FLEXURAL PROPERTIES OF HYBRIDIZED UNIDIRECTIONAL COMPOSITES

REINFORCED WITH BAMBOO AND BASALT FIBERS

By

Mario Calderon Cueva

A THESIS

Submitted

to Michigan State University

in partial fulfillment of the requirements

for the degree of

Materials Science and Engineering – Master of Science

2018

ABSTRACT

FLEXURAL PROPERTIES OF HYBRIDIZED UNIDIRECTIONAL COMPOSITES

REINFORCED WITH BAMBOO AND BASALT FIBERS

By

Mario Calderon Cueva

In the present work, the flexural properties of hybridized composites with bamboo and basalt as

reinforcing agents in an epoxy matrix (DGEBA + mPDA) are studied. Unidirectional composites

with different weight ratio content of fibers were mechanically evaluated using three-point bending

tests. The peak value for flexural modulus of the hybrid composites was observed in composites

with a weight ratio of 30/10 of bamboo to basalt, while composites with the highest basalt content

exhibited the maximum strength. The physical and chemical properties of bamboo and basalt fibers

were characterized with thermogravimetry. Bamboo fibers were processed from raw culms via

ammonia freeze explosion AFEX and with steam explosion. The adhesion of basalt fibers (without

sizing) with the matrix was investigated with single fiber fragmentation tests. Fracture surfaces

of the basalt, bamboo, and hybrid composites were evaluated to assess the synergistic behavior of

the combination of the bamboo and the basalt fibers in the matrix. The failure characteristics of

the samples subjected to flexural load were analyzed by evaluating the morphology of the fracture

surface using scanning electron microscopy. The results of this study show high values of flexural

properties of bamboo/basalt fiber composites, for the production of high-quality, environmentally-

benign, and sustainable materials. These properties could be optimized by the surface treatment

of basalt fibers to make the matrix and these fillers more compatible. Additionally, the parameters

for the AFEX treatment can be further investigated to improve the performance of bamboo as

reinforcement.

ACKNOWLEDGMENTS

My deepest appreciation to my advisor, Prof. Lawrence T. Drzal, for his support from the begin-

ning of my studies at Michigan State University. He is an inspiration for me for his knowledge, his

creative mind, and for encouraging me to think always think critically. My sincere appreciation to

Prof. Alex Zevalkink and Prof. Ramani Narayan for being part of my commitee. I thank them for

their time and effort for rewiewing this work.

I want to thank the Composite Materials and Structures Center staff for their support through-

out this process: Mike Rich, Brian Rook, Per Askeland, and Ed Drown. I am deeply grateful

for their advice and their constant willingness to provide me guidance in my experimental work.

Also, I would like to thank my colleages in my research group and my friends; David and Saeeda

Smiadak, Erik Stitt, and Eureka Pai. They have encouraged and motivated me greatly during this

process.

Finally, I want to thank my sister, my mom, and my dad -my eyes in the sky-. Their had work,

honesty, and tenacity are values that I admire greatly and have been of great help in my endeavour

as a researcher. My deep appreciation to my girlfriend, María Emilia, for always believing in me.

iii

TABLE OF CONTENTS

LIST OF TABLES .

LIST OF FIGURES .

.

.

.

.

.

.

.

.

.

.

KEY TO ABBREVIATIONS .

.

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vi

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

CHAPTER 1

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

. .
. .
.
.

1.1 Objectives . .
1.2 Materials .
. .

INTRODUCTION .
.
.
.
.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2.1 Matrix .
Fibers .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2.2
1.2.2.1 Bamboo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.2.2.2 Basalt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3 Overview of experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.1 Mechanical tests
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.2
. . . . . . . . . . . . . . . . . . . . . . . . . .
1.3.3 Morphological analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Physico-chemical analysis

.

1
2
3
3
3
3
4
5
5
5
5

CHAPTER 2 MATRIX CHARACTERIZATION . . . . . . . . . . . . . . . . . . . . . .
.
.
2.1
Introduction .
.
.
2.2 Epoxy resins .
.
2.3 Curing agents .
.
.
2.4 Curing process of epoxies .
2.5 Curing of epoxy resins with m-PDA . . . . . . . . . . . . . . . . . . . . . . . . .

7
7
7
8
9
9
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.5.1 Curing cycle . .

.

. .
. .
.

.
.
.

.
.
.

.
.
.

.
.
.

.

.

Introduction .

3.1 Bamboo fiber characteristics

CHAPTER 3 FIBER CHARACTERIZATION . . . . . . . . . . . . . . . . . . . . . . . . 11
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1.1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1.2 Bamboo constituents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1.3 Bamboo structure .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.1.4 Bamboo fiber harvesting techniques . . . . . . . . . . . . . . . . . . . . . 16
3.1.4.1 Mechanical extraction method . . . . . . . . . . . . . . . . . . . 17
3.1.4.2 AFEX extraction method . . . . . . . . . . . . . . . . . . . . . . 18
3.1.5 Experiments conducted on AFEX-treated fibers . . . . . . . . . . . . . . . 19
SEM images . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
TGA tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Flexural properties of single fiber bundles . . . . . . . . . . . . . 22
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.2 Basalt structure and constituents . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.3 Basalt fiber production . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.2 Basalt fiber characteristics
.

3.1.5.1
3.1.5.2
3.1.5.3

Introduction .

.

iv

3.2.4 Experiments conducted on basalt fibers

TGA tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Single fiber fragmentation test . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . 28
. 28
. 31

3.2.4.1
3.2.4.2

.
.

.
.

4.4.1
4.4.2
4.4.3

. . . . . . . . . . . . . . . . . . . .

CHAPTER 4 COMPOSITE CHARACTERIZATION . . . . . . . . . . . . . . . . . . . . 43
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
4.1 Neat epoxy characteristics
4.2 Procedure for the fabrication of composites
. 43
4.3 Evaluation of composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.3.1 Three-point bending tests . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Fiber distribution and void content assessment . . . . . . . . . . . . . . . . 48
4.3.2
4.4 Bamboo composites .
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Introduction .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Physical and mechanical properties . . . . . . . . . . . . . . . . . . . . . . 50
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Fracture analysis
.
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.5.1
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
.
4.5.2 Results on flexural tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.5.3
.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.6 Hybridized composites
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
Introduction .
.
Fracture analysis
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.6.2.1 Composites of Aw%=20 and Bw%=20 . . . . . . . . . . . . . . . 64
4.6.2.2 Composites of Aw%=30 and Bw%=10 . . . . . . . . . . . . . . . 67
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

.
Introduction .

4.5 Basalt composites .

4.6.3 Rule of mixtures

Fracture analysis

4.6.1
4.6.2

.
.

.
.

.

5.1 Mechanical behavior of composites . . . . . . . . . . . . . . . . . . . . . . . . .

CHAPTER 5 CONCLUSIONS AND FUTURE WORK . . . . . . . . . . . . . . . . . . . 73
. 73
5.1.1 Bamboo fiber performance . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.1.2 Basalt fiber performance . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Sizing of basalt fibers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Fiber distribution for improvement of flexural properties
. . . . . . . . . . 76
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
Impact tests .

5.2 Next steps .

5.2.1
5.2.2
5.2.3

. .

.

.

.

.

.

.

.

.

APPENDICES .

.

BIBLIOGRAPHY .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

v

LIST OF TABLES

Table 3.1 Drying of bamboo samples after pressure-chamber treatment . . . . . . . . . . . 18

Table 3.2 Density measurements for bamboo samples . . . . . . . . . . . . . . . . . . . . 22

Table 3.3 Composition of basalt fibers . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 25

Table 3.4 Physical and mechanical properties of basalt [a]

. . . . . . . . . . . . . . . . . . 26

Table 3.5 Single fiber fragmentation test of basalt fibers in epoxy (number of breaks

measured in a span of 18 mm)

. . . . . . . . . . . . . . . . . . . . . . . . . . . 37

Table 3.6 Single fiber fragmentation test of basalt fibers in epoxy (number of breaks

measured in a span of 22 mm)

. . . . . . . . . . . . . . . . . . . . . . . . . . . 38

Table 4.1 Conversion of weight to volume fraction of fibers in hybrid composites [a] . . . . 63

Table 4.2 Flexural strength and modulus calculated for basalt and bamboo fibers using

ROM [a] .

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

Table 5.1 Example of bamboo-reinforced composite (AFEX-treated) . . . . . . . . . . . . 84

vi

LIST OF FIGURES

Figure 2.1 DGEBA molecule[19]. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Figure 2.2 m-PDA molecule[59].

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8

9

Figure 3.1 Hierarchical structure of bamboo[44]

. . . . . . . . . . . . . . . . . . . . . . . 12

Figure 3.2 Fiber bundle micrograph (SEM) of mechanically treated bamboo. Note pres-

ence of parenchyma cells and residues of lignin/hemicellulose.

. . . . . . . . . 14

Figure 3.3 Vascular bundle of a bamboo culm.

Image taken with an optical micro-

scope.This image corresponds to the schematic figure 3.1(b) . . . . . . . . . . . 15

Figure 3.4 Micrograph of cross-section of bamboo fiber bundle embedded in a polymer,
composed of closely packed fibers. Arrow shows lumen. Circle displays a
single fiber of approx. 20µm[44]

. . . . . . . . . . . . . . . . . . . . . . . . . 16

Figure 3.5 SEM image of region of a bamboo fiber-bundle showing parenchyma cells.

Fiber treated with AFEX method. . . . . . . . . . . . . . . . . . . . . . . . . . 19

Figure 3.6 Thermogravimetric analysis on AFEX-treated fibers . . . . . . . . . . . . . . . 20

Figure 3.7 Thermogravimetric analysis on AFEX-treated fibers after drying for 2 hours

at 80ºC .

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Figure 3.8 Strength of bamboo fibers, extracted with pressurized water and rolling-

milled. Note significant change between the different samples . . . . . . . . . . 23

Figure 3.9 Elastic modulus of bamboo fibers, extracted with pressurized water and rolling-

milled. High spread of data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Figure 3.10 Junkers method for continuous fibers. Image taken from[33].

. . . . . . . . . . 27

Figure 3.11 Centrifugal method for basalt production. Figure taken from[33]

. . . . . . . . 28

Figure 3.12 TGA of basalt fibers showing a single-step mode of weight loss.

. . . . . . . . 30

Figure 3.13 Differential TGA for the sample of figure 3.12. The main weight loss occurs

at 222ºC. .

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. 31

Figure 3.14 Coupon dimensions for SFFT[39]. Black line represents the fiber.

. . . . . . . . 32

Figure 3.15 SFFT coupons mounted on metal surface for polishing.

. . . . . . . . . . . . . 33

vii

Figure 3.16 Micrograph of basalt fiber under polarized light in epoxy showing several

breaks. Image taken at saturation strain. Arrows indicate fiber breaks. . . . . . . 35

Figure 3.17 Region of basalt fiber break in epoxy at saturation strain under polarized

light. The fiber has experienced debonding at break.

. . . . . . . . . . . . . .

. 36

Figure 3.18 Breaks with respect to strain

. . . . . . . . . . . . . . . . . . . . . . . . . . . 39

Figure 3.19 Weibull fit for SFFT test of four samples . . . . . . . . . . . . . . . . . . . . . 41

Figure 3.20 Weibull fit for SFFT test of the other four samples . . . . . . . . . . . . . . . . 42

Figure 4.1 Mold scheme used to prepare flexural coupons.

. . . . . . . . . . . . . . . . . 46

Figure 4.2 Bamboo (AFEX) cut to a length of 12.7 cm.

. . . . . . . . . . . . . . . . . . . 47

Figure 4.3 Basalt fibers cut to a length of 12.7 cm. . . . . . . . . . . . . . . . . . . . . .

. 47

Figure 4.4 Basalt and bamboo fibers intertwined and cut to a length of 12.7 cm.

. . . . . . 48

Figure 4.5 Transmitted light micrograph of cross-section of bamboo composite (40 weight

% of fibers).1: resin, 2:hollow-vessel region, 3: fiber-sheath of sclerenchyma
cells, 4: parenchyma cells. Note good dispersion of fibers. . . . . . . . . . . . . 50

Figure 4.6 Surface of fiber bundles in composite (40 weight %) showing resin residues

on the fibers. .

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

Figure 4.7 Elastic modulus of bamboo-reinforced composites with different weight frac-

tion of fibers. .

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Figure 4.8 Flexural strength of bamboo-reinforced composites at several values of fiber

content.

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

Figure 4.9 Bamboo fiber bundle in tension side of composite (40 weight %) Note epoxy

on the surface of bamboo. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

Figure 4.10 Fracture surface of bamboo-reinforced composite. Resin deformation ob-

servable on top left side. Fibers broken, not pulled out of the matrix.

. . . . .

. 53

Figure 4.11 Fracture surface in center of sample. Different heights of fibers’ breaks indi-

cate energy absorption.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Figure 4.12 Fracture surface of bamboo-reinforced composite on tension side (40 weight
%). Hackle formation and river patters on resin surface. Debris indicate
tensile failure of fiber.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

Figure 4.13 Cross section of basalt composite (50 weight %). Note good dispersion of fibers. 56

viii

Figure 4.14 Elastic modulus of basalt-reinforced composites with different weight frac-

tion of fibers. Basalt fibers untreated.

. . . . . . . . . . . . . . . . . . . . . . . 57

Figure 4.15 Flexural strength of basalt-reinforced composites with different fiber volume

content.

.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Figure 4.16 Fracture surface of basalt composite (40 weight %) showing almost no residues

of resin on the surface of the fibers. . . . . . . . . . . . . . . . . . . . . . . . . 59

Figure 4.17 SEM micrograph of fracture surface of basalt composite (40 weight %) at the

edge on the specimen subjected to compression.

. . . . . . . . . . . . . . . . . 60

Figure 4.18 SEM micrograph of fracture surface of basalt composite (40 weight %) on

the side of the tension edge. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Figure 4.19 SEM micrograph of fracture surface of basalt composite (40 weight %) at the

middle of the sample.Note mixture of failure modes. . . . . . . . . . . . . . .

. 62

Figure 4.20 Flexural strength at different ratios of fiber weight percentage . . . . . . . . . . 63

Figure 4.21 Flexural modulus at different ratios of fiber weight percentage . . . . . . . . . . 64

Figure 4.22 Fracture surface showing the edge of the sample subjected to tension. Note

basalt fiber accumulation.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Figure 4.23 Hybrid composite 20%/20% AFEX/Basalt on tension edge. Bamboo fibers

have failed by break. River patterns on resin surface. . . . . . . . . . . . . . .

. 66

Figure 4.24 Hybrid composite 30%/10% AFEX/Basalt in the center of sample. Crack

propagation around the bamboo fiber bundle. Basalt fiber pulled-out.

. . . . . . 66

Figure 4.25 Hybrid composite 30%/10% AFEX/Basalt on tension edge. Note matrix de-

formation and voids formed

. . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Figure 4.26 Hybrid composite 30%/10% AFEX/Basalt on tension side. Note resin-rich

region and basalt fibers pulled out.

. . . . . . . . . . . . . . . . . . . . . . . . 68

Figure 4.27 Hybrid composite 30%/10% AFEX/Basalt in center of the sample. Failure

by delamination on the right side. No deformation of resin region.

. . . . . .

. 69

Figure 4.28 Flexural modulus comparison between predicted and calculated values in hy-

brid composites .

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Figure 4.29 Strength comparison between expected and observed values.

. . . . . . . . . . 72

Figure 5.1 Example for the estimation of fiber volume by using ImageJ . . . . . . . . . . . 83

ix

Figure 5.2 Foam structure of neat Acrodur 950L (BASF).

. . . . . . . . . . . . . . . . . . 86

Figure 5.3 Compression-moulded Acrodur-950L (BASF) resin with AFEX-treated bam-

boo fibers.

.

.

.

.

.

.

.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

x

KEY TO ABBREVIATIONS

EPONTM 828 Bisphenol A diglycidyl ether (DGEBA)

mPDA

TGA

SEM

AFEX

m-Phenylenediamine

Thermogravimetric analysis

Scanning electron microscopy

Ammonia freeze explosion process

DI-water

Deionized water

SFFT

IFSS

N

¯x

SD

SE

ROM

LCC

Single-fiber fragmentation test

Interfacial shear strength

Number of samples

Sample mean

(cid:114)

Standard deviation of the sample (s), given by s =
Standard error of the sample (s ¯x), given by s ¯x = s√
N
Rule of mixtures

Lignin-carbohydrate complex

i=1(xi− ¯x)2
∑N

N−1

xi

CHAPTER 1

INTRODUCTION

Composite materials are characterized by the combination of two or more types of materials to

create a material, that, if well designed, will exhibit the best features of each type. The convention
is to call composites the macroscopic combination of different materials[34]. Although there are

many different types of composite materials, there is a great interest in fiber-reinforcement of a

polymeric binder (a matrix). Fibers possess high length to diameter ratio, and excellent stiffness

and strength to density ratio. However, fibers (also called fillers in this context) are not useful if

they are not structurally stable. For this reason, it is necessary to embed the fibers in a matrix.

This type of materials are called fibrous composites. Fibrous composites consisting of carbon or

boron fibers in low-toughness epoxies were first studied in the 1960s in the United States for their
application in aerospace and defense areas[12]. Since then, many studies have been conducted to

improve the mechanical properties of the matrix, strengthen the fibers, improve the adhesion be-

tween the fillers and design and create better ways to produce synthetic fibers. The most commonly
used fibers are glass, comprising 95% of the reinforcement of thermosets and thermoplastics[48].

The issue with this type of synthetic fiber lies in the generation of waste for its production. For this

reason, the present study presents basalt fibers as an alternative to glass fibers. Basalt is a vulcanic
compound of several oxides that represents more than 30% of the earth’s crust[15]. The production

of these fibers does not require the addition of secondary products and the mechanical properties

are comparable to those of glass-fibers.

Additionaly, in an effort to create more eco-friendly products, as well as a way to enhance the com-

posite properties, cellulosic fibers are now being widely studied. Essentially, all natural fibers share

cellulose (the structural component) and lignin (the bonding agent between cellulose) in different

ratios and concentrations. Although natural fibers were already used in plastics (phenolic systems)
in the beginning of the 20th century, there is a an increasing interest in the implementation of ligno-

cellulosic fibers in recent decades, especially due to the imperative need to increase ductility of the

1

composite by the addition of fibers that possess higher elongation at break. Also, the use of natural

fibers is a way to tackle the environmental concern regarding the disposal of composites when they

have completed their life cycle, the increasing demand of resources by the population growth and

the potential economic growth of agricultural communities by the production of bio-based fibers.

Although some natural fibers have been developed to have comparable properties of glass fibers,

there are inherent limitations to the natural fibers, which will be discussed later in detail.

Since the ultimate goal is to produce an environmentally-friendly composite of excellent mechan-

ical properties, the need to produce a bio-degradable matrix of comparable characteristics of the

traditional polymers is still pending. A bio-degradable polymer is a system that is fully degraded
by aerobic (CO2), anaerobic (methane) processes, or by micro-organisms[21]. It is worth men-
tioning that not all bio-based polymers are biodegradable, and not all biodegradable polymers are

bio-based.

Although the present study investigates the flexural properties of bamboo fibers combined with

basalt fibers as fillers in a non-biodegrabale epoxy matrix, it should serve as a basis to demonstrate

the positive effect of combining these two types of fibers towards a sustainable final product that

could be potentially implemented on a large scale.

1.1 Objectives

The goal of the present work is to fabricate hybrid composites that are environmentally benign

while meeting the quality standards for structural and semistructural applications. This study aims

to identify the potential of combining natural fibers with synthetic fibers, referred to as hybridiza-

tion of composites, to achieve a synergistic effect in flexural properties in an epoxy matrix. For

this purpose, a series of tests were conducted to characterize the bamboo and the basalt fibers as

individual entities, as well as their behavior in the composites.

2

1.2 Materials

1.2.1 Matrix

The matrix selected for this work was an epoxy resin of DGEBA with 14.5 parts per hundred of

resin (phr) of mPDA. The selection of this thermosetting system was due to its high mechanical

properties, as well as the well-studied adhesion behavior with many fillers. All composites for

flexural tests were prepared with the same composition of the matrix under the same conditions,

and with the same commercial product (EPON 828 - from Miller Stephenson Company and mPDA

- from Sigma Aldrich). The matrix was cured at 75ºC for 2 hours in a convection oven, followed

by a post curing of 2 hours at 125ºC. The flexural modulus and strength reported for this polymeric
matrix are 2.9GPa and 95MPa, respectively[41]. This epoxy system has an elongation at break of
around 5% and a density of 1.2 g
cm3

[25].

1.2.2 Fibers

1.2.2.1 Bamboo

The bamboo fibers were received as untreated sections of the culm of 30cm long, with an outer

diameter of around 10cm and a inner diameter of approximately 8cm. The compoany Resource

Fiber LLC, based in Alabama, USA, provided the fibers. The species of the provided bamboo

culms are Phyllostachys edulis, also known as Moso. It was not possible to determine the age of

the bamboo, since the dimensions of the culm, as well as other features depend on the growing

conditions, such as the humidity of the region, the amount of rain the culm received, the sun expo-
sure, and the soil.[53]. There is an inherent variability of the mechanical and physical properties for

the untreated bamboo fibers. These characteristics depend on many factors, such as the age of the

specimen, the height, the position of the fiber with respect to radius of the bamboo, among others.

Consequently, the physical and mechanical characteristics of the fibers reported in the literature
show a range of values[17] [16]. In the present work, the term bamboo fiber is used interchangeably

3

with fiber-bundle. For bamboo, a fiber-bundle is composed by hundreds of individual fibers, that

cannot be separated by the processes used in this study. All bamboo specimens tested in this work

are fiber-bundles. When it is important to make a differentiation between fibers and fiber-bundles,

the fiber would be refered as individual fiber. This is the case of section 3.1.

• Density:0.8− 1.1 g
cm3

• Elongation at break 5− 10.2%

• Flexural strength 230− 388MPa

• Flexural modulus 11− 15GPa

1.2.2.2 Basalt

Basalt fibers are a product of processing . Basalt is dark, fine-grained, compact assembly of dif-

ferent compounds, such as magnesium-silicate and iron-silicate, and a low content of silica basalt

rock. Although this vulcanic rock is dark, the basalt fibers were brown gold in color. After the

curing process of the composites, the basalt fibers adopted a black color again. For the present

study, basalt fibers were provided by MAFIC Basalt (North Carolina, USA) as assembled-roving,

continuous fiber toes with a nominal fiber diameter of 13µm were not sized in any way, and the key
sizing element is reported as DI-water[42]. The following properties are reported by MAFIC[42]:

• Density:2.63 g
cm3

• Elongation at break 3.5%

• Tensile strength 3100MPa

• Tensile modulus 88− 92GPa

• Moisture content < 0.1%

4

1.3 Overview of experiments

The following experiments were conducted in the present study in order to characterize the physico-

chemical of the fibers, the neat epoxy, and the fabricated composites.

1.3.1 Mechanical tests

• Three-point bending test in compliance with ASTM 790 for all composites for the calculation

of flexural modulus and strength.

• Single fiber fragmentation test on basalt fibers in epoxy matrix for the evaluation of the

adhesion between the fibers and the epoxy.

• Static-load three point bending test in DMA apparatus for single fiber test on AFEX-treated

bamboo.

1.3.2 Physico-chemical analysis

• TGA on basalt and AFEX-treated fibers to determine the degradation temperatures of each

fibers as well as the water uptake. This test was also conducted to make an initial assessment

of the properties of hemicellulose, cellulose, and lignin for the AFEX-treated bamboo fibers.

• Density measurements of basalt and AFEX-treated fibers with pycnometer. The density of

bamboo specimens were measured before and after drying to determine the water absorption

of the fibers.

• Density measurements of basalt composites to confirm weight fraction.

1.3.3 Morphological analysis

• SEM on fracture surface for all flexural samples after testing to determine the failure mech-

anisms experienced by the composites.

5

• Transmission optical microscopy for basalt and bamboo composites to evaluate distribution

and void content.

• SEM on single bamboo fiber-bundles (mechanically and AFEX treated) to identify surface

roughness, and cell types present. Moreover, this test provided confirmation that the fiber-

bundles were not damaged by the extracting and handling.

• ImageJ digital image processing to evaluate density of bamboo composites and make an

estimation of the volume fraction.

6

CHAPTER 2

MATRIX CHARACTERIZATION

2.1 Introduction

The main purposes of a matrix (sometimes called resin binder) in a composite are to make the
composite rigid and to transfer the stresses to the fibers[45]. Hence, the adequate selection of the

polymer is crucial for the performance of the composite. The chosen polymer for the matrix of the

present study is a thermosetting epoxy resin consisting of Bisphenol A diglycidyl ether (DGEBA)

with m-Phenylenediamine (m-PDA) in a weight ratio of

DGEBA
mPDA

=

100g
14.5g

(2.1)

Thermoset systems are commonly used because they present several advantages:

• Low melt viscosity,

• Low temperatures of processing, and

• Lower cost (compared to thermoplastic resins)

Furthermore, epoxy resins are widely implemented in the composite fabrication due to their me-
chanical properties (even at humid and hot ambients), chemical resistance, and processability[31].

More importantly, these materials show good adhesion to several fibers.

2.2 Epoxy resins

The prefix Epoxy refers to compounds that contain oxygen (bonded to two other atoms in some

way) as bridge. Epoxies are polymers formed by the epoxide functional group. An epoxide is a

cyclic ether with a three-membered ring. The fundamental structure contains an oxygen atom

bonded with two carbon atoms of a hydrocarbon . Due to the strain of the three-membered

7

ring, this epoxide is very reactive compared to an acyclic ether[67]. The most common epoxy

resin is DGEBA (see fig. 2.1), which is a polymer formed by the reaction of epichlorohydrin (1-

Chloro-2,3-epoxypropane) and bisphenol A (2,2-Bis(4-hydroxyphenyl)propane). The formation of

DGEBA is a two-step reaction, starting with a reaction of epichlorohydrin with an active-hydrogen
compound, followed by the dehydrohalogenation with a base (usually Sodium of Potassium)[45].

Figure 2.1 DGEBA molecule[19].

2.3 Curing agents

Curing agents can be classified into three major categories for epoxy resins (based on the functional

groups): amine, acid anhydride, and acids. In the amine category, the aliphatic amine cures the

resin at room temperature, while aromatic amines react very slowly at room temperature and,
therefore, need higher temperatures to fully cure the epoxy[5].

The selection of the curing agent for the polymeric matrix not only affects the physical properties

of the cured epoxy, but other properties such as the curing rate and cure exotherm. The curing agent

selected for the present study was m-PDA (m-Phenylenediamine). Aromatic amines are chemical

compounds that posses one or more planar rings of atoms bonded covalently. These bonds are of
two different kinds. Amines derive from ammonia[9]. Although m-PDA has higher mechanical
properties when combined with DGEBA, it also possesses health concerns.[49]. Curing with amine

is accomplished by the reaction of the epoxite functional groups with the primary and secondary

amines.

8

Figure 2.2 m-PDA molecule[59].

2.4 Curing process of epoxies

During the curing process, the resin experiences changes in its physical and chemical properties:

The linear polymerization occurs, which is the formation of long chains (polymers) of covalently-

bonded structures (monomers). Also, cross-linking takes place, which is the formation of branched

structures due to bonding of the polymers formed at certain intervals. This process causes the

formation of a network structure. When the polymer is cooled down after the curing process,

the network structure remains because of the van-der-Waals forces that the polymers experience.

Additionally, the physical properties of the resin are changed with curing, namely the viscosity and
the temperature of the system[5].

2.5 Curing of epoxy resins with m-PDA

In order to achieve the desired properties, the stoichiometry of the epoxy resin with the curing agent

is crucial. Therefore, the phr (parts per hundred of resin) of the hardener has to be calculated. Phr

is defined as the ratio of the weight of resin (in grams) that contains one equivalent of epoxy

(Epoxy equivalent weight - EEW) and the amine hydrogen equivalent weight (AHEW). For the
EPON 828™it has been reported that the EEW of the DGEBA molecule lies between 185-195[18],

while the AHEW can be calculated by the molecular weight of m-PDA(cid:0)108 g

(cid:1) divided by the

mol

number of active hydrogen atoms in the amine functional group. With these values AHEW = 27

and, therefore,

phr =

EEW
AHEW

= 14.5[g]

9

(2.2)

It has been reported that a system of phr = 14.5 for m-PDA with DGEBA yields the smallest
heterogeneities (≈ 20nm in diameter) and also the lowest occurrence. This same study shows that
a heterogeneity in this polymer is a measure of a higher cross-linking density in region of lower
cross-linking degree[24].

2.5.1 Curing cycle

For the present study, all polymers were cured in the same way. A beaker containing the DGEBA

(EPON-828™) and another containing the m-PDA flakes (from Sigma-Aldrich) were placed at

an oven at 75ºC for one hour. Subsequently, the plastic beaker with DGEBA was pured into the

m-PDA container and vigorously mixed. The mixture was then placed in a vacuum oven with an

internal temperature of 75ºC and degassed. The degassing process removes the air bubbles trapped

in the liquid resin and also promotes the mixing of m-PDA with DGEBA. After three minutes to

five minutes, the beaker is removed from the vacuum oven and put in a convection oven (already

at a temperature of 75ºC). At this point, the curing takes place. The resin is held in the oven at

75ºC for two hours, upon completion, the temperature is increased to 125ºC. This step is called

post-curing. It has been shown that the density of cross-linking dramatically increases in epoxy

systems during the post-curing process and the glass transition temperature also reaches higher
values[26]. When the post-curing process has ended, the oven is turned off with the resin inside

until the temperature reaches 25ºC.

10

CHAPTER 3

FIBER CHARACTERIZATION

The quality of the fibers is a crucial parameter for the performance of the composite. Parameters

such as the cristallinity, surface morphology, defect concentration, temperature degradation, and

pre-treatment derive in the overall performance of the composite. The present chapter aims to

characterize the physical and mechanical properties of bamboo and basalt fibers to predict and

improve their behavior in the composite. Some bamboo fibers were treated via AFEX and others

were extracted in a combination of steam and rolling-milling.

3.1 Bamboo fiber characteristics

3.1.1

Introduction

Bamboo belongs to the grass family of Poaceae and the subfamily of Bambusoidaeae. It represents

a major economic commodity in several regions of the planet, such as India and China. Further-

more, these countries present a very large area of plantations. In the American continent, bamboo

is found from Chile to Mexico (neotropical woody bamboo), while the United States shows pres-

ence of north-temperate woody bamboo. This type of bamboo is also found in continental Africa,
Madagascar and Southeast Asia[69].

3.1.2 Bamboo constituents

Bamboo has three important constituents that are responsible for its features. Around 90% of

the mass of bamboo is holocellulose (hemicellulose and cellulose) and lignin. Cellulose varies
between 40% and 60%[2] depending on the species, age, and segment of the bamboo. α-cellulose

is the type of cellulose present in bamboo and it is has the highest degree of polymerization. This

type of cellulose has linear chains of β -1-4-linked glucose. Hemicelluose in bamboo is a xylan
with other components (uronic acid, arabinose, xylose) in a disordered way[70]. Holocellulose has

11

repeating subunits which promote hydrolysis. The third major component of bamboo is lignin,

which is a disordered polymer composed of phenylpropanoid sub-units. These sub-units are not

repeating, making hydrolisis difficult to be conducted.

3.1.3 Bamboo structure

Bamboo can be considered as a composite material, where the cellulose-microfibrils are embed-

ded in a hemicellulose-lignin matrix, called the lignin-carbohydrate complex (LCC). Figure 3.1(a)

shows a cross-section of a bamboo culm. The subsequent images ((b)-(e)) present in greater mag-

nification the hierarchical structure of the ligno-cellulosic structure.

Figure 3.1 Hierarchical structure of bamboo[44]

As mentioned in chapter 1, the present study used fiber bundles retrieved from the trunk of

bamboo of the Moso species. The trunk of bamboo, called the culm, is a hollow cylinder divided

12

by sections that go in transversal direction with respect to the culm. These transverse sections are

called the nodes, and they separate the internodes. The convention is to ennumerate the intern-

odes starting from the ground. For the present work, the fiber bundles were extracted from the

internodes at different heights of the bamboo. These fiber bundles are primarily oriented in the

axial direction of the bamboo. Fiber bundles, knwon also as sclerenchyma cells, are composed of

hundreds of individual cellulose fibers in a LCC complex. These fiber bundles have a diameter of a

few hundreds micrometers. They surround the hollow vessels, namely the protoxylem, phloem, and

metaxylem. These vessels transport water, nutrients, and products of photosynthesis between the

roots and the leaves. Fiber bundles, together with these vessels form a structure called a vascular

bundle - see figure 3.3-. The presence of vascular bundles increases from the inner to the outer side
of the culm, suggesting that the structural properties of bamboo are given by fiber bundles[17].

The amount of fiber bundles varies also as a function of the height of bamboo, such that, for Moso,

the fiber bundle density increases towards the internodes away from the ground. layer of the culm
has a greater density of individual fibers compared to the inner side[43]. Moreover, the distribution

of foam-like cells, known as parenchyma cells modify the density and the strength of the fibers.

These type of cells can be observed in figure 3.2, on the surface of the fiber bundle. Parenchyma

cells are present in bamboo along the fiber bundles, as displayed on the bottom right region of

figure 3.1, and their main purpose is the transport of nutrients. It is desireable to remove them to

increase the modulus of the composite. However, this can not always be done.

13

Figure 3.2 Fiber bundle micrograph (SEM) of mechanically treated bamboo. Note presence of
parenchyma cells and residues of lignin/hemicellulose.

Figure 3.3 displays the vascular bundle, consisting of fiber bundles for the structural stability

of bamboo, parenchyma cells for the transport of water and nutrients (in the radial direction), as

well as the phloem and the xylem for the same purpose in the vertical directions.

14

Figure 3.3 Vascular bundle of a bamboo culm. Image taken with an optical microscope.This image
corresponds to the schematic figure 3.1(b)

One important aspect of the presence of the hollow regions surrounded by fibers is that it

reduces the weight of the composite considerably. If figure 3.3 is further magnified, the so-called

lamellae are observed. A lamella is a pectin compound that holds cells together. This type of

structure is characterized by a concentric arrangement of cellulosic microfibrils oriented at different

angles with respect to the axial direction of the fiber bundle. This microstructure is displayed in

figure 3.1(c). It is proposed that thicker lamellae have microfibrils oriented close to 0º while thin
lamellae exhibit a 90º angle with respect to the axial direction of the fiber[? ]. In the same figure,

the S0 denotes the transition layer from the middle lamella to the lumen. The middle lamella is
the lignin rich region that bonds individual fibers together in bamboo. Lumen is an aqueous phase

that plays an important role in photosynthesis. The arrow in 3.4 shows the lumen of a particular

individual fiber, while the circled area shows an individual fiber.

15

Parenchyma cells 1 mm Protoxylem Metaxylem Fiber-bundle Fiber sheath Phloem with sieve Figure 3.4 Micrograph of cross-section of bamboo fiber bundle embedded in a polymer, composed
of closely packed fibers. Arrow shows lumen. Circle displays a single fiber of approx. 20µm[44]

Moreover, the differences in the sizes of the perimeter of the individual fibers in a fiber bundle

are believed to be responsible for the excellent strength of bamboo[70].

Going to higher magnifications -see figure 3.1(d)- each individual fiber is composed by a single-

crystal cellulosic structure, known as the microfibril. Each microfibril is embedded in a LCC.

Every microfibril is, in turn, built of cellulose nanofibers. They are believed to be between 3 and 5
nm in diameter[50].

3.1.4 Bamboo fiber harvesting techniques

The subject of how to harvest fibers is of great importance. Its primary objective is to achieve good

mechanical properties and to promote the adhesion of the fibers to the matrix.

There have been several studies regarding different techniques to retrieve fiber bundles. The most

important processes for this goal are a combination of chemical and mechanical steps to separate

the fiber bundles from other constituents. The most common process is the sodium hydroxide

treatment (NaOH), in which bamboo strips are placed in a NaOH solution of 1.5N, where N is
the normality (the number of OH- ions per 1 liter of solution). The bamboo samples are kept at

70ºC for several hours and then rolling-milled for several times with a progressive increase in the
pressure applied[66].

Another widely-implemented method, especially for the pulp extraction industry, is the so-called

16

steam explosion. This physicochemical technique relies on the repeated pressure release of a sealed

container with water. The main advantages are the low-energy consumption involved, as well as

the small amount of waste produced. Additionally, the effectiveness of the production of good-
quality fibers is based on the removal of lignin[58].

3.1.4.1 Mechanical extraction method

In the present study, two types of fiber extraction methods were implemented. In the first one,

raw bamboo culms of an external diameter of 10cm -without removing the bark- were placed in

a pressure chamber. 0.75L of DI-water were placed in the vessel.

It is important to note that

the bamboo samples were not fully immersed in this volume of water. The pressure vessel was

then sealed. The vessel was heated with an electric-resistance oven. Then, the temperature was

increased to 150ºC at a rate of 42ºC/h.The temperature of the vessel was electronically controlled

with a K-type thermocouple. Since the vessel wall has a thickness of 2cm, the rate was very low.

The temperature was kept at 150ºC for two hours. Finally, the heater was switched off and the

system was allowed to cool down naturally until the temperature reached 90ºC. At that point, the

release-valve was opened and the vessel was allowed to cool down to 30ºC. Then, the bamboo

samples were taken out and manually split in segments of approximately 1cm in width. The next

step was to pass the strips of bamboo through a manual rolling-mill (W Durston LTD, UK) three

times. Then, the gap between the wheels was reduced by 0.5 mm. When the sample presented

considerable resistance (by considering the force applied to the shaft) the gap was only decreased

by 0.25 mm. When the strips showed signs of damage on the surface, the roll-milling was stopped.

At this point the fiber-bundles were easily separated by hand. The bark of the samples was removed

because it does not contribute to the desired mechanical properties of the fibers in the composite.

The bamboo strips were placed in a convection oven at 80ºC for three hours to remove the water

absorbed. In order to determine the time required to dry the bamboo the weight loss was monitored

every hour. This test yielded the following results:

17

Table 3.1 Drying of bamboo samples after pressure-chamber treatment

Weight loss (%)

13.7
3.1
<1

3.1.4.2 AFEX extraction method

Time (h)

1
2
3

AFEX (Ammonia freeze explosion) treatment is a process in which the raw bamboo culms (with-

out bark) are soaked in liquid ammonia at 50ºC for approximately 15 minutes at a pressure of 15
atm (≈ 1.5 MPa). The pressure is reduced abruptly, causing a disruption of the fibrous structure
of bamboo due to the violent flash of ammonia[30]. Furthermore, mechanical pressure with steam
can be detrimental to the desired properties.[30] [11].

AFEX treatment was originally devised as a way to process cellulosic material for the hydrolisis

for feedstock purposes. The reason for this technology was that cellulose, although very abundant

in plants, is surrounded by lignin and hemicellulose. These two components are not easily hy-

drolized. The AFEX treatment provides an effective way to access cellulose for hydrolisis. With

this argument, the present work investigated the feasibility of AFEX-treatment as a way to produce

high quality fibers of bamboo. In order to understand the physical changes that this treatment pro-

duced to the fiber bundles, a thermogravimetric analysis was conducted. Although the results of

the AFEX-treated specimens vary strongly depending on the parameters of the reaction (tempera-

ture, pH, concentration), the result should be focused on the effectiveness of this treatment for fiber

harvesting with a small reduction of cristallinity. The degree of cristallinity is an indication of the

fiber strength and modulus. Some studies have shown that AFEX treatment reduces cristallinity
by less than 3%[57].

18

3.1.5 Experiments conducted on AFEX-treated fibers

3.1.5.1 SEM images

In order to observe the morphology of the fiber after the AFEX treatment, single fiber bundles

were coated with a 100nm-layer of platinum and observed in the SEM (EVO Model, ZEISS-

Germany). The samples revealed to possess a high surface roughness, and some segments dis-

played parenchyma cells (see figure 3.5).

Figure 3.5 SEM image of region of a bamboo fiber-bundle showing parenchyma cells. Fiber treated
with AFEX method.

3.1.5.2 TGA tests

The first test conducted to evaluate the fiber properties was a thermogravimetric analysis (TGA)

of the fibers in a nitrogen atmosphere. The weight of the sample was 30.23mg. The fibers were

chopped so that the could fit inside the alumina container. In order to avoid noise in the signal,

the fibers were slightly pressed to create a more compact structure. The results are shown in

figure 3.6. The apparatus (Q500, TA Instruments) was set to a ramp of 25ºC/min from room

temperature to 500ºC since natural fibers are expected to degrade in this temperature range. This

was confirmed as can be seen in the figure. From this figure, it is possible to observe that there is a
multiple step process of weight loss. Moreover, there is some water desorbed[71] (see small peak at

19

around 100ºC in the left plot of figure 3.6). Lignocellulosic materials can be characterized in their
thermal behavior by the zones they degrade[60]: the hemicelluloses region (245ºC–290ºC), the

cellulose region (290ºC–350ºC), and the lignin region (350ºC–500ºC), . Figure 3.6 shows that the

AFEX treated samples analyzed in this work possess a high concentration of cellulose (see peak

at 302ºC). Furthermore, the presence of hemicellulose is also high (peak at 280ºC). The region

of lignin degradation shows two peaks of a slow degradation. The reason for the appearance of

several peaks is because of the presence of different molecules present in lignin.

Weight loss (%) vs. temperature (ºC)

Differential weight loss (%/ºC)

vs. temperature (ºC)

Figure 3.6 Thermogravimetric analysis on AFEX-treated fibers

After this test, the effect of drying the sample was studied. A sample of 18.76mg of fibers

were dried in a convection oven for two hours. Upon completion of this step, they were placed in

the apparatus under the same conditions as for the undried fibers. The results of this experiment

presented in figure 3.7.

20

50100150200250300350400450500Tempertature (ºC)102030405060708090100110Weight (%)79%63%37%26%15%Not dried AFEX bamboo50100150200250300350400450Tempertature (ºC)00.10.20.30.40.50.60.70.80.91Weight loss (%/ºC)280302438481Not dried AFEX bambooWeight loss (%) vs. temperature (ºC)

Differential weight loss (%/ºC)

vs. temperature (ºC)

Figure 3.7 Thermogravimetric analysis on AFEX-treated fibers after drying for 2 hours at 80ºC

The results of this test are quite interesting from point of view of the effect of temperature on

the fibers. The widths and locations of the peaks remain very similar as for the case of the undried

test. However, there is a substantial reduction of the height of the cellulose peak which could mean

that the drying process has enabled the access to hemicellulose sites. This could be a potential

indication that the AFEX treatment can have a positive effect on the quality of the fibers.

Some studies suggest that it is not desired to remove high amounts of hemicellulose because the

cellulose nano-crystals are embedded in a lignin and hemicellulose complex. If the hemicellulose

is removed, the cellulose would not be able to carry the load applied because the bond-strength
would be considerably reduced[57].

Density measurements To further study the behavior of the AFEX-treated fibers, density mea-

surements in a pycnometer were performed. Samples of 2g-3g were placed in the apparatus and 10

measurements for every specimen were conducted. The following table shows the results of this

experiments.

21

0100200300400500600Tempertature (ºC)102030405060708090100110Weight (%)80%63%39%27%Dried AFEX bamboo for two hours50100150200250300350400450500Tempertature (ºC)00.10.20.30.40.50.60.70.80.91Weight loss (%/ºC)2783034234622 hours dried AFEX bambooTable 3.2 Density measurements for bamboo samples

(cid:17)

(cid:16) g

cm3

Sample characteristics

Density

AFEX culm dried for 2 hours

at 80ºC

AFEX fibers dry 2 hours at

80ºC

AFEX fibers dry 1 hour at

80ºC

AFEX fibers not dried

1.1355

1.1595

1.1874

1.2025

(cid:17)

(cid:16) g

cm3

Standard deviation
±0.0007
±0.0013
±0.0013
±0.0013

3.1.5.3 Flexural properties of single fiber bundles

The mechanical and physical properties of bamboo fiber bundles depend on many factors (e.g
species, age, region, etc.)[3] [32].

In a first attempt to reduce the discrepancies in the reported

flexural properties in the literature, single fiber bundles were analyzed via three-point bending

tests. The fiber bundles were extracted via the mechanical process explained previously. Each

fiber bundle was cut at a length of 10 mm and the diameter was measured every 2 mm under an

optical transmission microscope (Keyence Corp. of America, Torrance, CA, USA). Before the

test, the fibers were dried in a convection oven at 80ºC for four hours to remove the adsorbed

water and other volatiles. The three-point bending test was performed in a dynamic mechanical

analyzer (DMA Q800-TA Instruments, New Castle, DE, USA) under static load, with a span (L)

of 5mm. The cross-section of the fiber-bundles was assumed to be circular. This assumption yields

a moment of inertia about bending axis I:

I =

π
64

d4

(3.1)

where d is the mean diameter of the fiber bundle. The measured values that were found to be

outside the interquartile range were excluded. The elastic modulus and strength of the specimens

were calculated by using the equation for pure flexure where the load is applied in the midpoint of

the two supports. For the elastic modulus, the equation is

E =

mL3
48∆I

22

(3.2)

where ∆ is the deflection of the specimen in the midpoint between the two supports, and m is found

by computing the slope of the force (N) vs. displacement (mm) in the linear region of the plot.

The strength of the fiber was calculated using the equation

σ =

PL
4Z

(3.3)

where P is the applied static force, and Z the section modulus. For the case of a circular cross-

section

Z =

π
32

d3

(3.4)

Results The three following figures show the results of these calculations:

Figure 3.8 Strength of bamboo fibers, extracted with pressurized water and rolling-milled. Note
significant change between the different samples

23

d =255 m d=257m d=328m Average with SEBamboo single fiber three-point bending test024681012Modulus [GPa]Figure 3.9 Elastic modulus of bamboo fibers, extracted with pressurized water and rolling-milled.
High spread of data

As it is evident from the spread in the values for modulus and strength obtained in this test,

single fiber bundle flexural tests do not give representative values for the behavior of the fibers.

These values cannot be traced back to the diameter of the fiber bundle. Therefore, the approach of

using a large amount of fibers (by fabricating composites) was adopted.

24

d =255 m d=257m d=328m Average with SEBamboo single fiber three-point bending test050100150200250300Flexural strength [MPa]3.2 Basalt fiber characteristics

3.2.1

Introduction

The need to find alternatives to synthetic glass and carbon fibers with similar properties has lead

to the study of basalt fibers. These fibers present a series of desired properties such as high elas-

tic modulus, ultimate strength, stiffness, and flexibility. Perhaps the most attractive trait of these

vulcanic fibers in industrial application lies in the cost advantage in specific strength and specific
modulus when compared to other fibers used in composites[10]. A crucial trait for the implementa-

tion of basalt fibers lies in the lack of sizing agents for the fabrication (as opposed to glass fibers),
which make this process more environmentally friendly[61]. Additionally, basalt fibers are easier
to recycle, can be implemented in harsh environments and also resist higher temperatures[33].

3.2.2 Basalt structure and constituents

Basalt composition:

Compound

SiO2
Al2O3
Fe2O3
MgO
CaO
Na2O
K2O
TiO2

Table 3.3 Composition of basalt fibers

Composition (%)

B[b]

51.6-57.5
16.9-18.2
4.0-9.5
1.3-3.7
5.2-7.8
2.5-6.4
0.8-4.5

-

A[a]
52.8
17.5
10.3
4.63
8.59
3.34
1.46
1.38

[a]: taken from Jamshaid et al.[46], [b]: taken from Militky et al.[33]

3.2.3 Basalt fiber production

Compared to glass-fibers, the fabrication process of basalt fibers does not require an independent

addition of other elements. Furthermore, the generation of these fibers does not yield secondary

25

Table 3.4 Physical and mechanical properties of basalt [a]

Properties

Thermal

Max. application
Temperature (ºC)

Basalt filaments

980

820

1450
2.75
9-23
0.1

4840
89
3.15

Physical

Density

Sustained application

temperature (ºC)

Melting temperature (ºC)

Filament diameter (mm)
Humidity absorption (65%

(cid:17)

(cid:16) g

cm3

Mechanical

[a]: Adapted from[40]

RAH)

Tensile strength (MPa)
Elastic modulus (GPa)
Elongation at break (%)

products like glass-fiber production, which is another advantage of basalt vs. glass.

There are several methods for the fabrication of basalt fibers. Since basalt fibers used in this study

were provided as continuous fiber-toes, the process of fabrication of these types of filaments will

be discussed.

There are two main processes for the fiberization of basalt: the melt-blowing and the centrifugal-

multiroll system). In both technologies, basalt rocks are washed, crushed and melted in a container
at 1450ºC[14]. The melted lava is then fed to a series of rolls. In the melt-blowing technique (also

called the Junkers method, see figure 3.10), the melted lava is poured onto an accelerator cylinder

(while air is blown towards the melted basalt). The lava then passes through the gap between two

spinning cylinders roatating in opposing directions (fiberization cylinders). This process yields
fibers resulting in a diameter of 6− 10 µm.

26

Figure 3.10 Junkers method for continuous fibers. Image taken from[33].

In figure 3.10, the numbers represent the most important components of this process.

1 : Melted lava poured to the wheels

2 : Nozzles to blow air to cool down the lava

3 : Accelerating cylinder to feed basalt into the next wheels

4 : Fiberization cylinder, rotating in opposite directions

5 : Basalt fibers, as final product of this production method

The centrifugal-multiroll system, as displayed schematically in figure 3.11, produces continuous

fibers by passing the melted lava through several high-speed wheels in alternating direction of
rotation. The fibers achieve a final diameter of 11 µm and 400 filaments[6] per toe.

27

Figure 3.11 Centrifugal method for basalt production. Figure taken from[33]

As figure 3.11 shows, the centrifugal force exerted by the wheels to the melt forces the lava to

pass through the array of the centrifugal wheels. These spinning wheels have to be placed such

that they produce the desired diameter of the fibers.

3.2.4 Experiments conducted on basalt fibers

3.2.4.1 TGA tests

As table 3.4 shows, the fibers are composed by a series of oxides. Therefore, it is expected to

observe a high temperature degradation of the basalt. This behavior was confirmed by the ther-

mogravimetric analysis shown in figure 3.12. A sample of 18.32mg containing chopped fibers

was placed in the apparatus (QA500, TA Instruments, USA). The sample was heated in a N2-
atmosphere at a rate of 25ºC/min from room temperature to 600ºC. As figure 3.12 shows, the curve

can be regarded as a single-step mode. The value of the onset (i.e the temperature at which the

28

sample starts to lose weight) is at 222ºC. This value is similar to the results found in other stud-
ies[27]. Additionally, the weight loss is very small (around 0.6%), which is also in agreement with

the same study. It is important to note that the spectrum is characterized by a small fluctuating

pattern (noise). This feature is characteristic of samples that possess a low mass to volume ratio,

such as fibers. To improve the quality of the signal, the raw-data values obtained from the appara-
tus were filtered with a Gaussian window function.[8]. The two curves are superposed, showing a

good fit between them.

(cid:105)

(cid:104) %◦C

Figure 3.13 shows the derivative of weight change (W%) with respect to temperature variation
temperature in [◦C]. This gives a more precise indication of the
change in weight with the temperature. In this case, there are three clear peaks between 50ºC

in units of

vs.

and 125ºC. These maxima correspond to water desorbed. The height of the peaks in the order of

∆T

(cid:17)
(cid:16)∆W%
10−3(cid:104) %◦C

(cid:105)

integrated with the width of the peaks indicate a very small amount of water present

in the fibers (around 0.05%). The highest weight loss of basalt fibers is around 40ºC higher than

that of glass-fiber. This property of basalt is attributed to the spinning process and to the degree of
cristallinity[13].

29

Figure 3.12 TGA of basalt fibers showing a single-step mode of weight loss.

30

50100150200250300350400450500550600Tempertature (ºC)99.699.6599.799.7599.899.8599.999.95100100.05Weight (%)Raw dataData with gaussian fittingFigure 3.13 Differential TGA for the sample of figure 3.12. The main weight loss occurs at 222ºC.

3.2.4.2 Single fiber fragmentation test

Sample preparation In order to measure the adhesion of the basalt fibers with the epoxy matrix

(previously described in equation 2.2), eight coupons were prepared with the dimensions shown in

figure 3.14.

31

Figure 3.14 Coupon dimensions for SFFT[39]. Black line represents the fiber.

The mold used was made of flexible silicone, with an open end at the top. Before the mold

was used, it was thoroughly cleaned with acetone and dried under the hood for ten minutes. Then,

single basalt fibers were placed by hand on the mold cavities, such that the fibers would pass

through the sprue slot. Each fiber was fixed to the mold by applying small drops of rubber cement

(Elmer’s Products, USA) with a syringe on the far ends of each sprue slot. After the rubber cement

was applied, a stainless-steel spatula was used to roll the rubber cement and gently bring it to

the sprue slot. It is important to note that after this step, the fiber should not be in tension. The

mixed m-PDA and EPON-828™were degassed after being mixed (see chapter 2 for details) in a

beaker together. The mold was also degassed to eliminate the air bubbles on the surface of the

silicone-mold. The presence of bubbles in the cured coupon could create stress-concentrations in

the samples, enabling cracks to propagate, and ultimately, causing the coupon to break at a lower

stress. After the resin was degassed, a previously-heated glass-syringe was used to pour the resin

into the mold. This tool is utilized because of the small volume of the mold cavities, as well as the

fragility of one individual fiber. The mold with the resin was then placed in the oven for the curing

step. It is important to maintain the surface of the mold horizontal so that the liquid resin flows

evenly accross the cavity. When the samples are cured, they need to be polished by an automatic

polisher to homogeneize the thickness of the sample (on the top and bottom sides). This processes

ensures the surfaces are smooth so to avoid stress-concentrators. The eight samples were attached

to the metallic disc with double-sided adhesive tape as shown in figure 3.15 and mounted into a

metallurgical polisher machine (Struers Abramin Metal Polisher).

32

Figure 3.15 SFFT coupons mounted on metal surface for polishing.

The top surfaces of the samples have to face the polishing paper since the other side of the

coupons (in contact with the mold) are very regular. Additionally, the samples that are in contact

with one another should be oriented such that the end of greater thickness of one sample should be

next to the end of smaller thickness of the sample. This set-up derives in a homogeneous polishing

on all sides. Another important aspect to take into account is that the samples must be symmet-

rically oriented. This means that they have to be at the same distance from the midpoint of the

disc to apply equal pressure on each sample. The coupons were polished using 320-, 600-, and

1200-ANSI-grit sizes for approximmately one minute with a force of 3-5N. The samples were

then polished for 30 seconds with an alumina final polish abrasive.

Data collection One of the main parameters to assess the bonding strength of the fiber with the

matrix is the fiber average diameter. To accurately determine this parameter, a calibrated microm-

eter slide was placed on each sample and observed with an optical microscope. The image of the

scale on the slide of the microscope was projected to a real-time screen. Once this step was com-

33

pleted, the single fiber fragmentation test was conducted in the following way[54]:

The samples were placed under an optical microscope to measure the diameter of the embedded

fiber in the epoxy. For this purpose, a micrometer calibration slide was placed on the coupon and

the scalebar set on the screen of a real-time video apparatus. Then, the diameters were measured

at different points of the coupon and the averaged diameter was calculated (see tables 3.5 and 3.6).

The sample was fixed to the grips so that lateral displacements were prevented, but allowing the

sample to have some space to move upwards and downwards. After this step a drop of oil was put

on the middle of the sample. After the oil spread, a square glass slide of 18mm (see table 3.5) of

side for the first four samples of side was placed on top of the sample. This thin slide was placed

in the middle region of the coupon. For the second batch, a square glass slide of 22mm 3.5 was

used. The reason to place the glass is to define a region to analyze. The use of oil is to prevent

minimize the adhesion between the glass slide and the coupon, and thus, to measure the strain of

the sample without other effects. The samples were subjected to tension by manually turning a

screw on the tension stage in the axial direction, in increments of 0.0127mm. The extension was

determined by the observing the dial. After applying the stress, the sample was visually scanned

for fiber breaks and this number was recorded for each strain. By placing the sample between

cross-polarizer, the double-reffracted light, called birefringence patterns are observed in the vicin-

ity of every break. This effect called colloquially as photoelasticity originates from the distortion

of light in materials subjected to strain. Furthermore, the shape of these patterns provide informa-

tion about the adhesion between the fiber and the matrix and the failure mode of the fiber inside
the coupon[12]. When the number of breaks remained constant after the sample was subjected to

two increments of 0.0127mm, the sample was considered to have reached saturation. The break

patterns were then recorded (using a camera connected to an optical microscope under polarized

light). These break patterns are displayed in figures 3.17 at low magnification and 3.17 at high

magnification. The fiber breaks -see dark regions in figure 3.16- are surrounded by birefringence

(bright regions sorrounding the breaks). The horizontal length of the break is known as the break

gap. This length varies depending on the applied stress. Another crucial information that the shape

34

of the birefringence is the bonding with the matrix. A flat, elongated birefringence indicates that

the stress has been poorly transfered from the fiber to the matrix. This can be explained by the

behavior of the interface between the fibers and the matrix. After curing of the sample, the matrix
is bonded with the fiber at a molecular level[36]. When the sample is subjected to tensile stress,

the fiber should transfer this load to the epoxy. If the bonding of the fiber with the resin is high, the

interface region in the vicinity should show signs of stress. In the case of a basalt fiber in epoxy,

the birefringences indicate that the fiber has experienced debonding while breaking, i.e., there is a

poor adhesion of the fibers with the matrix. The debonding zone is referred as the bright regions

around a fiber break. Note that the fiber breaks appear as dark under polarized light because there

is no stress at those spots.

Figure 3.16 Micrograph of basalt fiber under polarized light in epoxy showing several breaks.
Image taken at saturation strain. Arrows indicate fiber breaks.

35

Figure 3.17 Region of basalt fiber break in epoxy at saturation strain under polarized light. The
fiber has experienced debonding at break.

The following tables (3.5 and 3.6) display the summarized results of the SFFT conducted in

this study:

36

Table 3.5 Single fiber fragmentation test of basalt fibers in epoxy (number of breaks measured in a
span of 18 mm)

Strain(cid:0)mm

mm

(cid:1)

Number of breaks

Sample 1

Sample 2

Sample 3

Sample 4

¯d = 11.67 µm[a]

¯d = 12.07 µm

¯d = 12.93 µm

¯d = 14.47 µm

0.00282
0.00353
0.00423
0.00494
0.00564
0.00635
0.00706
0.00776
0.00847
0.00917
0.00988
0.01058
0.01129
0.01199
0.01270
0.01341
0.01411
0.01482
0.01552

0
0
0
0
0
0
0
0
4
9
26
30
44
48
52
52
54
54

break

[a]: ¯d is the average diameter of the fiber measured at different points of the coupon

0
0
0
0
0
0
0
0
1
5
7
9
11
18
28
34
35

break

-

0
0
0
0
0
0
0
0
1
5
12
22
27
30

break

-
-
-
-

0
0
0
0
0
0
0
0
4
7
10
16
22

break

-
-
-
-
-

37

Table 3.6 Single fiber fragmentation test of basalt fibers in epoxy (number of breaks measured in a
span of 22 mm)

Strain(cid:0)mm

mm

(cid:1)

Number of breaks

Sample 5

Sample 6

Sample 7

Sample 8

¯d = 8.900 µm[a]

¯d = 11.03 µm

¯d = 11.03 µm

¯d = 10.80 µm

0
0
0
0
0
0
0
0
0
0
0
2
3
7
13
30
43
48
51
57

0
0
1
2
2
3
7
10
14
25
34
43
48

break

-
-
-
-
-
-
-

0.00231
0.00288
0.00346
0.00404
0.00462
0.00520
0.00577
0.00635
0.00693
0.00750
0.00808
0.00866
0.00924
0.00981
0.01039
0.01100
0.01154
0.01212
0.01270
0.01328
0.01385

0
0
0
0
2
2
8
10
21
27
39
57
62

break

-
-
-
-
-
-
-

0
0
0
0
8
18
27
39
49
51
55
57
60
60
-
-
-
-
-
-
-

[a]: ¯d is the average diameter of the fiber measured at different points of the coupon

break

The data from the tables 3.5 and 3.6 are summarized in the following image (figure 3.18).

38

Figure 3.18 Breaks with respect to strain

The shape of the behavior of the number of breaks with respect to the applied strain resemble
a Weibull distribution as reported in previous works[20] [55]. Furthermore, the distance of between

each break is not constant, as figure 3.16 shows. Therefore, the results are analyzed using the

following guidelines.

Data analysis The interfacial shear strength between the fiber and the epoxy matrix is given by
the equation proposed by Kelly and co-workers[35]:

τ =

σFd
2lc

39

(3.5)

020406080100Applied Stress [MPa]010203040506070Number of breaks in couponSample 1Sample 2Sample 3Sample 4Sample 5Sample 6Sample 7Sample 8where τ is the interfacial shear strength, σ the fiber tensile strength, and lc the critical length of the
fiber. lc is determined as

lc =

¯l

4
3

(3.6)

where ¯l is calculated as the gauge length divided by the number of breaks at the critical strain.
With this approach the IFFS gives a mean value of 9.60MPa with a SE of ±0.77MPa. Since the
critical length is subjected to considerable variation due to flaws of the fiber, handling of the spec-
imen, etc.[51], the critical length determination is established with a probablilistic approach[23].

Ultimately, this method is used estimation of the interfacial shear strength. The following equation
(as proposed by Drzal et al.[20]) predicts the IFFS of basalt in the epoxy matrix:

(cid:19)

(cid:18)
1− 1
α

τ =

εE f
2β

Γ

(3.7)

(3.8)

where ε is the measured strain in mm
mm, E f the nominal modulus of basalt (in MPa), β the shape
parameter and α the scale parameter of the Weibull distribution and Γ the gamma function[55].

This probabilistic approach yields a mean value for the interfacial shear strength of 9.73MPa with
a standard error of ±0.75. For the failure probability, a two-parameter Weibull probability density
function f (ε,α,β ) of the random variable ε of the form:

(cid:18)E f ε

(cid:19)α−1

(cid:20)

(cid:18)E f ε

(cid:19)α(cid:21)

exp

−

f (ε;α,β ) =

α
β

β

β

is used. The cumulative probability function F(ε;α,β ) for this distribution is obtained by inte-

grating equation 3.8. This means:

F(ε;α,β ) =

(cid:90) X

−∞

f (ε;α,β )dε

Last equation yields:

F(ε;α,β ) = 1− e(− ε

α )β

(3.9)

(3.10)

In order to assess the correctness of this distribution, the data obtained from the SFFT were plotted

in figures 3.19 and 3.20. These two plots were generated in MATLAB™and the plots are shown

40

in two separate figures for clarity. As it can be seen, the dots (which represent the actual data) are

in very good agreement with the predicted line (shown as discontinuous lines).

Figure 3.19 Weibull fit for SFFT test of four samples

41

2030405060708090100110Applied Stress [MPa]0.05  0.1   0.25  0.5   0.75  0.9   0.95  0.99  0.999 ProbabilityProbability plot for Weibull distributionWeibull sample 1Data sample 1Weibull sample 6Data sample 6Weibull sample 3Data sample 3Weibull sample 8Data sample 8Figure 3.20 Weibull fit for SFFT test of the other four samples

42

2030405060708090100Applied Stress [MPa]0.05  0.1   0.25  0.5   0.75  0.9   0.95  0.99  0.999 ProbabilityProbability plot for Weibull distributionWeibull sample 2Data sample 2Weibull sample 5Data sample 5Weibull sample 4Data sample 4Weibull sample 7Data sample 7CHAPTER 4

COMPOSITE CHARACTERIZATION

4.1 Neat epoxy characteristics

The resin selected for all of the composites used in this study was Ddiglycidyl ether of bisphenol A

(DGEBA - EPON 828 - Miller Stephenson Company,USA) with meta-phenylene diamine (mPDA)

as curing agent (Sigma Aldrich, USA). The ratio was 100/14.5 DGEBA/mPDA by weight. This

matrix in this stoichiometry was selected due to the fact that this epoxy has been very well char-
acterized for its mechanical and physical properties. The density of the matrix is 1.90 g
[38], a
cm3
flexural strength of 120MPa and elastic modulus of 3.2GPa[19]. The curing cycle of the mixture

was done 75ºC for one hour and a post-curing at 125ºC for two hours. The curing and post-curing

processes were conducted in a convection oven. Neither the resin nor the curing agent were treated

in any way. They were used as received.

4.2 Procedure for the fabrication of composites

The following method was implemented for the fabrication of all unidirectional flexural composites

analyzed in this study. The mould consisted of two stainless-steel plates of dimensions 150mm x

150m x 5mm. Two square pieces of aluminum foil were cut in dimension of 16cm x 16cm. The

aluminum foil was used to cover the plates, leaving 1 cm from each of the stainless-steel plate

so that the aluminum foil could be folded to leave a flat surface to be in contact with the resin.

After this step, both plates (covered with aluminum foil) were put between flat steel plates of

20kg each to improve flatness of the surfaces of the mold. After two hours under pressure, an U-

shaped high-temperature silicone rubber sheet is attached with high-temperature adhesive tape to

one side of the mold (see red U-shaped region in fig.4.1). Additionally, two silicone rubber strips

are adhered to the mold in order to set the with of the composites. When this step is completed,

43

mold release liquid is applied on the aluminum-covered stainless steel plate and to the silicone

rubber. The mold release has to be applied with a paper towel. The application of this liquid is

first applied vertically. After one minute, the mold release is dried with a paper-towel horizontally.

The mold release prevents the cured resin to adhere to the mold and to the rubber. It has to be

done under a specially designated hood in order to prevent contamination. After five minutes the

fibers (see brown region in fig.4.1) were manually aligned and held in place with two strips of

high-temperature adhesive on top and bottom (see blue features in fig.4.1). The other stainless-

steel plate was aligned and tightened with C-clamps on the four sides of the plates. The C-clamps

serve to compress the silicone rubber to achieve a gap of 3.2mm which is the desired thickness

(depth) for the dimension of the ASTM 790 standard. By doing this a mold with an open end on

top is achieved. The mold is placed in a convection oven at 75ºC for hour oriented as figure 4.1

shows.

Along with the mold, one plastic beaker with 100g resin (DGEBA) and another with 14.5g of the

curing agent (m-PDA) were placed in a convection oven for one hour at 75ºC. Then, the resin is

poured in the beaker containing the m-PDA and stirred for one minute, assuring that the resin on

the walls of the container are transferred to the other beaker. After this, the mixture of DGEBA

with m-PDA is brought back to the beaker that originally contained the resin. The liquid mixture

is stirred again for one approximately one minute. At this point, around 30g of the mixture is

poured into the mould, in both gaps between the silicone rubber and also over the fibers. The

mold should be oriented while pouring the mixture as shown in fig.4.1. The space between the

silicone rubber strips and the bottom of the U-shaped rubber enables the liquid resin to wet the

fibers. Then, the mold the the beaker containing the rest of the mixture are brought to a vacuum

oven at 75ºC, with the mold oriented as shown in the figure. When vacuum is applied the liquid

mixture wets the fibers and degasses the mixture. Vacuum should be applied for around three

to five minutes, until the bubbles that initially formed on the surface have grown and collapsed.

This step is necessary to remove air bubbles while promoting substantial mixing of the resin with
the curing agent[24]. The mold and the beaker were brought back to atmospheric pressure when

44

the small bubbles started to form at the bottom of the beaker and rise. This is an indication of

the process of evaporation of the constituents that may interphere with the curing process of the

epoxy and, ultimately, with the physical and mechanical properties of the composite. When the

degassing was completed, the liquid mixture was poured over the fibers several times, in order to

allow the fibers to be fully wetted. Additionally, the mixture was poured in both segments of the

mold without fibers to promote the flow of the resin to the fibers.

At this point, the mold with the resin was placed in a convection oven (previously set to 75ºC)

and kept at 75ºC for two hours. After this time has passed the temperature is increased to 125ºC

and kept for two hours for the post-curing process. When this cycle is complete the oven with

the sample inside should be turned off so that the internal temperature of the oven reaches room

temperature. This step takes approximately four hours.

The composites were then cut to specimens of dimensions to meet the standards for flexural testing
of reinforced plastics (ASTM-D790[63]). It was not necessary to polish the samples after before

testing them since the surfaces of the specimens were smooth and regular in cross-section.

4.3 Evaluation of composites

4.3.1 Three-point bending tests

The three-point bending tests were conducted in compliance with ASTM-D790[63] with a United

Testing Systems (UTS) apparatus model SFM-20. The following parameters were calculated for
each batch of samples in compliance with[63]:

• The span (L) was calculated as 16 times the width of the specimen

• The termination of the test when the midspan deflection has reached 0.05(cid:0)mm

mm

is calculated with the following equation:

D =

rL2
6d

45

(cid:1). This value

(4.1)

where r = 0.05 (strain in mm

mm), and d the thickness of the sample (referred as depth in ASTM-

D790).

(cid:1), calculated with
• The rate of the cross-section motion R in(cid:0)mm
where Z = 0.01 is the rate of straining of the outer fiber in(cid:0) mm

ZL2
6d

min

R =

(4.2)

(cid:1)

mm min

Figure 4.1 Mold scheme used to prepare flexural coupons.

46

15.015.02.01.01.011.5Figure 4.2 Bamboo (AFEX) cut to a length of 12.7 cm.

Figure 4.3 Basalt fibers cut to a length of 12.7 cm.

47

Figure 4.4 Basalt and bamboo fibers intertwined and cut to a length of 12.7 cm.

4.3.2 Fiber distribution and void content assessment

The following procedure was conducted to visually evaluate features of the composites such as the

distribution of the fiber-bundles, the presence of voids and the amount of the different components

of the bamboo fibers (see chapter 3). The flexural composites were cut through the thickness

(perpendicular to longest dimension of the sample) to obtain four samples of same dimensions.

This was done in order to have representative cross-section observations of the composite. After

the samples were cut, they were placed in a cylindrical holder for polishing, such that the cross-

sections would be exposed to the polishing surface. A long-cure epoxy in a weight ratio of 100g

resin to 14g of hardener (LC Epoxy, LECO Corporation, USA) was used to hold the pieces of the

sample in place. The sample was allowed to cure at room temperature overnight. After this, the

cured epoxy containing the sections of the composite were placed in a vibratory polisher (VibroMet

I™,Buehler, USA) for three hours.

48

4.4 Bamboo composites

4.4.1

Introduction

The next step of the present study was the fabrication (see section4.2 for details on the procedure)

of composites with AFEX-treated bamboo fillers in an epoxy-based matrix. There were three

batches produced with different bamboo weight percet composites with different fiber content

were fabricated. Additionally, a first attempt to use bio-based (Acrodur™-950L, BASF, Germany)

resins for the fabrication of bamboo-reinforced composites is reported in Appendix 5.2.3. These

composites were not further evaluated in the present study. All the following experiments were

conducted with epoxy resin (DGEBA + mPDA) as matrix.

The fabricated composites exhibited a good dispersion of the fibers as displayed in 4.5. This sam-

ple was prepared in accordance to the method described in section 4.3.2. The image was acquired

via a transmission optical microscope (Keyence, USA) in a grey-scale mode to improve the con-

trast between the fibers with the epoxy.

In order to understand the hybrid synergistic effect of the combination of basalt and bamboo fibers,

unidirectional bamboo samples were first prepared. This procedure was conducted in order to

understand first the interactions of the bamboo fibers with the matrix and then infere their per-

formance when combined with basalt. The current section studied the mechanical response to

three-point bending tests. It is expected that the increase of the bending strength of the composite
is, primarily, due to the interfacial shear strength of the fibers in the composite[66].

49

Figure 4.5 Transmitted light micrograph of cross-section of bamboo composite (40 weight % of
fibers).1: resin, 2:hollow-vessel region, 3: fiber-sheath of sclerenchyma cells, 4: parenchyma cells.
Note good dispersion of fibers.

Figure 4.6 Surface of fiber bundles in composite (40 weight %) showing resin residues on the
fibers.

4.4.2 Physical and mechanical properties

Mechanical properties[22]

50

1 2 3 4 Figure 4.7 Elastic modulus of bamboo-reinforced composites with different weight fraction of
fibers.

51

15 weight %40 weight %61 weight %61 weight %Bamboo fiber composites0510152025Flexural modulus [GPa]Figure 4.8 Flexural strength of bamboo-reinforced composites at several values of fiber content.

4.4.3 Fracture analysis

The bamboo fiber in fig.4.9 shows clear evidence of epoxy on the surface, a characteristic that

shows good interfacial bond strength. Furthermore, the matrix in fig.4.11 presents features of

river patterns which are evidence of high interfacial bonding. These features are also clear in

the resin-rich region of fig.4.12. The layered structure formed in this figure in the vicinity to the
fiber-bundles indicate fiber pull-out under tension[29].

52

15 weight %33 weight %39 weight %40 weight %Bamboo fiber composites050100150200250Flexural strength [MPa]Figure 4.9 Bamboo fiber bundle in tension side of composite (40 weight %) Note epoxy on the
surface of bamboo.

Figure 4.10 Fracture surface of bamboo-reinforced composite. Resin deformation observable on
top left side. Fibers broken, not pulled out of the matrix.

53

Compression Tension Figure 4.11 Fracture surface in center of sample. Different heights of fibers’ breaks indicate energy
absorption.

Figure 4.12 Fracture surface of bamboo-reinforced composite on tension side (40 weight %).
Hackle formation and river patters on resin surface. Debris indicate tensile failure of fiber.

54

Compression Tension Compression Tension 4.5 Basalt composites

4.5.1

Introduction

In order to investigate the behavior of basalt fibers as reinforcement agents in an epoxy-based ma-

trix, composites with different fiber content were fabricated. All the samples were produced using

the technique described in section 4.2. This technique was proved to be adequate for a good disper-

sion of this type of fibers as is observed in figure 4.13. The concentration of fibers in certain areas

can be explained by means of the fact that they were not treated in any way before the composite

fabrication, so a possible explanation is the poor adhesion between the fibers and the matrix. This

micrograph shows a region of the cross-section of a basalt-reinforced composite (50 weight percent

of fibers). This sample was prepared in accordance to the method described in section 4.3.2. The

image was acquired via a transmission optical microscope (Keyence, USA) in a mode of scales of

grey to enhance the contrast between the fibers (dark circular spots) and the light-grey matrix. The

two dark irregular spots that are visible in the image are due to the reflection of the light through

the specimen and do not represent a feature of the composite.

Since the ultimate goal of the present study is to understand the properties of the hybridized com-

posites (bamboo/basalt reinforced composites), it is imperative to understand the characteristics of

basalt composites. This section is dedicated to the understanding of the mechanical properties of

basalt-reinforced composites subjected to three-point bending tests with the hope to elucidate the

synergetic effect of the combination of bamboo and basalt as fillers.

55

Figure 4.13 Cross section of basalt composite (50 weight %). Note good dispersion of fibers.

4.5.2 Results on flexural tests

The three-point bending tests were conducted in compliance with the ASTM D-790 standard as

described in detail in section 4.3. A series of four samples for each batch of the same composition

was produced. The flexural strength and modulus were calculated. A summary of these properties

is presented in the figures 4.14 and 4.15.

56

Figure 4.14 Elastic modulus of basalt-reinforced composites with different weight fraction of
fibers. Basalt fibers untreated.

57

33 weight %39 weight %40 weight %Basalt fiber composites024681012141618Flexural modulus [GPa]Figure 4.15 Flexural strength of basalt-reinforced composites with different fiber volume content.

58

33 weight %39 weight %40 weight %Basalt fiber composites0100200300400500600Flexural strength [MPa]Figure 4.16 Fracture surface of basalt composite (40 weight %) showing almost no residues of
resin on the surface of the fibers.

4.5.3 Fracture analysis

The single-fiber fragmentation tests revealed that the bonding strength between the basalt and the

polymeric matrix was not high. This observation is also evident in the behavior of these fibers

in the flexural tests conducted for the basalt-reinforced composites. As can be observed in figure

4.17, the presence of voids on the fracture surface of the specimen indicate a poor adhesion of

the fibers with the epoxy. This means that the fibers have been pulled-out with low resistance.

Furthermore, the surfaces of the basalt fibers show no signs of resin, an additional characteristic

of poor fiber-matrix adhesion. Also, there are river patterns on the side of the fracture surface that

was subjected to tension. This is a characteristic of high strength of the fiber, but low bonding
strength between the filler and the matrix.[29].

59

50µm  Figure 4.17 SEM micrograph of fracture surface of basalt composite (40 weight %) at the edge on
the specimen subjected to compression.

Figure 4.18 shows the side of the fracture surface of a specimen subjected to a three point bend-

ing test. The area inside the circle shows a clear gap between the fiber and the matrix. This type

of low-adhesion characteristics are sources for crack propagations. Also, this region demonstrates
that a poor adhesion results in the formation of a layered fracture surface of the matrix[29].

60

Tension Compression Figure 4.18 SEM micrograph of fracture surface of basalt composite (40 weight %) on the side of
the tension edge.

Another important feature of the failure mode of the specimen is shown in figure 4.19, where

the matrix surface on the right shows evidence of polymer debris, demonstrating tensile failure of

the sample. This micrograph also shows that the fibers break at different lengths, suggesting that

some energy absorption has taken place. This is an important feature of basalt fibers.

61

Tension Compression Figure 4.19 SEM micrograph of fracture surface of basalt composite (40 weight %) at the middle
of the sample.Note mixture of failure modes.

4.6 Hybridized composites

4.6.1

Introduction

The specimens were prepared with the method reported in detail in section 4.2. The basalt and

bamboo fibers were manually placed in the composite, such that the two types of fibers would be

intermingled. This was achieved by alternating one toe of basalt fibers (approx. 400 fibers) with

one AFEX-treated bamboo fiber. This type of distribution of fibers is regarded as commingled.

In order to understand the synergetic effect of both types of fibers in the composite, this type of

fiber arrangement was chosen. This randomization of the fiber distribution was divised to make a

first assessment of the overall properties of the composites.For convinience, the weight fraction of

basalt fibers will be denoted as Bw% and the AFEX-treated bamboo fiber weight fraction as Aw%.
Since the main goal of the present study is to develop high-performing composites, adding basalt

fibers can be of great interest against the environmental degradation. It has been shown that the

mixing of natural fibers with more durable fibers (such as basalt) can mean a higher resistance to

62

Tension Compression aging effects of the composite[65]. [h]

Table 4.1 Conversion of weight to volume fraction of fibers in hybrid composites [a]

Sample
weight
percent
ratio

(AB/B) [b]

Bamboo

fiber
weight
(%)

Basalt
fiber
weight
(%)

Total fiber

weight
(%)

Bamboo

fiber
volume

(%)

Basalt
fiber
volume

(%)

Total fiber

volume

(%)

40/0
30/10
20/20
0/40

40
30
20
0

0
10
20
40

40
40
40
40

41
31
21
0
1.1595 g
cm3

(cid:18)

(cid:19)

0
5
10
23

41
36
31
23

and basalt fiber density of

(cid:18)

(cid:19)

[a]: Conversion made with density for dry AFEX-bamboo fibers of
2.6295 g
cm3

. [b]: AB/B denotes weight ratio of AFEX-treated bamboo to basalt fibers

Figure 4.20 Flexural strength at different ratios of fiber weight percentage

63

A-B:(40-0)A-B:(30-10)A-B:(20-20)A-B:(0-40)A:AFEX-treated bamboo fiber weight (%), B: Basalt fiber fiber weight (%)0100200300400500600Flexural strength [MPa]Figure 4.21 Flexural modulus at different ratios of fiber weight percentage

4.6.2 Fracture analysis

The following micrographs were taken in a scanning electron microscope (Model EVO-ZEISS,

Germany) under variable pressure mode. This mode was selected in order to avoid charging of the

non-conductive sample and therefore eliminate the need of coating. The following micrographs

show the fracture modes on the surface of the samples subjected to three-point bending test.

4.6.2.1 Composites of Aw%=20 and Bw%=20

In figure 4.22 it is possible to observe a void formation around the bamboo-rich region. Towards

the middle of the sample (on the left side of the figure) it is possible to identify a bundle of basalt

64

A-B:(40-0)A-B:(30-10)A-B:(20-20)A-B:(0-40)A:AFEX-treated bamboo fiber weight (%), B: Basalt fiber fiber weight (%)0510152025Modulus [GPa]fibers. This micrograph also demonstrates the break of bamboo fibers at different heights on the

side subjected to tension, indicating energy absorption of the sample. Furthermore, it is possible to

observe a crack almost perpendicular to the side of the sample on the side of tension, which may be
an indication of delamination that increases the energy absoprtion amount of the composite.[28].

There are other features of interest on the fracture surface. One of them is the presence of river

Figure 4.22 Fracture surface showing the edge of the sample subjected to tension. Note basalt fiber
accumulation.

patterns as shown in the bottom left region of figure 4.23 as well as hackles. This resin-rich region

demonstrates a tensile failure mode of the composite in the surroundings of bamboo fibers. A

good adhesion between the bamboo fibers and the matrix is observed due to the breaks of the

fibers (rather than pull-out features). This is not the case of basalt fibers in this hybrid composite

as can be seen in figure 4.24. In the polymer region between the basalt and the bamboo fibers there

is no sign of river patterns. Probably this is due to the poor bonding between the matrix and the

basalt fibers. There are traces of crack formation pn the top right side of the figure, around the

bamboo fibers. This is a remarkable characteristic of basalt fibers, since it hinders the micro-crack
propagation, as compared to bamboo[1]. Nontheless, the poor adhesion of the basalt to the matrix

65

Compression Tension has a negative effect in the transfer of loads (see pulled-out fibers in the figures of section 4.5).

Figure 4.23 Hybrid composite 20%/20% AFEX/Basalt on tension edge. Bamboo fibers have failed
by break. River patterns on resin surface.

Figure 4.24 Hybrid composite 30%/10% AFEX/Basalt in the center of sample. Crack propagation
around the bamboo fiber bundle. Basalt fiber pulled-out.

66

Compression Tension Compression Tension 4.6.2.2 Composites of Aw%=30 and Bw%=10

[h] In this section, the fracture surfaces of composites containing 30 percent in weight of bamboo

fibers (AFEX-treated) and 10 percent in weight of basalt are analyzed. This composition shows

the highest value of modulus (see figure 4.21). This can be explained by observing the basalt

composites and the bamboo composites. Since the bonding of bamboo with the epoxy matrix is

high, the load can be transferred to these fibers in the composites. However, the bamboo fibers are

prone to crack propagation (as can be seen in figures 4.22 and 4.23). On the other hand, basalt

fibers are not very well adhered to the matrix but they show no signs of micro-crack propagation

in the vicinity of the fiber bundles. Therefore, the combination of both types of fibers create a

positive effect in the overall modulus of the composite (see 4.21). This effect is evident in figure

4.25 where the matrix has undergone a great amount of deformation close to the bamboo fibers.

Also in the same figure, there is a formation of cracks on the left side of the micrograph, in a region

of bamboo predominance. This hypothesis is again confirmed in 4.27 where the resin surface is

almost undistorted (left side of the micrograph, close to the basalt fibers) while the right side of

the figures shows delamination along the bamboo fibers and traces of cracks. It appears that the

ratio of 30/10 weight percent of bamboo to basalt yields the highest stiffness of the composites.

However, for the case of strength, the hybridization does not seem to not have this synergetic

effect. This might occur because the strength of the composite is strongly dependant on the fiber

properties. Specially in the case of three-point-bending tests, the mechanisms of As a result of

the increased glass fibre content, the flexural strength increases with the addition of basalt fibers
because of the higher shearing resistance[47]. A similar result is found in the study conducted by
Sreekala et al[62], where the flexural strength of the hybrid composites increases with glass-fiber

content, while the modulus decreases. The authors argue that, in a hybrid composite system, the

fiber with the smallest elongation at break will determine the strength. This is also observed in the

present study.

67

Figure 4.25 Hybrid composite 30%/10% AFEX/Basalt on tension edge. Note matrix deformation
and voids formed

Figure 4.26 Hybrid composite 30%/10% AFEX/Basalt on tension side. Note resin-rich region and
basalt fibers pulled out.

68

Compression Tension Compression Tension Figure 4.27 Hybrid composite 30%/10% AFEX/Basalt in center of the sample. Failure by delami-
nation on the right side. No deformation of resin region.

4.6.3 Rule of mixtures

The rule of mixtures (ROM) is a mathematical tool to compute the properties of a composite,

based on the weighted average of the fiber volume fraction. The rule of mixtures, in its ideal case,
assumes no porosity[4]. Mathematically, the rule of mixture for composites with one type of fiber

is expressed as:

Xc = f X f + (1− f )Xm

(4.3)

where f is the volume fraction and X denotes the property. This could be density, flexural modulus,

flexural strength, to name a few. The terms Xc, Xf, and Xm, represent the property of the com-
posite, the fiber, and the matrix, respectively. Since bamboo fibers vary considerably in volume,

the weight of the fibers were measured for the fabrication of composites to make a reproduceable

method. For this reason, there is the need to convert from weight fraction of fibers in the composite

(w) to volume fraction (f). This is done with the ROM for densities (ρz), where the subscript z is

69

Compression Tension f,m, and z for the fibers, the matrix, or the composite. If in the following equation

f is expressed as:

ρc = f ρ f + (1− f )ρm

f =

Vf

Vf +Vm

(4.4)

(4.5)

where Vf +Vm is the total volume, or the volume of the composite, defined as Vc. With the defini-
tion of density ρ, as

and

rho =

m
V

w ≡

ρ fVf

ρ fVf + ρmVm

the volume fraction of fibers are calculated by:

(cid:17)
(cid:16) 1
1
w − 1

+ 1

f =

ρ f
ρm

(4.6)

(4.7)

(4.8)

For the calculation of hybrid composites, the ROM has to be slightly modified, as follows:

Xc = f (AXA + BXB) + (1− f )Xm

(4.9)

where A and B are the fraction coefficients for bamboo and basalt, respectively. XA and XB are the
properties for bamboo and basalt, respectively. Furthermore, the following equation holds always

true:

A + B = 1

(4.10)

Using the ROM for the results for flexural strength and modulus for bamboo and basalt composites,

shown in sections 4.5 and 4.4, respectively, the following properties are obtained: Using the results

Table 4.2 Flexural strength and modulus calculated for basalt and bamboo fibers using ROM [a]

Fiber
Bamboo
Basalt

Strength in MPa (SE)

Modulus in GPa (SE)

264.65 (22.145)
1769.5 (153.49)

29.896 (2.0462)
59.3402 (3.2419)

70

from table 4.2 as the expected values for the modulus and strength of bamboo and basalt fibers,

and the nominal flexural properties for the matrix indicated in chapter 1, the following plot was

generated:

Figure 4.28 Flexural modulus comparison between predicted and calculated values in hybrid com-
posites

71

A-B:(40-0)A-B:(30-10)A-B:(20-20)A-B:(0-40)A:AFEX-treated bamboo fiber weight (%), B: Basalt fiber weight (%)0510152025Modulus [GPa]Predicted valuesSE for predicted valuesExperimental valuesSE for experimental valuesFigure 4.29 Strength comparison between expected and observed values.

It is remarkable to observe the behavior of the hybrid composite at 30 weight percent bamboo

with 10 weight percent in terms of the modulus (see figure 4.28). That is the synergistic effect
that yields a modulus higher than predicted by ROM[37]. It would be necessary to conduct more

experiments to study this phenomenon in depth. For the case of figure 4.29, the predicted values

are in good agreement with the experimental ones. It is worth noting that the values taken for the

fiber modulus and strength are the average and do not include the standard deviation.

72

A-B:(40-0)A-B:(30-10)A-B:(20-20)A-B:(0-40)A:AFEX-treated bamboo fiber weight (%), B: Basalt fiber weight (%)0100200300400500600Flexural strength [MPa]Predicted valuesSE for predicted valuesExperimental valuesSE for experimental valuesCHAPTER 5

CONCLUSIONS AND FUTURE WORK

5.1 Mechanical behavior of composites

The present study has shown that the it is possible to achieve good flexural properties of bam-

boo/basalt hybrid composites in an epoxy resin. Furthermore, the synergetic effect of the com-

bination of these two types of fibers is presented. It has been shown that the flexural modulus is

increased by approximately 66%. Additionally, the failure characteristics were investigated in or-

der to understand the mechanisms involved and, ultimately, determine ways to improve the quality

of the composites.

Since the bamboo and the basalt fibers possess different physical and mechanical features, a series

of tests were conducted to maximize their performance in the composite. The AFEX treatment

was proven to be an effective, and environmentally-friendly way to process the bamboo fibers.

This treatment reuses up to 99% of the original ammonia, which conversely produces significant

waste. The bamboo fiber-bundles obtained show good adhesion to the matrix due to their shared

hydrophilicity character. The surface roughness of AFEX-bamboo is greater than other fibers,

which also adds to the good mechanical properties observed in the composites.

The TGA on the AFEX-bamboo revealed that the hemi-cellulose and the lignin present were re-

duced. This is important to a certain degree because the fiber bundles are left with more suitable

components (such as a greater degree of cellulose) for structural features. Although the water up-

take of water after the AFEX treatment was still high, the water desorption is quickier. This is of

particular interest for the industry, since it means shorter times of drying, less energy, and there-

fore, cost reductions.

In the case of basalt fibers, they did not undergo any treatment for the composite fabrication. As

the TGA figure shows, their water absorption is very small, as well as the degradation up to 600ºC.

Although the strength of the hybrid composites increased with the weight content of these fibers,

73

the single fiber fragmentation and SEM fracture surface show a poor bonding strength with the

epoxy.

In terms of the fabrication process of flexural coupons, a simple method was developed, that

yielded good wetting of the basalt and bamboo fibers and a minimal void content. An impor-

tant aspect of this technique is that it does not require mechanical pressure or vacuum. This is

translated into a reliable, easily reproduceable fabrication technique to make composites. This

resin-pouring method also assures a high degree of alignment and dispersion of the fibers, which

is vital for high-quality unidirectional coupons.

The SEM micrographs revealed that the failure mechanisms are different depending on the filler

used (basalt, bamboo, or hybrid fillers). In the case of bamboo, the bonding strength to the matrix

caused the flexural samples to form hackles and river patterns on the surface of the epoxy. Also, the

fibers broke at different heights, which is an indication of energy absorption by the system. This

characteristic is desireable in many structural and impact applications. In the case of basalt com-

posites the low adhesion of the fibers to the epoxy caused pull-outs of the fibers from the matrix.

This needs to be improved in order to utilize the intrisic mechanical properties of the fiber. Basalt

fibers are remarkable for the amount of energy they can absorb, which is why, partially, the matrix

in the vicinity of the fibers of the fracture surface show no matrix distortion. This characteristic can

be advantageous to avoid the micro-crack propagation and therefore, create stronger composites.

5.1.1 Bamboo fiber performance

The AFEX treatment conducted on the bamboo culms proved to be positive as a harvesting tech-

nique. After this pretreatment, the fiber bundles were easy to harvest without applying mechanical

pressure with a rolling-mill. The rolling-mill could deteriorate the cristallinity of the fibers and

create defects that possibly derive in lower strength. Moreover, thys type of fiber displayed good

adhesion with the epoxy matrix. In order to improve the performance of the fibers, the parameters

of the AFEX treatment can be studied. These parameters include the pH, the time of treatment and

the temperature. This procedure would allow to find the ideal hemicellulose and lignin content.

74

Additionally, fibers of smaller diameter could result in higher mechanical properties[52].

5.1.2 Basalt fiber performance

A good adhesion is crucial to the performance of the composite, and therefore there is the imper-

ative need to find surface treatments for the basalt fibers. In this study, it has been that the basalt

fibers did not ehxibit epoxy residues on their surfaces. This poor bonding strength also derived in

fiber pull-outs, which is detrimental to the overall performance of the composite. One important

feature of the AFEX treated fibers is the increase of surface roughness compared to the ones har-

vested with the pressure chamber. It is hypothesized that this surface characteristic is translated

into mechanical locking to the matrix which, in turn, creates a higher flexural modulus to the com-

posites.

5.2 Next steps

5.2.1 Sizing of basalt fibers

As mentioned before, the poor interfacial bonding between basalt and epoxy is detrimental for the

composite performance as a whole. There are several ways to increase the adhesion of these fibers

with the epoxy matrix. The nano-hybrid of silicon dioxide (SiO2) with epoxy as one candidate
to increase the bonding strength with the matrix and, consequently, increasing the strength of the
composite[68]. Another coupling agent for this purpose is silane. Silanes are covalently-bonded

compounds of silicon and hydrogen in the form: SinH2n+2. This is a very versatile compatibi-
lizer that can be used in different matrices. Mafic basalt fibers, among other companies, offer

silane-treated, continuous fibers, compatible for different polymeric matrices, such as polypropy-
lene, epoxy, vinyl ester, polyurethane, and polyamides (PET/PBT)[42]. The effectiveness of silane

treatments for the enhancement of flexural strength is attributed to the change in the pH of ionic

75

silane compounds when combined with the fibers[64].

5.2.2 Fiber distribution for improvement of flexural properties

In this work, the basalt and bamboo fibers were distributed randomly in the composite. However,

a more rational approach in order to increase the modulus is to place the stiffer fibers (i.e. basalt)

as far as possible from the neutral axis in the three point bending test. For this purpose, instead

of fabricating single lamina of hybrid composites, a laminate structure could be prepared. The

fabrication of one-fiber-type composites yields better control of the dispersion of fibers. After the

preparation of the laminae, they would be stacked in a sequence to form a laminate. A sandwich

arrangement for the laminate is desireable, where the laminae of basalt-reinforced composites lie
away from the neutral axis[37].

5.2.3

Impact tests

As the fracture surfaces of this study suggest, the energy absoprtion of the hybrid composites is

high even for the low adhesion strength of basalt with epoxy. The lack of crack propagation in the

environment of basalt fibers suggest that the here-studied composites can take an important amount

of energy before failing. To determine this hypothesis, the composites can be tested in Izod impact

tests. Furthermore, the behavior of the energy uptake of the materials can be assessed as a function

of the environment, by varying the temperature and humidity of the composites. Furthermore, the

change of the environment conditions of the tests could verify the claim that basalt fibers have

superior properties in terms of their dimensional stability when compared to glass fibers.

76

APPENDICES

77

APPENDIX A

MATLAB CODE FOR THERMOGRAVIMETRIC ANALYSIS

1

2

3

4

c l c

c l o s e

c l e a r

a l l

a l l

f i l e n a m e = ’ n o t d r i e d sample . t x t ’ ;

5 A = i m p o r t d a t a ( f i l e n a m e ) ;

6 %A = x l s r e a d ( f i l e n a m e ) ;

7 d e r i v = A ( : , 3 ) ;

8 T = A ( : , 1 ) ;

9

f i g = f i g u r e ;

10 D i s t a n c e =500;

11

[ peaks ,

l o c a t i o n s ] = f i n d p e a k s ( d e r i v , ’ MinPeakHeight ’

, 0 . 2 , ’ MinPeakDistance ’ , D i s t a n c e ) ;

12 h o l d on

13 %Temp=T ( l o c a t i o n s )

14 %D e r i v a t i v e = p ea ks

15 %[ pks ,

l o c s ]= f i n d p e a k s ( l o c a t i o n s , peaks , ’ Annotate ’ , ’

e x t e n t s ’ , ’ MinPeakHeight ’ , 0 . 5 ) ;

16 a= p l o t ( T , d e r i v ) ;

17 %s c a t t e r ( T ( l o c a t i o n s ) , d e r i v ( l o c a t i o n s ) , 4 0 , ’ f i l l e d ’ )

18

s c a t t e r ( T ( l o c a t i o n s ) , pe ak s )

19 n= l e n g t h ( pe ak s ) ;

20

i =1;

21 %h o r i z o n t a l =0* ones ( l e n g t h ( T ) ) ;
22 %p l o t ( T , h o r i z o n t a l )

78

23 Temp=T ( l o c a t i o n s ) ;

24 Pks= pe ak s ;

25 w h i l e i <=n

26 T e m p e r a t u r e =Temp ( i ) * ones ( l e n g t h ( Pks ) ) ;
27 H e i g h t =0: Pks ( i ) / ( l e n g t h ( T e m p e r a t u r e ) −1) : Pks ( i ) ;

28

29

30

p l o t ( Temperature , Height , ’−−’ )
t e x t ( T e m p e r a t u r e ( i ) , Pks ( i ) , i n t 2 s t r ( Temp ( i ) ) , ’

H o r i z o n t a l A l i g n m e n t ’ , ’ c e n t e r ’ , ’

V e r t i c a l A l i g n m e n t ’ , ’ bottom ’ )

i = i +1;

31 end

32 %p l o t ( p l o t

33 %d e r i v ( pks )

34 h o l d o f f

35 xlim ( [ 2 5 5 0 0 ] )

36 ylim ( [ 0 1 ] )

37 y l a b e l ( ’ Weight

l o s s

(%/C) ’ )

38 x l a b e l ( ’ T e m p e r t a t u r e (C) ’ )

39 g r i d

40

41

42

43

44

l e g e n d ( a , ’ Not d r i e d AFEX bamboo ’ )

f i g . P a p e r P o s i t i o n M o d e = ’ a u t o ’ ;

f i g _ p o s = f i g . P a p e r P o s i t i o n ;

f i g . P a p e r S i z e = [ f i g _ p o s ( 3 )

f i g _ p o s ( 4 ) ] ;

s a v e a s ( f i g , ’ \ \ t s c l i e n t \ home \ THESIS DOCUMENT\ images \

notdriedAFEX . pdf ’ )

79

45 %s a v e ( ’ \ \ t s c l i e n t \ home \ THESIS DOCUMENT\ images \

notdriedAFEX . pdf ’ )

80

APPENDIX B

ESTIMATION OF FIBER VOLUME FRACTION IN COMPOSITES WITH IMAGEJ

Motivation : Due to the inherent variability of the the dimensions and density of natural fibers,

the assumption of a cylindrical shape can yield erroneous calculations of the volume

fraction present in the fibers. Therefore, a first attempt to estimate the volume

fraction is presented in this section.

Procedure : The bamboo-reinforced composites were polished using the procedure described

in section 4.3. Then, the samples were observed under the light transmission mi-

croscope (Keyence, USA). The optical microscope needs to be calibrated and each

micrograph should show a scalebar. The images were saved as tiff (Tag Image File
Format) files for evaluation with the ImageJ software[56]. The micrographs were

evaluated using the following procedure:

– Open the desired file. Command: File→ Open...
– Improve contrast. Command: Process → Enhance contrast
– Convert image to scale of grey image type: Command Image → Type →

8bit

– Filter out wavelength of the image that correspond to the matrix. Command:
Image → Adjust → Threshold. The value for the threshold varies from
sample to sample. The user has to find the right setting to have adequate

results.

– Convert image to black and white image. Command: Process → Binary

→ Make binary

– Select option to measure area. Command: Analyze → Set measurements...

→ Select Area checkbox

81

– Calculate area (in units of pixels). Command: Analyze → Measure. This
last step will return the total dark area of the image. The total dark area of the

image, that corresponds to the fiber regions, is obtained by dividing this value

to the total area of the image (shown at the top left side of the image).

– Generate negative image. Command: Image → Lookup Table → Invert

LUT.

– Repeat area measure command. Compute area by dividing the value returned

by the software by the total area of the image (in pixels).

– By adding the two area measurements, the area fraction of the bamboo region

is computed.

Although the aforementioned procedure is straightforward, there is not one single

recipe for all samples since the contrast changes. Therefore it is advised to adjust

the optical parameters in ImageJ accordingly. This is specially important for com-

posites where the contrast between the matrix and the bamboo fibers is low. This

technique is not necessary for synthetic fibers because they are posses less varia-

tions in their diameter, shapes. Thus, in the case of basalt, the main use of this

procedure is to observe the fiber distribution. Also, the measurements for several

cross-sections of the same sample should be done in order to better estimate the

actual volume fraction. (see Table 5.1).

82

Original image

Bright-bamboo area

Dark-bamboo area Total area of bamboo

Figure 5.1 Example for the estimation of fiber volume by using ImageJ

83

1 mm Table 5.1 Example of bamboo-reinforced composite (AFEX-treated)

Sample

Count

Sample 1 -

Cross section 1

Sample 1 -

Cross section 2

Sample 1 -

Cross section 3

Sample 1 -

Cross section 4

Sample 1 -
Cross section

Average

164

169

130

295

-

Total area (in

pixels2)
98844

Average area
size in pixels2

602.707

99891

104381

106140

-

591.071

802.931

359.797

-

Area (%)

14.993

15.015

15.492

15.417

15.229

(SD=0.262)

84

APPENDIX C

BIODEGRADABLE POLYMERS AS MATRIX

The ultimate goal of this study is to promote the fabrication of high-performance composites with

environmental sustainability. For this purpose, the study of the environmentally-friendly resins

should be conducted. BASF has developed a series of acrylic resins that show interesting binding

properties where the byproduct of the curing process is water. The reports of this product state that

the cross-linking is achieved by the reaction of polycarboxylic acids with multifunctional alcohols.
The only by-product caused by curing is water[7]. In a first attempt to characterize this product, a

series of curing processes were conducted during this work. The neat resins exhibited a foam-like

structure after curing (see figure 5.2, making them very difficult to test them for their mechanical

aspects. The curing process for this sample was: 80ºC for two hours in a convection oven followed

by a temperature rate of 10ºC/min to 160ºC. The temperature was held at 160ºC for 20 minutes.

The compression-moulded composite fabrication with AFEX-treated yielded better results, as

shown in figure 5.3. For this specimen, a beaker with the 950L resin was heated for one hour at

80ºC. After this time, the resin was poured into a pre-heated mold at 160ºC and pressure of approx.

100lb was applied for 20 minutes. The system was allowed to cure naturally.

85

Figure 5.2 Foam structure of neat Acrodur 950L (BASF).

Figure 5.3 Compression-moulded Acrodur-950L (BASF) resin with AFEX-treated bamboo fibers.

86

10mm 10mm BIBLIOGRAPHY

87

BIBLIOGRAPHY

[1] Agarwal, B. D., Broutman, L. J., and Chandrashekhara, K. (2017). Analysis and performance

of fiber composites. John Wiley & Sons.

[2] Akinlabi, E. T., Anane-Fenin, K., and Akwada, D. R. (2017). Bamboo: The Multipurpose

Plant. Springer.

[3] Amada, S. and Untao, S. (2001). Fracture properties of bamboo. Composites Part B: Engi-

neering, 32(5):451–459.

[4] Ashby, M. and Bréchet, Y. (2003). Designing hybrid materials. Acta materialia, 51(19):5801–

5821.

[5] Balasubramanian, M. (2013). Composite materials and processing. CRC press.

[6] Basaltex (2015). Basalt products. http://www.basaltex.com/en/detail_detail20_32.aspx. Ac-

cessed: 2018-03-30.

[7] BASF (2017). Acrodur acrylic resins. https://www.basf.com/us/en/products-and-industries/

General-Business-Topics/dispersions/Products/acrodur.html. Accessed: 2018-01-15.

[8] Bendat, J. S. and Piersol, A. G. (2011). Random data: analysis and measurement procedures,

volume 729. John Wiley & Sons.

[9] Block, E. and Smith, P. A. (2018). Amine. https://www.britannica.com/science/amine. Ac-

cessed: 2018-03-30.

[10] Brik, V. B. (1997). Basalt fiber composite reinforcement for concrete. Technical report.

[11] Buschle-Diller, G., Fanter, C., and Loth, F. (1999). Structural changes in hemp fibers as a
result of enzymatic hydrolysis with mixed enzyme systems. Textile research journal, 69(4):244–
251.

[12] Carlsson, L. A., Adams, D. F., and Pipes, R. B. (2002). Experimental characterization of

advanced composite materials. CRC press.

[13] Carrillo, F., Colom, X., Sunol, J., and Saurina, J. (2004). Structural ftir analysis and thermal
characterisation of lyocell and viscose-type fibres. European Polymer Journal, 40(9):2229–
2234.

[14] Deák, T. and Czigány, T. (2009). Chemical composition and mechanical properties of basalt

and glass fibers: a comparison. Textile Research Journal, 79(7):645–651.

[15] Dhand, V., Mittal, G., Rhee, K. Y., Park, S.-J., and Hui, D. (2015). A short review on basalt

fiber reinforced polymer composites. Composites Part B: Engineering, 73:166–180.

[16] Dixon, P., Ahvenainen, P., Aijazi, A., Chen, S., Lin, S., Augusciak, P., Borrega, M., Sved-
ström, K., and Gibson, L. (2015). Comparison of the structure and flexural properties of moso,
guadua and tre gai bamboo. Construction and Building Materials, 90:11–17.

88

[17] Dixon, P. G. and Gibson, L. J. (2014). The structure and mechanics of moso bamboo material.

Journal of the Royal Society Interface, 11(99):20140321.

[18] Downey, M. A. and Drzal, L. T. (2014a).

Toughening of aromatic epoxy polymers
via aliphatic epoxy monomer addition: Optimized fiber-reinforced polymer composites for
lightweighting. SPE Automotive Composites Conference & Exhibition Troy, MI, USA.

[19] Downey, M. A. and Drzal, L. T. (2014b). Toughening of aromatic epoxy via aliphatic epoxy

copolymers. Polymer, 55(26):6658–6663.

[20] Drzal, L. T., Rich, M. J., and Lloyd, P. F. (1983). Adhesion of graphite fibers to epoxy

matrices: I. the role of fiber surface treatment. The Journal of Adhesion, 16(1):1–30.

[21] Farris, S., Schaich, K. M., Liu, L., Piergiovanni, L., and Yam, K. L. (2009). Development of
polyion-complex hydrogels as an alternative approach for the production of bio-based polymers
for food packaging applications: a review. Trends in food science & technology, 20(8):316–332.

[22] Faruk, O., Bledzki, A. K., Fink, H.-P., and Sain, M. (2012). Biocomposites reinforced with

natural fibers: 2000–2010. Progress in polymer science, 37(11):1552–1596.

[23] Gong, X., Arthur, J., and Penn, L. (2001). Strain rate effect in the single-fiber-fragmentation

test. Polymer composites, 22(3):349–360.

[24] Gupta, V., Drzal, L., Adams, W., and Omlor, R. (1985a). An electron microscopic study of

the morphology of cured epoxy resin. Journal of materials science, 20(10):3439–3452.

[25] Gupta, V., Drzal, L., Lee, C.-C., and Rich, M. (1985b). The temperature-dependence of
some mechanical properties of a cured epoxy resin system. Polymer Engineering & Science,
25(13):812–823.

[26] Gupta, V., Drzal, L., and Rich, M. (1985c). The physical basis of moisture transport in a

cured epoxy resin system. Journal of Applied Polymer Science, 30(11):4467–4493.

[27] Hao, L. and Yu, W. (2010). Evaluation of thermal protective performance of basalt fiber

nonwoven fabrics. Journal of thermal analysis and calorimetry, 100(2):551–555.

[28] Hariharan, A. B. A. and Khalil, H. A. (2005). Lignocellulose-based hybrid bilayer laminate
composite: Part i-studies on tensile and impact behavior of oil palm fiber-glass fiber-reinforced
epoxy resin. Journal of Composite Materials, 39(8):663–684.

[29] Harish, S., Michael, D. P., Bensely, A., Lal, D. M., and Rajadurai, A. (2009). Mechanical

property evaluation of natural fiber coir composite. Materials characterization, 60(1):44–49.

[30] Holtzapple, M. T., Jun, J.-H., Ashok, G., Patibandla, S. L., and Dale, B. E. (1991). The
ammonia freeze explosion (afex) process. Applied Biochemistry and Biotechnology, 28(1):59–
74.

[31] Hyer, M. W. (2009). Stress analysis of fiber-reinforced composite materials. DEStech Publi-

cations, Inc.

89

[32] Jain, S., Kumar, R., and Jindal, U. (1992). Mechanical behaviour of bamboo and bamboo

composite. Journal of Materials Science, 27(17):4598–4604.

[33] Jamshaid, H. and Mishra, R. (2016). A green material from rock: basalt fiber–a review. The

Journal of The Textile Institute, 107(7):923–937.

[34] Jones, R. M. (2014). Mechanics of composite materials. CRC press.

[35] Kelly, A. and Tyson, a. W. (1965). Tensile properties of fibre-reinforced metals: copper/tung-
sten and copper/molybdenum. Journal of the Mechanics and Physics of Solids, 13(6):329–350.

[36] Kim, B. W. and Nairn, J. A. (2002). Observations of fiber fracture and interfacial debond-
ing phenomena using the fragmentation test in single fiber composites. Journal of Composite
Materials, 36(15):1825–1858.

[37] Kretsis, G. (1987). A review of the tensile, compressive, flexural and shear properties of

hybrid fibre-reinforced plastics. Composites, 18(1):13–23.

[38] Kutz, M. (2015). Mechanical Engineers’ Handbook, Volume 1: Materials and Engineering

Mechanics. John Wiley & Sons.

[39] Lee, J. and Drzal, L. T. (2005). Surface characterization and adhesion of carbon fibers to

epoxy and polycarbonate. International journal of adhesion and adhesives, 25(5):389–394.

[40] Liu, Q., Shaw, M. T., Parnas, R. S., and McDonnell, A.-M. (2006). Investigation of basalt
fiber composite mechanical properties for applications in transportation. Polymer composites,
27(1):41–48.

[41] Ma, P. C., Kim, J.-K., and Tang, B. Z. (2006). Coating carbon nanotubes with silver nanopar-
ticles to get conductive nanocomposites. In Electronic Materials and Packaging, 2006. EMAP
2006. International Conference on, pages 1–4. IEEE.

[42] MAFIC (2018). Basalt continuous fibers. http://www.maficbasalt.com/blog/wp-content/

uploads/2016/06/Mafic-Assembled-Roving-TDS-1.pdf. Accessed: 2018-04-01.

[43] Malherbe, S. and Cloete, T. E. (2002). Lignocellulose biodegradation: fundamentals and

applications. Reviews in Environmental Science and Biotechnology, 1(2):105–114.

[44] Mannan, S., Knox, J. P., and Basu, S. (2017). Correlations between axial stiffness and mi-

crostructure of a species of bamboo. Royal Society open science, 4(1):160412.

[45] May, C. (1987). Epoxy resins: chemistry and technology. CRC press.

[46] Militky, J. and Kovacic, V. (1996). Ultimate mechanical properties of basalt filaments. Textile

Research Journal, 66(4):225–229.

[47] Mishra, S., Mohanty, A., Drzal, L., Misra, M., Parija, S., Nayak, S., and Tripathy, S. (2003).
Studies on mechanical performance of biofibre/glass reinforced polyester hybrid composites.
Composites Science and Technology, 63(10):1377–1385.

90

[48] Mohanty, A. K., Misra, M., and Drzal, L. T. (2005). Natural fibers, biopolymers, and bio-

composites. CRC press.

[49] National Center for Biotechnology Information (2000). M-phenylenediamine.

https://

pubchem.ncbi.nlm.nih.gov/compound/7935. Accessed: 2018-03-30.

[50] Parameswaran, N. and Liese, W. (1976). On the fine structure of bamboo fibres. Wood

Science and Technology, 10(4):231–246.

[51] Park, J.-M., Shin, W.-G., and Yoon, D.-J. (1999). A study of interfacial aspects of epoxy-
based composites reinforced with dual basalt and sic fibres by means of the fragmentation and
acoustic emission techniques. Composites Science and Technology, 59(3):355–370.

[52] Ray, D., Sarkar, B. K., Rana, A., and Bose, N. R. (2001). The mechanical properties of
vinylester resin matrix composites reinforced with alkali-treated jute fibres. Composites Part A:
applied science and manufacturing, 32(1):119–127.

[53] Resource Fiber LLC (2018).

Bamboo products.

http://www.resource-fiber.com/

why-bamboo. Accessed: 2018-04-01.

[54] Rich, M., Drzal, L., Hunston, D., Holmes, G., and McDonough, W. (2002). Round robin
assessment of the single fiber fragmentation test. In Procedings of the American Society for
Composites 17th Technical Conference.

[55] Santulli, C. (2004). Critical length measurements in carbon fibers during single fiber frag-

mentation tests using acoustic emission. Journal of materials science, 39(8):2905–2907.

[56] Schneider, C. A., Rasband, W. S., and Eliceiri, K. W. (2012). Nih image to imagej: 25 years

of image analysis. Nature methods, 9(7):671.

[57] Sella Kapu, N. and Trajano, H. L. (2014). Review of hemicellulose hydrolysis in softwoods

and bamboo. Biofuels, Bioproducts and Biorefining, 8(6):857–870.

[58] Shao, S., Jin, Z., Wen, G., and Iiyama, K. (2009). Thermo characteristics of steam-exploded

bamboo (phyllostachys pubescens) lignin. Wood science and technology, 43(7-8):643–652.

[59] Sigma-Aldrich (2000). m-phenylenediamine p23954.

https://www.sigmaaldrich.com/

catalog/product/aldrich/p23954?lang=en&region=US. Accessed: 2018-03-30.

[60] Singh, S., Simmons, B. A., and Vogel, K. P. (2009). Visualization of biomass solubilization
and cellulose regeneration during ionic liquid pretreatment of switchgrass. Biotechnology and
Bioengineering, 104(1):68–75.

[61] Singha, K. (2012). A short review on basalt fiber. International Journal of Textile Science,

1(4):19–28.

[62] Sreekala, M., George, J., Kumaran, M., and Thomas, S. (2002). The mechanical performance
of hybrid phenol-formaldehyde-based composites reinforced with glass and oil palm fibres.
Composites science and technology, 62(3):339–353.

91

[63] Standard, ASTM (2010). D790 (2010) standard test methods for flexural properties of unre-
inforced and reinforced plastics and electrical insulating materials. ASTM International, West
Conshohocken.

[64] Subramanian, R. and Austin, H. (1980). Silane coupling agents in basalt-reinforced polyester

composites. International Journal of Adhesion and Adhesives, 1(1):50–54.

[65] Thwe, M. M. and Liao, K. (2003). Durability of bamboo-glass fiber reinforced polymer

matrix hybrid composites. Composites science and technology, 63(3-4):375–387.

[66] Tokoro, R., Vu, D. M., Okubo, K., Tanaka, T., Fujii, T., and Fujiura, T. (2008). How to
improve mechanical properties of polylactic acid with bamboo fibers. Journal of materials
science, 43(2):775–787.

[67] Wade, L. G. (2011). Epoxide. https://www.britannica.com/science/epoxide. Accessed: 2018-

03-30.

[68] Wei, B., Cao, H., and Song, S. (2011). Surface modification and characterization of basalt
fibers with hybrid sizings. Composites Part A: Applied Science and Manufacturing, 42(1):22–
29.

[69] Yeasmin, L., Ali, M. N., Gantait, S., and Chakraborty, S. (2015). Bamboo: an overview on

its genetic diversity and characterization. 3 Biotech, 5(1):1–11.

[70] Youssefian, S. and Rahbar, N. (2015). Molecular origin of strength and stiffness in bamboo

fibrils. Scientific reports, 5:11116.

[71] Zhang, J., Feng, L., Wang, D., Zhang, R., Liu, G., and Cheng, G. (2014). Thermogravimetric
analysis of lignocellulosic biomass with ionic liquid pretreatment. Bioresource technology,
153:379–382.

92