PP composites containing

ammonium polyphosphate and expandable graphite as ﬁre retardants. The biocomposites

93

were analyzed using TGA, FTIR, cone calorimeter, limiting oxygen index, and glow wire
test. As expected ﬁre retardancy was better for samples containing ammonium
polyphosphate and expandable graphite, compared to neat biocomposite. Comparing

same amounts of both ﬁre retardants, graphite was better than ammonium polyphosphate.

94

References

l.

10.

11.

12.

13.

14.

15.

A. K. Mohanty, M. Misra, G. Hinrichsen, Macromolecular Materials and
Engineering, 276/277, (2000), 1-24.

A. K. Mohanty, M. Misra, L. T. Drzal, Journal of Polymers and the Environment, 10
(1/2), (2002), 19-26.

Plant/Crop-based Renewable Resources 2020, Available at:
http://www.oit.doe.gov/agriculture/pdfs/vision2020.pdf

The Technology Roadmap for Plant/Crop-based Renewable Resources 2020,
Available at: http://www.oit.doe.gov/agriculture/pdfs/ag25942.pdf

J. N. McGovern, In Encyclopedia of Polymer Science and Engineering, Vol.7, H. F.
Mark, N. M. Bikales, C. G. Overberger, G. Menges, J. I. Kroschwitz, Eds, John
Wiley & Sons; New York, (1987), 16.

R. M. Rowell, 1n Polymers and other advanced materials: emerging technologies and

business opportunities, P. N. Prasad, J. E. Mark, T. J. Fai, (Eds), Proceedings of the
3rd International conference on frontiers of polymers and advanced materials, 16-20
January 1995, Kuala Lumpur, Malaysia, New York, Plenum Press, 659-665.

E. S. Lipinsky, J. D. Ingham, DOE (U .S.A) report number ILA 207376 — A — H1.

J. L. Broge, Automotive Engineering International, October (2000), 120.

A. K. Bledzki, J. Gassan, Journal of Progress in Polymer Science, 24(2), (1999), 221.
T. Wang, L. T. Drzal, B. E. Dale, A. K. Mohanty, M. Misra, Unpublished results.

S. W. Beckwith, Composites Fabrication, November/December 2003, 12-16.
Available at: www.cﬁnagazine.com

L. T. Drzal, P. Herrera-Franco, 1H. Ho, In Test Methods for the Measurement of
Fiber-Matrix Adhesion. Comprehensive Composite Materials, A. Kelley; C. Zweben,
(Eds), Chapter 5, (2000).

Daimler Chrysler Corporation, Daimler Chrysler 2001 Environmental Report, Auburn
Hills, MI.

S. Biswas, A. Mittal, G. Srikanth, In Proceedings of the International Conference and
Exhibition on Reinforced Plastics, (2002), Indian Institute of Technology, Madras,
26-36.

J. Nickle, U. Riedel, Materials Today, 6(4), (2003), 44-48.

95

16.

l7.

l8.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Phenix Biocomposites. Company Homepage, Available at:
www.phenixbiocomposites.com

S. Biswas, G. Srikanth, S. Nangia, In Proceedings of the Composites Fabricators
Association’s Conference, (2001), Tampa, FL.

Dialrner Chrysler High Tech Report, 1999, 88

A. K. Mohanty, L. T. Drzal, “Biocomposites from Engineered Natural Fibers for
Housing Panel Applications”, NSF-PATH proposal (2001) (Award number 0122108)

T. Walkida and S. Tokino, Indian Journal of Fibre & Textile Research, 21, (1996),
69.

A. K. Mohanty, M. Misra, L. T. Drzal, Composite Interfaces, 8(5), (2001), 313-343.

J.S. Han, Fiber property comparison, Presented at TAAPI 1998 North America
Nonwood Symposium, Atlanta, GA, February 18-19, (1998)

R. M. Rowell, In Proceedings of the 3rd International conference on frontiers of
polymers and advanced materials, Polymers and other advanced materials: emerging
technologies and business opportunities, P. N. Prasad, J. E. Mark, T. J. Fai, (Eds),
16-20 January 1995, Kuala Lumpur, Malaysia, New York, Plenum Press, 659-665.

T. Peijs, e-Polymers, T_002, (2002), 1-12. Available at: www.c-polymers.com

U. Reidel, J. Nickel, Die Angewandte Makromolekulare Chemie, 272 (4756), (1999),
34-40.

S. Mishra, A. K. Mohanty, L. T. Drzal, M. Misra, S. Parija, S. K. Nayak, S. S.
Tripathy, Composites Science and technology, 63, (2003), 1377-1385.

Mwaikambo and Ansell, Die Angewandte Makromolekulare Chemie, 272 (4753),
(1999), 108-116.

E. T. N. Bisanda, Applied Composite Materials, 7, (2000), 331-339.

J. Rout, M. Misra, S. S. Tripathy, S. K. Nayak, A. K. Mohanty, Journal of Applied
Polymer Science, 84,(2002), 75-82.

A. Stamboulis, C.A. Baillie, T. Peijs, Composites: Part A, 32, (2001), 1105-1115.

G. G. Silva, D.A. De Souza, J.C. Machado, D.J. Hourston, Journal of Applied
Polymer Science,76, (2000), 1 197-1206.

J. Gassan, A. K. Bledzki, Journal of Applied Polymer Science, 71, (1999), 623-629.
A. Bismarck, A. K. Mohanty, I. Aranberri-Askargorta, S. Czapla, M. Misra, G.
Hinrichsen, J. Springer, Green Chemistry, 3, (2001), 100-107.

96

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

45.

46.

47.

48.

49

AK. Mohanty, L. T. Drzal, K.J. Ferguson, B. E. Dale, M. Misra, R. Schalek, ACS
Polymer Preprints, 43(2), (2002), 749-750.

R. K. Samal, M.C. Ray, Journal of Applied Polymer Science, 64(11), (1997), 2119-
2125.

C. Pavithran, P. S. Mukherjee, M. Brahmakumar, A. D. Damodaran, Journal of
Materials Science Letters, 7, (1988), 825-826.

C. Pavithran, P. S. Mukherjee, M. Brahmakumar, A. D. Damodaran, Journal of
Materials Science Letters, 6, (1987), 882-884.

K. Oksman, L. Wallstrom, L. A. Berglund, R. D. T. Filho, Journal of Applied
Polymer Science, 84, (2002), 2358-2365.

L. A. Belcher, L. T. Drzal, M. Misra, A.K. Mohanty, ACS Polymer Preprints, 42 (2),
(2001), 73-74.

L. A. Belcher, L. T. Drzal, M. Misra, A.K. Mohanty, Polymeric Materials: Science
and Engineering, 87, (2002), 256-257.

I. K. Varma, S. R. Anantha Krishnan, S. Krishnamoorthy, Composites, 20 (4), (1989),
383-388.

H. K. Mishra, B. N. Dash, S. S. Tripathy, B. N. Padhi, Polymer-Plastics T echnologv
and Engineering, 39(1), (2000), 187-198.

J. Rout, M. Misra, S. S. Tripathy, S. K. Nayak, A. K. Mohanty, Composites Science
and Technology, 61, (2001), 1303-1310.

. J. M. Zimmerman, N. S. Lousure, Journal of Advanced Materials, 30 (2), (1998), 32.

M. Z. Rong, M. Q. Zhang, Y. Liu, H. M. Yan, G. C. Yang, H. M. Zeng, Polymer
Composites, 23 (2), (2002), 182-192.

S. Mishra, M. Misra, S. S. Tripathy, S. K. Nayak, A. K. Mohanty, Polymer
Composites, xx (x), (2001), xx-xx.

T. H. D. Sydenstricker, S. Mochnaz, S. C. Amico, Polymer Testing, 22, (2003), 375-
380.

C. A. S. Hill, H. P. S. A. Khalil, Journal of Applied Polymer Science, 78, (2000),
1685-1697.

H. P. S. A. Khalil, H. Ismail, M. N. Ahmad, A. Arifﬁn, K. Hassan, Polymer
International, 50, (2001), 395-402.

97

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

AK. Mohanty, M. Misra, L. T. Drzal, In Proceedings of the 24th Annual Meeting of
the Adhesion Society—Adhesion Science for the 21St Century, February 25-28, (2001),
418-420.

D. Ray, B. K. Sarkar, A. K. Rana, N. R. Bose, Bulletin of Materials Sciences, 24 (2),
(2001), 129-135.

D. Ray, B. K. Sarkar, A. K. Rana, N. R. Bose, Composites: Part A, 32, (2001), 119-
127.

D. Ray, B. K. Sarkar, S. Das, A. K. Rana, Composites Science and Technology, 62,
(2002), 911-917.

J. Gassan, A. K. Bledzki, Polymer Composites, 20 (4), (1999), 604-611.

D. G. Hepworth, D. M. Bruce, J. F. V. Vincent, G. Jeronimidis, Journal of Materials
Science, 35 (2), (2000), 293-298.

J. Rout, M. Misra, S. S. Tripathy, S. K. Nayak, A. K. Mohanty, Polymer Composites,
22(4), (2001), 468-476.

K. Joseph, L.H. de Carvalho, Natural Polymers and Composites, Proceedings from
the Third International Symposium on Natural Polymers and Composites, Sao Pedro,
Brazil, May 14-17, (2000), 454-459.

N. P. Giacomini, A. L. Leao, A. M. Neis, Natural Polymers and Composites,
Proceedings from the Third International Symposium on Natural Polymers and
Composites, Sao Pedro, Brazil, May 14-17, (2000), 386-392.

L. A. Pothan, S.Thomas, Z. Oommen, Natural Polymers and Composites,
Proceedings from the Third International Symposium on Natural Polymers and
Composites, Sao Pedro, Brazil, May 14-17, (2000), 393-398.

K. Oksman, L. Wallstrom, L. Berglund, R. D. T. Filho, Natural Polymers and
Composites, Proceedings from the Third International Symposium on Natural

Polymers and Composites, and the Workshop on Progress in Production and

Processing of Cellulosic Fibres and Natural Polymers, Sao Pedro, Brazil, May 14-17,
(2000), 355-361.

G. B. Nando, B. R. Gupta, In Short Fiber-Polymer Composites, J. R. White and S. K.
De (Eds), Woodhead Publishing, Cambridge, 1996.

R. G. Raj and B. V. Kokta, Journal of Applied Polymer Science, 38, 1987, 1989.

A. R. Sanadi, D. F. Caulﬁeld, R. E. Jacobson, In Paper and Composites from Agro-
Based Resources, CRC Press, Boca Raton, 1996.

98

64. R. Rowell, A. R. Sanadi, D. F. Caulﬁeld, R. E. Jacobson, In Lignocellulosic- Plastics
Composites, A. L. Leao, F. X. Carvalho, and E. Frollini (Eds), USP and UNESP,
Brazil, 1997.

65. N. Pereira, M. L. Souza, J. A. M. Agnelli, L. H. C. Mattaso, In Fourth lntemational
Conference on Wood ﬁber-Plastic Composites, Forest Products Society, Madison,
(1997)

66. S. E. Selke, J Childress, In Wood Fiber/Polymer Composites: Fundamental Concepts,
Process, and Material Options, Forest Product Society, Madison, (1993).

67. J. X. Rietveld, M. J. Simon, In Wood Fiber/Polymer Composites: Fundamental
Concepts, Process, and Material Options, Forest Product Society, Madison, (1993).

68. K. C. Manikam, S. M. Diwan, S. Thomas, Journal of Applied Polymer Science, 60,
(1996), 1483.

69. R. Ali, S. Iannace, L. Nicolias, Journal of Applied Polymer Science, 88, (2003), 1637-
1642.

70. V. A. Alvarez, R.A. Ruscekaite, A. Vazquez, Jouranl of Composite Materials, 37(17),
(2003), 1575-1588.

71. Composites Technology Yellow Pages, (2000), 6.
72. Composite Industry News Letter, 10(2), March 7, 2000.

73. M. Malik, V. Chaudhary, I. K. Varma , Journal of Material Science-Reviews
Macromolecular Chemistry and. Physics, C40, 2&3, (2000), 139-165

74. H.V Boeing, Unsaturated polyesters: structure and properties, Elsevier Pub. Co.,
(1964)

75. G. Odian, Principles of Polymerization, Third Edition, John Wiley & Sons, Inc., New
York, (1991), 53

76. G. Giindiz, In Polymeric Materials Encyclopedia, J.C. Salamone (Ed.), 1], CRC
Press, Boca Raton,, FL, (1996), 8469-8476

77. C. H. Bamford, C. F. H. Tipper, Free-radical polymerization, Elsevier Pub. Co.,
(1976)

78. R. Burns, Polyester Molding Compound; Marcel Dekker: New York, (1982)

79. A. Fradet, P. Arlaud, In Comprehensive Polymer Science, G. Allen, J. C. Bevington,
(Eds), Vol. 5, Pergamon Press, Oxford, (1989), 331

99

80. J. Makhlouf, In Kirk-Othmer Encyclopedia of Chemical Technology, H. F. Mark, D.
F. Othrner, C. G. Overberger, G. T. Seaborg, (Eds), Vol. 18, John Wiley & Sons,
New York, (1982), 575

81. J. Selley, In Encyclopedia of Polymer Science and Engineering, H. F. Mark, N. M.
Bikales, C. G. Overberger, G. Menges. (Eds), Vol. 12, John Wiley & Sons, New
York, 1988, 256

82. M. R. Kamal, S. Sourour, Polymer Engineering and Science, 13(1), (1973), 59-64.
83. J. M. Salla, J. L. Martin, Thermochimica Acta, 126, (1988), 339-354.

84. J. M. Salla, X. Ramis, J. L. Martin, A. Cadenato, T hermochimica Acta, 134, (1988),
261-167.

85. M. Paci, E. D. Vecchio, F. Campana, Polymer Bulletin, 15, (1986), 21-27.
86. X. Ramis, J. M. Salla, Journal of Applied Polymer Science, 45, (1992), 227-236.

87. D. Rouison, M. Sain, M. Couturier, Journal of Applied Polymer Science, 89, (2003),
2553-2561.

88. A. Nzihou, P. Sharrock, A. Ricard, Chemical Engineering Journal, 72, (1999), 53-61.

89. J. L. Vilas, J. M. Laza, M. T. Gary, M. Rodriguez, L. M. Leon, Journal of Applied
Polymer Science, 79,(2001), 447-457.

90. Y-J. Huang, J-S. Leu, Polymer, 34(2), (1993), 295-304.

91. H. Yang, L. J. Lee, Journal of Applied Polymer Science, 79, (2001), 1230-1242.
92. H. Yang, L. J. Lee, Journal of Applied Polymer Science, 84, (2002), 211-227.
93. J. S. Chen, T. L. Yu, Journal of Applied Polymer Science, 69, (1998), 871-878.

94. V. Kosar, Z. Gomzi, Chemical and Biochemical Engineering Quarterly, 15 (3),
(2001), 101—108.

95. Y.-M. Yun, S.-J. Lee, K.-J. Lee, Y.—K. Lee, J .-D. Nam, Journal of Polymer Science:
Part B: Polymer Physics, 35, (1997), 2447—2456.

96. S. Ziaee, G. R. Palmese, Journal of Applied Polymer Science, 37, (1999), 725-744.

97. R. P. Brill, G. R. Palmese, Journal of Applied Polymer Science, 76, (2000), 1572-
1582.

98. B. L. Grunden, C. S. P Sung, Macromolecules, 36, (2003), 3166-3173.

99. R-H. Lin, J-H. Hsu, Polymer lntemational, 50, (2001), 1073-1081.

100

100. L. Shan, C. G. Robertson, K. N. E. Verghese, E. Burts, J. S. Rifﬂe, T. C. Ward,
K .L. Reifsnider, Journal of Applied Polymer Science, 80,(2001), 917-927.

101. C. D. Han, K-W. Lem, Journal of Applied Polymer Science, 28, (1983), 3155-3183.
102. C. D. Han, K-W. Lem, Journal of Applied Polymer Science, 28, (1983), 3185-3206.
103. C. D. Han, K-W. Lem, Journal of Applied Polymer Science, 28, (1983), 3207-3225.

104. C. Serre, M. Vayer, R. Erre, C. Ollive, Journal of Material Science, 34, (1999),
4203-4208.

105. C. Serre, M. Vayer, R. Erre, N. Boyard, C. Ollive, Journal of Material Science, 36,
(2001), 113—122.

106. C. Marieta, P. M. Remiro, G. Garmendia, I. Harismendy, I. Mondragon, European
Polymer Journal, 39, (2003), 1965- 1973.

107. S. A. Mazrouh, Pigment and Resin Technology, 30(2), (2001), 1 14-119.

108. Hangzhou JLS Flame Retardants Chemcial Co. Ltd., JLS-APP Series Flame
Retardants, Available at: www.jlschemical.com

109. M. Choate, Isola Laminate Systems, LaCrosse, WI, Presented at [PC Printed
Circuits EXPO® 2001

110. Avtec Industries, ﬁre retardant additive TSWBTM, Available at:
http://www.avtecindustries.com/tswb.html

111. V. A. Rosa, M. I. Felisberti, Journal of Applied Polymer Science, 81,(2001), 3272-
3279.

112. V. A. Rosa, J. Karger-Kocsis, M. I. Felisberti, Journal of Applied Polymer Science,
81, (2001), 3280-3289.

113. A. V. Rajulu, L. G. Devi, G. B. Rao, Journal of Applied Polymer Science, 89,
(2003), 2970-2972.

114. B. Gawdzik, T. Matynia, J. Osypiuk, Journal of Applied Polymer Science, 79,
(2001), 1201-1206.

115. K.-E. Min, Y.-G. Hwang, G.-Y. Choi, H.-G. Kim, W.-S. Kim, D.-H. Lee, L.-S. Park,
K.-H. Seo, I.-K. Kang, I.-R. Jun, J .-C. Lim, H.-C. Song, Journal of Applied Polymer
Science, 84, (2002), 735-740.

116. A. B. Cherian, E. T. Thacil, Progress in Rubber and Plastics Technology, 17(4),
(2001), 205-224.

101

117. A. B. Cherian, E. T. Thacil, Journal of Elastomers and Plastics, 36, (2003), 367-
380.

118. M. D. Thouless, J. Du, A. F. Yee, In Toughening of Plastics: Advances in Modeling
and Experiments, R.A. Pearson, H.-J. Sue, A. F. Yee, (Eds), American Chemical
Society Symposium Series 759, 2000, 71-85.

119. Y. Huang, D. L. Hunston, A. J. Kinloch, C. K. Riew, In Toughened Plastics 1:
Science and Engineering, C. K. Riew, A. J. Kinloch (Eds), Advances in Chemistry
Series 233, 1993,1-35.

120. El McGarry, R. Subramaniam, In Toughened Plastics II: Novel Approaches in
Science and Engineering, C. K. Riew, A. J. Kinloch (Eds), Advances in Chemistry
Series 252, 1996, 133-150.

121. U. Schulze, M. Skrifvars, N. Reichelt, H.—W. Schmidt, Journal of Applied Polymer
Science, 64(3), (1997), 527-537.

122. G. Hegemann, Macromolecular Symposium, 199, (2003), 333-342.

123. W. E. Douglas, G. Pritchard, British Polymer Journal, 15(1), (1983), 19-23.

124. Q. Guo, H. Zheng, Polymer, 40, (1999), 637-646.

125. Q. Guo, H. Zheng, Journal of Applied Polymer Science, 74, (1999), 322-327.
126. C. D. Han, K.-W. Lem, Journal of Applied Polymer Science, 28, (1983), 743-762.
127. C. D. Han, K.-W. Lem, Journal of Applied Polymer Science, 28, (1983), 779-793.
128. M. Rufﬂer, N. Vincent, G. Merle, Polymer Bulletin, 30, (1993), 111-118.

129. Y.-J. Huang, C.-C. Su, Journal of Applied Polymer Science, 55, (1995), 305-322.
130. Y.-J. Huang, C.-C. Su, Journal of Applied Polymer Science, 55, (1995), 323-342.
131. B. Sun, T. L. Yu, Journal of Applied Polymer Science, 57, (1995), 7-23.

132. Y.-J. Huang, C.-M. Liang, Polymer, 37 (23), (1996), 401-412.

133. R. Saito, W.-M. J. Kan, L. J. Lee, Polymer, 37 (16), (1997), 3567-3576.

134. W. Li, L. J. Lee, Polymer, 39 (23), (1998), 5677-5687.

135. Y.-J. Huang, W.-C. Jiang, Polymer, 39 (25), (1998), 6631-6641.

136. Y.-J. Huang, L.-D. Chen, Polymer, 39 (26), (1998), 7049-7059.

137. Y. N. Hsieh, T. L. Yu, Journal of Applied Polymer Science, 73, (1999), 2413—2428.

102

138. Y.-J. Huang, C.-J. Chu, J .-P. Dong, Journal of Applied Polymer Science, 78, (2000),
543—557.

139. Z. Zhang, S. Zhu, Polymer, 41, (2000), 3861—3870.
140. X. Cao, L. J. Lee, Journal of Applied Polymer Science, 82, (2001), 738—749.
141. X. Cao, L. J. Lee, Polymer, 44, (2003), 1893—1902.
142. X. Cao, L. J. Lee, Polymer, 44, (2003), 1507-1516.

143. N. Boyard, M. Vayer, C. Sinturel, R. Erre, D. Delaunay, Journal of Applied
Polymer Science, 88, (2003), 1258-1267.

144. Y.-J. Huang, T.-S. Chen, J.-G. Huang, F.-H. Lee, Journal of Applied Polymer
Science, 89, (2003), 3336—3346.

145. Y.-J. Huang, T.-S. Chen, J.-G. Huang, F.-H. Lee, Journal of Applied Polymer
Science, 89, (2003), 3347—3357.

146. K. Wang, J. H. Wang, Y. D. Zhu, Y. Xia, Key Engineering Materials, 249, (2003),
351-354.

147. Z. M. Xue, J. H. Wang, Q. P. Feng, Key Engineering Materials, 249, (2003), 343-
346.

148. L. H. Sperling, Interpenetrating Polymer Networks and Related Materials;
Plenum: New York, (1981).

149. B. Suthar, H. X. Xiao, D. Klempner, K. C. Frisch, Polymers for Advanced
Technologies, 7, (1996), 221-233.

150. L. H. Sperling, V. Mishra, Polymers for Advanced Technologies, 7, (1996), 197-208.

151. S. Gunanathan, R. Hariharan, M. Sarojdevi, Journal of Applied Polymer Science, 92,
(2004), 817—829.

152. Y. C. Chou, L.J. Lee, In Interpenetrating Polymer Networks, Advances in
Chemistry Series 239, D. Klempner, L. H. Sperling, L. A. Utracki, (Eds), American
Chemical Society, (1994), 305-331.

153. L. H. Sperling, In Interpenetrating Polymer Networks, Advances in Chemistry
Series 239, D. Klempner, L. H. Sperling, L. A. Utracki, (Eds), American Chemical
Society, (1994), 3-37.

154. I. Frichinger, S. Dirlikov, In Interpenetrating Polymer Networks, Advances in

Chemistry Series 239, D. Klempner, L. H. Sperling, L. A. Utracki, (Eds), American
Chemical Society, (1994), 517 -538.

103

155. I. Frichinger, P. Muturi, S. Dirlikov, In Interpenetrating Polymer Networks,
Advances in Chemistry Series 239, D. Klempner, L. H. Sperling, L. A. Utracki,
(Eds), American Chemical Society, (1994), 539-556.

156. G. Mehta, A. K. Mohanty, M. Misra, L. T. Drzal, Green Chemistry, 6(5), (2004),
254-258.

157. Alfred Thomas, In Ullmann's Encyclopedia of Industrial Chemistry, 2000, John
Wiley & Sons.

158. E. B. Cahoon, AgBioForum, 6(1&2), (2003), 11-13.

159. I. Frichinger S. Dirlikov, In Toughened Plastics 1 Science and Engineering,
Advances in Chemistry Series 233, C. K. Riew and A. J. Kinloch (Eds), American
Chemical Society, 1993, 451-489.

160. S. Dirlikov, I. Frischinger, and Z. Chen, In Toughened Plastics II Novel
Approaches in Science and Engineering, Advances in Chemistry Series 252, C. K.
Riew and A. J. Kinloch (Eds), American Chemical Society, 1993, 95-109.

161. P. Tan, In Interpenetrating Polymer Networks, Advances in Chemistry Series 239,
D. Klempner, L. H. Sperling, and LA Utrachi, (Eds), American Chemical
Society,l994, 595-614.

162. American Soybean Association, 2001. SoyStats 2001, Available at:
http://www.unitedsoybean.org/soystatsZOO1

163. USDA statistics on oil crops. Available at: www.usda.mannlib.comell.edu/data-
sets/ crops/89002/table35.xls

164. Linseed oil applications, Available at: www.ienica.net/crops/1inseed.pdf
165. J. Rosch, R. Mulhaupt, Polymer Bulletin, 31, (1993), 679-685.

166. H. Warth, R. Mulhaupt, B. Hofﬁnann, S. Lawson, Die Angewandte
Makromolekulare Chemie, 249, (1997), 79-92.

167. S. N. Khot, J. J. LaScala, E. Can, S. S. Morye, G. 1. Williams, G. R. Palmese, S. H.
Kusefoglu, R. P. Wool, Journal of Applied Polymer Science, 82, (2001), 703—723.

168. E. Can, S. Kusefoglu, R. P. Wool, Journal of Applied Polymer Science, 81, (2001),
69—77.

169. E. Can, S. Kusefoglu, R. P. Wool, Journal of Applied Polymer Science, 83, (2002),
972—980.

170. F. K. Li, R. C. Larock, Journal of Polymers and the Environment, 10 (1/2), (2002),
59-67.

104

171. F. K Li, R. C. Larock, Journal of Polymer Science Part B-Polymer Physics, 38(21),
(2000), 2721-2738 .

172. F. K Li, R C. Larock, Journal of Applied Polymer Science, 78(5), (2000), 1044-
1056.

173. F. K Li, R. C. Larock, Journal of Polymer Science Part B-Polymer Physics, 39(1),
(2001), 60-77.

174. M. A. Dweib, B. Hu, A. O’Donnell, H.W. Shenton, R.P. Wool, Composite
Structures, 63, (2004), 147—157.

175. Soueck J. Chen, M. D. Soucek, W. J. Simonsick, R. W. Celikay, Polymer, 43,
(2002), 5379—5389.

176. Z. G. Zong, M. D. Soucek, Y. Liu, J. Hu, Journal of Polymer Science: Part A:
Polymer Chemistry, 41, (2003), 3440—3456.

177. J.-P. Latere Dwan’Isa, A. K. Mohanty, M. Misra, L. T. Drzal, M. Kazemizadeh,
Journal of Polymers and the Environment, 11 (4), (2003), 161-168.

178. J.-P. Latere Dwan’Isa, A. K. Mohanty, M. Misra, L. T. Drzal, M. Kazemizadeh,
Journal of Materials Science, 39, (2004), 1887—1890.

179. J.-P. Latere Dwan’Isa, A. K. Mohanty, M. Misra, L. T. Drzal, M. Kazemizadeh,
Journal of Materials Science, 39, (2004), 2081—2087.

180. L. T. Drzal, J.-P. Latere Dwan'Isa, A. K. Mohanty, M. Misra, “Natural oil-based
polyols and methods for making those and their uses”, (2003), US Patent Pending

181. I. Javni, W. Zhang, Z. S. Petrovic, Journal of Applied Polymer Science, 90, (2003),
3333—3337.

182. I. Javni, W. Zhang, Z. S. Petrovic, Journal of Applied Polymer Science, 88, (2003),
2912—2916.

183. A. Guo, I. Javni, Z. S. Petrovic, Journal of Applied Polymer Science, 77, (2000),
467—473.

184. I. Javni, Z. S. Petrovic, A. Guo, R. Fuller, Journal of Applied Polymer Science, 77,
(2000), 1723—1734.

185. A. Guo, Y. J. Cho, Z. S. Petrovic, Journal of Polymer Science: Part A: Polymer
Chemistry, 38, (2000), 3900—3910.

186. Z. S. Petrovic, A. Guo, W. Zhang, Journal of Polymer Science: Part A: Polymer
Chemistry, 38, (2000), Vol. 4062—4069.

105

187. H. Uyama, M. Kuwabara, T. Tsujimoto, M. Nakano, A. Usuki, S. Kobayashi,
Chemistry of Materials, 15, (2003), 2492-2494.

188. A. Garg, K. Chandrashekhara,V. F 1anigan,S. Kapila, “Natural Fiber Reinforced
Soy-Based Composites”, Paper # [354b], Composites and Laminates Based on Forest
and Biobased Products 111, Forest Products Division, American Institute of Chemical
Engineers (AIChE) Annual meeting 2002, Indianapolis, IN

189. T. Eren, S. H. Kusefoglu, Journal of Applied Polymer Science, 91, (2004), 2700—
2710.

190. T. Eren, S. H. Kusefoglu, R. P. Wool, “Polymerization of Maleic Anhydride—
Modiﬁed Plant Oils with Polyols”, Journal of Applied Polymer Science, 90, (2003),
197—202.

191. H. Y. Gu, K. Ren, D. Martin, T. Marino, D. C. Neckers, Journal of Coatings
Technology, 74, 2002, 927.

192. H. Miyagawa, A. K. Mohanty, M. Misra, L. T. Drzal, “Biobased Epoxy/Layered
Silicate Nanocomposites: Fabrication and Properties Evaluation”, Green Chemistry,
to be submitted

193. H. Miyagawa, A. K. Mohanty, M. Misra, L. T. Drzal, ‘Thermo-Physical Properties
of Epoxy Containing Epoxidized Linseed Oil, Part 1: Anhydride-cured Epoxy”,
Macromolecular Materials and Engineering, to be provisionally accepted

194. H. Miyagawa, A. K. Mohanty, M. Misra, L. T. Drzal, “Thermo-Physical Properties
of Epoxy Containing Epoxidized Linseed Oil, Part 2: Amine-cured Epoxy”,
Macromolecular Materials and Engineering, to be provisionally accepted

195. H. Miyagawa, A. K. Mohanty, M. Misra, L. T. Drzal, “Fracture Toughness and
Impact Strength of Anhydride-cured Biobased Epoxy”, Polymer Engineering and
Science, to be submitted

196. H. Miyagawa, A. K. Mohanty, M. Misra, L. T. Drzal, “Biobased Epoxy-clay
Nanocomposites: Therrnophysical properties and fracture behavior Evaluation”,
Journal of Polymers and the Environment, to be submitted

197. H. Miyagawa, A. K. Mohanty, M. Misra, L. T. Drzal, “Effect of clay and alumina-
nanowhisker reinforcements on elastic modulus, impact strength and fracture
toughness of nanocomposites from biobased epoxy: A comparative study”, Journal
for Nanoscience and Nanotechnology, to be submitted ~

198. H. Miyagawa, A. K. Mohanty, M. Misra, L. T. Drzal, “Bio-based Epoxy/Clay
Nanocomposites as a New Matrix for CFRP”, Materials Science and Engineering
Part A-Structural Materials: Properties, Microstructure and Processing, In
preparation

106

199. H. Miyagawa, A. K. Mohanty, M. Misra, L. T. Drzal, “Bio-based epoxy, their
nanocomposites and methods for making those”, 2003, US Patent Pending

200. G. Mehta, A. K. Mohanty, M. Misra, L. T. Drzal, In preparation

201. D. Hull, T.W. Clyne, An introduction to composite materials, Cambridge; New
York, Cambridge University Press, 1996.

202. R. S. Pamas, Liquid Composite Molding, Hanser Gardner Inc., 2000

203. J. L. O’Dell, In Proceedings of the 4th international conference on wood ﬁber-
plastic composites, R. M. Rowell, A. R. Sanadi, (Eds), Forest Products Society,
Madison, WI, (1997), 280-285.

204. R. Rowell, J. L. O’Dell, R. K. Basak, M. Sarkar, In Proceedings of lntemational
seminar on jute and allied ﬁbres: changing global scenario; Calcutta, India. IJIRA
Association, (1997), 89-96.

205. M. M. Schwartz, Composite Materials Vol. II: Processing, Fabrication, and
Applications, Prentice Hall PTR, 1996.

206. H. W. Shenton III, R. P. Wool, B. Hu, A. O’Donnell, L. Bonnaille, E. Can, R.
Chapas, C. Hong, In M. Anson, J. M. Kao, E. S. S. Lam, (Eds), Advances in Building
Technology, Proceedings of the lntemational Conference on Advances in Building
Technology, 4-6 December 2002, Hong Kong, China, 1, (2002), 255-262.

207. M. A. Dweib, B. Hu, A. O’Donnell, H.W. Shenton, R.P. Wool, “All natural
composite sandwich beams for structural applications”, Composite Structures, 63,
(2004), 147—157.

208. M. A. Dweib, A. O’Donnell, B. Hu, L. Rosas, H. Shenton, R. P. Wool, In
lntemational Conference on Composite Materials, ICCM-14, July 14-18, 2003, San
Diego, CA, USA.

209. A. O'Donnell, M. A. Dweib, R. P. Wool, Composites Science and Technology,
64(9), (2004), 1 135-1145.

210. C. K. Hong, R. P. Wool, In lntemational Conference on Composite Materials,
ICCM-14, July 14-18, 2003, San Diego, CA, USA.

211. M. J. Quagliata, Development and Characterization of Biocomposite Cellular
Beams and Plates for Load-Bearing Components, MS thesis, Department of Civil and
Environmental Engineering, Michigan State University, 2003

212. C. J. Davis, R. P. Wood, E. R., Miller, “Glass ﬁber-reinforced sheet molding
compound”, US 3615979, (1971).

107

213. F. C. Peterson, L. P. Theard, “Therrnosetting sheet molding compounds”, US
3713927, (1973).

214. A. N. Piacente, US 3835212, “Resinous sheet like products”, (1974).
215. F. C. Peterson, L. P. Theard, “Sheet molding compounds”, DE 2357000, (1974).

216. T. Kishino, “Unsaturated polyester compositions for moldings”, JP 49076988,
(1974)

217. K. Horiuchi, T. Kamiya, Y. Ogawa, Y. Nakamura, M. Sugimori, “Unsaturated
polyester compositions”, JP 49107086, (1974).

218. K. Takaishi, K. Masamoto, K. Tagawa, “Molding of unsaturated polyesters”, JP
50005463, (1975).

219. M. Sekiguchi, K. Nomaguchi, H. Tanaka, “Molding resin compositions”, JP
50008882, (1975)

220. Y. Takayama, Y. Ichimura, T. Aoyagi, “Sheet molding compositions”, JP 50014792,
(1975)

221. M. Sato, T. Watanabe, W. Koga, “Sheet molding compositions”, JP 50036588,
(1975)

222. R. Kondo, K. Nakagawa, M. Fukuda, I. Kishi, T. Ohtsuki, “Fire retardant
thermosetting resin composition”, US 3931095, (1976). Kondo, Renichi; Nakagawa,
Koji; F ukuda, Makoto; Kishi, Ikuji; Ohtsuki, Tateki.

223. R. P. Wool, J. Lu, S. N. Khot, “Sheet molding compound resins from plant oils”,
US 2003088007, (2003).

224. B. van Voom, H. H. G. Smit, R.J. Sinke, B.de Klerk, Composites: Part A, 32
(2001),1271-1279.

225. D. N. Goswami, P. C. Jha, K. Mahato, K. K. Kumar, Popular Plastics & Packaging,
48(3), (2003), 68-71.

226. M. Kawamura, A. Nakai, H. Hamada, In lntemational Conference on Composite
Materials, ICCM-14, July 14-18, 2003, San Diego, CA, USA.

227. P. T. Odenberger, H. M. Andersson, T. S. Lundstrom, In lntemational Conference
on Composite Materials, ICCM-14, July 14-18, 2003, San Diego, CA, USA.

228. V. Feuillade, A. Bergeret, J-C. Quantin, A. Crespy, C. Metra, In lntemational

Conference on Composite Materials, ICCM-14, July 14-18, 2003, San Diego, CA,
USA.

108

229. V. Massardier-Nageotte, A. Maazouz, G. Peix, S. Bres, Polymer Testing, 22, (2003),
867-873.

230. L. M. Matuana, D. P. Kamdem, and J. Zhang, Journal of Applied Polymer Science,
80 (1 l), (2001), 1943-1950.

231. M. A. Kalnins, W. C. F eist, Forest Products Journal, 43(5), (1993) 55-57.
232. W. C. Feist, Forest Products Journal, 44(6), (1994) 15-23.

233. L. M. Matuana and D. P. Kamdem, 5th Paciﬁc Rim Bio-based Composites
Proceedings, 10- 14 December 2000, 644-651

234. L. Podgorski, M. Arnold, G. Hora, Coatings World, February 2003, 39-48
235. P. Chow, D. S. Bajwa, American Wood-Preservers Association Proceedings,

236. P. Nzokou, D. P. Kamdem, Journal of Topical Forest Products, 8 (2), 2002, 200-
209

237. W. C. F eist, Forest Products Journal, 40(7/8), (1990) 21-26.

238. N. Stark, L. M. Matuana, Proceedings of Society of Plastics Engineers Annual
Technical Conference ANTEC 2002, 2209-2213.

239. J. Gassan, A. K. Bledzki, Journal of Applied Polymer Science, 82, (2001), 1417—
1422.

240. B. Schartel, U. Braun, U Schwarz, S. Reinemann, Polymer, 44, (2003), 6241-6250.

109

CHAPTER THREE: EXPERIMENTAL METHODS

3. EXPERIMENTAL METHODS

3.1 Materials

The raw materials used for the fabrication of biocomposites were natural ﬁbers (from
plants) and matrix materials (Table 3.1.1 and 3.1.2). Non woven mats of industrial hemp
ﬁbers, chopped industrial hemp ﬁbers (green and retted), chopped ﬂax ﬁbers, kenaf
ﬁbers, chopped big blue stem grass, green ﬂax core, chopped henequen, chopped coir,
chopped pineapple leaf ﬁber, woven jute mats, chopped E- glass ﬁbers and non woven
glass mat were used as reinforcements for different composites. Two kinds of non-woven
hemp ﬁber mats, one containing 90% randomly oriented hemp ﬁbers and 10% poly
ethylene terephthalate (PET) as binder (see Figure 3.1.1.), and a second containing 100%
hemp ﬁbers was kindly provided by FlaxCraﬁ Inc. Kenaf ﬁbers were procured from
Texas Kenaf Industries, USA. Another kind of kenaf was obtained from China. Chopped
hemp, chOpped green hemp and chopped green ﬂax core ﬁbers were kindly provided by
FlaxCraﬁ Inc. Cresskill, USA. Jute mats were obtained from IJIRA, India. Henequen
ﬁbers were provided by Dr. Pedro Herrera-Franco of Mexico. Chopped E- glass ﬁbers
and non woven glass mat were supplied by Kemlite Inc. Joliet, USA. Rovings of E-glass
for SMC line were purchased from Owens Coming.

The natural ﬁbers were surface treated with various chemicals such as alkali,
acrylonitrile, silanes etc. Sodium hydroxide, gamma methacryloxy propyl trimethoxy
silane, amino propyl trimethoxy silane, ethanol, methanol, chloroform and acetone were

purchased from J. T. Baker. Acrylonitrile, maleic anhydride, methyl methacrylate, and

110

glycidyl mehtacrylate were obtained from Aldrich.

The polymer matrix used in this project is a combination of ortho unsaturated polyester
resin (UPE) and styrene, which was obtained from Kemlite Inc. The initiator used for

curing the resin was methyl ethyl ketone peroxide (MEKP), and the promoter used was

cobalt naphthenate (CoNap). Both MEKP and CoNap were obtained from Aldrich.

Pure double reﬁned soybean oil, castor oil and acrylated epoxidized soybean oil (AESO)
were procured from Aldrich. Methyl ester of soybean oil (MESO) was kindly provided
by AG Environmental Products. Epoxidized methyl linseedate (EML), soybean oil
phosphate ester polyol (SOPEP), methyl ester of epoxidized soybean oil (MEESO) and

epoxidized linseed oil (ELO) were obtained from Atoﬁna Chemicals.

All materials were used as such without further puriﬁcation.

Table 3.1.1: Reinforcements for composites made in this project

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Name of ﬁbers Composition Length of Density of Source of
fibers (mm) ﬁbers ﬁbers
Non-woven 90% hemp 10% 45-50 274 (gm/m2) F laxcraf’t Inc.
hemp mat-1 PET
Non-woven 100% hemp 45-50 600 (gm/m? F laxcraﬁ Inc.
hemp mat-2
Green hemp 100% hemp 6 1.2 (gm/m3) Flaxcraft Inc.
Retted hemp 100% hemp 6 1.2 (gm/m3) F laxcraﬁ Inc.
Big blue stem 100% BBSG 4 1.2 (gm/m3) Michigan,
grass (BBSG) USA
Green ﬂax core 50% ﬂax bast 1-2 1.2 (gm/m3) F laxcraft Inc.
50% ﬂax core
Kenaf 100% kenaf 6-8 1.2 (gm/m3) Flaxcraﬂ Inc.
Henequen 100% henequen 6-8 1.2 (gm/m3) Mexico
Coir 100% coir 6-8 1.2 (gm/m3) India
Jute mats 100% jute continuous 1.2 (gm/m3) India
Pineapple leaf 100% PALF 6-8 1.2 (gm/m3) India
ﬁber (PALF)

 

111

 

Table 3.1.1 (continued)

 

 

 

 

E-glass chopped 100% UPE 55-60 2.6 (gm/m3) Kemlite
sized glass Company Inc.
E-glass chopped 100% UPE 55-60 2.6 (gm/m3) Owens Corning
compatible
glass
E-glass mat 100% UPE 6—8 2.6 (gm/m3) Kemlite
sized glass Company Inc.
E-glass roving 100% UPE continuous 2.6 (gm/m3) Owens Corning
compatible
glass

 

 

 

 

 

 

 

 

Figure 3.1.1: Hemp mat with ﬁller (Bastrnat 115) obtained from FlaxCraft Inc.
Table 3.1.2: Materials and suppliers

Material Supplier
Kemlite Inc.
ketone

naphthenate

Methyl ester

(EML)
ester (SOPEP)
(MEESO) Products

 

112

Table 3.1.2 (Continued)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Poly(butadienemaleic anhydride) (PBMA) Sartomer Chemcial
Company
Epoxy resin (Epon 828) Shell Chemical Company

Sodium hydroxide J. T. Baker
y—methacryloxypropyl trimethoxy silane (MPS) Gelest Inc.
aminopropyltrimethoxy silane (APS) Gelest Inc.
Ethanol J .T. Baker
Methanol J .T. Baker

Chloroform Sigma Aldrich
Acetone J .T. Baker

Acrylonitrile (AN) Sigma Aldrich

Maleic anhydride (MA) Sigma Aldrich

Glycidyl mehtacrylate (GMA) Sigma Aldrich

Methyl methacrylate (MMA) Sigma Aldrich

Glacial acetic acid Sigma Aldrich

Dicumyl peroxide (DCP) Sigma Aldrich

Dueterated chloroform Sigma Aldrich

Polycaprolactone (PCL) Union Carbide
Polyoxypropylene triamine (J effamine-T403) Hunstman

meta-phenylene diamine (MPDA) Aldrich
3.2 Surface Modiﬁcation of natural ﬁbers

Prior to any treatment, non woven hemp mats and jute fabrics were cut into small
uniform rectangles of desires sizes. These mats and other chopped ﬁbers were dried in
the vacuum oven at the temperature of 80 OC and vacuum pressure of -30 in Hg (-102
kPa) for ﬁve hours before any treatment (see Figure 3.2.1). The pre treatment and post
treatment steps of these surface treatments are tabulated in Table 3.2.1. All surface

treated bioﬁbers were characterized and used for composite fabrication.

113

      

.. ”4.434 M ‘8'“.

Figure 3.2.1: Vacuum oven for drying bioﬁbers

3.2.1 Alkali treatment

A 5% solution of Sodium Hydroxide (NaOH) was made using pellets of NaOH in
deionised water (DI) water using a magnetic stirrer. The ﬁbers were placed in this
solution for 1 hour at room temperature. Then the ﬁbers were washed with tap water until
all alkali was removed from their surface. The ﬁbers were then neutralized in tap water
with 2 % glacial acetic acid, and tested for neutralization using pH paper to maintain a
pH of 6. This solution was then drained and the ﬁbers were washed with DI water and

then stored overnight under the hood followed by vacuum drying.

3.2.2 Acrylonitrile treatment
With DCP as initiator: The required amount of ﬁbers was vacuum dried prior to sizing.
The dried ﬁbers were soaked in a solution containing 3.0% acrylonitrile (AN), 0.5%

dicumyl peroxide and 96.5 % ethanol for 15 minutes.

114

With MEKP as initiator: The ﬁbers were soaked in a premixed solution containing 3.0%

acrylonitrile, 0.5% methyl ethyl ketone peroxide and 96.5% ethanol for 15 minutes.

3.2.3 Silane treatment

A 1 % gamma methacryloxy propyl trimethoxy silane (y-MPS) solution was made in
99% DI water and ethanol (1:1) , maintaining the pH of the solution by using 2 % glacial
acetic acid .The solution was continuously stirred for two hours in a closed lid plastic
container and kept under the hood. The vacuum dried ﬁber was soaked in this solution,
under the hood, for 1 hour. The solution was then drained from the ﬁbers, and kept for
curing in an air oven for 5 hours. Fresh solution was made each time this surface

treatment was used.

3.2.4 Steam treatment
Fibers were placed in a pressure explosion vessel (PARR) containing hot water, and
pressure (90-120 psi) was applied. The ﬁbers were thus steam treated (160-170 0C, 90-

120 psi) for 15 min.

3.2.5 Unsaturated polyester resin treatment

For unsaturated polyester resin (UPE) treatment, ﬁbers were soaked in a premixed
solution containing 5% of UPE and MEKP in acetone (UPE: MEKP = 99:1) for 30

minutes.

3.2.6 Maleic anhydride treatment

115

The ﬁbers were soaked in a premixed solution containing 3.0% maleic anhydride, 0.5%

dicumyl peroxide and 96.5% ethanol for 15 minutes.

3.2.7 Methyl methacrylate treatment
Natural ﬁbers were then soaked in a premixed solution containing 3.0% methyl

methacrylate, 0.5% dicumyl peroxide and 96.5% ethanol for 15 minutes.

3.2.8 Epoxy treatment

With Jeﬂamine: A 5% solution of epoxy resin and polyoxypropylene triamine
(stoichiometric amounts of epoxy resin and jeffamine were used) was made in acetone
with constant stirring using a magnetic stirrer. The ﬁber mats were placed in this solution
at room temperature for 1 hour.

With MPDA: A 5% solution of epoxy resin and meta- phenylene diamine was made
(stoichiometric amounts of epoxy resin and MPDA were used) in acetone with constant
stirring using a magnetic stirrer. The ﬁber mats were placed in this solution at room

temperature for 1 hour.

3.2.9 Glycidyl methacrylate treatment

The ﬁbers were soaked in a premixed solution containing 3.0% glycidyl methacrylate,

0.5% dicumyl peroxide and 96.5% ethanol for 60 minutes.

Hemp mats 1, and 2, and kenaf ﬁbers were treated with three different concentrations of

silane solution to ﬁnd an optimum silane treatment. In the case of all ﬁbers, 1% solution

116

of silane (1 % vol silane in 99 % vol solution of ethanol and DI water (50:50)) resulted in
highest mechanical properties compared to the other concentrations of silane solutions
(0.5 %, 1 %, and 2 %).

Table 3.2.1: Pre-treatment and post treatment for surface treatments of the natural ﬁbers

 

 

 

 

 

 

 

 

 

 

Fiber Surface Treatment
Name Preparation After Post
treatment- treatment-
Stepl Step 2
Non-woven hemp Alkali Dry the ﬁbers in Drain and dry Vacuum dry
mat-1, non-woven (NaOH) vacuum oven for for a day at ~80 0C,
hemp mat—2, 5 hours under hood. and -102
kenaf, kPa for 5
hours
Non-woven hemp Acrylonitrile Dry the ﬁbers in Drain and dry Vacuum dry
mat-l, non-woven (AN)-DCP vacuum oven for for a day at ~80 OC,
hemp mat-2, 5 hours under hood. and -102
kenaf, kPa for 5
hours
Non-woven hemp Acrylonitrile Dry the ﬁbers in Drain and dry Vacuum dry
mat-1, non-woven (AN)-MEKP vacuum oven for for a day
hemp mat-2, 5 hours under hood.
kenaf,
Non-woven hemp Silane(MPS) Dry the ﬁbers in Drain and dry Vacuum dry
mat-l, non—woven vacuum oven for for a day
hemp mat-2, 5 hours under hood,
kenaf, BBSG, followed by
green ﬂax core curing in air
oven.
Non-woven hemp Silane(APS) Dry the ﬁbers in Drain and dry Vacuum dry
mat-1, non-woven vacuum oven for for a day
hemp mat-2, 5 hours under hood,
kenaf, followed by
curing in air
oven.
Non-woven hemp Steam Dry the ﬁbers in Drain and dry Vacuum dry
mat-1, kenaf, vacuum oven for for a day
5 hours under hood.
Non-woven hemp UPE-MEKP Dry the ﬁbers in Drain and dry Vacuum dry
mat-l, non-woven vacuum oven for a day
hemp mat-2, under hood.
kenaf,

 

 

 

 

 

117

 

Table 3.2. 1 (continued)

 

 

 

 

 

 

Non-woven hemp Maleic Dry the ﬁbers in Drain and dry Vacuum dry
mat-l anhydride vacuum oven for for a day

5 hours under hood.
Non-woven hemp Methyl Dry the ﬁbers in Drain and dry Vacuum dry
mat-1 methacrylate vacuum oven for for a day

5 hours under hood.
Non-woven hemp Epoxy- Dry the ﬁbers in Drain and dry Vacuum dry
mat-1 Jeffamine vacuum oven for for a day

5 hours under hood.
Non-woven hemp Epoxy- Dry the ﬁbers in Drain and dry Vacuum dry
mat-1 MPDA vacuum oven for for a day

5 hours under hood.
Non-woven hemp Glycidyl Dry the ﬁbers in Drain and dry Vacuum dry
mat-1 methacrylate vacuum oven for for a day

5 hours under hood

 

 

 

 

 

 

3.3 Bioresin

Two approached were used for modifying vegetable oils to introduce reactive functional
groups in the unsaturated inert backbones of oils like soybean oil, castor oil. In the ﬁrst
approach, a reactive monomer (here acrylonitrile) was grafted on to the oils by using a
catalyst. In the second method, soybean polyols were reacted with a reactive monomer,
(maleic anhydride), to produce a UPE compatible resin.

3.3.1 Grafting OfMSO, Reﬁned Soybean Oil and Castor Oil

The vegetable oils were grafted with a reactive monomer in an effort to introduce
polymerizable groups on the backbone of natural oils. Acrylonitrile was selected as the
monomer, and benzoyl peroxide was selected as the initiator. The grafting was done in
the following manner:

A three-necked round-bottomed 500 m1 ﬂask was kept in the silicone oil bath. The oil

bath was placed on a hot plate/stirrer (Coming), and heated by an electric temperature

118

controller (Coming). The temperature was ﬁrst set at 100 0C, and required amount of
benzyl peroxide (BPO) with adequate vegetable oil was added to the ﬂask. This solution
was continuously mixed with the help of a magnetic stir bar. The mixture was heated for
15 minutes at 100 0C. It is believed that BPO degrades into free radicals around this
temperature. The set point was then changed to 70 0C (the boiling point of acrylonitrile is
77 0C). The ﬂask was taken out of the oil bath. When the oil bath reached the desired
temperature, required amount of acrylonitrile was added to it. The ﬂask was then lowered
into the oil bath. This mixture was heated for 4 hours at 70 0C. At the end of four hours,
the solution was removed from the ﬂask, weighed and stored in a container. The grafting
reaction was quantiﬁed by using FTIR. The various compositions, parameters, and results
of the grafted samples are tabulated in Table 3.3.1.

Table 3.3.1: Bioresins from grafting of vegetable oils

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AN 011
BPO content content content Oil Time
Sample (wt %) (wt %) (wt %) type Viscosity (hrs)
ANMSOBPl 0.1 5 94.9 MSO no change 2
little
ANMSOBP2 0.5 10 89.5 MSO increase 2
ANMSOBP3 0.5 10 89.5 MSO viscous 3
ANMSOBP4 0.1 5 94.9 MSO no change 4
ANMSOBPS 0.2 5 94.8 MSO increased 4
ANMSOBPWl 1 20 79 MSO solidiﬁed 1
very
ANMSOBPW2 0.5 10 89.5 MSO viscous 1
ANCASBPI 0.2 5 94.8 CAS no chage 4
AN SOYBPl 0.5 5 94.5 SOY no change 4
ANSOYBP2 0.2 5 94.8 SOY no change 4

 

*MSO=Methyl ester of soybean oil, CAS: Castor Oil, SOY=Reﬁned soybean oil

Development of Bioplastics

119

Bioplastics are plastic samples made from a bioresin, with or without a synthetic resin,
without any reinforcement. The following plastics were made with grafted oils, UPE,
MEKP, CoNap, and polybutadiene adducted maleic anhydride (PBMA). PBMA is a
liquid rubber, and was added to the bioresin to bring about an ease in phase separation
during curing.

Table 3.3.2: Bioplastics obtained from grafted oils

 

 

 

 

 

 

 

 

 

 

 

 

 

Grafted oil
Plastic content Grafted oil UPE PBMA MEKP CoNap
name (wt %) used (wt Vii (wt %) (wt %) (wt °/o)
GFTl 30 ANMSOBPI 70 O l 0.03
GFT2A 30 ANMSOBP2 70 0 1 0.03
GFT 28 20 ANMSOBP2 80 0 1 0.03
GFT 2C 20 ANMSOBP2 78 2 1.5 0.03
GFT 4A 20 ANMSOBP4 8O 0 1 0.03
GFT 4B 20 ANMSOBP4 78 2 1.5 0.03
GFT 13 20 ANSOYBPl 80 0 1 0.03
GFT 1A 20 ANSOYBP2 78 2 1.5 0.03
GFT 5B 20 ANMSOBPS 80 O l 0.03
GFT 5A 20 ANMSOBPS 75 5 1.5 0.03
MSO- None (pure
UPE 20 MSO) 75 5 1.5 0.03

 

 

 

 

 

 

 

 

Bioplastic samples were also made with MSO, MEESO, ESO, EML, and ELO.
Bioplastics were made by blending the UPE and the functionalized oils and curing them
in silicone molds or compression molding. Biocomposites were also made with bioﬁbers
and blends of UPE and functionalized oils as matrix by compression molding. The
bioplastic samples were characterized by FTIR, DSC, TGA, mechanical and

morphological analyses.

3.3.2 Grafting Soybean Oil Phosphate Ester Polyol (SOPEP) with maleic anhydride

120

The soybean oil phosphate ester polyol (SOPEP) was used as the starting material. The
OH content of this polyol is 122 mg of KOH. The polyol was reacted with maleic
anhydride in presence of a catalyst, N, N-Dimethylbenzylamine. The solution containing
polyol, maleic anhydride and catalyst were heated for some time.

It is expected that during this reaction maleic anhydride will react with OH groups of the
polyol, and will result in double bonds in the backbone of the oil. These double bonds can
easily be polymerized during curing of the resin.

Three different bioresin products were obtained by the following reactions:

a) OF T SOPEP] : 1 mole of pure SOPEP, 0.9 moles of pure maleic anhydride powder and
N,N—Dimethylbenzylamine as catalyst (5% of total weight) were reacted together in a
round bottomed three necked ﬂask at 80 0C for 4 hours. The resulting modiﬁed polyol
was brown in color and very viscous. The starting material was orange in color and less
viscous.

b) GFTSOPEP2: 1.0 mole soy polyol + 1.0 mole maleic anhydride + 5.0% by weight
catalyst

c) GFTSOPEP3: 1.0 mole soy polyol + 0.5 mole maleic anhydride + 5.0% by weight
catalyst

As a result of these experiments, the grafted polyol was optimized. The grafted polyols
were kept in a vacuum oven for 24 hours at a pressure of -103 kPa to get rid of any
unreacted maleic anhydride. There was no change in the FTIR spectra of the grated
polyol before and after being exposed to vacuum for 24 hours. This indicated that all
maleic anhydride in the reaction vessel had been used.

Development of Bioplastics

121

Bioplastics comprising of different weight fraction of the grafted soy polyols and UPE
resin were made. All plastics were made in small DMA coupons using conventional oven
curing. All the samples were cured at standard curing conditions. The samples having
grafted oil 30% or greater had to be post cured at 150 0C for 2 hours.

Compositions made were:

a) 50% Grafted Soy Polyol (GFTSOPEPl, GFTSOPEP2, or GFTSOPEP3) + 50% UPE
+ 3.0% MEKP + 0 .03% CoNp

b) 40% Grafted Soy Polyol (GFTSOPEPI, GFTSOPEP2, or GFTSOPEP3) + 60% UPE
+ 3.0% MEKP + 0.03% CoNp

c) 30% Grafted Soy Polyol(GFTSOPEP1, GFTSOPEP2, or GFTSOPEP3) + 70% UPE
+ 3.0% MEKP + 0.03% CoN

d) 20% Grafted Soy Polyol(GFTSOPEP1, GFTSOPEP2, or GFTSOPEP3) + 80% UPE
+ 3.0% MEKP + 0.03% CoNp

e) 10% Grafted Soy Polyol (GFTSOPEPl, GFTSOPEP2, or GFTSOPEP3) + 90% UPE
+ 3.0% MEKP + 0.03% CoNp

The samples made with 30%, 40% and 50% of GFTSOPEP2 had inferior properties.
Biocomposites were also made with bioﬁbers and blends of UPE and functionalized oils
as matrix by compression molding. The pure polyol was characterized using DSC, TGA
and FTR. The pure maleic anhydride and modiﬁed polyol were also scanned under
FT IR. The bioplastic samples were characterized by FTIR, DSC, TGA, mechanical and

morphological analyses.

3.4 Fabrication of composites

122

Composites and plastics were made by either conventional curing in an oven,
compression molding, sheet molding compounding, or vacuum assisted resin transfer
molding.

3.4.1 Silicone Molds

Aligned Fibers: For aligned ﬁber composites, combed unidirectional kenaf ﬁbers were
used. The biocomposites were made in coupons. In this study, composites with 30 vol. %
ﬁbers content were used for making the biocomposites. The fabrication of the composite
was done in the following way. The UPE was mixed well with the initiator MEKP and
the promoter CoNap in required amounts and then degassed at room temperature for 2
minutes. The ﬁbers were laid out on silicone molds after weighing. The resin was poured
under and over them; another mold was kept over this mold leaving enough space for air
bubbles to move out. Some heavy weights (3-5 lbs) were put over it, and it was oven
cured at 100 0C for 2hours, followed by 150 0C for 2 hours.

UPE Control: For making the control panel from neat resin, degassed UPE solution was
poured over degassed silicone molds and cured in a conventional oven using the same
temperature proﬁle.

Bioplastics: Derivitized vegetable oils were mechanically blended with UPE in the ratio
of 10%, 20%, 30%, 40%, or 50% functionalized oil, and 90%, 80%, 70%, 60%, or 50%
UPE, respectively by weight. Bioplastics were made by blending the UPE, the
functionalized oils and the curing agents (MEKP and CoNap), followed by curing them
in silicone molds in a conventional oven with the standard curing conditions. Five percent

by weight of PBMA was added to one of the blends of MESO and UPE. The resulting

123

plastics were cut into required shapes for various tests. Table 3.4.1 shows different

plastics and composites made using silicone molds.

 

Figure 3.4.1: Conventional oven for curing silicone molds

Table 3.4.1: Samples made using silicone molds

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Composition
Fiber UPE Vegetable Initiator Promoter
Oil

UPE Control 0 99.7 % wt 0 1.0 % 0.03% wt

wt

Aligned ﬁber

Water wasth kenaf- 30 % 70 % vol 0 1.0 % 0.03% wt

UPE vol wt
Alkali treated kenaf- 30 % 70 % vol 0 1.0 % 0.03% wt

UPE vol wt
AN-DCP treated kenaf— 30 % 70 % vol 0 1.0 % 0.03% wt

UPE vol wt
UPE-MEKP treated 30 % 70 % vol 0 1.0 % 0.03% wt

kenaf-UPE vol wt
Silane treated kenaf— 30 % 70 % vol 0 1.0 % 0.03% wt

UPE vol wt
Untreated kenaf- 30 % 50 % vol 20 % vol 1.5 % 0.03% wt

Bioresin-UPE vol wt
E-glass-UPE 30 % 70 % vol 0 1.0 % 0.03% wt

vol wt

Bioplastics“

 

 

 

 

 

 

124

Table 3.4.1 (continued)

 

 

 

 

 

UPE-MSO 90,80 10, 20 1.5 % 0.03% wt
wt
UPE-MSO-MA 90,80 10,20 1.5 % 0.03% wt
wt
80 20 (5% 1.5 % 0.03% wt
MEESO—PBMA PBMA) wt
80 20 1.5 % 0.03% wt
ESO-UPE wt
80 20 1.5 % 0.03% wt
EML-UPE wt

 

 

 

 

 

 

 

 

b.)

*Graﬁed bioresins-UPE have already been discussed in section 3. .
3.4.2 Compression Molding

For Mats: The required amounts of ﬁber mats were vacuum dried for 5 hours. The
polyester resin (UPE) was mixed well with MEKP and CoNap, in required amounts and
then degassed under vacuum at room temperature for 2 minutes. The ﬁber mats were
individually coated with the degassed resin. They were then placed between two
aluminum plates covered with teﬂon release sheets. The plates were placed in a
compression molding press (Carver ® Laboratory Press) and the composites were cured
at 80 psi for 2 hours at 100° C followed by 2 hours at 150° C.

For chopped ﬁbers: The polyester resin (UPE) was mixed well with the initiator, methyl
ethyl ketone peroxide (MEKP) and the accelerator, cobalt naphthenate (CoNp), in
required amounts and then degassed under vacuum at room temperature for 2 minutes.
The ﬁbers were mixed nicely with the degassed resin by hand. Uniform resin distribution
in the entire ﬁber mass was ensured. This ﬁber-resin mixture was then placed in a picture
frame mold which was kept between two aluminum plates covered with teﬂon release

sheets. The plates were placed in a compression molding press (Carver ® Laboratory

125

Press) and composites were cured at 80 psi for 2 hours at 100° C followed by 2 hours at
150° C. The biocomposite using untreated ﬁber mats was made in same way. The
resulting composite plaques were cut into required shapes for various tests. Bioplastics:
Derivitized vegetable oils were mechanically blended with UPE in the ratio of 10%, 20%,
30%, 40%, or 50% functionalized oil, and 90%, 80%, 70%, 60%, or 50% UPE,
respectively by weight. Bioplastics were made by blending the UPE, the ﬁmctionalized
oils and the curing agents (MEKP and CoNap), followed by curing them in silicone
molds in a conventional oven with the stande curing conditions. The resulting plastics
were cut into required shapes for various tests. Figure 3.4.2 shows a four platen Carver

lab press. The samples made with compression molding are described in Table 3.4.2 and

some are shown in Figure 3.4.3.

  

Figure 3.4.2: Carver compresion Press

126

Table 3.4.2: Samples made using compression molding

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sample Composition
Fiber UPE Natural Initiator Promoter
Oil (wt %) (wt %)
Non woven mats
Untreated hempmatl-UPE 30 % vol 70 % vol 0 1.0 % 0.03%
Alkali treated hempmatl- 30 % vol 70 % vol 0 1.0 % 0.03%
UPE
AN-DCP treated 30 % vol 70 % vol 0 1.0 % 0.03%
hempmatl -UPE
UPE-MEKP treated 30 % vol 70 % vol 0 1.0 % 0.03%
hempmatl-UPE
Silane treated hempmatl- 30 % vol 70 % vol 0 1.0 % 0.03%
UPE
Untreated hempmatl- 30 % vol 65 % vol 0 1.0 % 0.03%
Acrylonitrile(5%) in UPE 5% AN
MA treated hempmatl-UPE 30 % vol 70 % vol 0 1.0 % 0.03%
MMA treated hempmatl- 30 % vol 70 % vol 0 1.0 % 0.03%
UPE
EPON-MPDA treated 30 % vol 70 % vol 0 1.0 % 0.03%
hempmatl -UPE
EPON-T403 treated 30 % vol 70 % vol 0 1.0 % 0.03%
hempmatl -UPE
GMA treated hempmatl- 30 % vol 70 % vol 0 1.0 % 0.03%
UPE
Steam treated hempmatl- 30 % vol 70 % vol 0 1.0 % 0.03%
UPE
Untreated hempmatl- 30 % vol 50 % vol 20 % 2.0 % 0.03%
Bioresin-UPE
Untreated hempmatl- 30 % vol 40 % vol 30 % 2.0 % 0.03%
Bioresin-UPE
Untreated hempmatl-E- 30 % vol 70 % vol 0 1.0 % 0.03%
glass mat-UPE
Pure hemp mat-(hempmaIZ) I
Untreated pure hempmat2— 30 % vol 70 % vol 0 1.0 % 0.03%
UPE
Alkali treated pure 30 % vol 70 % vol 0 1.0 % 0.03%
hempmatZ-UPE
AN-DCP treated pure 30 % vol 70 % vol 0 1.0 % 0.03%
hempmatZ-UPE
UPE-MEKP treated pure 30 % vol 70 % vol 0 1.0 % 0.03%
hempmatZ-UPE
Silane treated pure 30 % vol 70 % vol 0 1.0 % 0.03%
hempmatZ-UPE

 

 

 

 

 

 

127

 

Table 3.4.2 (continued)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Untreated hempmat2- 30 % vol 50 % vol 20 % 2.0 % 0.03%
Bioresin-UPE
Untreated hempmat2- 3O % vol 40 % vol 30 % 2.0 % 0.03%
Bioresin-UPE
Untreated hempmatl-E- 3O % vol 70 % vol 0 1.0 % 0.03%
glass mat-UPE
E-glass mat-UPE 30 % wt 70 % wt 0 1.0 % 0.03%
E-glass mat-Bioresin-UPE 30 % wt 50 % wt 20 % 2.0 % 0.03%
E-glass mat-Bioresin-UPE 30 % wt 40 % wt 30 % 2.0 % 0.03%
Chopped ﬁbers
Hemp-UPE 35 % wt 65 % wt 0 1.0 % 0.03%
Kenaf-UPE 35 % wt 65 % wt 0 1.0 % 0.03%
Henequen-UPE 35 % wt 65 % wt 0 1.0 % 0.03%
BBSG-UPE 35 % wt 65 % wt 0 1.0 % 0.03%
Green ﬂax core-UPE 35 % wt 65 % wt 0 1.0 % 0.03%
Coir-UPE 35 % wt 65 % wt 0 1.0 % 0.03%
Indian grass-UPE 35 % wt 65 % wt 0 1.0 % 0.03%
E-glass-UPE 30 % wt 70 % wt 0 1.0 % 0.03%

 

*Bioplastics have already been discussed in section 3.3

128

 

 

Figure 3.4.3: Picture of bioplastic samples made using compression molding

129

(a)= Henequen-UPE, (b)=Kenaf-UPE , (c)=Green hemp-UPE , (d)=Coir-UPE ,(e)=E-
Glass-Bioresin-UPE, (f)=E-Glass-UPE-CaCO3, (g)= Hempmatl (Bastrnat 115)-E-Glass-
UPE, (h)= Hempmat1(pure hemp mat)-E-Glass-UPE

3.4.3 Sheet molding compound panel processing

A new process for large-scale production of biocomposites containing thermoset resins
and natural ﬁbers was designed. The equipment used for this processing is the common
industrial SMC line. Currently, natural ﬁbers reinforced thermoset composites are not
prepared using SMC. We aim to use SMC process to fabricate these natural ﬁber-
thermoset composites. In commercial SMC set-up, continuous glass ﬁbers rovings are fed
to a chopper, which cuts them to a 6 mm size, and they fall on the carrier ﬁlm, forming a
uniform layer of chopped glass ﬁbers. Since, natural ﬁbers cannot be obtained in a
continuous from, and making a continuous yarn or roving with these ﬁbers would be a
difﬁcult and expensive, chopped natural ﬁbers were used in this new process. The
chopped natural ﬁbers are dried for 3-4 hours to remove any ambient moisture, and then
fed to a calibrated vibratory feeder. The ﬁbers are ﬁrst fed into a ‘screw feeder, which
delivers these ﬁbers to the vibratory feeder; from which they fall onto the carrier ﬁlm at a
constant ﬂow rate.

The resin formulation was prepared in a Ross mixture. The Ross mixer was connected to
cooling water. The polyester and calcium carbonate were combined and mixed for 10
minutes. The Cobalt Naphthanate was added and mixed for 5 minutes. Finally the
MEKP was added and mixed for 5 minutes. The Ross mixing pot-cooling water was shut
off and disconnected. Resin was transferred into a 3.5 gal epoxy coated steel storage

container. The Ross mixer was then cleaned several times with acetone.

130

The twin-screw feeder and vibratory feeder were turned on. The SMC line was turned on
shortly before the upper and lower resin pots were ﬁlled and was run at an optimum
speed. The process was allowed to run to see what the product looked like after the
compaction rollers. The product appears uniform in thickness and the ﬁber distribution
in the product appears even. Several feet of product are produced at these settings.

The ﬁnished product was cut into 1 ft long segments and was placed ﬂat in sealed
polyethylene container. This container was labeled with the date and time and is
immediately put into a freezer for storage. Gel time of the resin formulation was
observed to be approximately 2 hr 30 min.

Various process parameters, for example, line speed, cutter pressure, idler roll pressure,
wipe blade gap height upper, wipe blade gap height lower, compaction roller pressure,
compaction roller gap, grip assembly, twin screw feeder, etc. were adjusted and
optimized for natural ﬁbers. The prepregs from SMC line were matured for desired time
and then compression molded. The molded samples were tested for various mechanical,
and thermal properties, in accordance with ASTM procedures.

Batch to batch repeatability was established after running numerous experiments with
glass ﬁbers and base resin, and adjusting all of the parameters.

The new SMC processing for natural ﬁbers can be done on normal commercial SMC
equipment with a minor change on the way ﬁbers are fed to the line. In industrial SMC
set-ups, the glass ﬁbers rovings are fed to a chopper, and chopped ﬁbers fall onto the
carrier ﬁhn. But the natural ﬁbers are not supplied as continous rovings or yarns, and
must be fed to the ﬁlm in a chopped form. Therefore, in the new set —up, the chopped

ﬁbers fall from a calibrated vibratory feeder onto the carrier ﬁlm, and get drenched in

131

resin ﬂowing from two resin pots, and advance to the compression rollers, and ﬁnished
product is obtained at the end of the line. The prepreg from SMC line is reﬁ-igerated for
certain time to reach desirable gelation. The gelled product is then compression molded
in variety of molds to get the desired shape.

SMC Line description

Figure 3.4.4 shows the schematic of the proposed invention. The process begins with an

upper and lower supply roll of carrier ﬁlm material. The upper and lower carrier ﬁlm is

SCHEMATIC OF CONTINUOUS BIOCOMPOSITE SHEET
MOLDING COMPOUND PANE(BCSMCP) MANUFACTURE

  
  
 
 
       
       
      

Feed Hopper RandomlyPROCESS Upper Carrier
Twin Screw Oriented heet Supply R0"
Feeder Fiber Being
FDC Deposited 2 Upper Final

   
  

On Top Of arrier Sheet

V Resin Upper Sheet , , Material
Lower .- :1 ‘_ . -sin resin Friction Cut To
Carrier i": ' 'ot Compaction Grip Le gth

Rollers hee

Figure 3.4.4: Scheme of BCSMCP process
fed beneath the upper and lower resin pots respectively, which deposits a pre-determined

thickness of resin onto the carrier ﬁlm by means of adjustable height wipe blades. The

lower carrier sheet then passes beneath the bio-ﬁber vibratory feeder, which deposits 3

132

calibrated weight % of bio-ﬁber uniformly onto the surface of the resin. The upper and
lower carrier sheets are then brought together creating a sandwiched SMC sheet material.
The sheet material is then fed through compaction rollers to help provide a uniform sheet
cross section and to also to provide some mixing forces to ensure the even distribution of
ﬁber in the resin. Grip wheels near the end of the line provide the necessary mechanical
force to pull the sheet material through the system at a controlled rate. The sheet material
is then cut to the desired length.
Feeder Description
The bioﬁber feeder system consists of four main components (see Figure 3.4.5):

1.) The material feed hopper

2.) The twin screw feeder

3.) The ﬁber distribution feed chute

4.) The vibratory feeder
Bioﬁber is put in the hopper for feeding. The ﬁber is then fed onto the ﬁber distribution
feed chute by the twin-screw feeder, which is calibrated to output the required ﬁber
weight/minute. The ﬁber distribution chute spreads the ﬁber out uniformly from the
width of the twin feeding screws to the width of the vibratory feeder. The vibratory
feeder then conveys the ﬁber to the opening at the end of the feeder tray providing the
uniform distribution of ﬁber onto the SMC sheet.
This new processing has been used to make various biocomposite samples with different
composition and weight fraction of natural ﬁbers and polyester resin. The process has
been optimized for natural ﬁbers after many experimentations and repetitions. We have

already established this process with glass ﬁbers, polyester sized glass ﬁbers, big blue

133

stem grass ﬁbers, silane sized big bluestem grass ﬁbers and hemp ﬁbers. So far, we have
used unsaturated polyester resin, calcium carbonate, methyl ethyl ketone peroxide and
cobalt naphthenate in the matrix for these composites. We have achieved very promising

results from these experiments.

BIOFIBER FEEDER SCHEMATIC

Material

 
 
   
 
     
     
     

Twin Screw Feeder

Suppon
Distribution
Feed Chute
Vibratory Feeder

  
 
   
   

Adjustable Vibratory
Height Lift Feeder

Table Drive
Motors

Figure 3.4.5: Scheme of vibratory feeder and screw feeder

 

 

Experiments with Glass ﬁbers:

Experiments were done with glass ﬁbers to ensure that the composites developed using
SMC process had consistent properties. An unsaturated polyester resin compatible glass
roving, called Owens Corning 985, was used for making composites. Other glass ﬁbers
like Owens Corning 485, and other sized glass ﬁbers were not compatible with polyester
resin. The major goal of the run was to produce .188 thick SMC using Owens Corning

985 E-glass and then to compression mold the resultant sheet material. The resulting

134

glass composites were of the following composition: glass ﬁbers 30 wt %, CaCO3 42 wt
%, and UPE 28 wt %

The runs were conducted using the following parameters:

Line speed - Dial setting of 1.5/ .3 in/sec

Glass cutter - Dial setting of 3.2/approx. 223 g/min (31% by weight)
Four tows were fed into the glass cutter to improve glass distribution
Cutter Pressure — 40 psi

Idler roll pressure — 56 psi

Wipe blade gap height upper - .060”

Wipe blade gap height lower - .060”

Compaction roller pressure — 30 psi

Compaction rollers shimmed to ensure a minimum .250 gap

Upper carrier sheet roller shimmed to a gap of .265

Upper resin tray assembly shimmed up .500

Grip assembly used with slight preload on roller for friction
Procedure

A total of 8971.3 g of resin was mixed shortly before the SMC run using the following
formulation:

40% Unsaturated Polyester Resin - 3580.92 g

60% Calcium Carbonate Filler - 5371.38 g

.03% Cobalt Naphthanate Promoter - 1.074 g

.5% MEKP - 17.90 g

135

The Ross mixer speed dial was set to 5.0. Cooling water to the mixing pot was connected
and turned on. The polyester and calcium carbonate were combined and mixed for 10
minutes. The Cobalt Naphthanate was added and mixed for 5 minutes. Finally the
MEKP was added and mixed for 5 minutes.

The Ross mixing pot cooling water was shut off and disconnected. Resin was transferred
into a 3.5 gal epoxy coated steel storage container. The Ross mixer was then cleaned
several times with acetone using blue WypAll Plus all-purpose wipes.

The SMC line was turned on shortly before the upper and lower resin pots were ﬁlled and
was run at a speed of .3 in/sec. The glass cutter was then turned on. The process was
allowed to run to see what the product looked like after the compaction rollers. The
product appeared uniform in thickness and the glass distribution in the product appeared
even. Several feet of product were produced at these settings.

From the point where the MEKP was introduced into the resin, a total time of
approximately 45 minutes elapsed including cleanup time. The ﬁnished product was cut
into 1 ft long segments and was placed ﬂat in sealed polyethylene container. This
container was labeled with the date and time and was immediately put into a freezer for
storage. Gel time of the resin formulation was observed to be approximately 2 hr 30 min.
After the prepreg was properly frozen, it was compression molded in the Carver Press.
The prepregs was placed in a picture frame mold, and placed between two aluminum
plates, lined with Teﬂon sheets. This sample was cured at around 80 psi for 2 hours at
100° C followed by 2 hours at 150° C. The resulting composite is cut into desired shapes

for various tests.

136

For making the control sample, a degassed UPE solution is poured into degassed silicone
molds and cured in a conventional oven at the same curing conditions.

Glass ﬁber composites were fabricated using SMC until consistent materials with same
mechanical properties were obtained. The parameters mentioned for this particular run
were obtained after optimization. Data from two SMC runs with glass ﬁbers, after all
optimizations are discussed in the upcoming section. Biocomposites with natural ﬁbers
were fabricated after success with glass composites.

Experiments with natural ﬁbers:

Experiments were done to produce biocomposites using the SMC line. The major goal of
these runs was to produce sheet material using natural ﬁbers as the reinforcement.

The natural ﬁbers were vacuum dried overnight at 80 ° C and -102 kPa, before being used
in the biocomposite fabrication. Some composites were made with untreated ﬁbers as
such, and some were made with chemically treated ﬁbers. In particular, silane treatment
was used on big blue stem grass and green ﬂax core. The silane treatment of these ﬁbers
was done in the following way:

A 1 % methacryloxypropyltrimethoxy silane (y-MPS) solution was made in 99% DI
water and Ethanol (1:1) , maintaining the pH of the solution at 4.0 by using 2 % Glacial
Acetic Acid .The silane solution was hydrolyzed for 2 hours by continuously stirring
using a magnetic stir bar. The ﬁbers were soaked in this solution for 1 hour .The solution
was then drained from the ﬁbers, and theyr were dried under the hood for at least one
day. They were then cured in an air oven for 5 hours at 80 0C. The ﬁbers were then dried

in a vacuum oven at 80 °C and - 30 in Hg.

137

When two ﬁbers were used for making hybrid biocomposites, the ﬁbers were mixed well
by continuous agitation, and the resulting uniform mixture was vacuum dried before
using to make biocomposites.

The natural ﬁbers also had to be introduced in to the SMC line in a different way as
compared to glass ﬁbers. They were fed by a combination of twin screw feeder and
vibratory feeder. This way of ﬁber introduction was optimized after various attempts with
vibratory feeding.

The resulting biocomposites contained 20 wt % ﬁber, 30 wt % CaCO3, and 60 wt %
UPE. This composition is different from that of glass composites discussed earlier,
because, natural ﬁbers have a tendency to soak up resin. As the natural ﬁbers are hollow,
when they are mixed with resin, they absorb resin in their interstitial paces. An
experiment was done to measure the amount of resin absorbed by big blue stern ﬁber. It
was found that 1 gram of ﬁber absorbed 1.85 grams of resin (average of three readings).
The experimental runs were conducted using the following parameters:

0 Line speed - Dial setting of 1.5 (.3 in/sec)=(18 in/min)

o K—tron twin screw feeder — Hi gear with a digital readout setting of 550 RPM

0 Vibratory feeder dial setting — 100%

o Feeders calibrated to approximately 101 g/min. (20% ﬁber content) over the 12” wide
resin coverage area

0 A modiﬁed ﬁber spreading dispersion tray was used to spread ﬁbers to the vibratory
feeder 18” width

0 Wipe blade gap height upper - .050”

0 Wipe blade gap height lower - .050”

138

o Compaction roller pressure/Open Position — 30 psi

0 Compaction rollers shimmed to ensure a minimum .250 gap

0 Upper carrier sheet roller shimmed to a gap of .265

0 Upper resin tray assembly shimmed up .500

0 Grip assembly used with slight preload on roller for ﬁ'iction

0 Big Blue Stem Grass ﬁber and Flax ﬁber was dried for approximately 16 hrs in a
vacuum oven at approximately 80 °C before run.

Procedure

A total of 6560g of resin was mixed shortly before the SMC run using the following
formulation:

60% Unsaturated Polyester Resin - 5182.76 g

30% Calcium Carbonate Filler - 1350.00 g

.03% Cobalt Naphthanate Promoter - 1.56 g

.5% MEKP - 26.06 g

The Ross mixer speed dial was set to 5.0. Cooling water to the mixing pot was connected
and turned on. The polyester and calcium carbonate were combined and mixed for 10
minutes. The Cobalt Naphthanate was added and mixed for 5 minutes. Finally the
MEKP was added and mixed for 5 minutes.

The Ross mixing pot cooling water was shut off and disconnected. Resin was transferred
into a 3.5 gal epoxy coated steel storage container. The Ross mixer was then cleaned
several times with acetone using blue WypAll Plus all-purpose wipes.

The SMC line was turned on shortly before the upper and lower resin pots were ﬁlled and

was run at a speed of .3 in/sec. Resin was then transferred to the resin pots. The K-tron

139

twin screw feeder and vibratory feeder were then turned on. The process was allowed to
run to see what the product looked like after the compaction rollers. The sheet material
looked very good. The modiﬁed feeder setup distributed ﬁber evenly and consistently.
Several feet of product was produced at these settings.

The resultant sheet material looked very good. Fiber dispersion uniformity from the
modiﬁed dispersion tray setup was good, but could be better if the house air pressure
could steadily supply more air pressure. The pneumatic rotary vibratory feeder on the
dispersion tray was set to 75 psi, but the house air supply had trouble keeping up pressure
after a period of time. A drop in pressure was not observed until the run was complete.
90 psi provided the best ﬁber distribution but was not sustainable for than a minute or so.
The ﬁnished product was cut into 1 ft long segments and was placed ﬂat in a sealed
polyethylene container. This container was labeled with the date and time and was
immediately put into a freezer for storage. After cooling in the freezer for several hours,
the sheet material was compression molded in the Carver press for mechanical testing.
Figures 3.4.6, 3.4.7 and 3.4.8 show the various aspects of BCSMCP processing
technique. In Figure 3.4.6, the mixing of resin components can be seen. In Figure 3.4.7,
the entire SMC processing of biocomposites is highlighted. In Figure 3.4.8, the method
for natural ﬁber feeding to the SMC line is shown.

Figures 3.4.9, 3.4.10, 3.4.11 and 3.4.12 showcase different composite samples made

using SMC line.

140

 

c) After mixing initiator and promoter (1) All resin components mixed in the Ross mixer

Figure 3.4.6: Mixing of resin components in the Ross mixer '

The different biocomposites made using BCSMCP processing were of the following

compositions:

141

88w 8on 0:3 Em .I. Ommm...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

02 8 8 o S 8 8:88: 88:5 9:8 88:5 a
09 o 8 o E 8 8:88: 88:5 89. 88:5 0
o? o 8 o 9 8 8:88: 88:5 98: 88:5 2
09 o 8 o 8 83 88:5 s.
02 o 8 o 8 8:88: 88:5 _
o? o 8 o 8 8.8 88:5 9
02 8 8 o 8 :9: 88mm ..
02 o 8 8 8 89mm _
02 o 8 8 e N: :9: 893 a8: 88:5 I
o? o 8 8 a 2 as ea 88:5 a8: 88:5 o
2: o 8 8 .a. m: :9: 98 88:5 98: 88:5 n
02 o 8 8 8 98 xn: 89m 88:5 w
o? o 8 8 N: m 98 xn: 893M989: 888 8mm 88: 888 o
o? o 8 8 N: m 98 xn: 89m 88:5 88 88:5 0
02 o 8 8 8 08m 88: n88 m
02 o 8 8 8 88 88:5 <
.\. .25: 3:; £835 «a; ma: 3 es 808 .x. E N Be“. #5 e .8: w Ban :8: 29:8

 

 

0:: 02m mamm: 288 8383. "min 03.5.

142

 

b) Resin pot] and vibratory feeder

a) Resin poured in resin pot 1

 

d) Uniform sheet of ﬁbers

0) Fibers falling from feeder

 

f) Fibers being sandwiched between resin

6) Resin ﬂowing down from pot 2

143

 

i) Sheet coming from last of compaction rollers j) Sheet falling down the ﬂoor

 

k) Pre-preg after the SMC line 1) sheet being cut to 1 ft long pieces

 

m) Rectangular pieces of pre-preg n) pre-preg ready to be stored in freezer

Figure 3.4.7: BCSMCP process for natural ﬁber-polyester resin composites, a) to n)

 

Figure 3.4.8: Feeding natural ﬁbers though a vibratory feeder

145

The chopped natural ﬁbers were dried for 3-4 hours to remove any ambient moisture, and
then fed to a calibrated vibratory feeder. The ﬁbers were ﬁrst fed into a screw feeder,
which delivers these ﬁbers to the vibratory feeder; from which they fell onto the carrier
ﬁlm at a constant ﬂow rate.

The resin formulation was prepared in a Ross mixture. The Ross mixer was connected to
cooling water. The polyester and calcium carbonate were combined and mixed for 10
minutes. The Cobalt Naphthanate was added and mixed for 5 minutes. Finally the
MEKP was added and mixed for 5 minutes. The Ross mixing pot-cooling water was shut
off and disconnected. Resin was transferred into a 3.5 gallon epoxy coated steel storage
container. The Ross mixer was then cleaned several times with acetone.

The twin—screw feeder and vibratory feeder were turned on. The SMC line was turned on
shortly before the upper and lower resin pots were ﬁlled and was run at an optimum
speed. The process was allowed to run to see what the product looked like after the
compaction rollers. The product appeared uniform in thickness and the ﬁber distribution
in the product appeared even. Several feet of product were produced at these settings.
The ﬁnished product was cut into 1 feet long segments and was placed ﬂat in sealed
polyethylene container. This container was labeled with the date and time and was
immediately put into a freezer for storage. Gel time of the resin formulation was
observed to be approximately 2 hours and 30 minutes.

Various process parameters, for example, line speed, cutter pressure, idler roll pressure,
wipe blade gap height upper, wipe blade gap height lower, compaction roller pressure,
compaction roller gap, grip assembly, twin screw feeder, etc. were adjusted and

optimized for natural ﬁbers. The prepregs from SMC line were matured for desired time

146

and then compression molded. The molded samples were tested for various mechanical
and thermal properties, in accordance with ASTM procedures. Batch to batch
repeatability was established after running numerous experiments with glass ﬁbers and

base resin, and adjusting all of the parameters.

  

.‘u

Figure 3.4.9: An E-Glass-CaC03-UPE composite

147

    

’ .\. ‘

Untreated big blue stem grass

(BB S G) _ UPE bio-composite Untreated green ﬂax core-UPE b10-

composite

   

30 weight % % calcium carbonate Untreated BBSG —F1ax-CaCO3-UPE
(CaC03)-70 Wt% UPE Bio—composite

Figure 3.4.10: Composite samples made using SMC line

148

 

13 % Hemp 7% Jute mats
20%CaCO3 60%UPE 20%CaC03 60%UPE

 

18 % Hemp 7% Jute mats 20%CaCO3 20% Hemp 10% Henequen 20%
5 5%UPE Bioresin 50%UPE

Figure 3.4.11: Composite samples made using SMC line

149

   

30 volume % hempmatl- -20 volume % 30 volume % hempmatl- -30 volume %
bioresin (from modiﬁed polyol)— —50 volum bioresin (from modiﬁed polyol)- -40
% UPE volume % UPE-

 

Hybrid biocomposite of 25% Hemp
and 10% Henequen

Hybrid biocomposite of 25% Kenaf
and 10% Henequen

Figure 3.4.12: Composite samples made using SMC line

150

The biocomposite and composites samples made suing SMC were characterized by

mechanical, thermal, and morphological tests.

3.4.4 Biobeams and bioplates

The work on bioplates and biobeams was done in collaboration with Mario Quagliata,
and Dr. Rigoberto Burgueﬁo of the Department of Civil and Environmental Engineering.
More details on this part of the work can be obtained ﬁom Mario’s M.S. thesis,
“Development and Characterization of Biocomposite Cellular Beams and Plates for
Load-Bearing Components”, Masters Thesis, Michigan State University, East Lansing,

MI; 2003. A brief description of this work is presented in the literature review.

3.5 Testing

To evaluate thermal, mechanical, morphological, physical and chemical nature of the
systems under study, various kinds of tests are carried on. Some tests are speciﬁcally
done on ﬁbers, and some on composites and plastics. All tests methods comply with

relevant standards such as ASTM (American Society of Testing Methods) standards.

3.5.] Natural ﬁber

The ﬁbers were characterized using Fourier Transform Infra Red Spectroscopy (FTIR),
Thermogravimetric analysis (TGA), Differential Scanning Calorimetry (DSC),
Environmental Scanning Electron Microscopy (ESEM) and X-Ray Photo Spectroscopy

(XPS).

151

3.5.1.1 Thermogravimetric analysis (TGA)

The thermal behavior of the surface modiﬁed ﬁbers was measured using
thermogravirnetric analyzer (Hi-Res TGA 2950, TA instruments, USA). TGA measures
the mass change of a sample as a ﬁinction of temperature. The TGA measuring cell may
be used for making content determinations, or to characterize phenomena such as

evaporation and drying, decomposition, oxidation, and oxidative stability.

 

Figure 3.5.1: Hi-Res TGA 2950 (TA instruments) in CMSC lab

The data from TGA in is the form of plots showing variation of sample weight and
derivative weight with temperature. The ﬁbers to be analyzed were ﬁnely chopped ﬁbers
and then kept in an Aluminum pan which goes inside the ﬁrmace. It was programmed to
go from room temperature to 600 °C at the rate of 20 °C per minute. The weight of the
ﬁbers used for analysis ranged from 7-19 mg. Dry nitrogen gas was circulated within the

test cell at a ﬂow rate of 60 cc/min.

3.5.1.2 Fourier Transform Infra Red Spectroscopy (FTIR)
Fourier transform infra red spectroscopy (FTIR) was used for characterizing the
constituent functional groups in the ﬁber. The ﬁbers were ﬁnely chopped into small

particles. These chopped ﬁbers and potassium bromide (KBr) were placed in a mortar in

152

the ration of 1:5. The solids were ground to make a homogeneous mixture using a pestle.
Approximately 200 mg of mixture was equally distributed in a pellet press. A pressure of
8000 psi was applied to the press for 5 minutes. The pellet of KBr and ﬁber was removed

from the press and scanned in a Perkin Elmer spectrum 2000 FT-IR.

  

Figure 3.5.2: Perkin Elmer spectrum 2000 FT-IR in CMSC lab

3.5.1.3 X—Ray Photo Spectroscopy (XPS)

In X-ray photoelectron spectroscopy (XPS), also called electron spectroscopy for
chemical analysis (ESCA), X-rays excite photoelectrons, and the emitted electron signal
is plotted as a spectrum of binding energies. Differing chemical states resulting from
compound formation are reﬂected in the photoelectron peak positions and shapes.
Spectral information is collected from a depth of 2-20 atomic layers, depending on the
material studied. The XPS technique involves the bombardment of a sample surface with
X-rays and the measurement of the concomitant photoemitted electrons. The
photoemitted electrons have discrete kinetic energies that are characteristic of the
emitting atoms and their bonding states. XPS goes beyond elemental analysis to provide
chemical information. It can distinguish chemical arrangements such as silicon-to-silicon

bonds from silicon—to-oxygen bonds. The strength of XPS is its ability to identify

153

different chemical states. This ability is useful in a range of physical studies, for example,
oxidation/corrosion products, adsorbed species, and thin-ﬁlm growth processes.
Applications of XPS include:

1) Chemical-state analysis: Evaluates valence states, bonding environments, and the
molecular composition of surface layers.

2) Elemental analysis: Identiﬁes elements from lithium to uranium, with detection levels
down to 0.5 at. %.

3) Imaging: Uses raster seaming to produce images with a spatial resolution of 26 pm.
4) Depth proﬁling: By sputtering material from a surface, generates compositional depth
proﬁles for materials up to 1 pm thick

5) Thin-ﬁlms: Frequently used for the analysis of surfaces of thin-ﬁlm materials.
6) Polymers: Especially valuable for analyzing functional groups in polymers and other
organic materials. (Particularly useful in this regard when used as a complementary tool
with static SIMS analysis)

7) Catalysts: Evaluates the surface of catalysts to determine reactive species.
8) Other materials: Valuable for chemical-state analysis of materials ranging from metals

to insulators to semiconductors

154

 

Figure 3.5.3: PHI 5400 ESCA (XPS) system in CMSC lab

The X-ray photoelectron spectra (XPS) of the untreated and surface treated ﬁbers were
determined by Per Askeland (PHI 5400 ESCA system). A non-monochromatic Mg
source was used with a take-off angle of 45 degrees. Fiber samples were carefully cut and
afﬁxed to a stainless steel sample holder by a molybdenum mask. Relative atomic
concentrations were determined by measuring the area under an element’s spectral
envelope and applying the appropriate correction factor.

XPS is an information rich method. The surface to be analyzed is ﬁrst placed in a
vacuum environment and then irradiated with photons. For XPS, the photon source is in
the X—ray energy range. The atoms comprising surface emit electrons (photoelectrons)
after direct transfer of energy from the photon to the core-level electron.

These emitted electrons are subsequently separated according to energy and counted.

The energy of the photoelectron is related to the atomic and molecular enviromnent from

155

which they originated. The number of the electrons emitted is related to the
concentration of the emitting atom in the sample. The most basic XPS analysis of a
surface will provide qualitative information on all the elements present (except H and
He). More sophisticated application of the method such as curve ﬁtting can yield much
detailed information about the chemistry, organization, and morphology of a surface. The
probe depth is less than 10mm so this method is very surface sensitive and is a great tool
to characterize the functional groups on the cellulose ﬁber surface.

Chemical information indicating changes in the treated ﬁber surface was elucidated by
curve ﬁtting the carbon ls (Cls) and oxygen ls (Ols) spectra. The Cls and 013 curves
were ﬁtted with a Lorentzian-Gaussian mix Voigt proﬁle using a nonlinear least-squares
curve-ﬁtting program. The resulting curve ﬁt has a level of experimental error around 5-

10%. All peaks are referenced to adventitious carbon at 284.6 eV.

3.5.1.4 Differential Scanning Calorimetry (DSC)

Thermal scans of untreated and surface treated ﬁber samples were done using Differential
Scanning Calorimetry (TA 2920 Modulated DSC, TA instruments, USA). DSC is used
for ﬁnding the variation of heat with temperature, which helps to get glass transition
temperature (Tg), degradation temperature, crystallization temperature, melting
temperature, heat of reaction etc. Differential scanning calorimetry is a technique used to
study the thermal transitions of a polymer. Thermal transitions are the changes that take
place in a polymer when it is heated. The ﬁbers were ﬁnely chopped and then placed in
an aluminum pan and weighed. The sample pan along with a reference aluminum pan

was kept inside the fumace. The DSC fumace was programmed to go from —60 °C to 300

156

°C at 10 0C per minute. The data from DSC is in the form of plots of heat ﬂow with

respect to temperature.

 

Figure 3.5.4: TA 2920 Modulated DSC (TA instruments) in CMSC lab

3.5.1.5 Environmental Scanning Electron Microscopy (ESEM)

The morphology of the untreated and surface treated ﬁbers was investigated using
Environmental Scanning Electron Microscopy (ESEM). The ESEM used for this work
was manufactured by Electroscan Corporation (ESEM Model no. 2020). It is equipped
with a Lanthiurn Hexaboride ﬁlament. Water vapor acts as the imaging gas. The ﬁber
samples to be examined were mounted on an aluminum stub using an adhesive, and
placed in the sample chamber. The imaging pressure (Chamber pressure) was set between
2-3 Torr. The working distance between the detector and the sample was set between 8—
10 mm. The accelerating voltage was set to 20 kV. The scan time was 2.1 seconds, the
condenser was set to 40% and the ﬁlament was heated to 1.86 A. The sample was
focused at different points in its area, and micrograph pictures were taken at different

magniﬁcations.

157

 

Figure 3.5.5: Sputter coater (left) and ESEM (right) Model no. 2020 at CMSC lab

3.5.2 Bioplastics and biocomposites

The plastics and composites are characterized using Fourier Transform Infra Red
Spectroscopy (FTIR), Thermogravimetric analysis (TGA), Differential Scanning
Calorimetry (DSC), thermo-mechanical analysis (TMA), Environmental Scanning
Electron Microscopy (ESEM), impact, ﬂexural, tensile, moisture absorption,
weatherability, and ﬂammability tests.

3.5.2.1 Thermogravimetric analysis (TGA)

The thermal behavior of the polyester resin and modiﬁed resin was measured using
thermogravirnetric analyzer (Hi-Res TGA 2950, TA instruments, USA). The samples to
be analyzed were either liquid or solid. A drop of liquid sample or a small piese of solid
samole were placed on tared aluminum pan. The sample went inside the TGA furnace. It
was programmed to go from room temperature to 600 °C at the rate of 20 0C per minute.
The weight of the samples used for analysis ranged from 10-30 mg. Dry nitrogen gas was

circulated within the test cell at a ﬂow rate of 60 cc/min.

3.5.2.2 Differential Scanning Calorimetry (DSC)

158

To study the curing characteristics of the thermoset polyester resin, blends of polyester
with initiator methyl ethyl ketone peroxide and promoter cobalt naphthenate were
scanned in DSC (TA 2920 Modulated DSC).

In order to obtain heat ﬂow curves for curing characterization, all experiments must be
done under isothermal conditions. The DSC cell should be allowed to stabilize at each
isothermal condition before introduction of the sample. Once an isothermal experiment is
over, the DSC must be cooled quickly to room temperature, and when stabilized, the
residual heat of reaction of the sample is measured with a constant heating rate until no
exotherm is observed. When using DSC for isothermal curing of thermoset resins, it is
assumed that the amount of heat generated is proportional to degree of cure, a (or the
extent of reaction) of the sample at that time. So, the rate of curing, da/dt, may be related
to rate at which heat is evolved, dQ/dt, by:

da/dt = (l/Qto, )(dQ/dt) -(1)

Therefore the relative degree of cure is determined by integrating equation (1) from 0 to
t:

a = (1/Qtot ) ol‘(dQ/dt)r dt -(2)

The total heat for curing reaction (Q10. ) is given by:

Qtot = Qiso + Q

Where Qiso is the heat generated during isothermal DSC runs at each temperature, and Qr
is the residual heat released when the sample is heated up to a high temperature (200°C)
at a constant heating rate. But, it is also important to run one sample from room
temperature to a high temperature (200°C) to see where the exotherm peak appears on the

heat ﬂow curve. This curve will also give Or.

159

A reference Aluminum pan and a weighed sample pan were kept inside the furnace,
which was programmed speciﬁcally according to the nature of the experiment. Heat ﬂow

was kept constant at 5 °C / min in all runs.

3.5.2.3 Dynamical Mechanical Analysis (DMA)

Dynamical Mechanical Analyzer measured the storage modulus, loss modulus and tan
Delta. The DMA apparatus is a TA DMA 2890 model, and the measurement was done at
the frequency of 1 Hz with a heating constant rate of 4 °C /min (cpm). Rectangular bars,
50mm X 12 mm X 3 mm were placed on the 3 point bending ﬁxture in the furnace and
heated from room temperature to 150 0C. A minimum of three specimens of each

composition were tested.

The heat deﬂection temperature (HDT) of the samples was determined using the same
machine at a heating rate of 2 °C /min. HDT is widely used in automotive applications
and represents the temperature at which the material deﬂects by 0.25 mm at an applied
force (three point bending arrangement) of 66 psi (ASTM D 648). A modiﬁed ASTM D
648 in a single cantilever mode was used for HDT measurement since the DMA could

only handle a smaller size of specimen as compared to that of ASTM D 648.

160

. w-.. ___-‘. J...-‘.e—»~J—:.

Figure 3.5.6: TA DMA 2890 in CMSC lab

 

3.5.2.4 Flexural testing
The biocomposites, bioplastics and control samples are used for ﬂexural tests complying
with ASTM D790 standards. A United Calibration Corp. SFM — 20 machine was used for

tensile and ﬂexural testing. System control and data analysis were performed using

Datum software.

  

Figure 3.5.7: UTS inﬂexural testing mode at CMSC lab

3.5.2.5 Impact testing

161

The biocomposites, bioplastics and control samples are used for notched Izod impact
tests complying with ASTM D 256 standards. The impact test was carried out using an
impact tester from Testing Machines Inc. 43-OA-01. Izod impact specimens with the
same dimension as indicated in ASTM D256 were tested with a 5.0 lb pendulum. The
dimension of the notched Izod impact specimens was 12.7 mm (length) x 12.7 mm
(width) x 10 mm (thickness), and a notch of 1.5 mm length and 0.25 mm radius was
marked along the thickness direction. The specimen was held as a vertical cantilever
beam and was impacted on the notched face by a single swing of the pendulum.

Therefore, the crack propagated from the tip of the notch. A minimum of three

specimens for each composition were tested.

 

Figure 3.5.8: Impact testing machine (left) and notching machine (right) at CMSC lab
3.5.2.6 Tensile testing
The biocomposites, bioplastics and control samples are used for tensile tests complying

with ASTM D638 standards. A United Calibration Corp. SFM — 20 machine was used for

162

tensile and ﬂexural testing. System control and data analysis were performed using

Datum software.

 

Figure 3.5.9: United Corporation “SFM-20” Test System

3.5.2.7 Thermo—mechanical Analysis (T MA)

The thermal mechanical analysis was done on a TMA 2980 (TA instruments) to
determine the coefﬁcient of thermal expansion (CTE) of each material type. The samples
were heated to 140 0C at a rate of 4 °C per minute. A plot of the change in length of the
sample versus the temperature was tracked by the software and used to compute the CTE

of the material.

163

 

Figure 3.5.10: TMA 2980 (TA instruments) in CMSC lab

3.5.2.8 Environmental Scanning Electron Microscopy (ESEM)

The impact-fractured surfaces of composites and plastics were investigated using
Environmental Scanning Electron Microscopy (ESEM). The ESEM used for this work
was manufactured by Electroscan Corporation (Model no. 2020). It is equipped with a
Lanthiurn Hexaboride ﬁlament. Water vapor acts as the imaging gas. The samples to be
examined were placed in a sample holder located in the sample chamber. The imaging
pressure (Chamber pressure) was set between 2-3 Torr. The working distance between
the detector and the sample was set between 8 —10 mm. The accelerating voltage was set
to 20 kV. The sample was focused at different points in its area, and micrograph pictures
were taken at different magniﬁcations. The fractured specimens were gold sputtered

before taking the images.

3.5.2.9 AFM

164

The atomic force microscope (AFM) is one of about many types of scanned-proximity
probe microscopes. All of these microscopes work by measuring a local property - such
as height, optical absorption, or magnetism - with a probe or "tip" placed very close to the
sample. The AFM works by scanning a ﬁne ceramic or semiconductor tip over a surface
much the same way as a phonograph needle scans a record. The tip is positioned at the
end of a cantilever beam shaped much like a diving board. As the tip is repelled by or
attracted to the surface, the cantilever beam deﬂects. The magnitude of the deﬂection is
captured by a laser that reﬂects at an oblique angle from the very end of the cantilever
(See Figure 3.5.11). A plot of the laser deﬂection versus tip position on the sample
surface provides the resolution of the hills and valleys that constitute the topography of
the surface. The AF M can work with the tip touching the sample (contact mode), or the

tip can tap across the surface (tapping mode).

posution sensutwe laser

. I/ l spring
~ ‘7 . cantilever

 
  
 

(a)

 

Figure 3.5.11: (a) Working of a contact mode AFM. (b) MultiMode Scanning Probe
Microscope (SPM) from Digital Instrument.

165

In this research, the bioplastic samples were evaluated using a multimode Scanning Probe
Microscope (SPM) (Figure 3.5.11). AF M imaging was conducted using a Nanoscope IV
atomic force microscope from Digital Instruments (Santa Barbara, CA) equipped with an
E scanner. Samples were mounted onto a stainless steel disk using a sticky tab (Latham,
NY). The microscope was allowed to thermally equilibrate for thirty minutes before
imaging. Scanning rates less than 1 Hz were used. Room temperature was maintained at
22 :l: 1 °C. Images were recorded in tapping mode using etched silicon probes (Digital
Instruments). The parameters, especially the set point and the gain, were adjusted to
obtain the best image resolution. For every sample, images were collected at different
locations to obtain reproducible and reliable images. The images are presented in top
view and the surface topography is expressed by different colors. Most images are
presented without treatment unless speciﬁed. Injection molded samples were used for
AFM imaging. The samples are cross, plane, and 45° titled cross-sectioned with a
diamond knife at room temperature. The cross section surface was polished with 4000#

grit from until a smooth surface was obtained.

3.5.2.10 FT IR

FTIR was used for studying the constituent functional groups in the various chemicals
used in the matrix and also to study the kinetics of curing. For ﬁnding the constituent
functional groups in a chemical, scans were made at room temperature with a drop of
liquid between two sodium chloride plates (NaCl). For studying the kinetics, isothermal
experiments were run at one speciﬁc temperature for 4 hours. The scans were taken every

5-10 minutes. One drop of the reaction mixture was placed between two undrilled NaCl

166

IR crystal windows, 2mm thick. The NaCl plates were then placed inside the IR heated
cell. The temperatures were: 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150 and 160 0C.

This kinetic study was done to determine the curing of the system at these conditions.

3.5.2.11 Moisture Absorption Test

Moisture absorption test was carried out to ﬁnd out the moisture uptake by the
composites and control. Moisture absorption tests determine the rate of absorption of
materials and also showcase the equilibrium moisture content of the samples. The
samples were dried until their weight stabilized, their weight was recorded, and then they
were placed in a relative chamber operating at 30 °C and 90% relative humidity. At
regular intervals, they were taken out of the chamber, wiped properly with dry paper
towels, and weighed again. After weighing the samples, they were reinserted in the
humidity chamber. Measurements were taken once a week until the material reached a

steady state (no longer absorbing moisture).

  

, ‘ 311 \, -.
Egg: \icz‘ii
Figure 3.5.12: Humidity chamber in CMSC lab

3.5.2.12 Weatherability Test

167

Biocomposite samples were placed in an accelerated weatherometer (QUV, Dr. Pascal
Kamdem’s laboratory, Department of Forestry, MSU) for weathering tests. The
accelerated weatherometer was set to the following cycle to expose the samples to UV,
condensation, and water spray: UV (340 nm) at 60°C for 2:30 hrs followed by water
spray for 0:30 hrs, repetition of preceding steps, 48 times, and ﬁnally condensation at
45°C for 24:00 hrs. Twelve cycles were repeated in the test, and the total duration for the
whole test was 2016 hours. The measurements for color, surface roughness test weight
change, and dynamical mechanical analysis were taken at regular intervals of time.
Sample preparation

Samples of biocomposites were cut into the dimensions of 76.2mm X 31.75 inch X 4
mm. The samples were conditioned at 65% RH and 25 °C after machining and polishing.
Thirteen specimen of each sample were used for these experiments.

Weatherometer (QUV)

The samples were subjected to weathering in a QUV accelerated weathering tester Model
QUV/Spray (Accelerated Weathering Tester, Q-Panel, Cleveland, OH, USA), which
allowed for water spray as well as condensation. The samples were subjected to
accelerated weathering by exposure to ﬂuorescent ultraviolet light radiation at a
wavelength of 340 nm and temperature of 60 °C for two and a half hours, followed by a
spray of water for half an hour, followed by condensation at 45 °C for twenty four hours.
The average irradiance was set to 0.85 W/m2 at a wavelength of 310 nm. The
temperature of the spray was set to 25 °C. The exposure times for the samples in this

study were 0, 500, 1000, 1500, and 2000 hours approximately. The measurements for

168

color, surface roughness test weight change, and dynamical mechanical analysis were

taken at regular intervals of time.

Fi

 

Figure 3.5.14: Open panel of QUV weatherometer

Color measurements

The color of the biocomposite samples was determined in accordance with the ISO-2470
standard, using a reﬂectometer (Datacolor international, Lawrenceville, NJ, USA) with
CIELAB system. The CIELAB system is characterized by three parameters, L*, a*, b*.

The L* axis represents the lightness, whereas, a* and b* axes are the chromaticity co-

169

ordinates. In the CIELAB co-ordinates, +a* is for the red, -a* for green, +b* for yellow, -
b* for blue and L* varies from 100 (white) to zero (black). L*, a*, and b* color co-
ordinates of each specimen were measured before and after the exposure to weathering
conditions. These values were then used to calculate the color change, AE* and the color

saturation, C* as a function of the weathering exposure duration.

 

Figure 3.5.15: CIELAB system for color measurement

Surface roughness measurement

A surface proﬁlometer system (HommelWerke, Germany) was used to evaluate
modiﬁcations on the biocomposites’ surface before and after weathering. The system was
equipped with a 25 mm long stylus, which traverses the surface and its vertical
displacement is converted into an electrical signal. The signal is ampliﬁed before it is
converted into a set of digital information. The digital information is transmitted to the
computer and the surface roughness parameters calculated from this information. The
stylus travel speed was set at 0.5 mm'1 across the biocomposites sample with a span of

25.6 mm. The maximum range was set at 800 um and an average of ﬁve measurements

170

was taken from each sample. Each sample was evaluated before and after the exposure to
weathering.

Several surface—texture parameters were obtained from this method: Ra, Rz, Rm, Rk, Rpk,
and Rvk. R3 is the average surface roughness and it represents the deviation from the
mean peak. RC. is often used to deﬁne surface roughness, but it does not differentiate
between the peaks and the valleys of a surface proﬁle. R2 is the mean of the peak-to-

valley height. Rmax is the maximum of the peak—to-valley height. Ra, R2, and Rmax were

measured to characterize the physical changes of the biocomposite sample surfaces.

 

Figure 3.5.16: Surface proﬁlometer for roughness measurement

Weight change
The weights of the samples were measured before and after the weathering exposure. The
samples were dried until their weight stabilized, and then their weight was recorded to

four signiﬁcant ﬁgures.

171

 

Figure 3.5.17: Weathered samples, on left, a hempmat-UPE biocomposite, on right, a
BBSG-GFC-UPE biocomposite

3.5.2.13 Flammability Test

Fire test of biocomposite samples was carried out in accordance with ASTM protocol.
The dimensions of the samples for the ﬁre test were: 0.5 inch X 0.125 inch X 2.5 inch.
Two to four samples were tested for each composite/plastic.

The samples tested were: UPE Control, untreated hemp mat (90 hemp+10 PET)—UPE,
acrylonitrile treated hemp mat (90 hemp+10 PET)-UPE, untreated hemp mat (pure)-
UPE +Styrene, UPE treated hemp mat (pure)—UPE +Styrene, untreated hemp mat(pure)-
Glass mat-UPE, untreated hemp mat(BM115)-Glass mat-UPE, untreated BBSG-CaCO3-
UPE, silane treated BBSG-CaCO3-UPE, silane treated green ﬂax core-CaCO3-UPE,
untreated BBSG& GFC-CaCO3-UPE, silane treated BBSG& GFC-CaCO3-UPE, jute-
Hemp-CaCO3(20)-UPE(55), E-glass mat-CaCO3(30)-UPE(50), CaCO3(30)-UPE,
untreated hemp mat(BM115)-bioresin 30-UPE 70, E-glass mat-Bioresin-UPE, E-glass

mat-UPE, Particle board, dry wall, and wood (douglas ﬁr).

172

CHAPTER FOUR: RESULTS ON ENGINEERED FIBERS

4.0 RESULTS ON ENGINEERED FIBERS

This chapter describes the results of the studies done to make an engineered bioﬁber by
surface modiﬁcations and hybridization.

4.1 Effect of surface modiﬁcation on hemp mat with ﬁller

The following surface treatments were done on hempmat ﬁbers: alkali treatment, 7-
methacryloxypropyltrimethoxy silane (y-MPS) treatment, acrylonitrile treatment, and
UPE-MEKP treatment. The composites were made by compression molding (process

depicted in Figure 4.1.1).

 
  
 
  
 
 
   
   

Aluminum

Platen

Teﬂon sheet

Matrix

'; Bioﬁbers
Picture frame

mold

Teﬂon sheet

   

 

 

Figure 4.1.1: Schematic representation of the compression molding process
Analysis of surface treated hemp ﬁbers
The surface treated ﬁbers were characterized with TGA, DSC, and XPS.

Thermogravimetric analysis (TGA)

173

Thermogravimetric curves for untreated and surface treated hemp ﬁbers are shown in
Figure 4.1.2 and 4.1.3. Table 4.1.1 summarizes the maximum decomposition
temperatures for untreated and surface treated hemp ﬁbers. Up to 200 °C, there was less
than 5 % loss in weight of untreated as well as surface treated hemp ﬁbers. After the
alkali treatment, the temperature at the maximum rate of decomposition of hemp ﬁber
increased, indicating that the alkali treatment lead to an enhancement in the thermal
stability of the hemp ﬁber, as has been noted by other authors [1, 2]. Weight loss of 10%
occurred between 288-332 0C for the untreated and surface treated ﬁbers, while a weight
loss of 20% occurred in the range of 329-366 °C, and weight loss of 30% was observed in

range of 346-378 °C for untreated and surface treated hemp ﬁbers.

 

    

 

 

 

 

— Untreated hemp ﬁber — Alkali treated hemp ﬁber
— Acrylonitrile treated hemp ﬁber — UPE treated hemp ﬁber
100 T5-——._' m w
80 ~
:3
3.." 60 -
:1
a
3 40 ~
20 ~
0 l l 1 r i
0 100 200 300 400 500 600
Temperature (0 C)

Figure 4.1.2: TGA of surface treated hemp ﬁbers

174

—Untreated hempﬁber —Al(alitreated lamp fiber
—Acrybnitrib treated lamp ﬁber — UPE-MEKP treawd hemp ﬁber

*

l

 

u—e
W

p—r
&
l

 

Derivative Weight (%/C)

 

 

 

 

0.6 4
0.2 4
P"
-0.2 t t . . ,
0 100 200 300 400 500 600
Temperature (0 C)

Figure 4.1.3: Derivative thermo-gravimetric analysis (DTGA) of surface treated hemp
ﬁbers

Table 4.1.1: Thermogravimetric results for surface treated hemp ﬁbers

 

 

 

 

 

 

 

 

 

 

 

 

10% Weight 20% Weight 30% Weight Max. degradation
loss (°C) loss (°C) loss (°C) temperature (°C)
A 311 348 363 381
B 332 366 378 395
C 295 339 357 383
D 288 329 346 369
Legend:

A= Untreated hemp ﬁber, B= Alkali treated hemp ﬁber, C= UPE-MEKP treated hemp
ﬁber, D= Acrylonitrile treated hemp ﬁber

Differential scanning calorimetry (DSC)
The DSC curves for untreated and surface treated ﬁbers expressed in terms of heat ﬂow
are shown in Figure 4.1.4. The melting point of cellulose has been reported to be near

300 °C [3]. However, a controversy surrounds the glass transition temperature (T g) of

175

cellulose, which has been reported between -30 to 160 0C [4]. The T8 for lignin has been

reported between 135- 172 °c [5,6].

— Untreated hemp rmt — Alkali treated hemp mat
— UPE treated hemp mat — Acrylonitrile treated hemp net

0.5

 

 

 

 

5
B
.9.
Fit
8 J
I l
l
i
l
-2.5 l f r 1 r
-60 -10 40 90 140 190
Temperature (°C)

Figure 4.1.4: DSC of surface treated hemp ﬁbers
It is observed from Figure 4.1.4 that the crest associated with weight loss due to

evaporation of water was found in all ﬁber samples, and ranged from 10-130 °C in hemp
ﬁbers, depending upon the degree of hydrogen bonding interaction [7]. Furthermore, this
transition shifted towards the right after surface treatment of the hemp ﬁber. It was also
observed that ﬁbers had less amount of moisture percentage because they had been pre-
dried before any kind of treatment. This ensured that the inherent moisture of the ﬁbers,
which is 12-14% of the total ﬁber weight, was not able to interact with the surface
treatments in any way. Therefore, the chemicals of the surface treatments had an
opportunity to react with the actual ﬁber surface and ﬁber constituents. This would lead

to a direct link between the chemicals used and the ﬁber constituents such as cellulose

176

and hemicelluse. The Tg for lignin and cellulose were not observed by this particular
method in DSC. Two exothermic peaks were observed by some researchers from 300 0C
to 500°C for untreated and surface treated natural ﬁbers, indicating the thermal stability
of the ﬁbers as a ﬁmction of the treatment [8]. However, we were not able to see such

peaks in out DSC therrnograms as our experiments were conducted only up to 300 °C.

X-ray photon spectroscopy (XPS)

XPS survey scans were taken for untreated as well as surface treated hemp ﬁbers. These
scans revealed the presence of carbon, oxygen, nitrogen, calcium, and silicon in the hemp
ﬁbers. Table 4.1.2 and 4.1.3 show the elemental composition and elemental ratios of the
hemp ﬁbers, respectively. Following an hour long treatment with alkali solution it is
observed that the carbon content increases, while oxygen and nitrogen content decrease.
After silane treatment, there is a marked increase in silicon as well as carbon content, and
a decrease in oxygen and nitrogen content. Again, after treatment with the matrix, there is
an increase in carbon content, while a decrease in oxygen and nitrogen content. However,
with acrylonitrile treatment, there is a decrease in carbon content and an increase in
nitrogen and oxygen content.

It is well known that natural ﬁbers consist of cellulose, hemi-cellulose and lignin [9].
Hemicellulose consists of a mixture of different sugars and other subsituents which are
soluble in water or bases. Lignin is similar to a highly unsaturated or aromatic polymer in
structure and has low oxygen to carbon ratio, and is partially soluble in water [9, 10]. Due
to the alkali treatment, a part of hemicellulose and lignin might have dissolved, and

washed away, leading to a decrease in oxygen content. The increase in carbon and silicon

177

content after silane treatment might be due to the attachment of bulky alkyl group and
silicon to the hemp ﬁber as a result of this treatment. The large increase in carbon content
of UPE treated hemp ﬁber can again be due to the large oligomer molecule of UPE
reacting with hydroxyl groups of ﬁber. Meanwhile, the increase in nitrogen content after
treatment with acrylonitrile might be due to grafting of acrylonitrile monomer to the ﬁber

surface.

Table 4.1.2: Elemental composition of surface treated hemp ﬁbers (from XPS analysis)

 

 

 

 

 

 

 

 

 

 

 

 

 

ClsL314] N1s[.499] Ols[.733] Sin[.368l Ca2p[1.927]
A 71.64 2.63 24.39 0.75 0.52
B 73.74 2.23 21.21 0.36 0.56
C 79.44 0.77 16.09 3.12 0.59
D 80.17 0.98 18.85
E 69.91 2.92 26.79 0.38
Legend:

A= Untreated hemp ﬁber, B= Alkali treated hemp ﬁber, C= Silane treated hemp ﬁber, D=

UPE-MEKP treated hemp ﬁber, E= Acrylonitrile treated hemp ﬁber

Table 4.1.3: Elemental ratios of surface treated hemp ﬁbers (from XPS analysis)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

C/O C/N C/Si C/Ca
A 2.94 27.24 95.52 137.77
B 3.48 33.07 - 131.68
C 4.94 103.17 25.46 134.64
D 4.25 81.81 - -
E 2.61 23.94 - 183.97
Legend:

A: Untreated hemp ﬁber, B= Alkali treated hemp ﬁber, C= Silane treated hemp ﬁber, D=
UPE-MEKP treated hemp ﬁber, E= Acrylonitrile treated hemp ﬁber

Reaction schemes
The schematic diagram of reactions that might result due to surface treatment of ﬁbers is

shown in Figure 4.1.5. Figure 4.1.6 and 4.1.7 show the reactions that might take place

178

after the surface treated ﬁbers react with UPE matrix, during curing, and produce three
dimensional networked structures. Conclusive proofs for these reactions will be

discussed in future work.

 

 

 

 

 

 

Na+OﬂiemﬂOH
(Alkali tpleated hemp ﬁber)
E 25 0C NaOH
H-C—CN Alkali (St)x
H2 UPE H .
Hf.“ A 25 0C H MEKP R'QOC —CHC00
<9 - OEHemaOH 0 -
CH2 HzCZSHf-mNIéUntreated hemp ﬁber) 25 C OJEHma_9_-_
o E11 ry m .
“ugh? Dicumylperoxide 25 0C Sllane (UPE-MEKP treated hemp ﬁber)
H
reg-(221s H2 H3CQSPCH3
{
H 211(2)! 3c OCH3
- - O
(Acrylonitrile treated hemp ﬁber) V
[ l HempT l]
..0 ,O ..O\
H (5.110 ' H. ‘.H H.9..H
HO— '

 

lee-Q

°=$=cuz

 

 

 

H3C 3C H3C
Heat -2H 0
l A) GHgimp ]
HO— 1 G i-OH
°=$=CH2°=$=C HST§=CH2
H3C H3C H3
(Silane treated hemp ﬁber)

Figure 4.1.5: Proposed reaction for surface treatment of hemp ﬁbers

179

was 0:8 80%.: Q80: 00800 mongesm C0 meta :8 00:08: 08035 6.3. 0.5mm..—

 

 

 

 

 

 

 

xcmv : 50 m :00
r - -ooomvloooo- -000WI0000-I-000EI0000 .. - -- 1m
. a 0800 05:
~00 05 28w 05% 0 m 89.: 89:: a a
_
NIUIVO mmolﬂo NE Iwuo Um:
N
w it... it... 1r.
moi 0WW0 : Ion 0W0 W
we Aw m 880 well «melon
a_ 828 mum: .95 m m
m 5.. - U _ ._ _
m M 7 . en. E

—--—-t

 

180

It csN

 

 

 

H (SOX' CI'Iz'C- CH2'C"'
(F 2 UPE, MEKP, CoNap H I CN H
H-g-CN Curing ‘R—_—_OOCC_CCOO- ————————
9H2 (,St)y
° 516mg?“ H—C-Ch
_'-2 8H2
I 2 CH2
H- C- CN
a 61%!th
. . CH
Ac lonrtnle treated hem ﬁber 1 2
( ry p ) H_ 9- CN
8H2
H-C- CN
R ————— —ooc$i——— Ecoo— - -
(St)z
UPE, MEKP, CoNap (Fux _-_
NaO'U-lemﬁIOH Curing ; R- _ _ _ 00 C C— ﬁCOO— ________
(Alkali treated hemp ﬁber) (St)y
Na ElemQO
R ----- -OOC$l—— ECOO— - -
(St)z
(ISOX (SPX H (St)x
R- OOC —CCOO-OOC COO-- -
R—OOCC—ﬁCOO— E H $_ﬁ
(St)y UPE MEKP C m 3’ (St)y O-‘Eliemp3—0
unng
O—El emﬁOH O—ElemﬂO (80y
(UPE-MEKP treated hemp ﬁber) R‘—00CC—ECOO-00Cﬁ'—8C00-
(St)z (St)z

Figure 4.1.7: Proposed reaction for curing of surface treated hemp ﬁbers and UPE

Optimization of ﬁber volume fraction
The ﬁber volume fraction was optimized by making bio-composites with non woven
hemp mat ﬁber volume fractions of 20 %, 30 % and 40 %, and testing the mechanical

properties of the resulting biocomposites. It should be noted that with 40 % volume

181

fraction of hemp ﬁbers, a consolidated biocomposite with a complete wet out could not
be manufactured. Figure 4.1.8 shows the result ﬁ'om ﬂexural test of ﬁber volume
fraction optimization study. The biocomposite with 30 volume (vol) % hemp ﬁbers had
higher bending strength and modulus of elasticity as compared to composite with 20 vol
% hemp ﬁbers as well as UPE control. The bending strength of 30 vol % hemp ﬁber
composite was 16 % higher than that of UPE control, and 9 % higher than 20 vol % hemp
ﬁber composite. Its modulus of elasticity was 150 % higher than the neat resin, and 45 %
higher than 20 vol % hemp ﬁber composite. Similar results were obtained from tensile
strength, tensile modulus and impact strength (data not shown). Therefore, it was decided
that 30 volume % of hemp ﬁbers were optimum for a biocomposite made of hemp mats

and UPE resin.

00

160

p—a
N
O
O\

OO
O

.5
Modulus of Elasticity, MOE(GPa)

A
O
N

Bending Stregnth, BS(MPa)

 

O

A B C

Figure 4.1.8: Optimization of ﬁber volume fraction by evaluation of mechanical
properties

Legend: A= UPE control, B: Untreated hemp mat (20 % vol)—UPE, C= Untreated hemp
mat (30 % vol)—UPE

182

Tensile Properties

Figure 4.1.9 shows the comparison of tensile properties of various composites. The
tensile properties of biocomposites with surface treated hemp ﬁbers were higher than that
of untreated hemp mat composite as well as that of neat resin. The tensile strength and
modulus of untreated hemp ﬁbers based biocomposite was 45 % and 325 % higher than
that of neat resin, respectively. Comparing surface treated ﬁber based composites, the
tensile strength of alkali treated ﬁber based biocomposite was 34 % higher than that of
untreated hemp ﬁber based composite, while that of silane treated ﬁber based
biocomposite was 48 % higher than that of untreated hemp composite. UPE-MEKP
treated hemp ﬁber based biocomposite has tensile strength of 57 % higher than that of
untreated hemp composite, and for acrylonitrile treated ﬁbers it was 80 % higher than
that of untreated hemp ﬁber based composite. In terms of tensile modulus, silane treated
ﬁber had enhancement of 6 % as compared to that of untreated hemp, while UPE-MEKP
treated ﬁbers had an enhancement of 4 %, and acrylonitrile treated ﬁbers had a 25 %
enhancement compared to that of untreated hemp.

The tensile strength of E-glass-UPE composite was 130 % higher, and the tensile
modulus was 70 % higher as compared to that of untreated hemp mat biocomposite at
same volume %. The hybrid E-glass-hemp mat-UPE composite had an increment of 76 %
in tensile strength, and 34 % in tensile modulus compared to that of untreated hemp mat
based biocomposite. The E-glass—UPE composite had 23 % higher tensile strength as
compared to that of hybrid composite of E-glass and hemp mats. On comparing speciﬁc

tensile strength (tensile strength divided by density of composite) and speciﬁc tensile

183

modulus (tensile modulus divided by density of composite) of all composites, it was
found that the biocomposites and glass composites were in the same range.

This increase in tensile strength and modulus of the chemically treated hemp ﬁbers based
biocomposites may be an outcome of the improved adhesion between the ﬁber and the
matrix. This improved adhesion might have enhanced the interfacial bonding and thus,
made it easier for the stress to be effectively transferred from the matrix to the ﬁber (as

depicted in reaction schemes 3, 4, and 5) [1 1-13].

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

El TS I TM
100 20

Q E
8 8°“ 9.
2

a 1 1 I {-1
" 60 e 1 a
€11 .I. I l '3
5 *1 ». '0
Q . , o
.2 ii L 1.; '12
12 20 - [3

O ' I r If I T I

 

 

 

A B C D E F G H

Figure 4.1.9: Comparison of tensile properties of surface treated composites

Legend: A: UPE control, B: Untreated hemp mat (30 % vol)—UPE, C= Alkali treated
hemp mat (30 % vol)—UPE, D=Silane treated hemp mat (30 % vol)—UPE, E= UPE-
MEKP treated hemp mat (30 % vol)—UPE, F=Acrylonitrile treated hemp mat (30 %
vol)-UPE, G=E-g1ass mat-UPE, H=E-glass mat-hemp mat-UPE

Flexural Properties
Flexural strength is a combination of the tensile and compressive strengths which directly

varies with the interlaminar shear strength. In ﬂexural testing various mechanisms such

184

as tension, compression, shearing etc. take place simultaneously. In a three point flexural
test, the failure occurs due to bending and shear failure. The bending strength and
modulus of elasticity of different composites are compared in Figure 4.1.10.

The flexural properties of biocomposites with surface treated hemp ﬁbers were higher
than that of untreated hemp mat composite; also, the flexural properties of all composites
were higher than that of neat resin. Comparing surface treated ﬁber based composites, the
modulus of elasticity as well as bending strength of alkali treated, silane treated and UPE-
MEKP treated hemp ﬁber based biocomposites lie in the same range. However as
discussed earlier [14], the modulus of elasticity and bending strength of acrylonitrile
treated ﬁbers based composites were 7%, and 35% higher than those of untreated hemp
ﬁber biocomposite respectively. As compared to UPE control, the surface treated ﬁber
based biocomposites had a 10-16% enhancement in bending strength, and 140-225%
increment in modulus of elasticity.

The bending strength of E-glass-UPE composite was 144 % higher, and the modulus of
elasticity was 56 % higher as compared to that of untreated hemp mat biocomposite. The
hybrid E-glass-hemp mat-UPE composite had a bending strength 83 % higher, and
modulus of elasticity 66 % higher than that of untreated hemp mat based biocomposite.
The E-glass-UPE composite had 25 % higher bending strength as compared to that of
hybrid composite of E-glass and hemp mats, but its modulus was 7 % lower than that of
hybrid composite.

On comparing speciﬁc bending strength and speciﬁc modulus of elasticity of all
composites, it was found that the biocomposites and glass composites were in the same

range.

185

The increment in the ﬂexural properties of biocomposites aﬁer surface treatment of the
hemp ﬁbers can be attributed to the modiﬁcations in the molecular level of the ﬁbers due
to chemical bonding between ﬁber components and the treatment reagents. Such linkage
might have lead to better interfacial bondage, better adhesion, and effective stress transfer

(as depicted in reaction schemes shown in Figure 4.1.5, 4.1.6, and 4.1.7) [1 1-13].

.88 A MOE

y—a
N

 

600
500 j i i

400 ~ I

ﬁ
00

300 -

  

I
.h

200 J

\\"\‘

  

100 *

Bendmg Strength, BS (MPa)
l
l-H
t-D-i
..
Modulus of Elastncrty, MOE (GPa)

 

 

   

l
C

 

Figure 4.1.10: Comparison of ﬂexural properties of surface treated composites
Legend: A: UPE control, B: Untreated hemp mat (3O % vol)—UPE, C= Alkali treated
hemp mat (3O % vol)—UPE, D=Silane treated hemp mat (3O % vol}UPE, E= UPE-
MEKP treated hemp mat (30 % vol)—UPE, F=Acrylonitrile treated hemp mat (3O %
vol)-UPE, G=E-glass mat-UPE, H=E-glass mat-hemp mat-UPE

Impact Strength

Impact strength is the ability of a material to resist the fracture under stress applied at
high speed. The impact properties of the composite are directly related to its overall
toughness. The ﬁbers play a very important role in the impact resistance of the composite

as they interact with the crack formation in the matrix and act as stress transferring

186

medium. It is a common observation that with any surface treatment, the ﬂexural and
tensile properties of the composite increase, but the impact strength (shown in Figure
4.1.11) will decrease.

The impact strength of all composites were higher than that of neat resin. The impact
strength of biocomposites with surface treated hemp ﬁbers was higher than that of
untreated hemp mat composite. As compared to neat resin, there was an increment of 82
% in impact strength of untreated hemp ﬁber based composites, 49 % for alkali treated
ﬁbers, 94 % for silane treated ﬁbers, 120 % for UPE—MEKP treated ﬁbers, and 180 % for
acrylonitrile treated hemp ﬁber based biocomposites. On comparing the impact strength
of untreated and surface treated biocomposites, it was found that the impact strength of
silane treated ﬁbers was 7 % more than that of untreated ﬁbers, while that of UPE-MEKP
treated ﬁbers was 21 % more than that of untreated ﬁbers, and that of acrylonitrile treated
ﬁbers was 54 % more than that of untreated ﬁbers.

The impact strength of E-glass-UPE composite was 16.3 times higher than that of neat
resin and 8.5 times higher than that of untreated hemp ﬁber based biocomposite. The
hybrid E-glass-hemp mat-UPE composite had impact strength 16.5 times higher than that
of neat resin and 8.6 times higher than that of untreated hemp ﬁber based biocomposite.
The impact strength of hybrid composite of E-glass mats and hemp mats was almost

same as that of E-glass-UPE composite (within error bars).

187

 

200

—a

M

O
i

100 j

Impact Strength (J/m)

U!
C
1

 

 

 

 

A B C D E F G H

Figure 4.1.11: Impact strength of surface treated composites

Legend: A= UPE control, B= Untreated hemp mat (30 % vol)—UPE, C= Alkali treated
hemp mat (3O % vol)—UPE, D=Silane treated hemp mat (3O % vol)—UPE, E= UPE-
MEKP treated hemp mat (3O % vol}—UPE, F =Acrylonitrile treated hemp mat (30 % vol)-
UPE, G=E-glass mat-UPE, H=E-glass mat-hemp mat-UPE

Dynamic Mechanical Analysis

Dynamic mechanical methods expose the specimen to periodic stresses. The polymer is
subjected continuously to forced oscillations, and the applied stress is sinusoidal with a
frequency. The deformation of ideal-elastic bodies follows the stress instantaneously but
that of viscoelastic polymer experiences a delay. The stress vector is assumed to be a sum
of two components: one component is in phase with the deformation; the other is not.
Each of these two components possesses a modulus. The real modulus (shear storage
modulus) 6' measures the stiffness and shape stability of the specimen, whereas the

imaginary modulus (shear loss modulus) G" describes the loss of usable mechanical

188

energy by dissipation into heat. The maximum of tan 8 as a function of temperature is
generally identiﬁed as the glass transition temperature T8, which is dependent on the
deformation rate. The damping properties of the material are related to energy absorption.
The area below tan delta is function of the absorption of the energy necessary to pass

ﬁ‘om glassy state to rubbery state.

 

 

— UPE control —— Untreated hemp rmt-UPE

—- Acrybnitrib treated hemp rmt—UPE — Akali treated herrp mat -UPE

—— UPE treated hemp mat-UPE — Glass-UPE

12000 JTV— ,. —‘#—~k 7— i ,, ——¥— —- u *ﬁiﬁii

a 3 I
a 9000 J .
u t l
E. | i
= i l
'8 6000 |
E l
0

°" 1
E 3000 l ‘
s l
m l \

0 l 1 1 ' v 1
20 40 60 80 100 120 140

Temperature (0 C)

Figure 4.1.12: Typical storage modulus curves of surface treated composites

The typical curves of storage modulus and tan delta for neat resin and composites can be
seen in Figures 4.1.12 and 4.1.13 respectively. The storage modulus decreased as a
function of temperature, as is commonly observed for composites. At higher temperatures,
all biocomposites plateau to the same value of modulus. The storage moduli of surface
treated hemp ﬁbers based biocomposites were higher than that of untreated hemp ﬁber
based biocomposite. A comparison of storage moduli of composites at 40 0C was made in

Figure 4.1.14. The storage modulus of biocomposites at 40 0C was enhanced by 110-

189

190 % compared to neat resin. Glass mat based composite had a storage modulus 307 %
higher than that of neat resin, 68 % higher than that of untreated hemp ﬁber based

composite, and 26% higher than that of hemp mat-glass hybrid composite at 40 0C.

 

 

 

 

 

— UPE control -- Untreated hempmt-UPE
— Acrybnirik treated herrp nut-UPE — Akali heated hemp mat-UPE
—— UPE heated lenp nut-UPE — Glass-UPE
0.5 *# *4 l
l
0.4 ~ 1
1
§ 0.3 . J
0
G l
E 0.2 .
4‘
if “N-
O T T —__T— ﬂ I I
20 40 60 80 100 120 140
Temperature (0 C)

Figure 4.1.13: Typical tan delta curves of surface treated composites

The loss modulus of composites was higher than that of the neat resin. This follows the
trend found previously for natural ﬁber-thermoset composites, where the loss modulus
increased after addition of ﬁbers to the plastic [15].

Over the entire range of temperature, tan 6 was highest for the neat resin due to huge
reduction in the storage modulus values at higher temperatures. The lower values of tan 8
for the biocomposite made with surface treated hemp ﬁbers suggest that there is less

damping in the chemically treated hemp ﬁber based composites. The tan 8 vs.

190

temperature plot for biocomposites as well as E-glass composite is similar. The tan
5curve of all composites shiﬁed towards right as compared to neat resin, but this shift
was signiﬁcant in case of glass based composite. The Tg of the neat resin was 95 0C. For

biocomposites, the Tg increased about 2-5 0C as compared to neat resin.

 

Storage Modulus at 40 0C (GPa)

 

 

 

2 - l+l
0 1 I
A B C D E F G H

Figure 4.1.14: Storage modulus of surface treated composites at 40 0C

Legend: A= UPE control, B= Untreated hemp mat (30 % vol)—UPE, C= Alkali treated
hemp mat (30 % vol)—UPE, D=Silane treated hemp mat (30 % vol)—UPE, E= UPE-
MEKP treated hemp mat (30 % vol)—UPE, F=Acrylonitrile treated hemp mat (30 %
vol)-UPE, G=E-g1ass mat-UPE

l

Morphology of fiber surface

The micrographs of untreated and surface treated hemp ﬁbers are shown in Figure 10.
The dish'ibution of the ﬁbers in the hemp mat was random, and uneven. The micrographs
in Figure 4.1.15 focus on single ﬁber surface. Fibrillation was observed in ﬁbers alter

surface treatment.

191

 

Figure 4.1.15: ESEM micrographs of surface treated hemp mat ﬁbers
a) Untreated hemp mat, magniﬁcation 1200 X, scale bar 45 um
b) Alkali treated hemp mat, magniﬁcation 900 X, scale bar 50 um
c) Silane treated hemp mat, magniﬁcation 900 X, scale bar 50 pm
(1) Acrylonihile treated hemp mat, magniﬁcation 1000 X, scale bar 45 pm

This could provide more anchorage for the matrix, and hence improve the strength of the
composite. In general, the surface of chemically treated ﬁbers looked different from that
of the untreated hemp ﬁber.

In the biocomposites, the ﬁber pull out was clearly observed (pictures not shown).
Biocomposite with untreated hemp ﬁbers showed poor interfacial bonding between the

ﬁber and matrix, which resulted in relatively clean surface over the pulled out ﬁbers due

192

to greater extent of delamination. In case of untreated ﬁber based biocomposites, shear
failure results in high degree of pull out. The adhesion between the ﬁber and the matrix
was enhanced in biocomposites with surface treated ﬁbers. The ﬁbers were covered with

mahix, and the ﬁber pull out was relatively smaller.

   
  
  
 
  
 
  
 
 
 
  

Natural ﬁber reinforced composite

 

Polymer microgels

 

Bulk polymer
Polymer chains

Polymer of different
properties

Sizings

Adsobed material

Bioﬁber chemistry

Bioﬁber topology Bioflber morphology Bul bjoﬁber Impurities Thermal
Chemical
Mechanical
Environments

Fiber-Matrix Interface

Figure 4.1.16: Schematic representation of the ﬁber matrix interface and various factors
affecting the ﬁber-matrix adhesion

Although there have been many studied on natural ﬁber thermoset composites, only a
few mention the use of non woven ﬁber mats as reinforcement in biocomposites [16-
19]. For useful composites, ﬁber—matrix adhesion has to be optimized for ensuring
good mechanical properties. The surface chemical modiﬁcations of natural ﬁbers like
dewaxing, alkali treatment, cyanoethylation, vinyl graﬁing and treatment with various

coupling agents, are some means to improve ﬁber matrix adhesion of the resulting

l93

biocomposites (as shown in Figure 4.1.16) [1 1-12, 20-26]. Surface modiﬁcation also
results in enhancement of the aspect ratio, improves the wettability of the ﬁbers, and
forms a strong interface between polar natural ﬁber and non polar matrix.

The adhesion between hemp ﬁbers and UPE matrix was increased by treatment of
hemp ﬁber surface with alkali, silane, UPE (matrix), and acrylonih'ile. The surface
treatment of hemp mats also resulted in higher mechanical and thermal properties.
Other modiﬁcations which could bring about the same effects are acetylation, bleaching,
UV/plasma, microwave, and steam explosion. It is our future plan to optimize these
treatments for natural ﬁbers. The gap between performance of glass based composites
and biocomposites was bridged by fabricating a hybrid composite comprising of glass
and hemp ﬁbers. In terms of speciﬁc modulus and strength, glass composites and

biocomposites were in the same range.

4.1 Effect of surface modification on pure hemp mat

The results of the following surface treatments done on pure hemp mat are discussed in
this section: alkali treatment, y-methacryloxypropyltrimethoxy silane (y-MPS) treatment,
and UPE-MEKP treatment.

Analysis of surface treated hemp fibers

The surface treated ﬁbers were characterized with TGA, DSC, and XPS.

Thermogravimetric analysis (TGA)

194

Thermogravimetric curves for untreated and surface heated hemp ﬁbers are shown in
Figure 4.2.1 and 4.2.2. Table 4.2.1 summarizes the maximum decomposition
temperatures for untreated and surface treated pure hemp ﬁbers. Up to 200 0C, there was
less than 8 % loss in weight of unheated as well as surface treated hemp ﬁbers. After the

alkali treatment, the temperature at the maximum rate of decomposition of hemp ﬁber

 

     
 

 

 

 

‘ l
801 1
§ 60 ~
:1
in .
g — Unheated pure hemp mt ‘
40 - — UPE-MEKP heated pure hemp mat 11
— Sihne heated pure herrrp net 1
l — Akali heated pure hemp rmt [
20 -
0 T T I 1 |
0 100 200 300 400 500 600
Temperature (0 C)

Figure 4.2.1: TGA of surface heated pure hemp ﬁbers

increased, indicating that the alkali treahnent lead to an enhancement in the thermal
stability of the hemp ﬁber, as has been noted by other authors [1, 2]. Weight loss of 10%
occurred between 278-308 0C for the unheated and surface treated ﬁbers, while a weight
loss of 20% occurred in the range of 323-359 0C, and weight loss of 30% was observed in

range of 342-373 0C for untreated and surface heated hemp ﬁbers. The maximum

195

decomposition temperature ranged from 362-3 92 0C.These results are comparable to

those of hemp mat with ﬁller, discussed in Section 4.1.

 

F 1
2,3 A —Unheatedpurehenprmt
—UPE-MEKP heated pue leap mat n
—Silareheatedpurehenprmt
1'84 —Akaliheatedpmelerrpnnt

 

 

 

Derivative Weight(%/“C)
3 CS

 

 

 

 

 

 

 

0.3 4
F —
-0.2 . r r . I .
20 120 220 320 420 520
Temperature (0 C)
gauze 4.2.2: Derivative thermo-gravimehic analysis (DTGA) of surface treated pure hemp

Table 4.2.1: Thermogravimeh'ic results for surface treated pure hemp ﬁbers

 

 

 

 

 

10% Weight loss 20% Weight loss 30% Weight loss Max. degradation
(°C) (°C) (°C) temperature (°C)
A 296 339 360 390
B 308 359 373 391.7
C 278 323 342 384.1
D 292 329 349 367.5

 

 

 

 

 

 

 

(Table 4.2.1 continued)
Legend: A= Untreated pure hemp ﬁber, B= Alkali treated pure hemp ﬁber, C= UPE-
MEKP treated pure hemp ﬁber, D= Silane heated pure hemp ﬁber

196

Differential scanning calorimetry (DSC)

The DSC curves for untreated and surface treated ﬁbers expressed in terms of heat ﬂow
are shown in Figure 4.2.3. It is observed from Figure 4.2.3 that the crest associated with
weight loss due to evaporation of water was found in all ﬁber samples, and ranged from
7—150 0C in pure hemp ﬁbers, depending upon the degree of hydrogen bonding
interaction [7]. Furthermore, this hansition shiﬁed towards the right aﬁer surface
treatment of the hemp ﬁber. It was also observed that ﬁbers had less amount of moisture
percentage because they had been pre-dried before any kind of treatment. This ensured
that the inherent moisture of the ﬁbers, which is 12-14% of the total ﬁber weight, was not

able to interact with the surface treahnents in any way. Therefore, the chemicals of the

 

— Unheated pure henp rmt — Alkali heated pure hemp net
— UPE-MEKP heated pure henp rmt — Sihne heated pure lemp mat
2 ,7, ——~ L,_Lﬂ

Heat Flow (mW)

 

 

 

'10 l l l 7 l l ‘17
-70 -20 30 80 130 180 230 280
Temperature (0 C)

Figure 4.2.3: DSC of surface treated pure hemp ﬁbers

surface treahnents had an opportunity to react with the actual ﬁber surface and ﬁber

constituents. This would lead to a direct link between the chemicals used and the ﬁber

197

constituents such as cellulose and hemicelluse. The T3 for lignin and cellulose were not
observed by this particular method in DSC. Exothermic peaks were not observed between
300 0C to 500 0C in our DSC thermograms as our experiments were conducted only up to

300 °C.

X-ray photon spectroscopy (XPS)

XPS survey scans were taken for untreated as well as surface treated pure hemp ﬁbers.
These scans revealed the presence of carbon, oxygen, nitrogen, calcium, and silicon in
the hemp ﬁbers. Table 4.2.2 and 4.2.3 show the elemental composition and elemental
ratios of the pure hemp ﬁbers, respectively. Following an hour long treatment with alkali
solution it is observed that the carbon content increases, while oxygen and nitrogen
content decrease. After silane treatment, there is a marked increase in silicon as well as
carbon content, and a decrease in oxygen and nitrogen content. However, after treatment
with the mahix, there is an increase in carbon content, while a decrease in oxygen and
nitrogen content.

Due to the alkali treatment, a part of hemicellulose and lignin might have dissolved, and
washed away, leading to a decrease in oxygen content. The increase in carbon and silicon
content aﬁer silane treatment might be due to the attachment of bulky alkyl group and
silicon to the hemp ﬁber as a result of this treatment. The large increase in carbon content
of UPE treated hemp ﬁber can again be due to the large oligomer molecule of UPE

reacting with hydroxyl groups of ﬁber.

198

Table 4.2.2: Elemental composition of surface treated pure hemp ﬁbers (from XPS

 

 

 

 

 

 

 

 

 

 

 

 

 

analysis)
C1s[.314] N1sL499] 0181.733] Si2p[.368] Ca2p[1.927]
A 74.48 2 22.52 0.75 0.47
B 76.38 1.95 19.9 1.19
C 79.44 0.77 16.09 3.12 0.59
D 81.46 1.47 15.52 1.14 0.41
Legend:

A= Untreated hemp ﬁber, B= Alkali treated hemp ﬁber, C= Silane treated hemp ﬁber, D=
UPE-MEKP treated hemp ﬁber

Table 4.2.3: Elemental ratios of surface treated pure hemp ﬁbers (from XPS analysis)

 

 

 

 

 

 

 

 

 

 

 

 

C/O C/N C/Si C/Ca

A 3.3 37.24 99.31 158.47

B 3.84 39.17 - 64.18

C 4.94 103.17 25.46 134.64

D 5.25 55.42 71.46 198.68
Legend:

A: Untreated hemp ﬁber, B= Alkali treated hemp ﬁber, C= Silane treated hemp ﬁber, D=
UPE-MEKP treated hemp ﬁber

Reaction schemes

The schematic diagram of reactions that might result due to surface treatment of ﬁbers is
shown in Figure 4.1.5. Figure 4.1.6 and 4.1.7 show the reactions that might take place
after the surface treated ﬁbers react with UPE matrix, during curing, and produce three
dimensional networked structures.

Conclusive proofs for these reactions will be

discussed in future publications.

Optimization of styrene content
The high density of the hemp ﬁber mats led to a resin impregnation problem, due to
which it was not possible to obtain fully consolidated biocomposites. This problem was

solved by adding extra styrene to the UPE resin to reduce the viscosity of the polymer

199

matrix. The styrene content was optimized by making biocomposites with non woven
pure hemp mat ﬁber, UPE, and styrene, with extra styrene weight percentage of 10%, 20
%, 30 % and 40 %, and testing the mechanical properties of the resulting biocomposites.
All of the biocomposites had 30 % volume of hemp ﬁbers; the change was only done to
the matrix which comprised 70% volume of the composites. Figure 4.2.4 shows the result
from flexural test of styrene optimization study. The biocomposite with 20 wt % extra

styrene had higher bending strength and modulus of elasticity as compared to composite

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

EIBSAMOE
100 8?
O-t
a? ‘ g

z 8
E. 80 I e68
W T T 2
an , J. .
4:“ 60 I T l ,4?
3'0 i 4.2
= a
2 40* a
e E
w _2 °

=

:3 204 g
5 E
m '6
O l l l 0%

 

A B C D

Figure 4.2.4: Optimization of styrene content by evaluation of flexural properties
Legend: A=30% untreated pure hemp mat-10% styrene, B: 30% untreated pure hemp
mat-20% styrene, C= 30% untreated pure hemp mat-30% styrene, D=30% untreated pure
hemp mat-40% styrene

with 10 wt % extra styrene as well as UPE control. The bending strength of 20 wt %
extra styrene-hemp ﬁber composite was 12 % higher than that of 10 wt % exha styrene
biocomposite, and its modulus of elasticity was 142 % higher than that of 10 wt % extra
styrene biocomposite. In the case of 30 wt % extra styrene-hemp ﬁber composite, the

bending strength was 36 % higher than that of 10 wt % extra styrene biocomposite, and

200

its modulus of elasticity was 115 % higher than that of 10 wt % extra styrene
biocomposite. Whereas, for the 40 wt % extra styrene-hemp ﬁber composite, the bending
strength was 36 % higher than that of 10 wt % extra styrene biocomposite, and its
modulus of elasticity was 112 % higher than that of 10 wt % extra styrene biocomposite.

Similar results were obtained from tensile strength, tensile modulus and impact strength.

The tensile strength of 20 wt % extra styrene-hemp ﬁber composite was 31 % higher than
that of 10 wt % extra styrene biocomposite, and its tensile modulus was 40 % higher than
that of 10 wt % extra styrene biocomposite (Figure 4.2.5). In the case of 30 wt % extra
styrene-hemp ﬁber composite, the tensile strength was 67 % higher than that of 10 wt %
extra styrene biocomposite, and its tensile modulus was 25 % higher than that of 10 wt %

extra styrene biocomposite. Whereas, for the 40 wt % extra styrene-hemp ﬁber

[:1 TS(MPa) o TM(GPa)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100
A i ‘6
“ a?
G-t 80* _ 9..
a f i 59
a: _ m
am“ i 4%
5 3'2
,_ -
a 40~ r E
Q) J: ‘L 20
“E 1. J. % " ‘5.
5 204 =- 5
E" *IE-t

0 l l l 0

A B C D

Figure 4.2.5: Optimization of styrene content by evaluation of mechanical properties
Legend: A=3 0% untreated pure hemp mat-10% styrene, B: 30% untreated pure hemp
mat-20% styrene, C= 30% untreated pure hemp mat-30% styrene, D=30% untreated
pure hemp mat-40% styrene

composite, the tensile strength was 20 % higher than that of 10 wt % extra styrene

201

biocomposite, and its tensile modulus was 22 % higher than that of 10 wt 0/o extra styrene
biocomposite.

The impact strength of 20 wt % extra styrene-hemp ﬁber composite was 23 % higher
than that of 10 wt % extra styrene biocomposite (Figure 4.2.6). In the case of 30 wt %
extra styrene-hemp ﬁber composite, the impact strength was 58 % higher than that of 10
wt % extra styrene biocomposite. Whereas, for the 40 wt % extra styrene-hemp ﬁber
composite, the impact strength was almost same as that of 10 wt % extra styrene

biocomposite within error bars.

 

 

 

 

A12
E

'3

.ﬂ

‘5‘ i

58

CI)

E f
a, I

54-

'O

O

.5

'D

0

2":

*50 . . .
Z

A B C D

Figure 4.2.6: Optimization of styrene content by evaluation of mechanical properties
Legend: A=30% untreated pure hemp mat-10% styrene, B= 30% untreated pure hemp
mat-20% styrene, C: 30% untreated pure hemp mat-30% styrene, D=30% untreated pure
hemp mat-40% styrene

Therefore, it was decided that 30 wt % extra styrene was optimum for biocomposites

made of pure hemp mats and UPE resin. For all biocomposites made henceforth with

pure hemp mat, 30 wt % extra styrene was added to the matrix.

202

Tensile Properties

Figure 4.2.7 shows the comparison of tensile properties of various composites. The
tensile properties of biocomposites with surface treated hemp ﬁbers were higher than that
of untreated hemp mat composite as well as that of neat resin. The tensile strength and

modulus of untreated pure hemp ﬁbers based biocomposite was 45 % and 222 % higher

200

..r r—b
00 N 0‘
C O O

Tensile Strength, TS (MPa)
4)
O

Tensile Modulus, TM (GPa)

 

A B C - D E F G

Figure 4.2.7: Comparison of tensile properties of surface treated pure hemp mat
based composites

Legend: A= UPE control, B: Untreated pure hemp mat (3O % vol)—UPE, C= Alkali
treated pure hemp mat (30 % vol)—UPE, D= UPE-MEKP treated pure hemp mat (3O %
vol)—UPE, E= Silane treated pure hemp mat (30 °/o vol)—UPE, F=E-glass mat-UPE,
G=E-glass mat-pure hemp mat-UPE

than that of neat resin, respectively. For alkali treated hemp ﬁbers based biocomposite,
the tensile strength and modulus was 81 % and 293 % higher than that of neat resin,
respectively. For UPE-MEKP treated hemp ﬁbers based biocomposite, the tensile
strength and modulus. was 70 % and 284 % higher than that of neat resin, respectively.
For silane treated hemp ﬁbers based biocomposite, the tensile strength and modulus was

62 % and 287 % higher than that of neat resin, respectively.

203

Comparing surface treated ﬁber based composites, the tensile strength of alkali treated
ﬁber based biocomposite was 25 % higher than untreated hemp ﬁber based composite,
while that of silane treated ﬁber based biocomposite was 12 % higher than untreated
hemp, whereas for UPE-MEKP treated hemp ﬁber based biocomposite it was 18 %
higher than untreated hemp ﬁbers. In terms of tensile modulus, alkali treated ﬁber had an
enhancement of 22 % as compared to untreated hemp, while silane treated and UPE-
MEKP treated ﬁbers had an enhancement of 20 % each compared to untreated hemp.

The tensile strength and modulus of E-glass-UPE composite was 232 % and 620 %
higher than that of neat resin, respectively. The tensile strength and modulus of hemp
mat-E-glass-UPE hybrid composite was 112 % and 471 % higher than that of neat resin,
respectively. The tensile strength of E-glass-UPE composite was 129 % higher, and the
tensile modulus was 124 % higher as compared to untreated hemp mat biocomposite. The
hybrid E-glass-hemp mat-UPE composite had an increment of 47 % in tensile strength,
and 78 % in tensile modulus compared to untreated hemp mat based biocomposite. The
E-glass-UPE composite had 56 % higher tensile strength and 26 % higher tensile
modulus as compared to hybrid composite of E-glass and hemp mats. On comparing
speciﬁc tensile strength (tensile strength divided by density of composite) and speciﬁc
tensile modulus (tensile modulus divided by density of composite) of all composites, it
was found that the biocomposites and glass composites were in the same range.

As discussed in Section 4.1, the increase in tensile strength and modulus may be
athibuted to the fact that due to chemical treatment of hemp ﬁber, the adhesion between
the ﬁber and the matrix is improved and the interfacial bonding is adequate making it

easier for the stress to be effectively transferred from the matrix to the ﬁber.

204

Flexural Properties

Flexural strength is a combination of the tensile and compressive strengths which directly
varies with the interlaminar shear strength. In ﬂexural testing various mechanisms such
as tension, compression, shearing etc. take place simultaneously. In a three point ﬂexural
test, the failure occurs due to bending and shear failure. The bending strength and

modulus of elasticity of different composites are compared in Figure 4.2.8.

 

 

 

       

 

400 20

EBSOMOE I;
' at
A a
£30m g
m 2
an Q
€2,200 ‘ é
a a
b c...
m G
31004 g,
'5 5
5 9 '8
=9 E

O T T I T
A B C D E F G

Figure 4.2.8: Comparison of ﬂexural properties of surface treated composites

Legend: A= UPE control, B= Untreated pure hemp mat (30 % vol)—UPE, C= Alkali
treated pure hemp mat (30 % vol)—UPE, D= UPE-MEKP treated pure hemp mat (30 %
vol)—UPE, E= Silane treated pure hemp mat (30 % vol)—UPE, F=E-glass mat—UPE,
G=E-glass mat-pure hemp mat-UPE

The ﬂexural properties of biocomposites with surface treated hemp ﬁbers were higher
than that of untreated hemp mat composite; also, the ﬂexural properties of all composites

were higher than that of neat resin.

205

Comparing surface treated ﬁber based composites, the modulus of elasticity as well as
bending strength of alkali treated, silane treated and UPE-MEKP treated hemp ﬁber
based biocomposites lie in the same range.

As compared to UPE control, the surface treated ﬁber based biocomposites had a 10-24%
enhancement in bending strength, and 140-172% increment in modulus of elasticity. The
bending strength of E-glass-UPE composite was 165 % higher, and the modulus of
elasticity was 273 % higher as compared to UPE control. The bending strength of E-
glass-UPE composite was 139 % higher, and the modulus of elasticity was 54 % higher
as compared to untreated hemp mat biocomposite. The hybrid E-glass-hemp mat-UPE
composite had a bending strength 70 % higher, and modulus of elasticity 280 % higher
than the UPE control. The hybrid E-glass-hemp mat—UPE composite had a bending
strength 53 % higher, and modulus of elasticity 57 % higher than the untreated hemp mat
based biocomposite. The E-glass-UPE composite had 56 % higher bending strength as
compared to hybrid composite of E-glass and hemp mats, but its modulus was same as
the hybrid composite within error bars. On comparing speciﬁc bending strength and
speciﬁc modulus of elasticity of all composites, it was found that the biocomposites and
glass composites were in the same range. As can be observed from Figures 4.2.7 and
4.2.8, the trends in ﬂexural and tensile properties were similar.

Again, the increment in the ﬂexural properties of biocomposites after surface treatment of
the hemp ﬁbers can be attributed to the modiﬁcations in the molecular level of the ﬁber

which lead to better interfacial bondage, better adhesion, and effective stress transfer.

Impact Strength

206

The ﬁbers play a very important role in the impact resistance of the composite as they
interact with the crack formation in the matrix and act as stress transferring medium. It is
a common observation that with any surface treatment, the ﬂexural and tensile properties

of the composite increase, but the impact strength (shown in Figure 4.2.9) will decrease.

 

      
 

 

 

 

 

1 80
E 3323‘ ""35:
3 150 i 552:5;
.t:
‘63
5 120 —
I:
m
3 90 r :2 2
~ 60* E
'° %
° E
5 E
a 3‘“ E
"3 I—I-l EI—I |—:|:—l E
*3 0 FBI , [*1 , T 1 . E
Z

A B C D E F G

Figure 4.2.9: Impact strength of surface treated composites

Legend: A: UPE control, B: Untreated pure hemp mat (30 % vol)—UPE, C= Alkali
treated pure hemp mat (30 % vol)—UPE, D= UPE-MEKP treated pure hemp mat (30 %
vol)—UPE, E= Silane treated pure hemp mat (30 % vol)—UPE, F=E-glass mat-UPE,
G=E-glass mat-pure hemp mat-UPE

The impact strength of biocomposites with surface treated hemp ﬁbers was higher than
that of untreated hemp mat composite; also, the impact strength of all composites were
higher than that of neat resin. As compared to neat resin, there was an increment of 8 %
in impact strength of untreated hemp ﬁber based composites, 36 % for alkali treated
ﬁbers, 30 % for silane treated ﬁbers, and 21 % for UPE-MEKP treated ﬁbers based
biocomposites. On comparing the impact strength of untreated and surface treated

biocomposites, it was found that the impact strength of alkali treated ﬁbers was 26 %

207

more than unheated ﬁbers, while that of UPE-MEKP heated ﬁbers was 21 % more than
unheated ﬁbers, and that of silane heated ﬁbers was 12 % more than untreated ﬁbers.

The impact strength of E-glass-UPE composite was 16.3 times higher than neat resin and
15 times higher than unheated hemp ﬁber based biocomposite. The hybrid E-glass-hemp
mat-UPE composite had impact strength nine times higher than neat resin and 8.27 times
higher than unheated hemp ﬁber based biocomposite. The impact shength of E-glass-
UPE composite was 73 % higher than that of hybrid composite of E-glass mats and hemp

mats.

Dynamic Mechanical Analysis

Dynamic mechanical methods expose the specimen to periodic stresses. The polymer is

 

 

— UPE corlrol — Untreated lerrp net-UPE
— Alkali treated lemp net-UPE — UPE-MEKP heated lenrp net-UPE
~—-—- Sihre heamd lenp net-UPE — Unheated herrp net- E-glass-UPE
\--——- E glass-UPE
12 T

l
10 e

W
1

Storage Modulus (GPa)
Os

 

 

 

 

 

 

 

 

4 ..

2 ~ I,“

0 l f l
20 60 Temnm (”C) 100 140

Figure 4.2.10: Typical storage modulus curves of surface heated composites

208

subjected continuously to forced oscillations, and the applied stress is sinusoidal with a
frequency. The deformation of ideal-elastic bodies follows the stress instantaneously but
that of viscoelastic polymer experiences a delay. The stress vector is assumed to be a sum
of two components: one component is in phase with the deformation; the other is not.
Each of these two components possesses a modulus. The real modulus (shear storage
modulus) 6' measures the stiffness and shape stability of the specimen, whereas the
imaginary modulus (shear loss modulus) G" describes the loss of usable mechanical
energy by dissipation into heat. The maximum of tan 8 as a function of temperature is
generally identiﬁed as the glass transition temperature Tg, which is dependent on the
deformation rate. The damping properties of the material are related to energy absorption.
The area below tan delta is function of the absorption of the energy necessary to pass
from glassy state to rubbery state.

The typical curves of storage modulus and tan delta for neat resin and composites can be
seen in Figures 4.2.10 and 4.2.11 respectively. The storage modulus decreased as a
function of temperature, as is commonly observed for composites. At higher
temperatures, all biocomposites plateau to the same value of modulus. The storage
moduli of surface treated hemp ﬁbers based biocomposites were higher than that of
untreated hemp ﬁber based biocomposite. A comparison of storage moduli of composites
at 40 0C was made in Figure 4.2.12. The storage modulus of biocomposites at 40 0C was
enhanced by 135-225 % compared to neat resin. As compared to neat resin at 40 0C,
there was an increment of 135 % in storage modulus of untreated hemp ﬁber based
composites, 177 % for alkali treated ﬁbers, 190 % for silane treated ﬁbers, and 169 % for

UPE-MEKP treated ﬁbers based biocomposites. On comparing the storage modulus of

209

untreated and surface treated biocomposites at 40 0C, it was found that the storage
modulus of alkali treated ﬁbers was 18 % more than unheated ﬁbers, while that of UPE-
MEKP treated ﬁbers was 15 % more than untreated ﬁbers, and that of silane treated
ﬁbers was 24 % more than untreated ﬁbers.

The storage modulus of E-glass-UPE composite at 40 0C was 307 % higher than neat
resin and 74 % higher than untreated hemp ﬁber based biocomposite. The hybrid E-glass-
hemp mat-UPE composite at 40 0C had storage modulus 225 % higher than neat resin and
39 % higher than unheated hemp ﬁber based biocomposite. Glass mat based composite

had a storage modulus 25 % higher than hybrid hemp ﬁber-glass based composite at 40

 

 

 

 

°C.

— UPE conhel — Unheated pure herrp net-UPE

— Akali heated hemp net-UPE — UPE-MEKP heated hemp net-UPE

w ~ Sihne heated hemp net-UPE — Unheated pure lenp net-E glass-UPE
E ghss-UPE

0.48 1 _L

0.36 -
:2
S
a 0.24 -
a
I-r

0.12 ~

0 T l I

 

0 100 140
Temperature (”C)

Figure 4.2.11: Typical tan delta curves of surface heated composites

210

The loss modulus of composites was higher than that of the neat resin. This follows the
trend found previously for natural ﬁber-thermoset composites, where the loss modulus

increased after addition of ﬁbers to the plastic.

 

 

 

 

 

 

 

 

4s
21 ‘5
O r
A G

Figure 4.2.12: Storage modulus of surface treated composites at 40 0C

Legend: A: UPE conh'ol, B= Untreated pure hemp mat (30 % vol)—UPE, C=
Alkali treated pure hemp mat (30 % vol)—UPE, D= UPE-MEKP treated pure hemp
mat (30 % vol)—UPE, E= Silane treated pure hemp mat (30 % vol)—UPE, F=E-
glass mat-UPE, G=E-glass mat-pure hemp mat-UPE

Over the entire range of temperature, tan 6 was highest for the neat resin due to huge
reduction in the storage modulus values at higher temperatures. The lower values of tan 8
for the biocomposite made with surface treated hemp ﬁbers suggest that there is less
damping in the chemically treated hemp ﬁber based composites. The tan 6 vs.
temperature plot for biocomposites as well as E-glass composite is similar. The tan delta

curve of all composites shifted towards right as compared to neat resin, but this shift was

211

very big in case of glass based composite. The Tg of the neat resin was 95 0C. For
biocomposites, the Tg increased about 2-5 0C as compared to neat resin.
Morphology of ﬁber surface

The micrographs of untreated and surface treated pure hemp ﬁbers are shown in Figure

 

Figure 4.2.13: ESEM micrographs of surface treated pure hemp mat ﬁbers
a) Untreated pure hemp mat, magniﬁcation 1000 X, scale bar 45 um
b) Alkali treated pure hemp mat, magniﬁcation 1000 X, scale bar 45 um
c) Silane treated pure hemp mat, magniﬁcation 1000 X, scale bar 45 pm
(1) Unsaturated polyester treated pure hemp mat, magniﬁcation 1000 X, scale bar 45
um

4.2.13. The dishibution of the ﬁbers in the hemp mat was random, and uneven. The

212

micrographs in Figure 4.2.13 focus on single ﬁber surface. Fibrillation was observed in
ﬁbers after surface treatment. This could provide more anchorage for the matrix, and
hence improve the strength of the composite. In general, the surface of chemically treated
ﬁbers looked different from that of the untreated hemp ﬁber.

In the biocomposites, the ﬁber pull out was clearly observed (Figure 4.2.14).
Biocomposite with untreated hemp ﬁbers showed poor interfacial bonding between the
ﬁber and matrix, which resulted in relatively clean surface over the pulled out ﬁbers due
to greater extent of delarrrination. In case of untreated ﬁber based biocomposites, shear
failure results in high degree of pull out. The adhesion between the ﬁber and the matrix
was enhanced in biocomposites with surface treated ﬁbers. The ﬁbers were covered with

matrix, and the ﬁber pull out was relatively smaller.

213

 

Figure 4.2.14: ESEM micrographs of tensile fractured surfaces of biocomposites from
pure hemp mat ﬁbers
a) Untreated pure hemp mat, magniﬁcation 100 X, scale bar 450 um
b) Alkali treated pure hemp mat, magniﬁcation 100 X, scale bar 450 um
c) Silane treated pure hemp mat, magniﬁcation 100 X, scale bar 450 um
d) Unsaturated polyester treated pure hemp mat, magniﬁcation 100 X, scale bar 450
um

4.2 Effect of surface modification on aligned kenaf fibers

214

The effects of the following surface treahnents on the thermal, mechanical and physical

porpoerties of the kenaf-UPE composites are discussed in this section: alkali h'eahnent, y-

 

yluxnnvwlh' ‘L y silane (y-MPS) heahnent, acrylonihile treatment, and

UPE-MEKP heatrnent.

Analysis of surface treated hemp ﬁbers
The surface treated ﬁbers were characterized with TGA, DSC, and XPS.
Thermogravimetric analysis (T GA)

Thermogravimeh'ic curves for untreated and surface treated hemp ﬁbers are shown in

 

— Water Washed Kemf — AkaIi heated Kenaf — UPE-MEKP
-— Sihne Treated Kemf -— AN -DCP Kenaf
100 H
9. l
80 ~
70 a

Weight(%)
Ur
C

 

 

 

40 ~
30 H
20 a
10 a
0 . , I f I i
0 100 200 300 400 500 600

Temperature (0 C)
Figure 4.3.1: TGA of surface heated kenaf ﬁbers

Figure 4.3.1 and 4.3.2. Table 4.3.1 summarizes the maximum decomposition

temperatures for unheated and surface treated kenaf ﬁbers. Up to 200 0C, there was less

215

than 7 % loss in weight of untreated as well as surface heated kenaf ﬁbers. After the
alkali heahnent, the temperature at the maximum rate of decomposition of kenaf ﬁber
increased, indicating that the alkali heahnent lead to an enhancement in the thermal
stability of the hemp ﬁber. Weight loss of 10% occurred between 272-310 0C for the
untreated and surface heated ﬁbers, while a weight loss of 20% occurred in the range of

313-353 °C, and weight loss of 30% was observed in range of 340-373 0C for untreated

 

 

 

 

and surface heated kenaf ﬁbers.
— Water Washed Kenaf — Akali Treated Kemf — UPE Treated Kenaf
—- Sihne Treated Kemf — Acrylonirih Treated Kemf
1.6 ”—7 -_, “—7- __,_ 1
1.4 ~ 1
Q 1.2 «
\° .
°: 1 - 1
l 1
if 0.8 . 1
o 0.6 r §
to... 1
.E .
a 0.2 T i
0a i
.02 l r . T . . . 4
0 100 200 300 400 500 600 700
Temperature (0 C)
Figure 4.3.2: Derivative thermogravirnetric analysis (DTGA) of surface heated kenaf
ﬁbers

The maximum decomposition temperatures ranged from 364-395 0C for untreated and

surface treated kenaf ﬁbers.

216

Table 4.3.1: Thermogravimetric results for surface treated kenaf ﬁbers

 

 

 

 

 

 

10% Weight 20% Weight 30% Weight Max. degradation

loss (°C) loss (°C) loss (°C) temperature (°C)
A 275 323 353 389
B 310 353 373 395
C 294 330 353 379
D 272 313 340 364
E 299 343 371 392

 

 

 

 

 

 

 

Legend: A: Water washed kenaf ﬁbers, B= Alkali treated kenaf ﬁbers, C= Acrylonihile
treated kenaf ﬁbers, D= UPE-MEKP treated kenaf ﬁbers, E: Silane treated kenaf ﬁbers

Differential scanning calorimetry (DSC)

The DSC curves for untreated and surface treated ﬁbers expressed in terms of heat ﬂow
are shown in Figure 4.3.3. It is observed from Figure 4.3.3 that the crest associated with
weight loss due to evaporation of water was found in all ﬁber samples, and ranged from
6-140 0C in kenaf ﬁbers, depending upon the degree of hydrogen bonding interaction.
Furthermore, this transition shifted towards the right after surface heahnent of the kenaf
ﬁber. It was also observed that ﬁbers had less amount of moisture percentage because
they had been pre-dried before any kind of treatment. This ensured that the inherent
moisture of the ﬁbers, which is 12-14% of the total ﬁber weight, was not able to interact
with the surface treatments in any way. Therefore, the chemicals of the surface treahnents
had an opportunity to react with the actual ﬁber surface and ﬁber constituents. This
would lead to a direct link between the chemicals used and the ﬁber constituents such as
cellulose and hemicelluse. The Tg for lignin and cellulose were not observed by this
particular method in DSC. No exothermic peaks were observed from 300 0C to 500 0C,

which are indicative of the thermal stability of the ﬁbers as a function of the treatment.

217

-— Water Wasted Kenaf — Alkali heated Kenaf — UPE-MEKP Kenaf
— Sihre Treated Kemf — AN-DCP Kemf

 

Heat Flow (mW)

 

 

 

-3 +——__ ﬂ . T .
-70 -10 50 110 170 230 290
Temperature(“ C)

Figure 4.3.3: DSC of surface heated kenaf ﬁbers

X-ray photon spectroscopy (XPS)

XPS survey scans were taken for unheated as well as surface heated kenaf ﬁbers. These
scans revealed the presence of carbon, oxygen, nitrogen, calcium, and silicon in the kenaf
ﬁbers. Table 4.3.2 and 4.3.3 show the elemental composition and elemental ratios of the
hemp ﬁbers, respectively. Following an hour long heahnent with alkali solution it is
observed that the carbon content increases, while oxygen and nitrogen content decrease.
After silane heahnent, there is a marked increase in silicon as well as carbon content, and
a decrease in oxygen and nihogen content. Again, after heahnent with the mah'ix, there is

an increase in carbon content, while a decrease in oxygen and nihogen content. However,

218

with acrylonitrile treatment, there is a decrease in carbon content and an increase in
nitrogen and oxygen content.

Due to the alkali heahnent, a part of hemicellulose and lignin might have dissolved, and
washed away, leading to a decrease in oxygen content. The increase in carbon and silicon
content after silane treatment might be due to the attachment of bulky alkyl group and
silicon to the hemp ﬁber as a result of this heahnent. The large increase in carbon content
of UPE treated hemp ﬁber can again be due to the large oligomer molecule of UPE
reacting with hydroxyl groups of ﬁber. Meanwhile, the increase in nitrogen content after
treatment with acrylonitrile might be due to grafting of acrylonitrile monomer to the ﬁber
surface.

Table 4.3.2: Elemental composition of surface treated hemp ﬁbers (from XPS analysis)

 

 

 

 

 

 

 

Cls[.314] N1s[.499j Ols[.733] Si2p[.368j Ca3pl1.927]
A 73.8 1.55 23.4 0.75 0.5
B 75.9 1.6 21.7 0.3 0.5
c 71.01 2.94 25.32 0.32 0.46
D 80.41 1.48 17.28 0.35 0.48
E 78.19 1.52 16.63 3.42 0.24

 

 

 

 

 

 

 

Legend: A: Water washed kenaf ﬁbers, B= Alkali treated kenaf ﬁbers, C= Acrylonihile
treated kenaf ﬁbers, D= UPE-MEKP treated kenaf ﬁbers, E= Silane treated kenaf ﬁbers

Table 4.3.3: Elemental ratios of surface treated hemp ﬁbers (from XPS analysis)

 

 

 

 

 

 

 

 

 

 

 

 

C/O C/N C/Si C/Ca
A 3.15 47.61 98.4 147.6
B 3.5 47.44 253 151.8
C 2.8 24.15 221.9 154.37
D 4.65 54.33 229.74 167.52
E 4.7 51.44 22.86 325.79

 

Legend: A: Water washed kenaf ﬁbers, B= Alkali treated kenaf ﬁbers, C= Acrylonihile
heated kenaf ﬁbers, D= UPE-MEKP treated kenaf ﬁbers, E= Silane treated kenaf ﬁbers

Flexural Properties

219

The effect of various surface treahnents on performance of biocomposites is illustrated in

Figure 4.3.4.The bending strength and modulus of elasticity all increase with use of

surface heated ﬁbers as compared to raw kenaf ﬁbers; also, the ﬂexural properties of all

composites were higher than that of neat resin. The highest values for shength and

modulus of the biocomposites are obtained for acrylonitrile treated kenaf ﬁber based

composite.

 

500
El BS(MPa) O MOE(Gpa)

400 4

 

 

 

 

 

 

 

 

  
   

  
       
     
     
         

\\\\\\\\\\\

\\

\\

  

 

100 .1
|-;-
O , 1 r - 1
B C

A D E

   

25

1
N
O

G
MOE(GPa)

~10

 

Figure 4.3.4: Comparison of flexural properties of surface treated kenaf based composites
Legend: A= UPE control, B= Water washed kenaf (30 % vol)—UPE, C= Alkali treated
kenaf (30 % vol)—UPE, D= Acrylonihile treated kenaf (30 % vol)—UPE, E= UPE-MEKP
heated kenaf (30 % vol)—UPE, F= Silane heated kenaf (30 % vol)-UPE, G=Aligned E-

glass-UPE

Comparing surface treated ﬁber based composites, the bending strength of alkali treated,

silane treated and UPE-MEKP heated kenaf ﬁber based biocomposites lie in the same

range. However, the modulus of elasticity and bending shength of acrylonitrile treated

ﬁbers based composites were 20 %, and 15 % higher than unheated kenaf ﬁber

220

biocomposite respectively. As compared to UPE control, the biocomposites had 12-28 %
enhancement in bending strength, and 125-175 % increment in modulus of elasticity.

The bending strength of water washed kenaf based biocomposite was 12 % higher, and
the modulus of elasticity was 128 % higher as compared to UPE control. The bending
strength of alkali heated kenaf-UPE biocomposite was 17 % higher, and the modulus of
elasticity was 146 % higher as compared to UPE control. The UPE-MEKP heated kenaf-
UPE biocomposite had the bending strength 21 % higher, and the modulus of elasticity
176 % higher as compared to UPE control. The silane treated kenaf-UPE biocomposite
had the bending strength 24 % higher, and the modulus of elasticity 161 % higher as
compared to UPE control. Whereas, the bending strength of acrylonitrile treated kenaf
based biocomposite was 28 % higher, and the modulus of elasticity was 172 % higher as
compared to UPE control.

The bending shength of alkali treated kenaf-UPE biocomposite was 5 % higher, and the
modulus of elasticity was 8 % higher as compared to water washed kenaf based
biocomposite. The UPE-MEKP heated kenaf—UPE biocomposite had the bending
strength 8 % higher, and the modulus of elasticity 20 % higher as compared water
washed kenaf based biocomposite. The silane treated kenaf-UPE biocomposite had the
bending shength 11 % higher, and the modulus of elasticity 15 °/o higher as compared to
water washed kenaf based biocomposite.

The bending strength of E-glass-UPE composite was 313 % higher, and the modulus of
elasticity was 5.7 times higher as compared to UPE control. The bending strength of E-
glass—UPE composite was 270 % higher, and the modulus of elasticity was 194 % higher

as compared to water washed kenaf based biocomposite. The E-glass-hemp mat-UPE

221

composite had a bending strength 222 % higher, and modulus of elasticity 147 % higher
than the acrylonihile treated kenaf—UPE biocomposite.

On comparing speciﬁc modulus of elasticity (modulus divided by the density of
composite) of all composites, it was found that the biocomposites and glass composites
were in the same range (Figure 4.3.4). Whereas, the speciﬁc bending strength of all
biocomposites were in the same range, while that of aligned E-glass composites was

higher than that of biocomposites.

 

200 25
Speciﬁc Strength A Speciﬁc Modulus

160 e

..-

N

O
1

BS(MPa/g/cc)
8
MOE(GPa/g/cc)

40—

 

 

 

 

 

Figure 4.3.6: Speciﬁc F lexural Properties of Kenaf-UPE composites,
Legend: A= Water wasth kenaf (30 % vol)—UPE, B= Alkali treated kenaf (30 % vol)—
UPE, C= Acrylonitrile treated kenaf (30 % vol)—UPE, D= UPE-MEKP treated kenaf
(30 % vol)—UPE, E= Silane treated kenaf (30 % vol)-UPE, F=A1igned E-glass-UPE

The increment in the ﬂexural properties of biocomposites after surface heahnent of the

hemp ﬁbers is again attributed to the modiﬁcations in the molecular level of the ﬁber

which led to better interfacial bondage, better adhesion, and effective stress transfer.

222

Impact Strength

Acrylonihile treated kenaf ﬁbers based biocomposites also show the highest impact
strength compared to other biocomposites (shown in Figure 4.3.5). The impact strength
of all composites was higher than that of neat resin. The notched Izod impact strength of
biocomposites was enhanced by 330-420 % compared to neat resin. As compared to UPE
control, there was an increment of 385 % in impact strength of water washed kenaf ﬁber
based composites, 331 % for alkali heated ﬁbers, 352 % for silane heated ﬁbers, 375 %
for UPE-MEKP heated ﬁbers based biocomposites, and 420 % for acrylonihile treated
ﬁbers based biocomposites. On comparing the impact shength of water washed and
surface heated biocomposites, it was found that the impact shength of alkali heated
ﬁbers was 11% less than untreated ﬁbers, while that of UPE-MEKP heated ﬁbers was
2 % less than unheated ﬁbers, and that of silane heated ﬁbers was 7 % less than untreated
ﬁbers. On the other hand, the impact strength of acrylonihile heated kenaf ﬁber based
biocomposite was 7 % higher than water washes kenaf composite.

The impact shength of E-glass-UPE composite was 57.7 times higher than neat resin and
11.1 times higher than water washed kenaf ﬁber based biocomposite. The acrylonitrile
heated kenaf ﬁber based biocomposite had impact strength 10.3 times lower than the

aligned E-glass based composite.

223

 

600

500 e

.h

0

O
1

Impact Strength(J/m)
N b.)
o o
o o

p—a

O

O
l

 

 

           

;

 

A B C D E F G

Figure 4.3.5: Notched Izod Impact strengths of kenaf-UPE composites

Legend: A= UPE conhol, B= Water washed kenaf (30 % vol)—UPE, C= Alkali heated
kenaf (30 % vol)—UPE, D= Acrylonihile heated kenaf (30 % vol)—UPE, E= UPE-MEKP
heated kenaf (30 % vol)—UPE, F= Silane heated kenaf (30 % vol)-UPE, G=Aligned E-
glass-UPE

Dynamic Mechanical Analysis (DMA)

The typical curves of storage modulus and tan delta for neat resin and composites can be
seen in Figures 4.3.6 and 4.3.7 respectively. The storage modulus decreased as a
function of temperature, as is commonly observed for composites. At higher temperatures,
all biocomposites plateau to the same value of modulus. The storage moduli of surface
heated kenaf ﬁbers based biocomposites were higher than that of untreated kenaf ﬁber
based biocomposite.

The acrylonitrile heated kenaf ﬁber based biocomposite has the highest storage modulus.
At 40 0C, the storage modulus of all composites and biocomposites is greater than that of

UPE conhol (see Figure 4.3.8). The storage modulus of biocomposites at 40 0C was

224

25 ——r——

 

— UPE control

——~ Water Washed Kenaf-UPE
— Alcali heated Kemf-UPE

— UPE/MEKP heated Kenaf-UPE
— Silane Treated Kemf-UPE

— AN/DCP heated Kemf-UPE
15 - —Glass—UPE

20*

Storage Modulus (GPa)

 

 

 

Temperature (0 C)
Figure 4.3.6: Typical storage modulus curves of surface heated kenaf composites
enhanced by 170-270 % compared to neat resin. As compared to neat resin at 40 0C, there

was an increment of 204 % in storage modulus of water washed kenaf ﬁber based
composites, 175 % for alkali heated ﬁbers as well as UPE-MEKP heated ﬁbers based
biocomposites, 192 % for silane treated ﬁbers, and 270 % for acrylonihile heated kenaf
ﬁber based biocomposites. On comparing the storage modulus of unheated and surface
heated biocomposites at 40 0C, it was found that the storage modulus of acrylonihile
heated kenaf ﬁber was 21 % more than unheated ﬁbers.

The storage modulus of E-glass-UPE composite at 40 0C was 6.7 times higher than neat
resin and 152 % higher than unheated hemp ﬁber based biocomposite. Glass ﬁber based
composite had a storage modulus 109 % higher than acrylonihile heated kenaf ﬁber

based biocomposite at 40 0C.

225

The storage modulus of the UPE conhol and the UPE resin reinforced with kenaf ﬁbers
at 40 0C follow the same hend as the ﬂexural modulus. At higher temperatures, the
storage modulus values of the composites and plastic decreased. Treating the surface
kenaf ﬁber with alkali, silane, acrylonihile and UPE led to an improvement in the

mechanical and thermal properties of the biocomposite.

 

 
   
  

 

 

 

0.5 — —.
— UPE Corlrol
0.45 J -—-- Water Washed Kernf-UPE
— Altali Treated KemflUPE
0'4 i -— UPE/MEKP Treated KeranUPr-z
0 35 4 — Sihre Treated KemflUPE
=3 ' — Acrybnrrre Treated Kemf-UPE
g 0.3 1 — Glass-UPE
5
I- o.25 1
0.2 ~
0.15 1
0.1 1
0.05 -
0 f I 1 ﬁ ﬂ 1 7*
0 20 40 60 80 100 120 140 160

Tenperature(n C)
Figure 4.3.7: Typical tan delta curves of surface heated kenaf composites
The loss modulus of composites was higher than that of the neat resin. This follows the

hend found previously for natural ﬁber-thermoset composites, where the loss modulus
increased after addition of ﬁbers to the plastic.

The maximum of tan 6 as a ftmction of temperature is generally identiﬁed as the glass
hansition temperature T8, which is dependent on the deformation rate. The damping

properties of the material are related to energy absorption.

226

The area below tan delta is function of the absorption of the energy necessary to pass
from glassy state to rubbery state. Over the entire range of temperature, tan 5 was highest
for the neat resin due to huge reduction in the storage modulus values at higher
temperatures. The lower values of tan 6 for the biocomposite made with surface heated
kenaf ﬁbers suggest that there is less damping in the chemically heated hemp ﬁber based
composites. The tan 5 vs. temperature plot for biocomposites as well as E-glass

composite is similar. The tan 5 curve of all composites shifted towards right as compared

N
(Ir

 

N
O
1

p—o
LII
1

10*

Storage Modulus (GPa) at 40 0C

 

 

 

Figure 4.3.8: Storage modulus of Kenaf-UPE composites at 40 0C

Legend: A= UPE conhol, B= Water washed kenaf (30 % vol)—UPE, C= Alkali heated
kenaf (30 % vol)—UPE, D= Acrylonitrile heated kenaf (30 % vol)—UPE, E= UPE-MEKP

heated kenaf (30 % vol)—UPE, F= Silane heated kenaf (30 % vol)-UPE, G=Aligned E-
glass-UPE

to neat resin, but this shift was signiﬁcant in case of glass based composite. The T8 of the
neat resin was 95 0C. For biocomposites, the Tg increased about 2-5 0C as compared to

neat resin.

227

The improvement in properties after surface treatment can be explained by the fact that
UPE resin sizing creates a coating on the uneven ﬁber surface thereby enhancing the
scope for improved ﬁber-matrix adhesion. The resin moves to all the free spaces
available in the ﬁber. When the composite is made, these chemically modiﬁed ﬁbers
have a better chance to adhere to the mahix. Therefore, the adhesion between ﬁber and
mahix is improved because resin can penetrate well through these ﬁbers, and the wet out

is better.

Morphology of ﬁber surface

The rrricrographs of unheated and surface treated kenaf ﬁbers are shown in Figure 4.3.9
and 4.3.10. The ESEM micrographs of impact fractured surface of UPE heated kenaf-
UPE composite shows smaller ﬁber pull out and good adhesion between ﬁber and matrix
as compared to water washed kenaf-UPE based biocomposite (Figure 4.3.10).
Unmodiﬁed composite shows poor interfacial bonding between the ﬁber and mahix. In
case of unheated ﬁber-polyester composites, shear failure results in high degree of pull
out. The ESEM micrographs quantify the claims that adhesion and interface bond

between ﬁber and mahix is improved by surface modiﬁcations.

228

 

Figure 4.8.9: SEM micrographs of surface heated kenaf ﬁbers
a) Water washed kenaf ﬁber, scale bar 20 um
b) Silane treated kenaf ﬁber, scale bar 20 um
c) Alkali heated kenaf ﬁber, scale bar 20 pm
(1) Acrylonihile heated kenaf ﬁber, scale bar 20 um

229

 

Figure 4.3.10: SEM nricrographs of surface treated kenaf ﬁbers based composites
a) Untreated kenaf-UPE, scale bar 200 um
b) Alkali heated kenaf-UPE, scale bar 200 mm
c) Water washed kenaf-UPE scale bar 200 pm
(1) Silane heated kenaf-UPE, scale bar 200 um
e) Acrylonitrile treated kenaf-UPE, scale bar 200 um
i) UPE-MEKP treated kenaf-UPE, scale bar 200 pm

230

REFERENCES

1.

D. N. Mahato, B.K. Mathur, S. Bhattacherjee, Indian Journal of Fiber Textile
Research, 20, (1995), 202.

D. S. Varma, M. Varma, I. K. Varma, T hermochimica Acta, 108, (1986), 199.
D.W. Van Krevelan, Properties of Polymers, 3rd Ed.; Elsevier: Amsterdam, 1990.

J. Brandrup, E. H. Immergut, Eds. Polymer Handbook, 3rd Ed; Wiley: New York,
1989.

D. Feldman, D. Banu, M. J. Khoury, Applied Polymer Science: Part C, 11,
(1965), 27.

H. Vazquez-Torres, G. Canche-Escamilla, C. A. Cruz-Ramos, Journal of Applied
Polymer Science, 45, (1992), 633.

G. Goulart Silva, D. A. De Souza, J. C. Machado, D.J. Hourston, Journal of
Applied Polymer Science, 76, (2000), 1197.

L.Y. Mwaikambo, M. P. Ansel], Die Angewandte Makromolekulare Chemie, 272,
(1999), 108.

A. K. Mohanty, M. Misra, G. Hinrichsen, Macromolecular Materials and
Engineering, 276/277, (2000), 1.

10. N. E. Zafeiropoulos, P. E. Vickers, C. A. Baillie, J. F. Watts, Journal of Material

Science, 38, (2003), 3903.

11. A. 1K. Mohanty, M. Misra, L. T. Drzal, Composite Interface, 8(9), (2001), 313.

12. A. K. Bledzki, J. Gassan, Progress in Polymer Science, 24(2), (1999), 221.

13. K. G. Satyanarayana, K. Sukumaran, A. G. Kulkami, S. G. K. Pillai, P. K.

Rohatgi, Composites, 17(4), (1986), 329.

14. G. Mehta, A. K. Mohanty, M. Misra, L. T. Drzal, Journal of Material Science, 39,

(2004), 2961.

15. A. K. Rana, B.C. Mitra, A.N. Banerjee, Journal of Applied Polymer Science, 71,

(1999), 531.

16. J. M. Zimmerman, N. S. Lousure, Journal of Advanced Materials, 30(2), (1998),

32.

231

17. S. N. Khot, J. J. LaScala, E. Can, S. S. Morye, G. 1. Williams, G. R. Palmese, S. H.
Kusefoglu, R. P. Wool, Journal of Applied Polymer Science, 82, (2001), 703.

18. M. A. Dweib, B. Hu, A. O’Donnell, H. W. Shenton, R. P. Wool, Composite
Structures, 63, (2004), 147.

19. A. O’Donnell, M.A. Dweib, R. P. Wool, Composites Science and Technology, 64,
(2004), 1135.

20. D. Ray, B. K. Sarkar, A. K. Rana, N. R. Bose, Composites: Part A, 32, (2001),
119.

21. J. Rout, S. S. Tripathy, M. Misra, A. K. Mohanty, S. K. Nayak, Journal of
Applied Polymer Science, 84, (2002), 75.

22. J. Rout, S. S. Tripathy, M. Misra, A. K. Mohanty, S. K. Nayak, Composites
Science and Technology, 61, (2001), 1303.

23.J. Rout, M. Misra, S. S. Tripathy, S. K. Nayak, A. K. Mohanty, Polymer
Composites, 22(4), (2001), 468.

24. C. Pavithran, P. S. Mukherjee, M. Brahmakumar, Journal of Reinforced Plastic
sand Composites 10(1), (1991), 91.

25. J. Gassan, A. K. Bledzki, Polymer Composites, 18(2), (1997), 179.

26. J. Gassan, A. K. Bledzki, Composites Science and Technology, 59(9), (1999),
1303.

232

CHAPTER FIVE: RESULTS ON MATRIX MODIFICATION

5.0 RESULTS ON MATRIX MODIFICATION

5.1 Curing Kinetics
The curing kinetics of the cross linking reaction of polyester resin and styrene were

studied by DSC and FTIR. Following are the results obtained from these experiments.

5.1.] DSC

Dynamic and isothermal experiments were carried out in DSC to study the curing
characteristics of the thermoset polyester resin, blends of polyester with initiator
methyl ethyl ketone peroxide and promoter cobalt naphthenate.

In order to obtain heat ﬂow curves for curing characterization, some experiments
were done under isothermal conditions. The DSC cell was allowed to stabilize at each
isothermal condition before inhoduction of the sample. Once an isothermal
experiment was over, the DSC was cooled quickly to room temperature, and when
stabilized, the residual heat of reaction of the sample was measured with a constant
heating rate until no exotherm is observed. During isothermal curing of thermoset
resins, it was assumed that the amount of heat generated is proportional to degree of
cure, or (or the extent of reaction) of the sample at that time.

So, the rate of curing, da/dt, was related to rate at which heat is evolved, dQ/dt, by:

dG/dt = (l/Qtot )(dQ/dt) -(1)

233

Therefore the relative degree of cure was determined by integrating equation (1) from
O to t:

a = (l/Qtot ) ol‘(dQ/dt)r dt -(2)
The total heat for curing reaction (Qto, ) was given by:

Qtot = Qiso+Qr -(3)

Where Qiso was the heat generated during isothermal DSC runs at each temperature,
and Qr was the residual heat released when the sample is heated up to a high
temperature (200°C) at a constant heating rate.
Some samples were also tested in dynamic DSC where, the sample was subjected to a
program which went from room temperature to a high temperature (200-350 0C) to
see the appearance of the exotherm peak on the heat ﬂow curve. This curve also gives
Or.
The plots from two different kinds of DSC experimental runs: isothermal runs, and
temperature sweeps, are showcased in Figures 5.5.1 to Figure 5.1.6. The ﬁrst two
ﬁgures are curing curves of temperature sweep for UPE conhol. The temperature
sweeps were done at heating rates of 5 OC/min and 10 0C/min. The rest three ﬁgures
are isothermal runs for UPE conhol at 50-1600C. For isothermal runs, the experiment
was carried out for at least 2 hours. The data from isothermal runs can also be plotted
as dQ/dt versus time. The last ﬁgure (Figure 5.1.6) shows the residual plots of the
isothermal experiments.
It can be observed from Figure 5.1.1 and 5.1.2 that the exothermic peak for curing of

UPE lies between 35 0C and 150 OC. Therefore, the curing cycle of UPE needs to be

234

set up in this temperature range. This particular range of temperature has also been

used by several authors for the curing of systems containing UPE [1-19].

 

DJ
1

Heat Flow (mW)

 

 

-l 1 1

 

 

I I I I

20 40 60 80 100 120 140 160
Temperature (0 C)

Figure 5.1.]: Temperature sweep for curing UPE from 25 0C to 160 0C

 

114

\l
1

Heat Flow (mW)
b.)

 

 

 

 

'5 I I I TI T I
O 50 100 150 200 250 300 350
Temperature (°C)

Figure 5.1.2: Temperature sweep for curing UPE from 35 0C to 300 0C

235

Figure 5.1.3, 5.1.4, and 5.1.5 show the isothermal cure plots of UPE at 50 °C, 60 °C,
70 °C, 80 °C, 90 °C, 100 °C, 110 °C, 120 °C, 130 °C, 140 °C, 150 °c and 160 °C. On
comparing these curves is observed that the heat ﬂow curve for curing UPE at 50 0C
has the lowest height. On increasing the temperature, the peak becomes narrower, and
increases in height. This indicates that the conversion of the UPE curing reaction is
increasing with temperature because the heat ﬂow curves are directly proportional to
dQ/dt, which in turn is directly proportional to ﬁactional conversion. The same hend
was observed in Figures 5.1.4 and 5.1.5. An interesting point to note ﬁom all these
ﬁgures was that the curves for 80 0C and higher, have a very steep initial slope. This
behavior again points out that the UPE systems at such temperatures were getting

cured very quickly.

——50C 60C —70C —80C

 

Heat Flow (mW) .

 

 

 

 

 

4000 6000 8000
Time (seconds)
Figure 5.1.3: Isothermal DSC cure of UPE at 50, 60, 70, 80 0C

236

—-90C --100C -- 110C -—120C
12

 

 

00
L

Heat Flow (mW)
.5

 

 

 

 

 

 

0 2000 4000 6000
Time (seconds)

Figure 5.1.4: Isothermal DSC cure of UPE at 90, 100, 110, 120 0C

 

 

 

 

 

 

—130C —-14OC —150C —160C
10 —
3
5
6 ..
5
E:
8
I t
2 _
f-
-2 T 1 r m
-500 500 1500 2500 3500
Time (seconth)

Figure 5.1.5: Isothermal DSC cure of UPE at 130, 140, 150, 160 0C
Figure 5.1.6 shows the residual heat curves for UPE systems at 50 °C, 130 0C, 140 0C,
150 0C, and 160 °C. Residual runs were done in order to ﬁnd out the heat of reaction

of any unreacted monomer that might not have consumed during the isothermal run at

237

a particular temperature. From the Figure 5.1.6, it can be observed that there was no
visible exothermic peak in the heat flow curves at all temperatures. The residual
temperature sweeps at other temperatures have not been shown. In fact, the residual
heat ﬂow curve in most cases was a ﬂat line. Therefore, the residual heat of reaction,
if any, was very small. Also, there was very little unreacted monomer present after

the isothermal curing had been performed for two hours.

 

 

 

 

 

 

 

 

 

—50C —110C ~130C —l40C --150C —l60C
2
0 _.
5
e -2-
B
E
ii '4 8
ﬂ:
-6 a
-8 r 1 r r 1
20 70 120 170 220 270

Temperature (0 C)
Figure 5.1.6: Residual heat ﬂow (temperature sweep) DSC UPE curing

Figures 5.1.7 and 5.1.8 show the comparison of isothermal heat of reaction and
fractional conversion at all temperatures studied in DSC. The curves in these two

plots were ﬁtted to obtain some kinetic constants.

238

800

 

 

 

 

y = 1.233 x + 523.49
R2 = 0.9858

700 I
30
H
E
o 600 —
.2
o

500 ~

400 . . . a r

50 70 90 110 130 150 170

Temperatum0 C)

Figure 5.1.7: Comparison of isothermal heat of reaction at different temperatures

 

  
   

1 a y = 0.0012 x + 0.8222
1?.2 = 0.9685

(1 (Conversion)
O
\O
.b
1

 

 

 

0.92 e
0.9 e
0.88 e
0.86 f I I T I I
50 70 90 110 130 150 170
Temperature (0 C)

Figure 5.1.8: Comparison of maximum conversions at different temperatures

The isothermal heat of reaction and the conversion are related to temperature by the
following equations:

Qiso = a + b T ‘(4)

a=c+dT -6)

239

The value of constants a, b, c and d was found out by plotting Q15, versus temperature,
and or versus temperature. Therefore, a = 523.49, b =1.233, c = 0.8222, and d =
0.0012. Thus, from our DSC data,

Q... = 523.49 + 1.233 T —(4)

a = 0.8222 + 0.0012 T -(5)

For an autocatalytic reaction, Kamal and Sourour approximated the DSC cure

reaction with the following empirical equation [1].

"CI—6: : (k1 +k2am Xau "aln

d’ -(6)

where m, n are the order of the reaction, kl and k2 are rate constants, and oru is the
maximum conversion.

And k= k0 exp (Ea/RT) -(7)

where E, is activation energy, R is universal gas constant and T is the absolute
temperature. In order to ﬁnd the values of k], k2, k0, Ea, m and n, a numerical method
has to be used.

From the isothermal plots, it is seen that conversion is directly dependent on
temperature. The higher the temperature, the better the conversion. The highest
conversion is seen at 160 0C. Similar results and conclusions were arrived at by
several authors [1-19].

DSC only gives an overall conversion of the UPE curing reaction. But, the UPE

system is very complex and the double bonds of UPE and styrene are consumed in

240

separate events in the curing reaction. Thus, to ﬁnd out the individual conversions of

the UPE and styrene, the curing reaction was studied by the use of FTIR.

5.1.2 F T IR

Isothermal experiments were carried out in F TIR to study the curing characteristics of
the UPE system which consisted of 99.7% of unsaturated polyester resin, 1.0% of
methyl ethyl ketone peroxide (MEKP) and 0.03% of cobalt naphthenate (CoNap).
The isothermal experiments were run at each temperature for four hours. The scans
were taken every 5-10 minutes. One drop of reaction mixture was placed between
two undrilled NaCl IR crystal windows, 2mm thick. The NaCl plates were then
placed inside the IR heated cell.

The major peaks to look forwarding the FTIR were: at 980 CH=CH2 of UPE (due to
=CH out of plane deformation), at 910 CH=CH— of styrene (due to =CH out of plane
deformation), at 842 C-H of UPE (due to CH2 out of plane wag), and at 700 C-H of
styrene (due to CH out of plane deformation) [20]. The peaks at 980 and 910 are
changing due to reaction, but the peaks at 842 and 700 remain constant through out
the experiment. Thus, normalized conversion of UPE and styrene can be found by the
formulae below:

<1 UPE (t) = 1' (AbS(t)980/ AbS(t=0)980) (Ab5(l=0)842/ AbS(t)842) '(8)

or 5,, (t) = 1- (Abs(t)910/ Abs(t=0)910) (Abs(t=0)7oo/ Abs(t)700) -(9)

Figures 5.1.9 and 5.1.10 show the FTIR specha of unreacted UPE and styrene
respectively, at room temperature. The peaks mentioned in the above paragraphs can

be clearly noted in these two ﬁgures.

241

was 83 mo 86on ”FE ”a.~.m 25E

 

 

 

 

 

cone—=53?»
oov com 00.1 83 oovm comm oovM ooom
_ _ _ p h _ L Owl
0.0.0
..2 0
C910 :0 i0 0
fo r 3
1
m
P. foo N
no; « m.
, _ £0qu .. w
Amozvxo
Cumv IIOHT—U 1000
om

 

 

242

oeocbm 0.5; we 88on Eula "3. am 95$...—

 

 

 

 

 

 

 

ATEoV a3855>3$
con 03 _ ooﬂ cog comm comm 00 H m 00m m comm
: _ _ L _ _ _ _ _ 0
lllll 1» 11,1;
1 om l
m
m
. ow m.
m
u
.. 8 a
%
. ow
ooﬁ

 

 

243

 

0
(It
1

—t=0min —t=250rnin

r“
25~ \ "I
T

h Ur
M M
1 n

Transmittance (%)
DJ
LII

 

 

 

 

 

\1
15 r
5 j I I T I I I\
1050 1000 950 900 850 800 750 700 650

Wave nunber (cm!)
Figure 5.1.1]: FTIR spectra of UPE system at 150 0C at t=0 minutes and t= 250 minutes.

Figure 5.1.11 shows the FTIR specha of UPE system at 150 0C at t=0 minutes and t= 250
minutes. At t=0 minutes, when the reaction has not yet started, the peaks of UPE system
at 980 and 910 cm'1 look very distinct. However, after t=250 minutes, when the reaction
has been completed, the peaks at 980 and 910 cm'1 are no longer visible. The FTIR
specha of UPE system at other temperatures, at the beginning and end of the reaction
look similar to Figure 5.1.11, and thus have not been shown.

Figures 5.1.12 through 5.1.23 show the fractional conversion of C=C bonds of UPE as
well as styrene at 50 °C, 60 °C, 70 °C, 80 °C, 90 °C, 100 °C, 110 °C, 120 °C, 130 °C, 140
0C, 150 0C and 160 0C respectively. The peak heights were normalized and calculated by
using the PerkinElmer Spechum 2000 software. The values of fractional conversion were

calculated using equations 7 and 8 on peak heights of all FTIR specha taken for four

hours.

244

 

 

 

 

 

 

 

Fractional Conversions at 50 0C

0-8 ‘ — UPE —— Styrene
,5 0.6 ~
(0
8
>
:1
5 0.4 ~

0.2 -

O I T I I
0 50 100 150 200 250

Time (min)

Figure 5.1.12: Fractional conversion at 50 0C from in situ FTIR study

 

 

 

 

 

 

 

 

Fractional Conversions at 60 0C I
v/
-— UPE -— Styrene
E
.2
8
>
:1
O
U
0 I l I I
0 50 100 150 200 250

Time (min)
Figure 5.1.13: Fractional conversion at 60 0C ﬁom in situ FTIR study

In Figure 5.1.12, fractional conversions of double bonds of UPE and styrene at different

times are shown for an isothermal experiment at 50 0C. The conversion curves were very

245

smooth, and the initial slopes of both curves were very gradual. At the beginning of the
reaction, the conversion of UPE was higher than that of styrene, but after 50 minutes, the
conversion of double bonds of styrene was more than those of UPE.

In Figure 5.1.13, fractional conversions of double bonds of UPE and styrene at different
times are shown for an isothermal experiment at 60 0C. Again the curves were smooth,
and the initial slope was gradual. As seen in ﬁgure 5.1.12, initially the conversion of UPE
double bonds were higher, however after 50 minutes, the conversion of double bonds of

styrene was more than those of UPE.

 

 

 

 

 

 

 

 

 

l
0.8 — —-UPE —-Styrene
t:
30.6 - 0
E 70 C
E
°0.4 i
U
0.2 r
O a ﬁ I I I
0 50 100 150 200 250 300

Time (nrinutes)

Figure 5.1.14: Fractional conversion at 70 0C from in situ FTIR study

Figure 5.1.14 shows fractional conversions of double bonds of UPE and styrene at 70 0C.
It can be seen that the initial slope became less gradual and more steep, compared to
Figures 5.1.12 and 5.1.13. Also, the initial conversion of UPE double bonds was higher,
however after about 20 minutes; the conversion of double bonds of styrene was more

than those of UPE.

246

 

 

 

 

 

 

 

0-8 ‘ l —UPE -—Styrene
0

g 0.6 — 80 C
5
g 04 -
O
o

0.2 -

0 I I ﬁr I I
—5 45 95 145 195 245

Time (nu'nutes)

Figure 5.1.15: Fractional conversion at 80 0C from in situ FTIR study

 

 

 

 

 

1
08 _, "— UPE — Styrene
0
90 C
.5 0.6 1
g 0.4 e
U
0.2 J
O I I I I I
0 50 100 150 200 250 300

Time (nrinutes)

Figure 5.1.16: Fractional conversion at 90 0C horn in situ FTIR study

Figure 5.1.15 shows fractional conversions of double bonds of UPE and styrene at 80 OC.
The initial slope of both curves became very steep, compared to Figures 5.1.12, 5.1.13
and 5.1.14. The initial conversion of UPE and styrene double bonds was almost the same.
The maximum conversion of styrene was higher than that of UPE, as already discussed

by many authors [15-18].

247

The fractional conversion of UPE and styrene double bonds at 90 0C is shown in Figure
5.1.16. The initial slope of both curves became very steep, compared to Figures 5.1.12,
5.1.13 and 5.1.14. The initial conversion of UPE and styrene double bonds was almost

the same.

 

 

9
oo
1

I — UPE —— Styrene

100 °C

9
ON
1

Conversion
.9
h
1

.C’
N
1

 

 

O 1 1 1 r 1

0 50 100 150 200 250 300
Time (nrinutes)

 

Figure 5.1.17: Fractional conversion at 100 0C from in situ FTIR study

 

 

 

 

 

 

 

1.2
Fractional conversions at 110 0C — UPE — Styrene
1 T /
0.8 -
t:
.2
t": 0.6 —
>
t:
O
U
0.4 —
0.2 ~
0 I I I I I I
-10 40 90 140 190 240 290 340

Time (min)

Figure 5.1.18: Fractional conversion at 110 0C from in situ FTIR study

248

Figure 5.1.17 shows fractional conversions of double bonds of UPE and styrene at 100
0C. The initial slope of both curves was very steep, similar to Figures 5.1.15, and 5.1.16.
The initial conversion of UPE and styrene double bonds was almost the same. The
maximum conversion of styrene was higher than that of UPE.

The fractional conversion of UPE and styrene double bonds at 110 0C is shown in Figure
5.1.18. The initial slope of both curves, again, became very steep. The initial conversion

of UPE and styrene double bonds was almost the same.

 

 

 

 

 

 

 

1 - Fractional conversions at 120 0C —UPE — Styrene
0.8 —
8
g 0.6 e
>
t:
O
U 0.4 -
0.2 4
O 1 1 1 1 1 1
-10 4O 90 140 190 240 290

Time (nrin)
Figure 5.1.19: Fractional conversion at 120 0C from in situ FTIR study
Figure 5.1.19 shows fractional conversions of double bonds of UPE and styrene at 120
0C. The initial slope of both curves was very steep, similar to Figures 5.1.16, 5.1.17, and

5.1.18. The initial conversion of UPE and styrene double bonds was almost the same. The

maximum conversion of styrene was higher than that of UPE.

249

1.2

 

Fractional Conversions at 130 0C — UPE — Styrene

l; ,-

p—l
l

 

 

Conversion
.6 .o .o
A as oo
1 L l

.o
N
1

 

 

 

0 I I I I I
- 5 45 95 45 195 245

 

Time (rrrin) 1

Figure 5.1.20: Fractional conversion at 130 0C from in situ FTIR study

 

1.2
Fractional conversions at 140 0C -— UPE — Styrene

W

1_.

 

.0
00
1

Conversion
o
ex
1

 

 

 

 

0.4 ~
0.2 -
O I I T I l I
- 10 40 90 140 190 240 290

Time (nrin)
Figure 5.1.21: Fractional conversion at 140 0C from in situ FT IR study

In Figure 5.1.20, fractional conversions of double bonds of UPE and styrene at 130 0C are

shown. The conversion of double bonds of styrene was more than those of UPE. At this

250

temperature, the conversion rates were much higher than at 50 or 60 OC. The curves
reached their maxima very quickly, in a matter of minutes.

Figure 5.1.21 shows fractional conversions of double bonds of UPE and styrene at 140
0C. The initial slope of both curves was, again, very steep. The initial conversion of UPE

and styrene double bonds was almost the same. The maximum conversion of styrene was

 

 

 

 

 

 

 

 

 

higher than that of UPE.
1.2
Fractional conversions at 150 0C — UPE — Styrene
1 I
08 A K7
=
.2
g
g 0.6 —
Q
U
0.4 i
0.2 r
O I I I I r
-10 40 90 140 190 240

Time (min)

Figure 5.1.22: Fractional conversion at 150 0C from in situ FTIR study

251

—s
N

 

 

 

 

 

 

 

 

 

 

Fractional Conversion of UPE at 160 oC __ UPE —— Styrene
1 -—4
__._/
0.8 J
:2
.2
a“:
> O 6 ‘
=
G
U
0.4 ~
0.2 ~
0 I I I I
-5 45 95 145 195 245
Time (nin)

Figure 5.1.23: Fractional conversion at 160 0C from in situ FTIR study

In Figure 5.1.22, fractional conversions of double bonds of UPE and styrene at different
times are shown for an isothermal experiment at 150 0C. The initial slopes of the curves
are again, very sharp. The conversion of double bonds of styrene was more than those of
UPE. This temperature was 90 0C higher than the one shown in Figure 5.1.12; therefore
the conversion rates are much higher. The curve reaches its maxima very quickly.

In Figure 5.1.23, fractional conversions of double bonds of UPE and styrene at different
temperatures are shown for an isothermal experiment at 160 0C. The conversion of
double bonds of styrene was more than those of UPE. Here again, the temperature was
110 0C higher than the one shown in Figure 5.1.12; therefore the conversion rates were

much higher. The maximum of the curves was reached in a few minutes.

Table 5.1.1: Comparison of maximum conversion at different temperatures from FTIR
experiments

252

 

Temperature Time a max 0. max

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(C) (minL UPE STY
50 220 0.87 0.89
60 220 0.88 0.90
70 26] 0.89 0.91
80 240 0.90 0.92
90 253 0.90 0.93
100 271 0.90 0.93
110 280 0.91 0.94
120 300 0.91 0.94
130 240 0.91 0.95
140 285 0.91 0.97
150 240 0.92 0.97
160 242 0.92 0.99

 

 

 

 

 

 

In Table 5.1.1, a comparison is made of the maximum fractional conversions of double
bonds of UPE and styrene at different temperatures measured at the end of experiment.
As expected, conversions are directly depending on temperature, and they increase with
increment in temperature. The conversion of double bonds of styrene is more than those

of UPE. This has been already reported in the literature [1-19].

5.2 Bioresins and bioplastics

The results of matrix modification experiments by introducing a derivatized vegetable oil
in the UPE curing system are described in this section. The first part deals with the results
from bioplastic obtained by grafting of acrylonitrile on vegetable oil like soybean, castor,
and MSO. The second part discusses the results of bioplastics obtained after graﬁing
soybean polyol with maleic anhydride.

5. 2. I Bioplastics made by grafting of vegetable oils with acrylonitrile

253

The bioresins formed were characterized by FTIR, TGA, DSC, and the bioplastics
developed from the bioresins were analyzed by mechanical, thermal and morphological
tests.

FTIR

The bioresin obtained after reacting acrylonitrile with natural oils, were characterized by
FTIR to validate the proof of reaction. FTIR scans of pure acrylonitrile, pure vegetable
oils, and, modiﬁed vegetable oils were taken and compared. Figure 5.2.1 through 5.1.7
show the above mentioned FTIR spectra of acrylonitrile, pure natural oils, as well as
grafted oils. In Figure 5.2.1, the spectrum of pure acrylonitrile is seen. The main peaks to
be observed are: CEN stretch from 2260-2200 cm'l, C=C stretch of vinyldene from 1665-
1620 cm'l, and C-C-CEN bend from 580-530 cm". On comparing all of the spectra, it
was seen that while the peak at 2200 cm'1 was not found in pure castor oil, pure soybean
oil and pure MSO, it could be easily spotted on grafted castor oil, grafted soybean oil and
grafted MSO. Also, the peaks at 1620 cm'1 and 530 cm'1 increased in height in grafted
MSO, and grafted castor oil, compared to pure MSO and pure castor oil (Figure 5.2.2,

5.2.4, 5.2.5, and 5.2.7).

254

160

 

 

 

 

— AN Neat
120 4
g?
0
g 80 4
g
E
'8 40 ~
a
h
[-
0 I I I I l
4000 3400 2800 2200 1600 1000 400

Wavenunber (cm")

Figure 5.2.1: FTIR spectra of pure acrylonitrile

 

 

 

0.8
Castor oil

a 0.7 ~
3,
0
g 0.6 ~
a
f
a 0.5 7
3E

0.4 ~

0.3 I I I T I I I l

3900 3400 2900 2400 1900 1400 900 400
Wave nunirer (cm")

Figure 5.2.2: FTIR spectra of pure castor oil

255

0.65

 

 

 

 

SoybeanOil

0.55 ~
E
8
:1 0.45 -
a
.n
In
8
2 0.35 —

0.25 I I r I I I I

3900 3400 2900 2400 1900 1400 900 400
Wave nunirer (cm!)

Figure 5.2.3: FTIR spectra of pure soybean oil

40

 

— MSO neat

DJ
0
1

20*

 

 

Transmittance (%)

p-d
O
l

 

 

0 I I I I r

4000 3400 2800 2200 _1 1600 1000 400
Wavenunber(cm )

 

Figure 5.2.4: F TIR spectra of pure methyl ester of soybean oil

256

 

20 , —ANCASBP2

Transmittance (%)
M

 

 

 

-10 f I I T If

4000 3400 2800 2200 1600 1000 400
Wavenunber (cm")

Figure 5.2.5: FTIR spectra of grafted castor oil

 

 

 

 

1.1 - SOY-BPO-AN
g 0.9 ~
0 1
U
:1 I
3
f 0.7 ~
8
a
<
0.5 r
0.3 I I I I I I I
3900 3400 2900 2400 1900 1400 900 400
Wave nunber (cm'I)

Figure 5.2.6: FTIR spectra of grafted soybean oil

257

OO
O

 

 

 

 

 

 

-—ANMSOBP3
§60 —

40 ~
:3
E
m
g 20 ~
I-

0 ‘I I W I I

4000 3400 2800 2200 1600 1000 400

Wavenumbe r(cm'l)
Figure 5.2.7: F TIR spectra of grafted methyl ester of soybean oil

Reaction scheme

Figure 5.2.8 shows the probable reaction scheme of the grafting process of vegetable oil
with acrylonitrile. This reaction mechanism has not been conﬁrmed yet, but it might be

the way acrylonitrile grafted to the unsaturated backbone of natural oils.

258

——

WEW I

Vegetable Oil '
1 Catalyst

(Free radical)

Temperature

+ CH2 =CH—GEN
(Acrylonitrile)

H2C

H i—CN

H2O

 

HC—CN

Acrylated Oil
Figure 5.2.8: Scheme showing probable reaction between vegetable oil and reactive
monomer, resulting in modified vegetable oil

Notched Izod impact strength
Figure 5.2.9 shows the notched Izod impact strength of the bioplastics. Impact energy is

the energy absorbed into the specimen during the impact event divided by its cross-
sectional area. For the bioplastics, the impact strength increased after addition of

biobased oil into the matrix (Figure 5.2.9). The impact strength of plastic containing

259

MSO was 25 % higher than that of neat UPE resin, and impact strength of plastic

containing EML is 96 % higher than neat UPE resin.

 

60 4
{I [:1 BS (MPa) A MOE(GPa)

I _

LII
O
1

A
o
1

W
O
L

 

N
O
L

 

 

 

Bending Strength (MPa) .
8
Di

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 r
A

Figure 5.2.9: Impact strength of the bioplastics

Legend: A= UPE control, B: MSO-UPE, C= GFT 2C-UPE, D= GFT ZB-UPE, E= GFT
2A-UPE, F= GFT 4A-UPE, G= GFT 4B-UPE, H= GFT 5A-UPE, I= GFT SB-UPE, J=
MSO-PBMA-UPE, K= BML-UPE

Flexural properties

Figure 5.2.10 shows the bending strength and the elastic modulus of the bioplastics. The
trend in flexural properties was opposite to that of impact strength. All bioplastics had
lower bending strength and elastic modulus as compared to neat UPE resin (Figure
5.2.10). The bending strength of plastic containing modiﬁed MSO was 20% lower than
strength of neat resin, and its modulus was 30% lower than that of neat resin. In the case
of biocomposites made with blends of oils and resin, the impact strength increased, but

the ﬂexural properties decreased.

260

 

p—A
A

H
N
l

(,I

Impact Strength (J/m)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 5.2.10: F lexural properties of the bioplastics

Legend: A: UPE control, B= MSO-UPE, C= GFT 2C-UPE, D= GFT 2B—UPE, E= GFT
2A-UPE, F= GFT 4A-UPE, G= GFT 4B-UPE, H= GFT 5A-UPE, I= GFT SB-UPE, J=
MSO-PBMA-UPE, K= EML—UPE, L=AESO-UPE

Biocomposites made with bioresins

Biocomposites were also made by impregnating hemp matl and kenaf ﬁbers with blends
of natural oils and UPE.

Impact strength of biocomposites

The impact strength of biocomposite containing hemp ﬁbers in a matrix of bioresin, was
96 % more than that of biocomposite made with hemp ﬁbers and UPE resin (Figure
5.2.11). Similarly, the kenaf based biocomposites gained about 95% strength compared
to a biocomposite containing kenaf ﬁbers in UPE matrix.

Flexural properties of biocomposites

However, the bending strength of the hempmatl biocomposite with bioresin was 20%

lower than that of biocomposite with neat resin and the modulus of elasticity for this

261

biocomposite was 7 % higher than that of biocomposite with neat UPE resin (Figure
5.2.12). In case of kenaf composite too, the bending strength decreased (20%) but the

elastic modulus was higher (9%) on introduction of bioresin in the polymer matrix.

40

I
10~ ‘
, L ll
0 I l i l '1
A C

B

 

DJ
0
1

Impact Strength (J/m)
N
O

 

 

 

 

 

 

 

 

D E F

Figure 5.2.11: Impact strength of biocomposites containing bioresins
Legend: A= UPE control, B: EML-UPE, C= Untreated hempmatl-UPE, D= Untreated
hempmatl-EML-UPE, E= Untreated kenaf-UPE, F= Untreated kenaf-EML-UPE

 

 

     

 

160 8
ElBS (MPa) OMOE(GPa) I Q
2 . I e
5312M ~ 6 E
a
E“ 80 ~ — 4 m
g I “5
U) m
vs 5
E” 40 — , — 2 '5
E a i
a: I
0 I I I I I 0

 

A B C D E F

Figure 5.2.12: Flexural properties of biocomposites containing bioresins
Legend: A: UPE control, B= EML-UPE, C= Untreated hempmatl-UPE, D= Untreated
hempmatl -EML—UPE, E= Untreated kenaf-UPE, F = Untreated kenaf-EML-UPE

262

Dynamic mechanical analysis of bioplastics

Following the trend of ﬂexural properties, the storage modulus for bioplastics decreased
in the same manner as the neat resin. As seen with bending strength and modulus of
elasticity, there was a 20—30% reduction in the storage modulus of the bioplastic (ﬁgure
not shown). This result was expected because vegetable oils are intrinsically low modulus
materials because of their molecular structure, and when they are added to the polyester
matrix, they lower the stiffness of the resulting material. The glass transition temperature
of neat UPE resin, obtained from a tan delta plot was 95 0C (ﬁgure not shown). For all
bioplastics, the T g decreased by 10—12 0C as compared to neat UPE resin. The glass
transition temperatures of plastics based on vegetable oils have been reported from 240 to

80 0C in the literature [21-23].

Morphology of bioplastics

      

  
   

/

’II
If.

~' \ . i I ‘ 7 ~. .
Figure 5.2.13: ESEM micrographs of impact fractures surface of bioplastic
containing MSO-PBMA-UPE, a) Magniﬁcation 2000X, Scale 25 um, b) 250 X,
Scale 200 um

 

263

ESEM micrographs of the impact fractured surfaces of the bioplastics show phase

    

Figure 5.2.14: ESEM micrographs of impact fractures surface ofbioplastic
containing MSO-UPE, a) Magniﬁcation 300 X, Scale 150 um b) Magniﬁcation
2000X, Scale 25 um

separation (Figures 5.2.13 and 5.2.14). These pictures showed discrete microstructures of
the dispersed phase which are scattered throughout the entire surface. SEM images
(Figure 5.2.15) of the impact ﬁ'actured surfaces of plastic samples showed contrast
between the neat UPE resin and bioresin (MSO-PBMA-UPE). Phase separation was
observed in the bioplastics samples.

From ESEM and SEM images, it was observed that addition of small amount of liquid
rubber, PBMA, improved the distribution of the second phase in the bioplastics, by

making it more uniformly spread out over the entire surface.

264

 

265

Flatten

 

Figure 5.2.16: AFM picture of neat unsaturated polyester resin (deﬂection image, scan

size: 5 umX 5 pm)

AFM images (Figures 5.2.16—18) also validated the existence of two phases in bioplastic
samples, and a single phase in neat polyester resin. As can be seen in Fig. 5.2.17, there
was uniformity in the distribution of hemispherical shaped craters of the second phase.
According to all ESEM, SEM and AFM images, the second phase seemed to be present
in craters or holes, which contain crosslinked molecules of the MESO, UPE and
polystyrene. They appeared in different sizes, but had the same content. In Figure 5.2.17,
the circular parts contain the second phase, which is the methyl ester of soybean oil. This
picture specially looked at the smaller circular particles, and represented different sized
domains distributed uniformly across the entire matrix. Figure 5.2.18 shows the inside
view of one such crater of the second phase. It was different in structure compared to the

neighboring UPE resin.

266

Flatten

 

Figure 5.2.17: AFM picture of bioplastic (MSO-UPE-PBMA) (deﬂection image, scan
size: 60 um X 60 um)

A lamellar structure was observed inside the circular crater, validating the difference in
mechanical properties of the UPE and MESO. The granular looking objects in Figure
5.2.16 (neat UPE), were probably micro gels formed during curing. It can be inferred that
the lamellar structured second phase particles, absorbed the impact energy, and thus

improved the impact strength of brittle neat polyester resin.

267

Flatten

 

Figure 5.2.18: AFM picture of bioplastic (MSO-UPE—PBMA) (deﬂection image, 9 urn X
9 um)

Most unmodiﬁed thermosets are brittle materials at ambient temperature. Consequently,
most polymers have to be impact modiﬁed in order to satisfy end-use requirements for
rigid applications. In most cases, the problem can be solved by incorporating rubber
domains into the polymer matrix. However, the nature and the size of the elastomeric
phase have to be adapted to each polymer matrix [24]. In most cases, it is also desirable
to use the minimum amount of rubber modiﬁer, in order not to affect other physical
properties such as modulus. The most important characteristics of an impact modiﬁer are:
rubber glass transition temperature, rubber domain particle size in the matrix, quality of
dispersion, and adhesion to the polymer matrix. Historically, several technical approaches
for toughening have been developed, which can be divided into three categories all based

on the incorporation of an elastomeric damping phase: elastomer introduction during

268

polymerization, dispersion of a thermoplastic elastomer phase during compounding
(uncrosslinked), and incorporation of elastomeric core-shell particles.

Traditionally, elastomeric additives have been used for toughening in the free radical
cross-linking of unsaturated polyester resin [25, 26]. But polyester blends with vegetable
based oils can also achieve the same purpose. The blends of oils and polyester resin form
semi immiscible systems. Here, the oil phase provided the rubbery structures after curing.
The oil phase was acting as an impact modiﬁer; hence, it absorbed impact energy and
delayed catastrophic failure.

The main toughening mechanisms which have been identiﬁed for thermoplastics as well
as thermosets are namely: crazing of the polymer matrix, shear yielding of this matrix,
and cavitation of the rubber phase. Depending on the polymer system, either a single
mechanism or a combination of different mechanisms will be activated [27].

The oil phase of the system consisted of small discrete rubbery particles, with an average
size of 5—15 mm. They were randomly distributed in the glassy brittle polyester resin
matrix. As reported in the literature about rubber toughened thermosets, these oil phase
particles relieved the constraints in the matrix through principal mechanisms of cavitation
and forming shear bands [28]. This phase separation phenomenon is similar to that in the
low proﬁle mechanism of unsaturated polyester resins.

The miscibility and interfacial properties of additive and resin blends play a major role in
the toughening process. Improvements in the fracture impact of thermoset polyester resin
can be obtained by dispersing elastomer particles with diameters from 0.5 to 5 mm in the
blends. The smaller particles produce more microcracks in the matrix and enhance the

fracture—impact energy. The additive is immiscible with the resin and gets phase

269

separated from the matrix during curing. Unsaturated polyester resins systems are very
reactive materials, which are initially in a liquid state, and become solid after curing.

The morphology of the polyester resin systems is decided by thermodynamics and
polymerization kinetics. Phase separation brings about a change in polydispersity.
Usually, the size of additive particles in the matrix is 1—40 microns. A way to reduce this
size is to decrease interfacial tension between two phases. Adding a small percentage of
coupling agent like maleic anhydride solves this problem. In this system, poly-
(butadiene—maleic anhydride) brought about uniformity in the dispersion, as well as
uniformity in sizes of the second phase. Therefore, smaller microcracks were formed in
the matrix leading to an increase in impact strength. These microcracks were produced
during the large volumetric shrinkage which takes place during polymerization.
Microvoids also occurred around the ﬁbers in the composite.

Microvoid formation is an important phenomenon for systems consisting of blends of
bioresin and UPE. It compensates for resin shrinkage, but induces intrinsic brittleness.
Increased additive content in the UP resin should produce higher impact strength, but if
the microvoid content is also enhanced, impact strength is reduced. But in this system,
the reverse effect was observed. Even after addition of 20% bioresin into the polyester
matrix, the impact strength was improved, with a little sacriﬁce of modulus of elasticity.
Well-dispersed particles in the resin matrix induced homogeneous distribution of internal
stresses due to network formation. Thus, low energy impact fractures in the unsaturated
polyester resin composites were eliminated by the use of this additive in the system. The
addition of the blends of derivatized oils and UPE in the matrix of the natural ﬁbre

biocomposites lead to an increase in the impact strength of the composites.

270

Thus, using bioresins as polymer matrices provides a two fold beneﬁt over thermoset
resins in biocomposites. Their presence can improve the impact strength of the resulting

bioplastic, as well as produce a material with higher biobased content.

5.2.2 Bioplastics made by reaction of vegetable oils with maleic anhydride

Bioresins were also made by chemical modiﬁcation of the vegetable oil (SOPEP) with a
functional reactive monomer (MA). The OH content of this polyol is 122 mg of KOH.
The pure polyol was characterized using mass spectroscopy (MS), nuclear magnetic
resonance (NMR), DSC, TGA and FTIR. The pure maleic anhydride and modiﬁed polyol
were also scanned under FTIR. The bioplastics developed from the bioresins were

analyzed by mechanical, thermal and morphological tests. These results are discussed in

this section.

 

 

 

 

 

 

 

TGA
1.6
1 J

_ t 1.2 2
0.8 ~ ——Werght% é
°\e —Derivative Weight q_ 0 8 a
E O 6 - ' ’33
.2» B
‘9 __ o
3 0.4 a 0'4 .3
ﬂ
\ e
0.2 ~ * 0 5

0 I I I I I -O.4

0 1 00 200 300 400 500 600
Temperature (0 C)

Figure 5.2.19: TGA of pure soybean oil phosphate ester polyol (SOPEP)

271

Figure 5.2.19 shows the curves from therrno gravimetric analysis of pure SOPEP. As can
be seen from the graph, pure polyol is stable until 200 oC. The maximum decomposition
temperature of pure SOPEP is 400 0C.

DSC

Figure 5.2.20 shows the curves from differential scanning calorimetry of pure SOPEP.

The DSC plot shows no crystallization peaks.

 

 

 

 

 

 

 

-I I

3

.2

In.

*5 -5 -‘

0

I
-7 -. "
'9 I I I I I I I

-70 -20 30 80 130 180 230 280

Tenperature (0 C)
Figure 5.2.20: DSC of pure soybean oil phosphate ester polyol (SOPEP)

FTIR
The modiﬁed polyol obtained after reacting maleic anhydride with soybean oil polyol,
were characterized by FTIR to validate the proof of reaction. FTIR scans of pure maleic

anhydride, pure soybean oil, pure soybean polyol, and, modiﬁed polyol were taken and

compared. Figure 5.2.21 through 5.2.24 show the FT IR spectra of MA, pure soybean oil,

272

pure SOPEP, and grafted SOPEP respectively. In Figure 5.2.21, the spectrum of pure MA
is seen. The main peaks to be observed are: C=O symmetric stretch from 1765-1725 cm’1
(1772), C=C stretch from 1680-1630 cm‘1 (1631), CH out of plane deformation from
730-665 cm", C-O-C anti-symmetric stretch from 1280-1070 cm“, coo— at 1567 cm"

and OH at 3600 cm".

In pure SOPEP (Figure 5.2.23), a high peak of OH is seen at 3467 cm'l which is not seen
in pure soybean oil (Figure 5.2.22). In modiﬁed SOPEP (Figure 5.2.24), the OH peak at

3460 cm'1 is small compared to that in pure SOPEP. Other signiﬁcant peaks seen in

modiﬁed SOPEP are: C=O at 1725 cm", C=C at 1643 cm’l, and COO- at 1547 cm’l.

 

Absorbance
N I»
in w in
l 1 l

N
L

 

 

 

p—o
VI
1_

 

 

1 I I I I I I I7

3900 3400 2900 2400 1900 1400 900 400
Wavenurrber

Figure 5.2.2]: FTIR spectra of pure maleic anhydride (MA)

273

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.59 -
2 0.49 I
a
.9
h
8
I:
<1
0.39 ~
0.29 r I I I I I I
3900 3400 2900 2400 1900 1400 900 400
Wavenunirer
Figure 5.2.22: FTIR spectra of pure soybean oil
1.05 ~
3 I
§085 ~
I-
8
2
0.65 -
0.45 I I I I I I T
3900 3400 2900 2400 l 900 1400 900 400
Wavenumber
Figure 5.2.23: FTIR spectra of pure SOPEP
4.5
3.5 r
8
E
N
£
a 2 5 r
.e
<
1.5 ~
0.5 I I I T I T l
3900 3400 2900 2400 l 900 1400 900 400

Wavenunber
Figure 5.2.24: FTIR spectra of modiﬁed SOPEP

274

Reaction mechanism

Figure 5.2.25 shows the probable reaction scheme of the grafting reaction of soybean
oil phosphate ester polyol with maleic anhydride. This reaction mechanism has not
been conﬁrmed yet, but it might be the way maleic anhydride attacks the OH groups
of SOPEP, and gets grafted.

DMA of bioplastics

Dynamic mechanical analysis was done on bioplastics with modiﬁed polyols
GFTSOPEPI, GFTSOPEP2, and GFTSOPEP3. Storage Modulus, tan delta and cross-
linking density of the bioplastics were found out. Figures 5.2.26, 5.2.27, and 5.2.28
show the typical storage modulus curves of bioplastics made using GFTSOPEPl,
GFTSOPEP2, and GFTSOPEP3. Typical tan delta curves of bioplastics made using
GFTSOPEPl, GFTSOPEP2, and GFTSOPEP3 are shown in Figures 5.2.29, 5.2.30,

and 5.2.31.

275

I OH OH
o—-< OH OH +

OH OH O OH OH

Soybean Polyol 0” 0“

Maleic Anhydride

N,N-Dimethylbenzylaminel T~= 80 0 C

OOH
OH
COOH

OH O

O I_\/\/\I/H\/Y\/\/l\/
m 0 0 COOH
O
HO

Modiﬁed Soybean Polyol

Figure 5.2.25: Proposed reaction mechanism for grafting of SOPEP with MA

The bioplastics samples with 30%, 40% and 50% GFTYSOPEP 2 had physical
problems and couldn’t be tested. The bioplastic samples with 30%, 40% and 50%
GFTSOPEP 3 were post-cured. For ﬁnding out the T8 of the bioplastic samples with
GFTSOPEP2 and GFTSOPSP3, the DMA was carried out from -20 “c to 160 °C.

In Figure 5.2.26, the storage modulus of bioplastics samples containing 40% and 50%
of GFTSOPEP], dramatically decreases as compared to samples with 10%, 20% and

30% GFTSOPEPI. The storage modulus of samples with containing 40% and 50% of

276

GFTSOPEPI is very low as compared to neat UPE, while the bioplastics containing
10%, 20%, and 30% GFTSOPEPI (8.5-11.5 GPa at room temperature) have very
higher storage modulus compared to UPE control (2.3 GPa at room temperature).
Similarly, in the case of GFTSOPEP3 bioplastics, the 10%, 20%, and 30%
GFTSOPEP3 samples have very higher storage modulus compared to UPE control
(Figure 5.2.27). And, for bioplastics containing GFTSOPEP2, the samples with 10%
and 20% GFTSOPEP2 samples have very high storage modulus compared to UPE
control (Figure 5.2.28).

Since ﬂexural and tensile moduli follow the same trend as storage modulus, it can be
argued that the bioplastics containing 10, 20, and 30% of GFTSOPEP], 2, and 3 have
very high elastic moduli and tensile moduli compared to UPE control.

The glass transition temperatures (T8) of the bioplastics samples were evaluated from

the maxima of the tan delta plots (Figures 5.2.29, 5.2.30, and 5.2.31).

277

— UPE Control

-— 10% GFTSOPEP] -—- 20% GFTSOPEPI

30% GFTSOPEP] — 40% GFTSOPEPI —- 50% GFTSOPEP]
12 '

Storage Modulus (GPa)

.b

4

TA; _I._

O

 

 

20

 

 

70 Temperature(0C) 120

Figure 5.2.26: Storage modulus of bioplastics made with GFTSOPEP]

l

I—l

Storage Modulus (GPa)

2

0

A

N

0

 

 

 

 

 

— UPE control — 10% GFTSOPEPZ
—— 20% GFTSOPEPZ — 30% GFTSOPEPZ
I
20 70 0

Temperature (°C) 12

Figure 5.2.27: Storage modulus of bioplastics made with GFTSOPEP2

278

-— UPE Control — 10% GFTSOPEP3 — 20% GFTSOPEP3
— 30% GFTSOPEP3 — 40% GFTOSPEP3 — 50% GFTSOPEP3
l6

 

y—e
N

Storage Modulus (GPa)
0°

 

 

 

 

 

0 -1 0 40 90 1 40
Temperature (11 C)
Figure 5.2.28: Storage modulus of bioplastics made with GFTSOPEP3
— UPE Control — 10% GFTSOPEP] -— 20% GFTSOPEPI
”023.0% GFTSOPEP] -- 40% GFTSOPEPI — 50% GFTSOPEP]

Tan Delta

 

 

 

 

20 70 Temperature (°C) 120
Figure 5.2.29: Tan delta of bioplastics made with GFTSOPEP]

279

— UPE control — 10% GFTSOPEP2
—— 20% GFTSOPEP2 — 30% GFTSOPEP2

0.5

0.4

Tan Delta
.0
U)

.0
N

0.1

 

20 70 Temperature (°C) 120

Figure 5.2.30: Tan delta of bioplastics made with GFTSOPEP2

 

-—- UPE Control — 10% GFTSOPEP3 — 20% GFTSOPEP3
30% GFTSOPEP3 — 40% GFTSOPEP3 — 50% GFTSOPEP3
0.6
0.4
5
'5
a
:I‘
a
I-
0.2
0
— 10 40 140

Temperature (0090
Figure 5.2.31: Tan delta of bioplastics made with GFTSOPEP3

280

Figures 5.2.32, 5.2.33, and 5.2.34 show the T3 of the bioplastic samples made from all
the grafted polyols. From Figure 5.2.32 it was seen that the glass transition
temperature increases after addition of grafted soy polyol for 10, 20 and 30%
GFTSOPEPI, but for samples with 40% and 50% GFTSOPEP] was not determined
from this experiment. The highest T8 was observed for samples with 20%
GFTSOPEP]. In the case of bioplastics with GFTSOPEP2 too, the sample with 20%
GFTSOPEP2 registered the highest Tg (Figure 5.2.33). The highest Tg in case of
bioplastics containing GFTSOPEP3 was demonstrated by the sample with 20%
GFTSOPEP3 (Figure 5.2.34). The T8 of samples with 40% and 50% GFTSOPEP was
below the T3 of neat UPE. The low T3 of the samples with 40 and 50% grafted polyol
hits at higher ﬂexibility and higher mobility of these samples, which was also

physically evidenced.

 

108

 

 

 

92r

Glass transition temperature () C)

 

 

 

 

 

 

 

 

 

 

oo
00
I

 

A B C D

Figure 5.2.32: Glass transition temperature of bioplastics made with GFTSOPEPI
Legend: A=UPE control, B=10% GFTSOPEPl-UPE, C=20% GFTSOPEPl-UPE,
D=30°/o GFTSOPEPl-UPE

281

 

h—I
N
O

100 ~ _I_
—17 ,

00
O
1

 

 

 

 

 

 

 

 

 

Glass transition temperature () C)
Ch
o

 

 

40 r
p' .
20‘ ‘ i ,. a
‘ £4 F "ﬁt“ 5;:
"’.°'. I ' ‘ mu
0 l I
A B C D

Figure 5.2.33: Glass transition temperature of bioplastics made with GFTSOPEP2
Legend: A=UPE control, B= 10% GFTSOPEP2-UPE, C: 20% GFTSOPEP2-UPE,
D= 30% GFTSOPEP2-UPE

 

 

 

 

 

 

 

 

 

 

 

 

 

120
.9
0 ’E *1
5 +
E —-:r:—
a 80‘
E
Q)
[-
ﬂ
.2
'2 40— ‘ l
a “,I I» ”t I
2 I; ”a I “I
0 1' ii- Tu
0 I 4 I I I" I I
A B C D E F

Figure 5.2.34: Glass transition temperature of bioplastics made with GFTSOPEP3

Legend: A=UPE control, B=10% GFTSOPEP3—UPE, C= 20% GFTSOPEP3—UPE,
D: 30% GFTSOPEP3-UPE, E= 40% GFTSOPEP3—UPE, F= 50% GFTSOPEP3-UPE
Figures 5.2.35, 5.2.36, 5.2.37 show the cross link density of the bioplastic samples.

An important property from a characterization standpoint of crosslinked networks is

the number of crosslinked sites and the molecular weight or length of chain between

282

sites. The crosslink density of bioplastics samples were measured in the rubbery
region, 40 OC above the Tg of the material, by the following formula:
DC: E, /(3*R*T) —(10)

where, uc=crosslink density, R=universal gas constant, T=Tg+40 0C, and BI: storage
modulus at temperature T.

The change in glass transition temperature is directly proportional to crosslink
density. As compared to neat polyester resin, 10, 20 and 30% GFTSOPEP] samples
show a large increment in the cross linking density. With 40 and 50% GFTSOPEP],
cross link density couldn’t be found out because their Tg was not in the range on the
temperature proﬁle of the DMA test (room temperature to 1600C @ 4 0C/min). For
ﬁnding out Tg of these samples, the DMA will have to carried out from a much lower
temperature, like -20°C to 1600C. Similarly, in the case of GFTSOPEP2 bioplastics,
the 10%, 20%, and 30% GFTSOPEP2 samples have very higher cross linking density
compared to UPE control (Figure 5.2.36). And, for bioplastics containing
GFTSOPEP3, the samples with 10%, 20%, and 30% GFTSOPEP3 samples have very
high cross linking density compared to UPE control (Figure 5.2.37). These results are
in coherence with the storage modulus results. The high values of cross link density
of the bioplastic samples demonstrate a tightly wound rigid network of the polymer

formed by blending bioresins with UPE.

283

0.016

 

0.012 -

0.008 1

Crosslink densrty(mol/nf)
.o
8
A

 

 

0-

 

A B C D

Figure 5.2.35: Crosslinking density of bioplastics with GFTSOPEPI
Legend: A=UPE control, B=10% GF'I‘SOPEPl-UPE, C= 20% GFTSOPEPl-UPE,
D= 30% GFTSOPEPl-UPE

 

 

 

 

0.012

“E

=I

e0.008 ~

5

E

E

1:

30.004 -

E

8

l-

U

0 - I I
A B C D

Figure 5.2.36: Crosslink density of bioplastics made with GFTSOPEP2
Legend: A=UPE control, B=10% GFTSOPEP2-UPE, C: 20% GFTSOPEP2-UPE,
D= 30% GFTSOPEP2-UPE

284

0.016

 

F3

0

id

N
1

20162261

*3‘

O

'o

if
1

Cross link density (mol/rrf)
2
O
00
an

. .7" EDI":
Ti .

a
Y
33"..

 

 

 

0 _
A B C D E F
Figure 5.2.37: Crosslink density of bioplastics made with GFTSOPEP3
Legend: A=UPE control, B=10% GFTSOPEP3-UPE, C= 20% GFTSOPEP3-UPE,
D= 30% GFTSOPEP3-UPE, E= 40% GFTSOPEP3-UPE, F= 50% GFTSOPEP3-UPE
Impact strength of bioplastics
The notched Izod impact strength of the bioplastic samples made from all the grafted

polyols are shown in Figures 5.2.38, 5.2.39, and 5.2.40.

20

 

H
II!
1

Impact Strength (J/m) ,
8

 

 

 

S—I
ii -*
0 I I I I I
A B C D E F

Figure 5.2.38: Impact strength of bioplastics made with GFTSOPEPl
Legend: A=UPE control, B=10% GFTSOPEPl-UPE, C: 20% GFTSOPEPl-UPE,
D= 30% GFTSOPEPl-UPE, E: 40% GFTSOPEPl-UPE, F: 50% GFTSOPEPl-UPE

285

 

12 J --

 

 

Impact Strength (J/m)

 

 

 

 

 

 

A B C

Figure 5.3.39: Impact strength of bioplastics made with GFTSOPEP2
Legend: A=UPE control, B=10% GFTSOPEP2-UPE, C= 20% GFTSOPEP2-UPE

 

 

 

 

 

20‘ "
516‘
E
B
312“ ‘I'
E
o
3:
m 1
*5 8~
«I
a.
E
4_
O I

 

 

 

 

A B C D E F

Figure 5.2.40: Impact Strength of bioplastics made with GFTSOPEP3

Legend: A=UPE control, B=10% GFTSOPEP3-UPE, C= 20% GFTSOPEP3-UPE,
D= 30% GFTSOPEP3-UPE, E= 40% GFTSOPEP3-UPE, F= 50% GFTSOPEP3-UPE
It is well known that impact strength and storage modulus (DMA) track trends

opposite to each other. It was earlier seen that storage modulus of 10%, 20% and 30%

286

GFTSOPEPI bioplastic samples was much higher than UPE control, while that of
40% and 50% GFTSOPEPI samples was much lower than the storage modulus of
neat resin (Figure 5.2.26). From Figure 5.2.37 it was observed that impact strengths
of 10%, 20%, and 30% GFTSOPEP] is much lower than neat UPE , while that of
40% and 50% GFTSOPEPl is higher than that of neat UPE. This trend was also seen
in impact strength of bioplastics samples of GFTSOPEP2 and GFTSOPEP3. Thus,
the ﬂexible samples containing 40% and 50% grafted polyols were much higher in

toughness, and low in stiffness.

Morphology of bioplastics samples

287

‘ 3.0“- ”A;
. - A a

I

Figure 5.2.41: ESEM micrographs of impact fractures surface of bioplastic
containing

a) 10%GFTSOPEPl-UPE, Magniﬁcation 3000 X, Scale bar 15 um

1)) 20% GFTSOPEPl-UPE, Magniﬁcation 300 X, Scale bar 150 um

c) 30% GFTSOPEPl-UPE, Magniﬁcation 800 X, Scale bar 50 um

d) 40% GFTSOPEPl-UPE, Magniﬁcation 285 X, Scale bar 150 um

 

Figures 5.2.41 and 5.2.42 show the ESEM micrographs of impact fractures surfaces

of bioplastics made from grafted polyols. In all of the bioplastics examined under

ESEM, phase separation was observed between grafted soy polyol rich phase

UPE rich phase. But the surface features changed with different compositions.

288

and

 

Figure 5.2.42: ESEM micrographs of impact fractures surface of bioplastic
containing

a) 10%GFTSOPEP3-UPE, Magniﬁcation 400 X, Scale bar 100 um
b) 20% GFTSOPEP3-UPE, Magniﬁcation 400 X, Scale bar 100 um
c) 30% GFTSOPEP3-UPE, Magniﬁcation 600 X, Scale bar 200 um
(I) 40% GFTSOPEP3-UPE, Magniﬁcation 300 X, Scale bar 150 um
In all of the bioplastics, circular or spherical shapes are observed on the surface. In
bioplastics with 10 and 20% modiﬁed polyol, there were only circular trenches,

where the dispersed phase might be present. The diameter of these circular objects

varied greatly, but these were present all over the entire fractured surface. In

289

bioplastics with 30 % modiﬁed polyol, circular as well as spherical objects were seen
on the surface. In bioplastics with 40 and 50% modiﬁed polyol, there are more
spherical objects and lesser circular objects. The size of the spherical looking balls
also varied over the entire surface, but the distribution was very uniform.

Figures 5.2.43 through 5.2.46 show the AFM images of a sample containing 20%
GFTSOPEP3 in a UPE matrix. The images for the GFTSOPEP3 samples show phase
separation. The height features are indicators of the second phase in these pictures.
The picture focuses on a relatively empty portion of the matrix, where small particles
of second phase, which is modiﬁed polyol, can be seen, strewn about in the entire
area. The morphology of this sample can be clearly observed from TV images of this

sample shown below in Figures 5.2.47.

Flatten

 

0 7.5.0 50.0
Figure 5.2.43: AFM picture of 20% GFTSOPEP3-UPE (deﬂection image, 50 um X
50 Inn)

290

 

Flatten

50.0

25.0

 

O
0 25.0 50.0

Figure 5.2.44: AFM picture of 20% GFTSOPEP3-UPE (deﬂection image, 50 um X
50 Ian)

Flatten

 

o 5.0 10 . o 15 . 0
Figure 5.2.45: AFM picture of 20% GFTSOPEP3-UPE (Z scale: 1000 nm, scan size,
16 mm X 16 pm)

291

Flatten

 

0 5.0 10.0 15.0

Figure 5.2.46: AFM picture of 20% GFTSOPEP3-UPE (deﬂection image, 16 um X
16 Inn)

 

Figure 5.2.47: Picture of 20% GFTSOPEP3—UPE on TV screen

Figure 5.2.44 focused on an area devoid of circular geometries, and looks at second
phase particles distributed on the UPE matrix. Figure 5.2.45 showed the top view of
a spherical object on the matrix. This object looks like a mercury drop on a solid
surface, but is composed of both modiﬁed polyol and UPE resin. The concave shape

of the spherical object can be easily seen. This is a height image, while Figure 5.2.46

292

is the deﬂection image of the same object. The composition of the spherical object
was clearly observed here.
Thus phase separated bioplastic samples resulted in changes in mechanical, thermal

and morphological characteristics of the polymers formed.

293

References

1. M. R. Kamal, S. Sourour, Polymer Engineering and Science, 13(1), (1973), 59—64.
2. J. M. Salla, J. L. Martin, Thermochimica Acta, 126, (1988), 339-354.

3. J. M. Salla, X. Ramis, J. L. Martin, A. Cadenato, Thermochimica Acta, 134,
(1988), 261-167.

4. M. Paci, E. D. Vecchio, F. Campana, Polymer Bulletin, 15, (1986), 21-27.
5. X. Ramis, J. M. Salla, Journal of Applied Polymer Science, 45, (1992), 227-236.

6. D. Rouison, M. Sain, M. Couturier, Journal of Applied Polymer Science, 89,
(2003), 2553-2561.

7. A. Nzihou, P. Sharrock, A. Ricard, Chemical Engineering Journal, 72, (1999), 53-
61.

8. J. L. Vilas, J. M. Laza, M. T. Gary, M. Rodriguez, L. M. Leon, Journal of Applied
Polymer Science, 79,(2001), 447-457.

9. Y-J. Huang, J-S. Leu, Polymer, 34(2), (1993), 295-304.

10.H. Yang, L. J. Lee, Journal of Applied Polymer Science, 79, (2001), 1230-1242.
11.H. Yang, L. J. Lee, Journal of Applied Polymer Science, 84, (2002), 211-227.
12.J. S. Chen, T. L. Yu, Journal of Applied Polymer Science, 69, (1998), 871-878.

13.V. Kosar, Z. Gomzi, Chemical and Biochemical Engineering Quarterly, 15 (3),
(2001), 101—108.

l4.Y.-M. Yun, S.-J. Lee, K.-J. Lee, Y.—K. Lee, J.-D. Nam, Journal of Polymer
Science: Part B: Polymer Physics, 35, (1997), 2447—2456.

15.8. Ziaee, G. R. Palmese, Journal of Applied Polymer Science, 37, (1999), 725-744.

16R. 11’. Brill, G. R. Palmese, Journal of Applied Polymer Science, 76, (2000), 1572-
1582.

17B. L. Grunden, C. S. P Sung, Macromolecules, 36, (2003), 3166-3173.
18.R-H. Lin, J-H. Hsu, Polymer lntemational, 50, (2001), 1073-1081.

19.L. Shan, C. G. Robertson, K. N. E. Verghese, E. Burts, J. S. Rifﬂe, T. C. Ward, K
.L. Reifsnider, Journal of Applied Polymer Science, 80,(2001), 917-927.

294

20.1nfrared and Raman Spectroscopy, Chapter 7, Group Frequencies: Infrared and
Raman, 159-182. Part B: Series Volume: 1 E. G. Brame; J. G. Grasselli, Print
Published1977

21. F. Li and R. C. Larock, J. Polym. Sci., Part B: Polym. Phys., 2001, 39, 60—77.
22.13. Can, S. Kusefoglu and R. P. Wool, J. Appl. Polym. Sci., 2001, 81, 69.

23.1. Frichinger and S. Dirlikov, in Toughened Plastics I Science and Engineering,
ed. C. K. Riew and A. J. Kinloch, 1993, (Advances in Chemistry, No. 233), American
Chemical Society, Washington, DC, pp. 451—489.

24.B. Cherian and E. T. Thachil, Progr. Rubber Plastics T echnol., 2001, 17(4), 205—
224

25. L. Suspene, Y. S. Yang and J. P. Pascault, in Toughened Plastics I Science and
Engineering, ed. C. K. Riew and A. J. Kinloch, 1993, (Advances in Chemistry, No.
233), American Chemical Society, Washington, DC, pp. 163—188.

26. F. J. McGarry and R. Subramanium, in Toughened Plastics [1 Novel Approaches
in Science and Engineering, ed. C. K. Riew and A. J. Kinloch, 1996, (Advances in
Chemistry Series, No. 252), American Chemical Society, Washington, DC, pp. 133—
149.

27.C. B. Bucknall, Toughened Plastics, Applied Science Pub. Ltd., London,

1977.

28. A. F. Yee and R. A. Pearson, J. Mater. Sci., 1986, 21, 2462.

295

CHAPTER SIX: RESULTS ON NEW PROCESSING

6.0 RESULTS ON NEW PROCESSING

The results obtained from testing biocomposites manufactured with a new SMC
technique are described in this chapter.

6.] Biocomposites from SMC

The new SMC processing for natural ﬁbers can be done on normal commercial SMC
equipment with a minor change on the way ﬁbers are fed to the line. In industrial SMC
set-ups, the glass ﬁbers rovings are fed to a chopper, and chopped ﬁbers fall onto the
carrier ﬁlm [1-11]. But the natural ﬁbers are not supplied as continuous rovings or yarns,
and must be fed to the ﬁlm in a chopped form. Therefore, in the new set —up, the chopped
ﬁbers fall from a calibrated vibratory feeder onto the carrier ﬁlm, and get drenched in
resin ﬂowing from two resin pots, and advance to the compression rollers, and ﬁnished
product is obtained at the end of the line. The prepreg from SMC line is refrigerated for
certain time to reach desirable gelation. The gelled product is then compression molded
in variety of molds to get the desired shape.

For glass composites:

The glass ﬁber polyester resin composites from SMC line were tested for mechanical and
thermal properties including, bending strength, modulus of elasticity, storage modulus
and tan delta.

The ﬂexural properties of glass composites are shown in Figure 6.1. There was
consistency in the data for samples B and C, which represented glass composites made
using SMC processing. This was a veriﬁcation of the fact that SMC process had been

optimized for Glass-UPE-CaCO3 system. The bending strength and modulus of elasticity

296

for samples B and C are almost same at 125 MPa, and 12.5 GPa respectively. The
bending strength of glass composites was 32 % more than that of UPE control, while

their modulus of elasticity was 370 % more than neat polyester resin.

a BS (Mpa) o MOE(GPa)

 

 

 

 

 

 

 

 

 

 

 

 

 

160 20
a 33
°- T T — 16 Q
E. 120 - a
'5 '5
g I I 12 '5
g 80 ~ II-“I’
(D _ 8 n—
O
.3 g
.2 40 ‘ _ 4 3
iii 5 g

o I I 0

A B C

Figure 6.1: F lexural properties of glass composites

Legend: A= UPE control, B= SMC Composite 1, C: SMC Composite 2

The storage modulus of glass ﬁber—UPE composites at 40 0C is shown in Figure 6.2.
Again, both B and C have almost the same values of storage modulus. As compared to
UPE control, glass composites had very high storage modulus, (1600 % more than that of
neat polyester) which reﬂected high stiffness of glass ﬁbers.

Figure 6.3 shows a typical dynamical mechanical analysis plot for glass ﬁber composites.
As is common in all thermoset systems, the storage modulus decreased with increasing
temperature. The glass transition temperature of this composite was 110 0C. As compared

to storage modulus, loss modulus was very low over the entire range of test temperature.

297

 

60 I“,-

w
o
1

   
 
 
 

«h
o
1

 

Storage Modulus (GPa)
N (a)
O O

...L
0
L1—

 

A B C
Samples

Figure 6.2: Storage modulus of glass composites at 40 0C
Legend: A= UPE control, B= SMC composite l, C= SMC composite 2

 

 

 

— Storage Modulus —- Loss Modulus — Tan Delta

60 0.3
A 50 ~
(I
6‘5
;' 40 ‘I r 0.2
2 a
§ 30 - a
a E
a. 20 ~ — 0.1
E
m 10 -

0 e 0

 

 

 

1 00
Temperature (°C)
Figure 6.3: DMA plot for glass composites 1

The data for glass ﬁber composites conﬁrmed that consistent materials had indeed been
produced, meaning thereby, that the process parameters also had been optimized. Thus,

the next step was development of SMC processed natural ﬁber composites.

298

For natural fibers:

The thermal properties of natural ﬁbers used for making biocomposites are shown in
Figures 6.4, 6.5 and Table 6.1. Figure 6.4 shows the plots from DSC for ﬁve ﬁber
samples. All the ﬁbers showed same transition from -60 0C to 300 0C. A trough was
observed in all of the ﬁve samples. This ranged from 74 0C for untreated BBSG, 94 0C
for untreated green ﬂax core, 106 0C for untreated hemp, 83 0C for silane treated BBSG,
to 99 0C for silane treated green ﬂax core. This was because of evaporation of water from
natural ﬁbers. This phenomenon is very common for natural ﬁbers. From this data, it was
interpreted that after silane treatment of ﬁbers, the water evaporation occurs at a higher
temperature as compared to that for untreated ﬁbers.

Figure 6.5 shows TGA plots for the same ﬁve ﬁber samples which were analyzed in
DSC. Here too, all ﬁbers showed same kind of transition. The maximum degradation
temperatures were found out from the TGA plots and are listed in Table 6.1. The
maximum degradation temperature for BBSG and green ﬂax core increased after silane
treatment. The percentage weight at 600 0C ranged from 18 % for untreated BBSG to
22.5 % for silane treated BBSG and silane treated green ﬂax core. At higher
temperatures, big blue stem grass was more stable than ﬂax, as can be seen from the

weight % curves.

299

 

Haet Flow (mW)
L

do

 

-12

 

 

 

-70

30 130 230
Temperature (°C)

Figure 6.4: DSC plots of natural ﬁbers used for SMC line
Legend: A= Untreated big blue stem grass, B= Untreated green ﬂax core, C=Untreated
hemp, D= Silane treated big blue stem grass, E= Silane treated green ﬂax core

—A —B -—-C ——D E

 

Weight (%)
O)
O

 

 

 

 

40 r
20 r
0 Ai——— WiiI If I I F I I A
0 1 00 200 300 400 500 600
Temperature (°C)

Figure 6.5: TGA plots of natural ﬁbers used for SMC line

300

Legend: A= Untreated big blue stem grass, B= Untreated green ﬂax core, C=Untreated
hemp, D= Silane treated big blue stem grass, E= Silane treated green ﬂax core

Table 6.1: Maximum degradation temperatures for natural ﬁbers

 

 

 

 

 

 

Samples Td,max (C)

A 305.47
B 305.79
C 307.96
D 326.54
E 307.21

 

 

 

 

Legend: A: Untreated big blue stem grass, B= Untreated green ﬂax core, C=Untreated
hemp, D= Silane treated big blue stem grass, E= Silane treated green ﬂax core

Table 6.2 and 6.3 show the results from XPS of natural ﬁber samples. These results also
depict changes occurring in the ﬁbers after surface treatment. According to Table 6.2,
untreated BBSG and green ﬂax core, contain no silicon, while the presence of silicon was
found in silane treated BBSG and green ﬂax core. As compared to untreated BBSG and
green ﬂax core, silane treated BBSG and green ﬂax core had a decreased concentration
level of carbon, and an increased concentration of oxygen.

Table 6.3 shows the ratio of atomic concentrations of C/O, C/Si and C/N in all four
samples. The C/O ratio decreased while moving from untreated BBSG to silane treated
BBSG, and from untreated ﬂax core to silane treated ﬂax core. Between silane treated
BBSG and silane treated ﬂax core, C/Si ratio was higher for silane treated ﬂax core.

Table 6.2: Atomic concentrations on the surfaces of ﬁbers used for SMC line

 

 

 

 

 

 

C1s[.314] le[.499] Ols[.733] Si2p[.368]
A 85.77 2.55 11.68 0
B 84.55 1.46 13.99 0
C 84.26 1.24 12.3 2.2
D 81.66 1.51 15.57 1.25

 

 

 

 

 

 

Legend: A= Untreated big blue stem grass, B= Untreated green ﬂax core, C=Silane
treated big blue stem grass, D= Silane treated green ﬂax core

301

Table 6.3: Ratio of atomic concentrations of ﬁbers

C/Si C/O C/N
7.34332 33.6353
6.0436 57.911
38.3 6.85041 67.9516

65.328 5.2447 54.0795
Legend: A: Untreated big blue stem grass, B= Untreated green ﬂax core, C=Silane
treated big blue stem grass, D= Silane treated green ﬂax core

 

 

 

 

 

 

 

 

 

 

U003>

 

Due to the large number of composites, the mechanical and thermal properties of these
composites have been divided into two groups, and each group is individually discussed.
The groups are: composites containing 20% calcium carbonate, and the composites
containing no calcium carbonate at all. The properties of the group with CaCO3 are ﬁrst
discussed.

The tensile strengths and moduli of SMC produced biocomposites containing calcium
carbonate, are shown in Figure 6.6. The bars represent tensile strength and the points
denote tensile modulus.

The tensile strength of silane treated big blue stem biocomposite is 29% more than
strength of untreated big blue stem (BBSG) biocomposite. While, the tensile strength of
silane treated big blue stem & green ﬂax core biocomposite, is 25% more than strength of
untreated big blue stem & green ﬂax core biocomposite. And, the tensile strength of
untreated treated big blue stem biocomposite is 12% more than strength of untreated big
blue stern & ﬂax biocomposite. The strength of untreated jute-hemp (25 wt %) hybrid
biocomposite is 70% higher than that of untreated BBSG biocomposite. The strength of
untreated green ﬂax core biocomposite is 40% lower than that of untreated BBSG
biocomposite. The strength of untreated jute-hemp (25 wt %) hybrid biocomposite is

20% higher than that of untreated jute-hemp (20 wt %) hybrid biocomposite. The strength

302

of chopped E-glass composite is 20% higher than that of hybrid of E-glass mat-Hemp
composite. The strength of chopped E-glass composite is 170% higher than that of
untreated BBSG biocomposite. The strength of E-glass mat (30wt%)-Bioresin (20wt %)-
is 40% higher than that of chopped E-glass composite.

In case of tensile modulus, silane treated big blue stem biocomposite, has a modulus
value 11.5% more than strength of untreated big blue stem biocomposite. The tensile
modulus of silane treated big blue stem & green ﬂax core biocomposite, is 12% more
than that of untreated big blue stem & green ﬂax core-UPE-CaCO3. And, the tensile
modulus of untreated treated big blue stem biocomposite is 22% more than modulus of
untreated big blue stem & green ﬂax core biocomposite. The tensile modulus of untreated
jute-hemp (25 wt %) hybrid biocomposite is 65 % higher than that of untreated BBSG
biocomposite. The modulus of untreated green ﬂax core biocomposite is 27 % lower than
that of untreated BBSG biocomposite. The modulus of untreated jute-hemp (25 wt %)
hybrid biocomposite is 28 % higher than that of untreated jute-hemp (20 wt %) hybrid
biocomposite. The modulus of chopped E-glass composite is 18 % higher than that of
hybrid of E-glass mat-Hemp composite. The modulus of chopped E-glass composite'is
100% higher than that of untreated BBSG biocomposite. The modulus of E-glass mat
(30wt%)-Bioresin (20wt %)-is 60% lower than that of chopped E-glass composite.

The highest tensile strength was of the samples containing E-glass mat (30wt%)-Bioresin
(20wt %). The chopped E-glass composite and E-glass mat-hemp hybrid biocomposite
had second and third highest tensile strengths, respectively. The highest tensile modulus
was of the samples containing chopped E-glass. The E-glass mat-hemp hybrid

biocomposite and untreated jute-hemp (25 wt %) biocomposite had second and third

303

highest tensile moduli, respectively. The low values of tensile strengths and moduli of

composites containing big blue stem grass and grass ﬂax core were because of short

length of these ﬁbers.

51 TS (MPa) 0 TM (GPa)
70 14

 

O)
O
1

U"
o
1

Tensile Strength(MPa)
Tensile Modulus(GPa)

 

 

 

 

 

   

s
A G

Figure 6.6: Tensile properties of biocomposites

Legend: A=UPE Control, B= Untreated BBSG-Caco3-UPE, C= Silane treated BBSG-
CaCO3-UPE, D= Untreated ﬂax&BBSG—CaCO3-UPE, E=Silane treated Flax &BBSG-
CaCO3-UPE, F= Jute-Hemp(25%wt)-CaCO3-UPE (SMC), G= Untreated core ﬂax-
CaCO3-UPE (SMC), H= Jute-hemp(20 wt %)- CaCO3-UPE (SMC), I= E-glassmat-
Hemp(20 wt%)— CaCO3-UPE (SMC), J= Chopped glass(20 wt%)-CaCO3-UPE (SMC),
K= E-glass mat(30wt%)-Bioresin(20wt %)-UPE

The bending strengths and moduli of elasticity of SMC produced biocomposites
containing calcium carbonate, are shown in Figure 6.7. The bars represent bending
strength and the points denote modulus of elasticity.

The bending strength of silane treated big blue stem-UPE-CaCO3, is 15% more than
strength of untreated big blue stem-UPE-CaCOg. While, the bending strength of silane
treated big blue stem & ﬂax -UPE-CaCO3, is 10% more than strength of untreated big

blue stem & ﬂax-UPE-CaCO3. And, the bending strength of untreated treated big blue

304

stem-UPE-CaCO3, is 6% more than strength of untreated big blue stem & ﬂax-UPE-
CaCO3. The strength of untreated jute-hemp (25 wt %) hybrid biocomposite is 165%
higher than that of untreated BBSG biocomposite. The strength of untreated green ﬂax
core biocomposite is 7% lower than that of untreated BBSG biocomposite. The strength
of untreated jute-hemp (25 wt %) hybrid biocomposite is 40% higher than that of
untreated jute-hemp (20 wt %) hybrid biocomposite. The strength of chopped E-glass
composite is 14% lower than that of hybrid of E-glass mat-Hemp composite. The strength
of chopped E-glass composite is 215% higher than that of untreated BBSG biocomposite.
The strength of E-glass mat (30wt%)-Bioresin (20wt %)-is 40% lower than that of
chopped E-glass composite composite.

In case of modulus of elasticity, silane treated big blue stem-UPE-CaCO3, has a modulus
value 21% more than strength of untreated big blue stem-UPE-CaCO3. The modulus of
elasticity of silane treated big blue stem & ﬂax -UPE-CaCO3, is 17% more than that of
untreated big blue stem & ﬂax-UPE-CaCO3. And, the modulus of elasticity of untreated
treated big blue stem-UPE-CaCO3, is 5% more than modulus of untreated big blue stem
& ﬂax-UPE-CaCO3. The modulus of elasticity of untreated jute-hemp (25 wt %) hybrid
biocomposite is 102 % higher than that of untreated BBSG biocomposite. The modulus
of elasticity of untreated green ﬂax core biocomposite is 17 % lower than that of
untreated BBSG biocomposite. The modulus of elasticity of untreated jute-hemp (25 wt
%) hybrid biocomposite is 25 % higher than that of untreated jute-hemp (20 wt %) hybrid
biocomposite. The modulus of elasticity of chopped E-glass composite is 5 % higher than
that of hybrid of E-glass mat-Hemp composite. The modulus of elasticity of chopped E-

glass composite is 126% higher than that of untreated BBSG biocomposite. The modulus

305

of elasticity of E-glass mat (30wt%)-Bioresin (20wt %)-is 55% lower than that of
chopped E-glass composite composite.

The highest bending strength was of the samples containing E-glass mat-hemp hybrid.
The chopped E-glass composite and untreated jute-hemp (25 wt %) hybrid biocomposite
had second and third highest bending strengths, respectively. The highest modulus of
elasticity was of the samples containing chopped E-glass. The E-glass mat-hemp hybrid
biocomposite and untreated jute-hemp (25 wt %) hybrid biocomposite had second and
third highest moduli of elasticity, respectively. The bending strengths and moduli of

elasticity followed the same trend as tensile strengths and moduli.

a BS (MPa) 0 MOE(GPa)

 

   

160 12
140 -« L 10
120 ~

100 - {7 i 8

80—
60- i

i
i
88
20-
0 -. .

A B C D E

Bending Strength(MPa)
Modulus of Elasticity (Gpa)

 

 

 

’///////////////.Z
///////////////ﬂ/’IZ

 

 

\

 

'11
Q
I
c.
K

Figure 6.7: Flexural properties of biocomposites

Legend: A=UPE Control, B= Untreated BBSG-Caco3-UPE, C= Silane treated BBSG-
CaCO3-UPE, D= Untreated ﬂax&BBSG-CaCO3-UPE, E=Silane treated Flax &BBSG-
CaCO3-UPE, F= Jute-Hemp(25%wt)-CaCO3-UPE (SMC), G= Untreated core ﬂax-
CaCO3-UPE (SMC), H= Jute-hemp(20 wt %)- CaCO3-UPE (SMC), I= E-glassmat-
Hemp (20 wt%)— CaCO3-UPE (SMC), J= Chopped glass(20 wt%)-CaCO3-UPE (SMC),
K= E-glass mat(30wt%)-Bioresin(20wt %)-UPE

Figure 6.8 shows the impact strength of composites of SMC produced biocomposites

containing calcium carbonate. The impact strength of silane treated big blue stem-UPE-

306

CaCO3, is 65% less than strength of untreated big blue stem-UPE-CaCO3. While, the
impact strength of silane treated big blue stem & ﬂax -UPE-CaCO3, is 20% less than
strength of untreated big blue stem & ﬂax-UPE-CaCOB. And, the impact strength of
untreated treated big blue stem-UPE-CaC03, is 66% less than strength of untreated big
blue stem & ﬂax-UPE-CaCO3. The strength of untreated jute-hemp (25 wt %) hybrid
biocomposite is 86% higher than that of untreated BBSG biocomposite. The impact
strength of untreated green ﬂax core biocomposite is 20% lower than that of untreated
BBSG biocomposite. The impact strength of untreated jute-hemp (25 wt %) hybrid
biocomposite is 20% higher than that of untreated jute-hemp (20 wt %) hybrid
biocomposite. The impact strength of chopped E-glass composite is 83% higher than that
of hybrid of E-glass mat-Hemp composite. The impact strength of chopped E-glass
composite is 1330% higher than that of untreated BBSG biocomposite. The impact
strength of E-glass mat (30wt%)—Bioresin (20wt %) is 7% higher than that of chopped E-
glass composite composite.

The impact strengths of composites followed a pattern completely opposite to that of
bending and tensile strengths. This is a common behavior for ﬁber reinforced plastics.
The highest impact strength was of the samples containing of E-glass mat (30wt%)-
Bioresin (20wt %). The chopped E-glass composite, and E-glass mat-Hemp hybrid
composite had second and third highest impact strength, respectively. The impact
strengths of composites containing big blue stem grass and grass ﬂax core were very
small, because, these ﬁbers very small in length. In particular, the length of BBSG ﬁbers

was about 4 mm, and that of green ﬂax core was 1mm.

307

 

 

 

 

 

200 - Notched Izod Impact Strength
’5‘
51504
5
O)
5
a
,"3 100 -
0
a
a.
E
50 -
0 atammmgagl,
A B C D E F G H

Figure 6.8: Impact properties of biocomposites

Legend: A=UPE Control, B= Untreated BBSG-Caco3-UPE, C= Silane treated BBSG-
CaCO3-UPE, D= Untreated ﬂax&BBSG-CaCO3-UPE, E=Silane treated Flax &BBSG-
CaCO3-UPE, F= Jute-Hemp(25%wt)-CaCO3-UPE (SMC), G= Untreated core ﬂax-
CaCO3-UPE (SMC), H= Jute-hemp (20 wt %)- CaCO3-UPE (SMC), I= E-glassmat-
Hemp(20 wt%)- CaCO3-UPE (SMC), J= ChOpped glass(20 wt%)-CaCO3-UPE (SMC),
K= E-glass mat(30wt%)-Bioresin(20wt %)-UPE

The storage modulus of composites of SMC produced biocomposites containing calcium
carbonate, are shown in Figure 6.9. The storage modulus of silane treated big blue stem-
UPE-CaCO3, at 40 0C, is 17.5% more than that of untreated big blue stem-UPB-CaCO3.
The modulus of silane treated big blue stern & ﬂax -UPE-CaCO3 is 13 % more than

modulus of untreated big blue stem & ﬂax-UPE-CaCO3. The modulus of untreated

308

treated big blue stem -UPE-CaCO3 is 26 % more than modulus of untreated big blue stem
& ﬂax-UPE-CaCO3. The modulus of silane treated big blue stem -UPE-CaCO3 is 32 %
more than modulus of silane treated big blue stem & ﬂax-UPE-CaCO3. The modulus of
untreated jute-hemp (25 wt %) hybrid biocomposite is 45 % higher than that of untreated
BBSG biocomposite. The modulus of chopped E-glass composite is 105% higher than
that of untreated BBSG biocomposite. The modulus of E-glass mat (30wt%)-Bioresin
(20wt %) is 55 % lower than that of chopped E-glass composite.

The data for storage modulus followed the same trend as tensile modulus and modulus of

elasticity.

Storage Modulus at 40 °C
1 2000

 

10000 ~
8000 a
6000 _

4000 ~

2000 — I
0 _ :'.:":.l

A B C D E F G H

Storage Modulus (MPa)

 

 

Figure 6.9: Storage modulus of biocomposites at 40 0C

Legend: A=UPE Control, B= Untreated BBSG-CaCO3-UPE, C= Silane treated BBSG-
CaCO3-UPE, D= Untreated ﬂax&BBSG-CaCO3-UPE, E=Silane treated Flax &BBSG-
CaCO3-UPE, F= Jute-Hemp(25%wt)-CaCO3-UPE (SMC), G= Untreated core ﬂax-
CaCO3-UPE (SMC), H= Jute-hemp(20 wt %)- CaCO3-UPE (SMC), I= E-glassmat-
Hemp(20 wt%)- CaCO3-UPE (SMC), J= Chopped glass(20 wt%)-CaCO3-UPE (SMC),
K= E-glass mat(30wt%)-Bioresin(20wt %)-UPE

309

The tensile strengths and moduli of SMC produced biocomposites containing no calcium
carbonate, are shown in Figure 6.10. The bars represent tensile strength and the points
denote tensile modulus.

The tensile strength of untreated hemp (25 vol %)-UPE biocomposite is 145 % more than
strength of untreated henequen (25 vol %)-UPE biocomposite. The tensile strength of
untreated kenaf (25vol %)-UPE biocomposite, is 138 % more than strength of untreated
henequen-UPE biocomposite. And, the tensile strength of untreated hemp-henequen -
UPE biocomposite is 107 % more than strength of untreated henequen (25 vol %)-UPE
biocomposite. The strength of untreated kenaf-henequen-UPE hybrid biocomposite is 13
% lower than that of untreated kenaf-UPE biocomposite. The strength of untreated hemp-
henequen (30 wt %)- bioresin (20 wt%) hybrid biocomposite is 10% lower than that of
untreated hemp biocomposite. The strength of untreated henequen biocomposite is 38 %
lower than that of neat polyester resin. The strength of untreated hemp biocomposite is 51
% higher than that of neat polyester resin. The strength of untreated kenaf biocomposite
is 47 % higher than that of neat polyester resin. The strength of untreated hemp-henequen
hybrid biocomposite is 27 % higher than that of neat polyester resin. The strength of
untreated kenaf-henequen hybrid biocomposite is 27% higher than that of neat polyester
resin. The strength of untreated hemp-henequen (30 wt %)-bioresin (20 wt %) hybrid
biocomposite is 35 % higher than that of neat polyester resin.

In case of tensile modulus, the modulus of untreated hemp (25 vol %)-UPE biocomposite
is 104 % more than modulus of untreated henequen (25 vol %)-UPE biocomposite. The
tensile modulus of untreated kenaf (25vol %)-UPE biocomposite, is 125 % more than

modulus of untreated henequen-UPE biocomposite. And, the tensile modulus of untreated

310

 

hemp-henequen -UPE biocomposite is 120 % more than modulus of untreated henequen
(25 vol %)-UPE biocomposite. The modulus of untreated kenaf-henequen-UPE hybrid
biocomposite is 24 % lower than that of untreated kenaf-UPE biocomposite. The
modulus of untreated hemp-henequen (30 wt %)-bioresin (20 wt %) hybrid biocomposite
is 14 % lower than that of untreated hemp biocomposite. The modulus of untreated
henequen biocomposite is 195 % higher than that of neat polyester resin. The modulus of
untreated hemp biocomposite is 500 % higher than that of neat polyester resin. The
modulus of untreated kenaf biocomposite is 560 % higher than that of neat polyester
resin. The strength of untreated hemp-henequen hybrid biocomposite is 550 % higher
than that of neat polyester resin. The modulus of untreated kenaf-henequen hybrid
biocomposite is 402 % higher than that of neat polyester resin. The modulus of untreated
hemp-henequen (30 wt %)-bioresin (20 wt %) hybrid biocomposite is 415% higher than
that of neat polyester resin.

The highest tensile strength was of the samples containing untreated hemp ﬁbers. The
untreated kenaf biocomposite and untreated hemp-henequen hybrid biocomposite had
second and third highest tensile strengths, respectively. The highest tensile modulus was
of the samples containing untreated kenaf ﬁbers. The untreated hemp-henequen hybrid
biocomposite and untreated hemp biocomposite had second and third highest tensile
moduli, respectively. The low values of tensile strengths and moduli of composites

containing big blue stem grass and grass ﬂax core were because of short length of these

ﬁbers.

311

a TS (MPa) 9 TM (GPa)

80 14
A 12...
E E
.C U)
+0 3
g 8 a
£40 103
CD 6 E
2 .2
'17) 4 "a
:20 C
Q 0
1- 2 1-

0 0

 

B C D E F G

Figure 6.10: Tensile properties of biocomposites

Legend: A=UPE Control, B= Untreated henequen (25 vol %)-UPE, C= Untreated hemp
(25 vol %)-UPE, D=Untreated kenaf (25vol %)-UPB, E= Untreated hemp-henequen (25
vol %)-UPE, F= Untreated kenaf- henequen (25 vol %), G= Untreated hemp-
henequen(30 wt %)- Bioresin (20 wt %)-UPE

The bending strengths and moduli of elasticity of SMC produced biocomposites
containing no calcium carbonate are shown in Figure 6.11. The bars represent bending
strength and the points denote modulus of elasticity.

The bending strength of untreated hemp (25 vol %)-UPE biocomposite is 28 % more than
strength of untreated henequen (25 vol %)-UPE biocomposite. The bending strength of
untreated kenaf (25vol %)-UPE biocomposite, is 12 % more than strength of untreated
henequen-UPE biocomposite. And, the bending strength of untreated hemp-henequen -
UPE biocomposite is 7 % less than strength of untreated henequen (25 vol %)-UPE
biocomposite. The bending strength of untreated kenaf-henequen-UPE hybrid

biocomposite is 11 % lower than that of untreated kenaf-UPE biocomposite. The bending

strength of untreated hemp-henequen (30 wt %)-bioresin (20 wt %) hybrid biocomposite

312

is 20 % lower than that of untreated hemp biocomposite. The bending strength of
untreated henequen biocomposite is 18 % lower than that of neat polyester resin. The
bending strength of untreated hemp biocomposite is 4 % higher than that of neat
polyester resin. The bending strength of untreated kenaf biocomposite is 8 % lower than
that of neat polyester resin. The bending strength of untreated hemp-henequen hybrid
biocomposite is 23 % lower than that of neat polyester resin. The bending strength of
untreated kenaf-henequen hybrid biocomposite is 27% lower than that of neat polyester
resin. The bending strength of untreated hemp-henequen (30 wt %)-bioresin (20 wt %)
hybrid biocomposite is 17 % lower than that of neat polyester resin.

In case of modulus of elasticity, the modulus of untreated hemp (25 vol %)-UPE
biocomposite is 66 % more than modulus of untreated henequen (25 vol %)-UPE
biocomposite. The modulus of elasticity of untreated kenaf (25vol %)-UPE biocomposite,
is 70 % more than modulus of untreated henequen-UPE biocomposite. And, the modulus
of elasticity of untreated hemp-henequen -UPE biocomposite is 16 % more than modulus
of untreated henequen (25 vol %)-UPE biocomposite. The modulus of untreated kenaf-
henequen-UPE hybrid biocomposite is 38 % higher than that of untreated kenaf-UPE
biocomposite. The modulus of untreated hemp-henequen (30 wt %)- bioresin (20 wt%)
hybrid biocomposite is 24 % lower than that of untreated hemp biocomposite. The
modulus of untreated henequen biocomposite is 94 % higher than that of neat polyester
resin. The modulus of untreated hemp biocomposite is 222 % higher than that of neat
polyester resin. The modulus of untreated kenaf biocomposite is 230 % higher than that
of neat polyester resin. The strength of untreated hemp-henequen hybrid biocomposite is

125 % higher than that of neat polyester resin. The modulus of untreated kenaf-henequen

313

hybrid biocomposite is 170 % higher than that of neat polyester resin. The modulus of
untreated hemp-henequen (30 wt %)-bioresin (20 wt %) hybrid biocomposite is 145 %
higher than that of neat polyester resin.

The highest bending strength was of the samples containing unheated hemp ﬁbers. The
untreated kenaf biocomposite and untread hemp-hennequen hybrid biocomposite had
second and third highest bending strengths, respectively. The highest modulus of
elasticity was of the samples containing untreated kenaf ﬁbers. The untreated hemp
biocomposite and untreated kenaf-henquen hybrid biocomposite had second and third
highest moduli of elasticity, respectively. The bending strengths and moduli of elasticity

followed the same trend as tensile strengths and moduli.

521 as (MPa) 9 MOE (GPa)

 

 

 

        
     
       
       
     
       
       
 

 

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

120 13
A T - 11 ‘3‘
g 100 r I g
‘2' T a: :-: r 9 V
5 80 — to»: T 3?
O) :e:e:e: I- 0
5 5:4: 7 '77,
e 50 “ It??? 2
”e. r5 w

b I I.
c 40 - 9:40: o
‘6 f 1360301 F 3 m
c 3"“ 2
8 20 C e _ 1 1:3
0
0 I I I I I I -1 5

A B C D E F G

Figure 6.11: Flexural properties of biocomposites

Legend: A=UPE Control, B= Untreated henequen (25 vol %)—UPE, C= Untreated hemp
(25 vol %)-UPE, D=Untreated kenaf (25vol %)-UPE, E= Untreated hemp-henequen (25
vol %)-UPE, F= Untreated kenaf- henequen (25 vol %), G= Untreated hemp-
henequen(30 wt %)- Bioresin (20 wt %)-UPE

314

Figure 6.12 shows the impact strength of composites of SMC produced biocomposites
containing no calcium carbonate. The impact strength of untreated hemp (25 vol %)-UPE
biocomposite is 50 % less than strength of untreated henequen (25 vol %)-UPE
biocomposite. The impact strength of untreated kenaf (25vol %)-UPE biocomposite, is 60
% less than strength of untreated henequen-UPE biocomposite. And, the impact strength
of untreated hemp-henequen -UPE biocomposite is 30 % less than strength of untreated
henequen (25 vol %)-UPE biocomposite. The impact strength of untreated kenaf-
henequen—UPE hybrid biocomposite is 9 % higher than that of untreated kenaf-UPE
biocomposite. The impact strength of untreated hemp-henequen (30 wt %)-bioresin (20
wt %) hybrid biocomposite is 16 % lower than that of untreated hemp biocomposite. The
impact strength of untreated henequen biocomposite is 430 % higher than that of neat
polyester resin. The impact strength of untreated hemp biocomposite is 170 % higher
than that of neat polyester resin. The impact strength of untreated kenaf biocomposite is
115 % higher than that of neat polyester resin. The impact strength of untreated hemp-
henequen hybrid biocomposite is 270 % higher than that of neat polyester resin. The
impact strength of untreated kenaf-henequen hybrid biocomposite is 135 % higher than
that of neat polyester resin. The impact strength of untreated hemp-henequen (30 wt %)-
bioresin (20 wt %) hybrid biocomposite is 82 % higher than that of neat polyester resin.

The impact strengths of composites followed a pattern completely opposite to that of
bending and tensile strengths. This is a connnon behavior for ﬁber reinforced plastics.
The highest impact strength was of the samples containing untreated henequen ﬁbers.
This result is not surprising because, it is known that leaf ﬁbers have high toughness and

low stifﬁress, while, bast ﬁbers have low toughness and high stiffness. A hybrid bio-

315

composite of 25 wt % untreated hemp and 10 wt % untreated henequen had second

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

J I
\ ‘, "
a ﬁ‘} D
v t» »
ccc
c b»),
H ccc
l)».
c»
h)».
ccc ‘ V
>>») a»):
ccc c c
_ >5») rs»:
ccc c c
’c’c’c’ ’c’c c.
H r»:» I a)»:
ccc ccc
>>>: : c or»:
ff)" ‘. 9 T 033‘:
“ ccc ccc ccc c c
0 ‘j t»); a). v v a»): b» )
ccc cccc I c ccc c c
b)»; to» a 3)): b )
a ccc cccc Iccc ccc ccc c (
l:)) v): >») ’DDI b»): a r
a ccc cccc Iccc ccc ccc c 4
tr») 5)» r)» a»): >)>: ) )
ccc cccc Iccc ccc ccc cccc
__‘ l»); a») a») a)»: >9») ))
ccc cccc Iccc ccc ccc cccc
>D)’ DD) .9: r»): bra» r»)
— c c ccc cccc [ccc ccc ccc cccc
o :1 >9)» at» a)» a»): p»): c».
ccc ccc cccc hccc ccc ccc cc 4
c)»: >3») or) 9»: a)» baa» r»)
ccc ccc cccc iccc ccc ccc cccc
a)»: l”’ a») )b) a») >))) ))
0 ccc 11; ccc ccc ccc ccc ccc

Figure 6.12: Impact properties of biocomposites

Legend: A=UPE Control, B= Untreated henequen (25 vol %)-UPE, C= Untreated hemp
(25 vol %)-UPE, D=Untreated kenaf (25vol %)-UPE, E= Untreated hemp-henequen (25
vol %)-UPE, F= Untreated kenaf- henequen (25 vol %), G= Untreated hemp-
henequen(30 wt %)- Bioresin (20 wt %)-UPE

The storage modulus of composites of SMC produced biocomposites containing no
calcium carbonate, are shown in Figure 6.13.

The storage modulus of untreated hemp (25 vol %)-UPE biocomposite is 22 % more than
storage modulus of untreated henequen (25 vol %)-UPE biocomposite. The storage
modulus of untreated kenaf (25vol %)-UPE biocomposite, is 30 % more than storage
modulus of untreated henequen-UPE biocomposite. The storage modulus of untreated

henequen biocomposite is 50 % more than that of neat polyester resin. The storage

modulus of untreated hemp biocomposite is 83 % higher than that of neat polyester resin.

316

 

Storage Modulus at 40 °C

 

          

 

         

sawswwawswawxc. . .«wﬂﬁﬁﬁﬂﬁﬁiﬂ
"v“vvvvvvvvvvw..vvvvvvvRVX
.3... V V»«vwvv«V3»(«vvvvvvvvvvwx
cvsvvvvv.«V«xv»@«vvvvvvvvwx
. VNVNVNVNVNVNVNVNVNVNVNVNVNVNVNVNVNVNM

0000000000000000000000000000

 
 
 
 

     

 

 
    
  
  

    

 
     

.aaaauacxxaawxxxxxxxxx
“~vaVvvvvvvvvwvwvcxx«em
.H .\«««\«««««««««««««««»«««««««««««««V«
«w~w~w~w~wswawswswstwswswswwswswswxwxwawt
.\./A/A/A/«VNVA/A/A/AlA/A/A/«VA/A/«VA/A/A/A4.

        
  

 
   

    

ccccc
O

000000000000

000000000000

O 0 o c

o c a

oooooooo

 

 

 

 

 

The data for storage modulus followed the same trend as tensile modulus and modulus of
7000

The storage modulus of untreated kenaf biocomposite is 93 % higher than that of neat

polyester resin.
elasticity.

LL
_ .i q _ _ q
0 0 0 0 0 0 0
0 0 0 O 0 0
0 0 0 0 0 0
6 5 4 3 2 1
was: «228.2 322m

Untreated henequen (25vol%)-UPE, C

Untreated kenaf (25%vol)-UPE

Storage modulus of biocomposites at 40 0C

Unsaturated polyester resin, B

SMC line. They are: untreated BBSG-UPE-CaCO3, silane treated BBSG- UPE-CaCOg,

Figure 6.14 shows the ESEM micrographs of four biocomposite samples made using
untreated BBSG-green ﬂax-UPE-CaCO3, silane treated BBSG-green ﬂax-UPE-CaCO3.
All the micrographs are at the magniﬁcation of 100 X and scale bar of 450 pm. The

Untreated hemp (25%vol)-UPE, D

Figure 6.13
Legend: A

ESEM pictures were taken from the tensile fractured surfaces of these composites. In all
the pictures, ﬁber pull out was observed. The length of the ﬁber pulled out became
317

shorter after the chemical treatment, as is seen in b) and d).

 

c) Untreated BBSG-GFC-UPE-CaCO3 d) Silane treated BBSG-GFC-UPE—CaCO3

Figure 6.14: ESEM micrographs of SMC biocomposites, at magniﬁcation of 100 X and
scale bar of 450 um

CONCLUSIONS

Biocomposites have been successfully made using natural ﬁbers, unsaturated polyester
resin, and bioresin by sheet molding compound panel processing. These biocomposites

were made in the same SMC equipment, which is used to fabricate glass-polyester

318

composites. As a comparison, we also fabricated glass-polyester composites on this
equipment. The biocomposites were made on this SMC line after a few minor
adjustments. Instead of using the traditional ﬁber feeding system, we used a screw feeder
and a vibratory feeder to supply natural ﬁbers to the set-up. Consistent and repeatable
results were obtained showing that this process is consistent and can be used for
fabrication of bio-composites.

We have also found that glass-UPE composites have almost same speciﬁc strength and
modulus as natural ﬁber-UPE composites. But, with optimization of the entire BCSMC
process, use of engineered natural ﬁbers, and inclusion of desirable additives, we seek to
achieve best mechanical, thermal and physical properties as comparable as to glass based
SMC, and thus replace/substitute glass-UPE composites with natural ﬁber biocomposites.
We aim to use biocomposites sheet molding compound panel processing (BCSMC) for
fabrication of biocomposites composed of natural ﬁbers and unsaturated polyester resin.
Chopped natural ﬁbers like, hemp, kenaf, pineapple leaf ﬁber, glass ﬁbers, hybrid ﬁbers,
will be used to reinforce unsaturated polyester resin and bioresins in high speed sheet
molding compound panel processing. This process would result in continuous and high
volume manufacture of biocomposites. As a result, the industrial scale production of
biocomposites would be possible. This will lead to accessibility of environmental goods
for multiple uses in automotives, buildings as well as in ﬁirniture industries.

The newly developed process focuses on large-scale production of biocomposites
containing thermoset resins and natural ﬁbers. The equipment used for this processing is
the common industrial SMC line. Currently, natural ﬁbers reinforced thermoset

composites are not prepared using SMC [12-14]. Our processing aims to use SMC

319

process to fabricate these natural ﬁber-thermoset composites. In commercial SMC set-up,
continuous glass ﬁbers rovings are fed to a chopper, which cuts them to a 6 mm size, and
they fall on the carrier ﬁlm, forming a uniform layer of chopped glass ﬁbers. Since,
natural ﬁbers cannot be obtained in a continuous from, and making a continuous yarn or
roving with these ﬁbers would be a difﬁcult and expensive, we make use of chopped
natural ﬁbers in this new process.

Biocomposites have been successfully made using natural ﬁbers, unsaturated polyester
resin, and bioresin by sheet molding compound panel processing. These biocomposites
were made in the same SMC equipment, which is used to fabricate glass-polyester
composites. As a comparison, we also fabricated glass-polyester composites on this
equipment. The biocomposites were made on this SMC line after a few minor
adjustments. Instead of using the traditional ﬁber feeding system, we used a screw feeder
and a vibratory feeder to supply natural ﬁbers to the set-up. Consistent and repeatable
results were obtained showing that this process is consistent and can be used for
fabrication of biocomposites.

We have also found that glass-UPE composites have almost same speciﬁc strength and
modulus as natural ﬁber-UPE composites. But, with optimization of the entire BCSMC
process, use of engineered natural ﬁbers, and inclusion of desirable additives, we seek to
achieve best mechanical, thermal and physical properties as comparable as to glass based
SMC, and thus replace/substitute glass-UPE composites with natural ﬁber biocomposites.
A novel high volume processing technique named ‘biocomposite stampable sheet
molding compound panel’ (BCSMCP) manufacturing process was developed so as to

mimic the continuous sheet molding compound (SMC) as is currently used in making

320

glass ﬁber-polyester resin composites. Natural ﬁber-unsaturated polyester resin
biocomposites were fabricated using the biocomposite stampable sheet molding
compound panel’ (BCSMCP) manufacturing process.

The natural ﬁbers used for making the biocomposites using this process were: big blue
stern grass, green ﬂax core, hemp, henequen, kenaf, coir, ﬂax and jute. For biocomposite
fabrication using SMC line, 20% calcium carbonate by weight was added to the matrix as
a ﬁller. The natural ﬁber content was 20% by weight in biocomposites with big blue stem
grass, silane treated big blue stem grass, silane treated big blue stem grass and green ﬂax
core, and, hemp and jute. Unsaturated polyester resin content in the formulations was
60% by weight. The highest ﬁber content for biocomposites, achievable using SMC
process with UPE and CaCO3 in the matrix was 25 "/0 by weight, achieved in the case of
a hybrid of untreated jute and hemp. Hybrid biocomposites were made by combining,
hemp with jute mats, hemp with kenaf, hemp with henequen, big blue stern grass with
green ﬂax core, silane treated big blue stern grass with silane treated green ﬂax core,
hemp with sisal. A hybrid composite was made by combining E-glass mats with hemp.
To reduce the amount of UPE in the matrix, 20% by weight of bioresin was added to a
hybrid biocomposite with hemp and henequen. The bioresin was a soybean oil phosphate
ester polyol modiﬁed with maleic anhydride. The mechanical properties of hybrid
biocomposites with BBSG and green ﬂax core increased after chemical treatment with
1% methacryloxypropyltrimethoxy silane (y-MPS). For the SMC samples containing
calcium carbonate, the highest tensile strength was of the samples containing E—glass mat
(30wt%)-Bioresin (20wt %). For the SMC samples containing calcium carbonate, the

chopped E-glass composite and E-glass mat-hemp hybrid biocomposite had second and

321

third highest tensile strengths, respectively. For the SMC samples containing calcium
carbonate, the highest tensile modulus was of the samples containing chopped E-glass.
For the SMC samples containing calcium carbonate, the E-glass mat-hemp hybrid
biocomposite and untreated jute-hemp (25 wt %) biocomposite had second and third
highest tensile moduli, respectively. For the SMC samples containing calcium carbonate,
the highest bending strength was of the samples containing untreated hemp ﬁbers. For the
SMC samples containing calcium carbonate, the untreated kenaf biocomposite and
untreated hemp-henequen hybrid biocomposite had second and third highest bending
strengths, respectively. For the SMC samples containing calcium carbonate, the highest
modulus of elasticity was of the samples containing untreated kenaf ﬁbers. For the SMC
samples containing calcium carbonate, the untreated hemp biocomposite and untreated
kenaf-henequen hybrid biocomposite had second and third highest moduli of elasticity,
respectively. For the SMC samples containing calcium carbonate, the highest impact
strength was of the samples containing of E-glass mat (30wt%)-Bioresin (20wt %). For
the SMC samples containing calcium carbonate, the chopped E-glass composite, and E-
glass mat-Hemp hybrid composite had second and third highest impact strength,
respectively. For the SMC samples containing no calcium carbonate, the highest tensile
strength was of the samples containing untreated hemp ﬁbers. For the SMC samples
containing no calcium carbonate, the untreated kenaf biocomposite and untreated hemp-
henequen hybrid biocomposite had second and third highest tensile strengths,
respectively. For the SMC samples containing no calcium carbonate, the highest tensile
modulus was of the samples containing untreated kenaf ﬁbers. For the SMC samples

containing no calcium carbonate, the untreated hemp-henequen hybrid biocomposite and

322

untreated hemp biocomposite had second and third highest tensile moduli, respectively.
For the SMC samples containing no calcium carbonate, the highest bending strength was
of the samples containing untreated hemp ﬁbers. For the SMC samples containing no
calcium carbonate the untreated kenaf biocomposite and untreated hemp-henequen
hybrid biocomposite had second and third highest bending strengths, respectively. For the
SMC samples containing no calcium carbonate, the highest modulus of elasticity was of
the samples containing untreated kenaf ﬁbers. For the SMC samples containing no
calcium carbonate, the untreated hemp biocomposite and untreated kenaf-henquen hybrid
biocomposite had second and third highest moduli of elasticity, respectively. For the
SMC samples containing no calcium carbonate, hybrid biocomposite of 25 wt %
untreated hemp and 10 wt % untreated henequen had second highest impact strength. For
the SMC samples containing no calcium carbonate, untreated henequen biocomposite had

the highest impact strength.

6.2 Biobeams and bioplates from VARTM

The work on bioplates and biobeams was done in collaboration with Mario Quagliata,
and Dr. Rigoberto Burgueﬁo of the Department of Civil and Environmental Engineering.
More details on this part of the work can be obtained from Mario’s M.S. thesis,
“Development and Characterization of Biocomposite Cellular Beams and Plates for
Load-Bearing Components”, Masters Thesis, Michigan State University, East Lansing,

MI; 2003. A brief description of this work is presented in the literature review.

References

323

1. C. J. Davis, R. P. Wood, E. R., Miller, “Glass ﬁber-reinforced sheet molding
compound”, US 3615979, (1971).

2. F. C. Peterson, L. P. Theard, “Thermosetting sheet molding compounds”, US
3713927, (1973).

3. A. N. Piacente, US 3835212, “Resinous sheet like products”, (1974).
4. F. C. Peterson, L. P. Theard, “Sheet molding compounds”, DE 2357000, (1974).
5. T. Kishino, “Unsaturated polyester compositions for moldings”, JP 49076988, (1974).

6. K. Horiuchi, T. Kamiya, Y. Ogawa, Y. Nakamura, M. Sugimori, “Unsaturated
polyester compositions”, JP 49107086, (1974).

7. K. Takaishi, K. Masamoto, K. Tagawa, “Molding of unsaturated polyesters”, JP
50005463, (1975).

8. M. Sekiguchi, K. Nomaguchi, H. Tanaka, “Molding resin compositions”, JP
50008882, (1975)

9. Y. Takayama, Y. Ichimura, T. Aoyagi, “Sheet molding compositions”, JP 50014792,
(1975)

10. M. Sato, T. Watanabe, W. Koga, “Sheet molding compositions”, JP 50036588,
(1975)

11. R. Kondo, K. Nakagawa, M. Fukuda, I. Kishi, T. Ohtsuki, “Fire retardant
thermosetting resin composition”, US 3931095, (1976). Kondo, Renichi; Nakagawa,
Koji; Fukuda, Makoto; Kishi, Ikuji; Ohtsuki, Tateki.

12. R. P. Wool, J. Lu, S. N. Khot, “Sheet molding compound resins from plant oils”, US
2003088007, (2003).

13.B. van Voorn, H. H. G. Smit, R.J. Sinke, B.de Klerk, Composites: Part A, 32
(2001),1271-1279.

14. D. N. Goswami, P. C. Jha, K. Mahato, K. K. Kumar, Popular Plastics & Packaging,
48(3), (2003), 68-71.

324

CHAPTER SEVEN: RESULTS ON DURABILITY

7.0 Results on weathering, moisture absorption, and ﬂammability

The biocomposites were exposed to moisture absorption tests, accelerated weathering
and ﬂammability tests to assess their durability. The results of these tests are
illustrated in the following sections.

7.1 Moisture Absorption

The samples of bioplastics and biocomposites were subjected to controlled moisture
absorption analysis in a humidity chamber. The percent weight gained by the samples
was plotted against square root of time. The test was continued until the curve
plateued off, and the equilibrium moisture content was achieved. Figures 7.1.1, 7.1.2,
and 7.1.3 show the moisture absorption characteristics of the bioplastic samples made
from grafted soybean polyol, compression molded hempmatl samples, and

biocomposites manufactured using SMC, respectively.

 

 

 

 

 

 

 

 

 

— 10% B'nresin --- 20% Bioresin — 30% Bioresin
0.1

g 0.06 .
E 0.02 .
a
CD
ﬁ-ODZ ~
0
B

-0.06 -

-0.1 T r l I
O 10 20 30 0.5 40 50
Square root of time (hr)

Figure7.1.l: Moisture absorption of bioresins

325

 

 

 

 

 

 

0.9
—— Untreated hempmat-UPE — AN treated hemprmt-UPE
0.7 ~
$5
8 0.5 ~
.5
a
U
E 0.3 ~
.29
Q)
3
0.1 7
“0.1 I I r Ti
0 10 4O 50

20 , 30 05
Square Root of Time (hrs)

Figure7.1.2: Moisture absorption of biocomposites of hemp mat

 

0.7
— Silane treated BBSG-CaCO3-UPE

—— Sihne treated BBSG & Green Flax Core-CaCOS-UPE

 

0.5 *

 

Weight Gained (%)
O
U)

0.1 ~/

-0.1 r r r I
0 10 4O 50

 

 

 

20 3O 0 5
Square Root of Time (his) '

Figure7.1.3: Moisture absorption of SMC manufactured biocomposites
From Figure 7.1.1, it was seen that the bioplastics samples made using 10%, 20%,

and 30% GFTSOPEP3, lost weight in the beginning of the experiment until 100 hours,

326

and then started gaining weight. The equilibrium moisture content for these samples
was about 0.06 % moisture. This shows very high hydrophobicity of the samples
made using bioresins, because under the same conditions, UPE control would have
the equilibrium moisture content of about 1 %. However, the initial weight loss of
these bioplastics samples was not understood. Possible explanations for this cause
were either leaching out of bioresin from the plastic samples, or presence of any
unreacted bioresin in the plastic. To investigate these claims, thermal analysis of the
samples was done after their initial exposure to humidity chamber.

Figure 7.1.2 shows the moisture absorption characteristics of biocomposite samples
made with untreated and acrylonitrile treated hempmatlina UPE matrix. The samples
started gaining weight right ﬁom the beginning of the experiment, and continued
doing so until the equilibrium was reached. This tendency was due to hydrophilic
nature of the ﬁbers due to the presence of many OH groups in the cellulosic backbone
of bioﬁbers. Water molecules get hydrogen bonded to the hydroxyl groups within the
ﬁber cell wall of the bioﬁbers. The equilibrium moisture content for the samples with
untreated hemp matl was 0.7 %, while that for the one with acrylonitrile treated
hempmatl was about 0.3 % moisture. Thus, it was observed that in addition to
increment in mechanical and thermal properties, acrylonitrile treatment leads to less
moisture absorption, giving rise to a more stable composite.

In Figure 7.2.3 the plot of weight gain % versus time is seen for silane treated BBSG-
CaCO3-UPE and silane treated BBSG & green ﬂax core-CaCO3-UPE, both of which
were processed by BCSMCP. It was seen that these biocomposites too started gaining

weight right from the beginning of the experiment, and continued doing so until the

327

equilibrium was reached. This weight gain has been observed by many authors, and
was due to hydrophilic nature of the ﬁbers due to the presence of many OH groups in
the cellulosic backbone of bioﬁbers [1-9]. The equilibrium moisture content for the
samples with silane treated BBSG was 0.28 %, while that for the one with silane
treated BBSG & green ﬂax core was about 0.55 % moisture. Here, the equilibrium
moisture content of silane treated BBSG & green ﬂax core was more than that of
silane treated BBSG. Green ﬂax core, which constitutes 50 % ﬁber and 50 % core, is
more hydrophilic than grass because of the presence of powdered core. This led to

higher weight increase in the hybrid biocomposite containing grass and green ﬂax.

Thermal analysis of bioplastics

The bioplastics which lost weight in the initial period of moisture absorption test were
thermally investigated by TGA and DSC. These samples were at room temperature
before the test. Figure 7.1.4 and 7.1.5 show the plot from TGA analysis and DSC
analysis of the bioplastics. Table 7.1.1 shows the maximum degradation temperature
of the bioplastics examined by TGA. It was observed that on increasing the amount of
bioresin in the bioplastic, the maximum degradation temperature decreased. For a
bioplastic with 10% GFTSOPEP3, the maximum degradation temperature was 346
0C; for a bioplastic with'20% GFTSOPEP3, it was 341 OC; and for a bioplastic with
30% GFTSOPEP3 it was 333 0C. The initial decomposition temperature of the
bioplastics was measured from Figure 7.1.4. The initial decomposition temperature
clearly became lower with an increase of bioresin concentration. The reduction of the

initial decomposition temperature is indicative of the existence of unreacted

328

constituents. Generally, thermoset polymers having higher cross-link density show

higher maximum decomposition temperature.

 

 

 

 

 

120 T ___,_L_ 1.20
30% Biores'n 30% Bioresil !

100 , "\11) 20% Bioresin T 1-00
20% B'nres'n - l A
11‘10 ' ' 1 U
T lO/oBroresn + 0.80 3
.\ so a . L
:7 ' 10% Bioresin j I 5
§ i i 0.60 g

In

a: 60 I , 1 i.
:5 I I 0.40 .5
.39 | a
g 40 i N " l 020 .E,
\ i . a

20 4 —* "‘f K \ i I 0.00

O + I I I I . I I .020

O 100 200 300 400 500 600 700
Temperature (0 C)

Figure 7.1.4: TGA of bioplastics

 

 

 

 

 

 

 

2- _ :
2 - f T 1
§ 10%Bioresin
s 1 ..
B
a 0? 20%Bioresin 1
’5 -1 “I ‘i - l
I -2 ~ 30%B'nres'm
-3- wt __..,_,,_ l
l I
-4 if I ﬂ I i
-70 30 130 230 330
Temperature (0 C)

Figure 7.1.5: DSC of bioplastics

329

Figure 7.1.5 shows the best ﬂow curves from -60 to 300 OC, obtained from DSC.
There were no exothermic peaks in the DSC curves of any of the bioplastic samples.
This indicated the absence of any unreacted bioresin in the bioplastic samples.

Table 7.1.1: Maximum degradation temperature of the bioplastics

 

 

 

 

 

Maximum decom osition
Sample temperature ( C)
A 346.15
B 341.23
C 332.66

 

 

 

Legend: A= 10% Bioresin, B=20% Bioresin, and C=30% Bioresin

7.2 Accelerated weathering

The accelerated weathering of the biocomposite samples was done to study the effect
of harsh weather elements on the properties of biocomposites. The exposed samples
were evaluated by color test, surface roughness, weight change, and dynamical
mechanical analysis. The results from these analyses are presented below.

Color change

+A +8 +C —x—D

 

 

 

 

0 x r I l T I I

 

15 30 45 6O 75 90
Time (Days)

Figure 7.2.1: Change in color parameter ‘L’ over time for biocomposites

330

Legend: A=Untreated hempmatl-UPE, B=Acrylonitrile treated hempmatl-UPE,
C=Silane treated BBSG-CaCO3-UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE

+A +B +C —x—D

 

 

 

 

 

0 15 30 45 60 75 90
Time (Days)
Figure 7.2.2: Change in color parameter ‘a’ over time for biocomposites
Legend: A=Untreated hempmatl -UPE, B=Acrylonitrile treated hempmatl -UPE,
C=Silane treated BBSG-CaCO3—UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE

+A +8 +C —x—D
15

 

10*

 

 

 

 

0 15 3O 45 60 75 90
Time (Days)
Figure 7.2.3: Change in color parameter ‘b’ over time for biocomposites
Legend: A=Untreated hempmatl-UPE, B=Acrylonitrile treated hempmatl -UPE,
C=Silane treated BBSG-CaCO3-UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE

331

—I—A +8 +C —x—D

 

 

 

 

 

O x ‘1‘ I I T I l

15 3O 45 . 60 75 90
Time (Days)

Figure 7.2.4: Change in color parameter ‘E’ over time for biocomposites

Legend: A=Untreated hempmatl-UPE, B=Acrylonitrile treated hempmatl-UPE,
C=Silane treated BBSG-CaCO3-UPE, D= Silane treated BBSG & Green Flax Core-
CaCOB-UPE

Figures 7.2.1 to 7.2.4 show the change in various color parameters of the
biocomposite samples with exposure to accelerated weathering.

In the CIELAB system, there are three parameters for color, L*, a*, b*. The L* axis
represents the lightness, whereas, a* and b* axes are the chromaticity coordinates.
While +a* is for the red, -a* is for green, +b* for yellow, -b* for blue, and L* varies
from 100 (white) to zero (black). The changes in the values of L, a, b were used to
calculate dE*, which is the cumulative color change.

The values of dL increased from 0 at beginning of the experiment to 50 for samples
exposed for 85 days. This was because the samples were changing color from
greenish brown to white after being exposed to UV, rain, condensation and humidity.
The combination of water, oxygen and UV irradiation promotes the change of color

of these biocomposite samples. The change in L was less pronounced for samples

containing BBSG and green ﬂax core which were processed in SMC.

332

The values of da decreased from 0 for all samples. The values of db changed from 0
to positive in case of hemp matl composites, but changed from 0 to negative in case
of BBSG and GFC biocomposites. The overall color change of the biocomposites, dE,
increased from 0 to 50 for all biocomposites. Here again, the dE values of
biocomposites with hemp matl were higher than those of samples containing BBSG
and green ﬂax core which were processed in SMC. The color change of untreated
hemp matl based biocomposites was slightly higher than that of the acrylonitrile
treated hemp matl based biocomposite.

Weight change

Figures 7.2.5 shows the plot of percentage weight gain with exposure time for
biocomposite samples. All of the samples were gaining weight in the beginning of the
weathering test, until about 30 days into the test. The weight gain was due to

hydrophilicity of the bioﬁbers.

 

Weight Gain (%)

 

 

 

 

-5 T I I I

0 20 40 80 100

60
Time (days)

Figure 7.2.5: Change in weight over time for biocomposites

333

Legend: A=Untreated hempmatl-UPE, B=Acrylonitrile treated hempmatl -UPE,
C=Silane treated BBSG-CaCO3-UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE

Aﬁer 30 days of exposure, the biocomposite samples started losing weight and
continued doing so until the end of the exposure time. This weight loss was due to the
biodegradation of the biocomposites, which began aﬁer 30 days of exposure to
weathering. However this weight loss was less than 2% for samples with untreated
hempmatl and silane treated BBSG. The weight loss was less than 5% for
biocomposite sample with silane treated BBSG and GFC, while, it was about 1.25%
for the sample with acrylonitrile treated hemp matl. Therefore, it was established that
acrylonitrile treatment of the hemp ﬁbers makes them more stable compared to
untreated hemp based ones. This could be due to improved adhesion between ﬁbers
and the matrix, and better interfacial bonding in a composite made with surface
treated ﬁbers. The values of weight loss was highest for the biocomposite containing

silane treated BBSG and GF C was again because of presence of core particles which

do not have the characteristics as the ﬁbers.

Surface roughness

Figures 7.2.5, 7.2.6, and 7.2.7 show the comparison of surface roughness parameters
of the biocomposite samples as a function of exposure time. In general, the rouglmess
of the composite samples increased with exposure time. Of all surface roughness
parameters, the values of Ra were smallest, followed by those of R2, and ﬁnally those
of Rm. The roughness parameters of the samples with untreated hemp matl were the

highest, followed by samples with silane treated BBSG and GFC, followed by sample

334

with acrylonitrile treated hemp matl. The biocomposite samples with silane treated

BBSG had the lowest values of surface roughness parameters.

 

 

 

 

9A IB AC oD
M
12~ ‘
A
A10~
‘1 o
E 6~ ‘ A o I
41‘ . ’ I
I I .
2’ . O O
OI I l I I
0 20 40 6O 80 100

Tirm (days)

Figure 7.2.6: Change in roughness parameter R, over time for biocomposites
Legend: A=Untreated hempmatl -UPE, B=Acrylonitrile treated hempmatl-UPE,
C=Silane treated BBSG-CaCO3-UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE «

 

 

 

 

0A IB AC CD
1%
m— A
A
E
t; 60 F .
E A ‘ .
N 40 ‘ I
m
. I
.4 C
20 z : . .
O l I I I I
20 40 60 80 100
Time (days)

Figure 7.2.7: Change in roughness parameter Rz over time for biocomposites

335

Legend: A=Untreated hempmatl-UPE, B=Acrylonitrile treated hempmatl -UPE,
C=Silane treated BBSG-CaCO3-UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE

 

 

 

 

0A IB AC 0D
1%
100 — A ‘
E
g 80 ~ ‘ A
o
E 60 + I
a 40 . 3
a o 0 O
20 I I 9
O T I I I T
O 20 40 60 80 100
Time (days)

Figure 7.2.8: Change in roughness parameter Rmax over time for biocomposites
Legend: A=Untreated hempmatl -UPE, B=Acrylonitrile treated hempmatl —UPE,
C=Silane treated BBSG-CaCOB-UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE

The samples get washed over by water, irradiated by UV, and react with ambient
oxygen while in the accelerated weathering chamber. These conditions lead to
breaking of the linkages between the ﬁbers and the matrix, and the biocomposite
surface stats crumbling. On increasing the exposure time of weathering, this spreads
though the thickness of the composite. The reactions involved in the weather
degradation are oxidation, reduction, dehydration, hydrolysis, swelling, shrinking,
freezing, and cracking.

The cell wall polymers responsible for the moisture sorption of bioﬁbers are:
hemicellulose, accessible cellulose, non-crystalline cellulose, lignin, and crystalline

cellulose. The cell wall polymers accountable for the ultraviolet degradation are:

lignin, hemicellulose, accessible cellulose non-crystalline cellulose, and crystalline

336

cellulose. While, the cell wall polymers responsible for the thermal degradation
properties of bioﬁbers are: hemicellulose, cellulose, and lignin. The strength of the
bioﬁbers is controlled by crystalline cellulose, matrix (non-crystalline cellulose +
hemicellulose + lignin), and lignin. The biological degradation if cell wall of
bioﬁbers is inﬂuenced by hemicellulose, accessible cellulose, and non—crystalline
cellulose.

Dynamic mechanical analysis

Figures 7.2.9 and 7.2.10 show the storage modules at 40 0C and the T8 of the
biocomposites as a function of the exposure time of weathering, respectively,
analyzed by DMA. With the physical and chemical changes occurring in the samples
in the course of artiﬁcial weathering, it was no surprise to see that the values of
storage modules at 40 0C and T3 of the biocomposites decreased as the weathering
time increased. However, this decrease in modulus was reasonably small, as was seen

by other authors earlier [10-12].

 

 

 

 

AA OB IC OD
10
E
g 8 . . . O .
3 A A A
2 4 . 0 O Q Q
8’»
E
s 2‘
m
0 I I I I
0 20 40 60 80 100
Time(Days)

Figure 7 .2.9: Change in storage modulus over time for biocomposites

337

Legend: A=Untreated hempmatl-UPE, B=Acrylonitrile treated hempmatl -UPE,
C=Silane treated BBSG-CaCO3-UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE

 

 

 

 

AA 0B IC OD
97
I
:8. 95 I ‘ .
an I .
E- 0 ‘ A
0 I
93 « . I
O
91 T I I I
O 20 40 . 6O 80 100
Tune (Days)

Figure 7.2.10: Change in Tg over time for biocomposites

Legend: A=Untreated hempmatl-UPE, B=Acrylonitrile treated hempmatl-UPE,
C=Silane treated BBSG-CaCO3-UPE, D= Silane treated BBSG & Green Flax Core-
CaCO3-UPE

The highest storage modulus was of the sample containing acrylonitrile treated hemp
matlbased biocomposite, followed by untreated hempmatl , silane treated BBSG, and
silane treated BBSG & green ﬂax core. The same trend was observed for glass
transition temperatures of biocomposites. The overall decrease in Tg for all samples
was less than 3 0C for the entire exposure time. The overall decrease in modulus of
surface treated hemp ﬁbers based biocomposites was 8.6% over the entire weathering
exposure time, 10% for silane treated BBSG-UPE, 10.3 % for untreated hempmatl-
UPE, and 11.5% for silane treated BBSG and GFC based biocomposite. The change
in storage modulus was related to change in weight of these samples.

The tendency of bioﬁbers to absorb moisture causes off-gassing (void formation)

during compounding. This results in a molded article with a microstructure having

338

variable porosity and resembling that of high-density foam. The pores formed will act
as stress concentration points which then lead to an early failure of the composite
during loading. Another major drawback of using bioﬁbers as reinforcing agent is the
high moisture absorption of the ﬁbers due to hydrogen bonding of water molecules to
the hydroxyl groups within the ﬁber cell wall. This leads to a moisture build-up in the
ﬁber cell wall (ﬁber swelling) and also reduction in the ﬁber-matrix interface bonding.
This is responsible for changes in the dimensions of bioﬁber-based composites,
particularly in the thickness and the linear expansion due to reversible and irreversible
swelling of the composites. As a consequence, the ﬁber-matrix adhesion is weak and
the dimensional stability of bioﬁber-based composites particularly for outdoor
applications will be greatly affected.

Nature builds lignocellulosic resources from carbon dioxide and water, and it has all
the tools to recycle them back to the starting chemicals. Possible ways of degradation
include biological, thermal, aqueous, photochemical, chemical, and mechanical
means of degradation. In order to produce cellulose ﬁber-based composites with a
long service life, the degradation processes caused by nature need to be retarded. One
way of preventing or slowing down the natural degradation process is by modifying
the cell wall chemistry of the material which is responsible for many of its properties.

This can be accomplished by chemical modiﬁcation of the ﬁbers.

339

References

1. J. Gassan, A. K. Bledzki, Journal of Applied Polymer Science, 82, (2001), 1417—
1422.

2. J. Gassan, Composites: Part A, 33, (2002), 369-374.

3. S. Mishra, A. K. Mohanty, L. T. Drzal, M. Misra, S. Parija, S. K. Nayak, S. S.
Tripathy, Composites Science and technology 63, (2003), 1377-1385.

4. J. Rout, M. Misra, S. S. Tripathy, S. K. Nayak, A. K. Mohanty, Journal of
Applied Polymer Science, 84,(2002), 75-82.

5. A. Bismarck, A. K. Mohanty, I. Aranberri-Askargorta, S. Czapla, M. Misra, G.
Hinrichsen, J. Springer, Green Chemistry, 3, (2001), 100-107.

6. J. Rout, M. Misra, S. S. Tripathy, S. K. Nayak, A. K. Mohanty, Composites
Science and Technology, 61, (2001), 1303-1310.

7. T. H. D. Sydenstricker, S. Mochnaz, S. C. Amico, Polymer Testing, 22, (2003),
375-380.

8. H. P. S. A. Khalil, H. Ismail, M. N. Ahmad, A. Arifﬁn, K. Hassan, Polymer
International, 50, (2001), 395-402.

9. J. Gassan, A. K. Bledzki, Polymer Composites, 20 (4), (1999), 604-611.

10. L. M. Matuana, D. P. Kamdem, and J. Zhang, “Photoageing and Stabilization of
Rigid PVC/Wood-Fiber Composites,” Journal of Applied Polymer Science, 80 (11),
(2001), 1943-1950.

11. L. M. Matuana and D. P. Kamdem, 5'” Pacific Rim Bio-based Composites
Proceedings, 10-14 December 2000, 644-651

12. N. Stark, L. M. Matuana, Proceedings of Society of Plastics Engineers Annual
Technical Conference ANT EC 2002, 2209-2213

340

CHAPTER EIGHT: CONCLUSIONS AND FUTURE WORK

8.] CONCLUSIONS

Environmental and economic consciousness around the world had led to a revolution in
interest in the use of renewable and sustainable materials for a variety of applications.
Biocomposites, which comprise of biodegradable or synthetic polymers reinforced with
natural ﬁbers, are one example of this new class of sustainable products.

The objective of this research was to develop novel low-cost biocomposite panels with
desired properties by incorporation of natural ﬁbers in an unsaturated polyester resin
matrix so that these biocomposites could then be used in structural and building
applications. Three interconnected parts of this project were completed in order to
achieve the research objective and included: i) design of engineered natural ﬁbers, ii)
modiﬁcation of the polymer matrix, and iii) development of a new process for
continuous production of biocomposites. In addition, the durability of these biobased

materials was investigated as well.

Conclusions: 1. Engineered Bioﬁbers
The objective of this part of the work was to design the ﬁbers responsible for the
reinforcement of the composite. Engineered bioﬁbers with optimized ﬁber—matrix
adhesion are required for good mechanical, thermal and physical properties. The
optimum ﬁber volume fraction was found to be 30% based on a comparison of the
mechanical properties of composites with three different volume fractions of ﬁbers.
The surface chemical modiﬁcations of natural ﬁbers like alkali treatment, vinyl grafting

and treatment with various coupling agents, were some means we employed to improve

341

ﬁber matrix adhesion of the resulting biocomposites. Surface modiﬁcation also resulted
in enhancement of the aspect ratio, improved the wettability of the ﬁbers, and formed a
strong interface between polar natural ﬁber and non polar matrix. An acrylonitrile
treatment for the hempmatl ﬁbers produced a 80% higher increase in tensile strength
over untreated hempmatl ﬁber based composites, a 430% increase in tensile modulus
over the UPE control, and a 53% increase in impact strength over the untreated
hempmatl-UPE composite.

The gap between performance of glass based composites and biocomposites was
bridged by fabricating a hybrid composite comprising of glass and bioﬁbers. Hybrid
bioﬁbers were also custom made to give maximum stiffness and toughness by blending
different weight fractions of different untreated or surface treated bioﬁbers (bast, leaf,
seed, hit, or grasses). In terms of speciﬁc modulus and strength, glass composites and
biocomposites were in the same range. Hybridization of the composite through
combining E-glass mats with hempmatl led to a 23% decrease in tensile strength, a
25% decrease in bending strength, and a 26% reduction in storage modulus at 40 OC
compared to a composite of 100% E-glass ﬁber mat in the UPE matrix. The speciﬁc
elastic modulus of a biocomposite containing acrylonitrile treated hempmatl ﬁbers was
70% greater than that of glass-UPE composite. Hybrid bioﬁbers can be designed to
meet the desired requirements of a particular application.

From the various experimental results, it can be concluded that the ﬁber-matrix adhesion
for a natural ﬁber reinforced composite system can be improved by the introduction of
chemical bonding at the interface. The effect of increasing the surface area and hence the

interfacial friction was not as signiﬁcant compared to the effect of increasing chemical

342

bonds across the ﬁber-matrix interface. It can be inferred that the interfacial adhesion is
directly related to the amount of reactive ﬁinctional groups, in this particular case the
hydroxyl content on the bioﬁber surface. When the possibility of chemical bonding is
removed from the ﬁber matrix interface, the interfacial adhesion no longer relates to the
interfacial chemistry and becomes a pure mechanical phenomenon.

It is also important to note that the composite properties can also depend on various other
factors such as void content and processing conditions. Good interfacial adhesion
between the natural ﬁber and matrix polymer can only be signiﬁcant when other major
factors are Optimized.

In summary, biocomposites were successfully made using various natural ﬁbers with
different surface treatments. An increase in mechanical and thermal properties is seen for
all surface treated ﬁber based biocomposites most probably due to the formation of
covalent bonds across the ﬁber-matrix interface. The UPE and alkali treatments are low
cost treatments, which increase the performances of the resulting biocomposites. The
speciﬁc modulus of this type of biocomposite is comparable to that of conventional glass
ﬁber-UPE composites. Therefore, these biocomposites have the potential to replace glass-

based composites.

Conclusions: 11. Matrix Modiﬁcation

The second major portion of this project was directed at modiﬁcation of the petroleum-
based polymer matrix by incorporation of substantial concentrations of natural oil based
resins to increase its biocontent and to improve the matrix properties. Polymerizable

bioresins were developed from natural oils by chemical modiﬁcation (grafting).

343

Bioresins were prepared from vegetable based oils by two methods: i) grafting vegetable
oils with acrylonitrile, and ii) grafting maleic anhydride onto the natural polyols.
Utilization of bioresins as polymer matrices provides a two fold beneﬁt over thermoset
resins in biocomposites. Their presence can improve the toughness of brittle unsaturated
polyester resin as well as produce a material with higher biobased content. The bioresins
produced can be ‘tuned’ to produce a range of bioplastics which varied from high in
stiffness to high in toughness by changing the amount of bioresin introduced into the
UPE matrix.

The kinetic parameters of the UPE as well as UPE-bioresin curing reaction were
investigated by kinetic studies. The fractional conversion of styrene double bonds was
higher than those of UPE. The reaction rate of curing increases with an increase in the
temperature. 100% conversion was not observed in any of the curing systems because of
the entrapment of monomers in crosslinked segments of the matrix however, fractional
conversion of styrene double bonds was 99% at a temperature of 160 °C, and for UPE
double bonds was 92% at a temperature of 160 0C.

In all bioplastics, phase separation was observed between the bioresin component and the
UPE component. The impact strength of the bioplastics and biocomposites was increased
by 100% by acrylonitrile grafting to the natural oils. At the same time, the bending
strength reduced by 20% (by adding 20% bioresin). Alternatively, maleic anhydride was
grafted onto soybean polyols. The bioplastic samples containing 40% and 50% grafted
polyols had higher toughness, and lower stiffness, while bioplastic samples containing
10%, 20%, and 30% grafted polyols had a very high modulus (from 2.8 GPa for UPE

control to 10 GPa for 10%GFTSOPEP1 at 40 OC) and very low impact strength.

344

Conclusions: 111. New Processing

The objective of developing a new continuous biocomposites sheet molding compound
panel (BCSMCP) processing was to create a process for making biocomposites that could
be scaled up to commercialization levels and at the same time take advantage of the
unique properties of these biocomposites. The BCSMCP method is a variation of the
traditional SMC processing method and was invented to enable high volume industrial
scale continuous production of biocomposites using any type of natural ﬁber, UPE and
even bioresins. After the SMC process was optimized for natural ﬁber-polyester resin
composites, several composites made with various untreated and surface treated chopped
hybrid ﬁbers, bioﬁbers, calcium carbonate, UPE, and bioresin were successfully

produced.

As part of this research, a new ﬁber feeder was designed to handle bioﬁbers and produce
a uniform distribution of chopped bioﬁbers onto the SMC line. A hybrid composite of E-
glassmat-hemp was made in the BCSMCP which had a tensile strength only 20% lower
than a chopped E-glass composite, a tensile modulus only 8% lower than, and a bending
strength only 4% lower, the tensile modulus only 5 % lower than chopped E-glass
composite, also made in the BCSMCP. The chopped E-glass composite had the tensile
strength 170% higher than untreated BBSG biocomposite, and had a storage modulus
105% higher than untreated BBSG biocomposite at 40 0C (untreated BBSG biocomposite

had lowest properties of all SMC composites).

345

The new BCSMCP process can easily be integrated to any existing SMC infrastructure

supporting the commercialization of biocomposites by SMC.

Conclusions: Durability

The objective of this part of the project was to evaluate performance and longevity of
bioplastic and biocomposite samples after exposure to the weather elements. The
durability of bioplastics as well as biocomposites were evaluated by moisture absorption
tests, accelerated weathering and ﬂammability tests. Moisture absorption tests indicated
that the equilibrium moisture content was 0.06 % for these polyester bioplastics, and was
less than 0.7 % for biocomposites.

The environmental exposed biocomposites changed color from brown to white as the
exposure time increased. The weight of the samples ﬁrst increased (up to 1% in 30 days),
and then decreased (up to 5% in 90 days). Surface roughness of the samples increased
with exposure time, and warpage was also observed. The maximum loss in storage
modulus after moisture equilibration and subsequent drying was about 12% for a sample
containing silane treated BBSG-green ﬂax core hybrid. Surface treated ﬁber based
composites performed better in terms of these characteristics compared to untreated ﬁber
reinforce composites. These tests suggest that biocomposites can be used for interior
applications for prolonged periods of time without excessive degradation due to

environmental conditions.

General Conclusions

346

Biocomposites were made using various types of natural ﬁbers with different surface
treatments by use of different processing techniques. Hybrid composites were made by
combining different varieties of bioﬁbers in different weight fractions, and also by
combining bioﬁbers and E-glass ﬁbers. Biocomposites could be produced within 30-80%
of the mechanical and thermal properties of glass composites in most cases. Parity in
properties could be achieved by hybridization of the bioﬁber and glass ﬁbers based
composites.

The impact strength and modulus could be improved by adding different types of
bioresins into the UPE matrix resulting in a phase separated material. Incorporating
certain bioresins in the matrix increased the biomass content of composites up to 60%.
Continuous production of biocomposites was demonstrated by utilization of the SMC
process. Engineered ﬁbers were designed to compete with glass ﬁber based SMC
composites. The mechanical properties of biocomposites do not change signiﬁcantly on

accelerated weathering.

Through the integration of all parts of this project, we were able to develop new cost
effective biocomposites for future applications in structural and building industries. Our
ultimate goal of replacing existing glass ﬁber-polyester composite panels currently used
in the housing industry by biocomposites has been demonstrated. Biocomposites have

the potential to replace glass ﬁber-UPE composites.

8.2 FUTURE WORK

347

We have been successful in making structural biocomposites for future applications,
however, some more additional research needs to be completed before this technology is
placed in widespread use.

The adhesion between ﬁbers and the matrix must be improved by surface treatment.
Alternative surface treatments such as acetylation, bleaching, UV/plasma, and microwave
should be investigated and optimized for each type of natural ﬁber in the future.

Bioresins having different chemistries can be developed from natural oils. These
bioresins then must be polymerized by blending with synthetic resins or by themselves to
increase the amount of ‘green’ material in a thermoset composite. Further experiments
need to be aimed at studying in detail the behavior of the second phase, the actual
mechanism of toughening and control of the dispersed phase in the bioresin-UPE systems.
The fracture studies must be performed in further detail to ﬁnd the critical parameters like
ﬁber-matrix adhesion and fracture dynamics. Alternative strategies to improve the
toughness of brittle UPE resins by adding rubbery particulates also should be investigated.
The curing times of the UPE or bioresin-UPE systems also must be optimized to get
maximum conversion in shortest amounts of time. There should be some solutions to the
volume shrinkage problems of the UPE and bioresin-UPE systems. The amount of
renewable and biodegradable materials in the biocomposites and bioplastics must be
increased to at least about 60% of the entire weight of the product. One option to
consider would be the incorporation of a small volume ﬁaction of economic
nanoparticles into the bioﬁber-bioresin-UPE system, to make a bio-nano-composite, with
the added advantages of increased stiffness, strength, and heat resistance; and decreased

moisture absorption, ﬂammability, and permeability.

348

An effort must be made to study in detail the various parameters that affect the properties
of composites in hybrid ﬁbers systems. A universal model needs to be developed which
allows for inclusion of natural ﬁber properties to estimate the theoretical values of the
biocomposite properties.

A cost analysis for the biocomposite sheet molding compound panel processing should be
performed, and compared with conventional SMC technique processing glass-UPE
composites.

Finally, collaboration with our industrial partners, F laxCraft Inc. and Kemlite Chemical
Company Inc. needs to be enhanced to promote and commercialize this technology in

housing panel applications.

349

 

EEEEEEEEEEEEEEEEEEEEEEEEEEE

IIIIIIIIIIIIIIIIIIIII

   

”J_-_-_-_-_-_-_-_-_-_-j_-,