Characterization of CFRP and GFRP Composite Materials at High Strain Rate Tensile Loading PDFDrive (SFILE

CHARACTERIZATION OF CFRP AND GFRP COMPOSITE MATERIALS AT
HIGH STRAIN RATE TENSILE LOADING
A Thesis by
Anand B. Deshpande
Bachelor of Engineering, Dr. B.A.M. University, India, 2002
Submitted to the Department of Mechanical Engineering

And the faculty of Graduate School of
Wichita State University
in partial fulfillment of
the requirements for the degree of
Master of Science
December 2006
© Copyright 2006 by Anand B. Deshpande
All Rights Reserved
Note that thesis and dissertation work is protected by copyright, with all rights reserved. Only
the author has the legal right to publish, produce, sell, or distribute this work. Author
permission is needed for others to directly quote significant amounts of information in their
own work. Limited amounts of information cited, paraphrased, or summarized from the work
may be used with proper citation of where to find the original work.
CHARACTERIZATION OF CFRP AND GFRP COMPOSITE MATERIALS AT HIGH
STRAIN RATE TENSILE LOADING
The following faculty members have examined the final copy of this thesis for form and
content and recommend that it be accepted in partial fulfillment of the requirements for
the degree of Master of Science, with a major in Mechanical Engineering.
__________________________________
Hamid M. Lankarani, Committee Chair
___________________________________
K. Suresh Raju, Co-Chair and Committee Member
___________________________________
Bob Minaie, Committee Member
iii
DEDICATION
To my parents and friends

ACKNOWLEDGEMENTS
I would like to present my sincere gratitude towards my advisor Dr. Hamid
Lankarani for his continuous support and guidance in completing my thesis. In all the
difficult times he supported me like a friend and cared like a parent and truly showed me
that creativity can only flourish if the mind is set free. I learned a great deal of things
from him in my entire journey of MS.
It is difficult to overstate my gratitude to my co-advisor Dr. K. Suresh Raju, who
has always helped me understanding technical details. For his valuable help, patience,
encouragement, I owe him the deepest gratitude. It’s hard to express the thanks
towards him in words.
I thank Dr. Bob Minaie for reviewing my manuscript and helping me to improve
this thesis. Special gratitude is given to Dr. Chandrashekhar Thorbole for his timely
guidance and support.
Most of my experiments would not have been possible without help from NIAR
laboratories.
I owe the deepest appreciation to my colleagues Vivek Mariyanna, Satish
Dandayudhapani, Sandeep Shetty, Juan Acosta Felipe, Manoj Varma, and Mohan
Ghimire and engineers from the National Institute for Aviation Research, especially for
Kim Leng Poon, Siddartha Arood for their kind assistance in carrying out experiments
and helping with various applications.
I am indebted to my many colleagues and friends for providing a stimulating and
fun environment to learn and grow. I wish to thank my best friends from Wichita State
University; Vinay Bhamare, Amit Yeole, Ashwin Sheshadri, Divakara Basavaraju, and
v
Govind Pillai for helping me to get through difficult times and for their emotional support,
entertainment and caring.
Lastly and most importantly, I want to thank my parents who have been a source
of encouragement and inspiration throughout my life.
Finally, I would like to thank all the direct and indirect supports that helped me
complete this thesis.
vi
ABSTRACT
High strength-to-weight ratio, directional strength and stiffness are the significant
factors, forcing polymer composites into the Aerospace, Marine and Automotive
industries. Due to these major factors fuel efficiency and crashworthiness properties are
the significant outcomes from use of these advanced materials. The work presented in
this thesis investigates the experimental study of tensile properties of in-plane polymer
matrix composite materials in Quasi-Static and High Strain Rate tests. Behavior of
Glass fiber-reinforced (GFRP) and Carbon fiber reinforced (CFRP) composite materials
is studied. The test coupons are balanced and symmetric in fiber orientation with
respect to the test direction. With a MTS 810 high stroke rate test machine the related
experiments are carried out to find out the mechanical properties of test specimens. The
test coupons are tested for quasi-static test of 0.05 in/min and at high speed of up to
500 in/s. All specimens are tested till total failure to point up the effect of high strain rate
on failure strength.
In this work, a new method to obtain stress-strain curves for the tensile tests is
proposed. The strain rate nature of composite laminates in tensile loadings clearly show
that unlike in metals these materials do not exhibit the constant strain rate behavior in
case of high strain rate tests. Throughout the test, the strain rate values change due to
the dynamics of the system and directional stiffness of the composite laminates. In case
of 0º fiber oriented specimens, the fiber properties dominate the matrix properties as
fiber strength is much higher than that of matrix materials. For different fiber orientations
of the laminates the strain rate varies for the same stroke rate tests as the matrix
material starts playing role in case of higher fiber angles.
vii
The outcome shows that high stroke rates have a considerable effect on the
properties of the composites materials. The increment of the failure strength with high
stroke rate is proportional to the strain rate. In the future developments the stress-strain
curves obtained from these various tensile tests can be used to insert in a finite element
code to develop a material model for computational simulations.
viii
TABLE OF CONTENTS
Chapter Page
1. INTRODUCTION .................................................................................................. 1
1.1 Simple Tensile Test ................................................................................... 2

1.2 Important Factors in Tensile Testing of Composite Materials .................... 4
1.2.1 Deformation of Mechanical Links .................................................. 4
1.2.2 Stress Concentration .................................................................... 4
1.2.3 Alignment Precision ...................................................................... 5
1.2.4 Strain Measurement...................................................................... 6
2. BACKGROUND AND LITERATURE REVIEW ..................................................... 7
2.1 General Terms and Purpose of Tensile Testing ........................................ 7

2.2 Literature Review ....................................................................................... 8
3. PROBLEM STATEMENT AND SOLUTION APPROACH .................................. 20
3.1 Introduction .............................................................................................. 20

3.2 Statement of Problem .............................................................................. 21
3.3 Solution Approach ................................................................................... 22
4. MATERIAL SPECIFICATION AND TEST SET-UP ............................................ 23
4.1 Material Specifications ............................................................................. 23

4.2 Static Testing ........................................................................................... 24
4.3 High Speed Testing ................................................................................. 25
5. EXPERIMENTAL RESULTS .............................................................................. 29
5.1 Test Results of Newport PWCF 0º specimens ......................................... 29

5.2 Test Results of Newport PWCF 15º, 30º, and 45º specimens ................. 30
5.3 Stress-Strain Plots for Newport PWCF (0º, 15º, 30º, 45º) ........................ 33
Specimens in Static Loading
5.4 Stress-Strain Behavior for Newport PWCF (0º) Specimens in ................. 35
Higher Stroke Rates
5.5 Failure Patterns of Newport PWCF (0º) Specimens in Different .............. 38
Stroke Rates
6. METHODOLOGY FOR ISO STRAIN RATE CHARACTERIZATION.................. 41
6.1 Introduction .............................................................................................. 41

6.2 Variable Strain Rate Behavior .................................................................. 42
ix
TABLE OF CONTENTS (continued)
Chapter Page
7. CONCLUSIONS AND FUTURE RECOMMENDATIONS ................................... 49
7.1 Conclusions ............................................................................................. 49

7.2 Future Recommendations........................................................................ 50
REFERENCES .............................................................................................................. 51
APPENDIX ................................................................................................................. 55
x
LIST OF TABLES
Table Page
1. Summary of Systems Used for Dynamic Loading ................................................ 9
2. Summary of Test Data for Newport PWCF Material (0º) .................................... 29
3. Summary of Test Data for Newport PWCF Material (15º, 30º, 45º) .................... 30
4. Summary of Test Data for Newport Glass Fiber Material (0º, 15º, 30º, 45º) ....... 70
xi
LIST OF FIGURES
Figure Page
1. General specimen geometry................................................................................. 3
2. Modified specimen geometry .............................................................................. 21
3. Static Test Set-up ............................................................................................... 24
4. Set-up of mechanical grips and slack inducer mechanism ................................. 26
5. Strength comparison for Newport PWCF (0º) specimens on the ........................ 31
basis of stroke rates
6. Strength distribution over the range of fiber orientation angles for ..................... 32
Newport PWCF material in quasi-static stroke rates
7. Stress-Strain behavior in quasi-static test for Newport PWCF (0º) material ....... 33
8. Stress-Strain behavior in quasi-static test for Newport (15º) material ................. 34
9. Stress-Strain behavior in quasi-static test for Newport PWCF (30º) material ..... 34
10. Stress-Strain behavior in quasi-static test for Newport PWCF (45º) material ..... 35
11. Stress-Strain behavior of Newport PWCF (0º) material in 1 in/s stroke rate ....... 36
12. Stress-Strain behavior of 0º Newport PWCF in 10 in/s stroke rate ..................... 36
13. Stress-Strain behavior of Newport PWCF (0º) material in 100 in/s stroke rate ... 37
16. Failure mode of Newport PWCF (0º) material in static stroke rate ..................... 38
17. Failure mode of Newport PWCF (0º) material in 1 in/s stroke rate ..................... 39
18. Failure mode of Newport PWCF (0º) material in 10 in/s stroke rate ................... 39
19. Failure mode of Newport PWCF (0º) material in 100 in/s stroke rate ................. 39
xii
LIST OF FIGURES (continued)
Figure Page
20. Failure mode of Newport PWCF (0º) in 250 in/s stroke rate ............................... 40
22. Strain-Time data for Newport PWCF (0º) material for 100 in/s test .................... 43
23. Strain Rate-Strain data for Newport PWCF (0º) material for 100 in/s test .......... 44
24. Family of Stress-Strain Rate curves for different strain intervals ........................ 45
25. Family of Stress-Strain Rate curves for different strain intervals ........................ 46
26. Family of Stress-Strain curves for Newport PWCF (0º) material at ................... 47
different strain rates
27. Family of Stress-Strain curves for Newport Glass fiber (0º) material at .............. 48
different strain rates
28. Stress-Strain behavior of Newport PWCF (15º) in 1 in/s stroke rate .................. 56
31. Stress-Strain behavior of Newport PWCF (15º) in 10 in/s stroke rate ................ 58
34. Stress-Strain behavior of Newport PWCF (15º) in 100 in/s stroke rate .............. 59
xiii
Figure Page
43. Failure mode of Newport PWCF (15º) in static stroke rate ................................. 64
44. Failure mode of Newport PWCF (15º) in 1 in/s stroke rate ................................. 64
46. Failure mode of Newport PWCF (15º) in 100 in/s stroke rate ............................. 65
xiv
Figure Page
61. Stress-Strain behavior of 0º Newport Glass-Fiber in Quasi-static test ................ 71
62. Stress-Strain behavior of 15º Newport Glass-Fiber in Quasi-static test .............. 72
65. Stress-Strain behavior of Newport Glass-fiber (0º) in 1 in/s stroke rate .............. 74
66. Stress-Strain behavior of Newport Glass-fiber (0º) in 10 in/s stroke rate ............ 75
67. Stress-Strain behavior of Newport Glass-fiber (0º) in 100 in/s stroke rate .......... 75
72. Stress-Strain behavior of Newport Glass-fiber (15º) in 100 in/s stroke rate ........ 78
80. Stress-Strain behavior of Newport Glass Fiber (45º) in 1 in/s stroke rate ........... 83
xv
Figure Page
85. Failure mode of Newport Glass Fiber (0º) in static stroke rate ........................... 86
86. Failure mode of Newport Glass Fiber (0º) material in 1 in/s stroke rate.............. 86
87. Failure mode of Newport Glass Fiber (0º) in 10 in/s stroke rate ......................... 86
88. Failure mode of Newport Glass Fiber (0º) in 100 in/s stroke rate ....................... 87
89. Failure mode of Newport Glass Fiber (0º) material in 250 in/s stroke rate.......... 87
91. Failure mode of Newport Glass Fiber (15º) in static stroke rate ......................... 88
94. Failure mode of Newport Glass Fiber (15º) in 100 in/s stroke rate ..................... 89
xvi
Figure Page
109. Family of Stress-Strain Curves for Newport PWCF (30º) material ...................... 94
110. Family of Stress-Strain Curves for Newport Glass Fiber (30º) material ............. 94
111. Family of Stress-Strain Curves for Newport PWCF (45º) material ..................... 95
112. Family of Stress-Strain Curves for Newport Glass Fiber (45º) material .............. 95
xvii
CHAPTER 1
INTRODUCTION
A composite material is formed by the combination of two or more different
materials to form a new material with improved material properties. The oldest
composite material is wood consisting of cellulose fibers in a lignin matrix. When
composites are selected over traditional materials such as metal alloys it is usually
because of one or more of the following advantages:
1. Weight:
Light weight
Weight distribution
2. Strength and Stiffness:
High strength-to-weight ratio
Directional strength and/or stiffness
3. Surface Properties:
Corrosion resistance
Weather resistance
Tailored surface finish
4. Thermal Properties:
Low thermal conductivity
Low coefficient of thermal expansion
1
Fiber-reinforced composite materials are characterized by specific stiffness and
strength exceeding that of similar metal structures. With emphasis on light weight
vehicles, the use of composite materials in aerospace and automotive structures has
created a need to further understand the energy absorption characteristics of composite
materials.
High strength and light weight remain the best combination that is forcing
composite materials in the new field from last few decades. Even if bulk of a metal or
other material provides a certain degree of safety but at the same time it can have
heavy weight parts and consequently high energy transfer in case of safety in terms of
weight. Composite materials are extensively used in aerospace, marine and automotive
industries due to the need for increased fuel efficiency, corrosion resistance and fatigue
resistance. In recent years of materials research work high strain rate testing of
laminated composites gained a significant importance as these materials are
prominently used in lightweight structural applications and there are many cases in
which the mechanical properties of composite materials are notably reliant on the strain
rate. The ultimate strength for composite materials show significant increment
compared to the value under static loading conditions.
1.1 Simple Tensile Test
A thin flat strip of material having a constant rectangular cross section, as shown
in Figure1 is mounted in the grips of a mechanical testing machine and monotonically
loaded in tension while recording the load. The ultimate strength of the material can be
determined from the maximum load carried before failure. If the coupon strain is
2
monitored with strain or displacement transducers then the stress-strain response of the
material can be determined [1].
Tab Area Gage Section
Overall Length
Thickness Tab Bevel Angle
Figure 1. General specimen geometry.
Tensile testing is the most fundamental type of mechanical testing and applies a
proof load to a specimen past the yield point to failure. This test method is intended to
construct tensile property data for specified material, structural design and analysis. As
per the ASTM standards for tensile testing of the composite materials, the test
specimens are balanced and symmetric laminates towards to the test direction.
Influential factors for the tensile response of the composites are: materials, methods of
material preparation, ply stacking sequence, specimen conditioning, specimen
alignment and gripping and condition of testing. Specimens may be dog bone tensile
bars or dumbbell shape or straight sided specimens with end tabs. Any tough adhesive
system that can meet the test environments can be used with a minimum, uniform
bondline thickness to reduce the undesirable stress concentrations while bonding the
tabs to the test coupons [1]. Tensile tests provide following mechanical properties in the
test direction:
3
Ultimate tensile strength
Ultimate tensile strain
Young’s Modulus
Poisson’s ratio
1.2 Important Factors in Tensile Testing of Composite Materials:
1.2.1 Deformation of Mechanical Links
High performance composites, such as those made from carbon or ceramic
fibers, are generally very stiff and strong, but also quite brittle, making them especially
sensitive to non-uniform loading during axial tests. This means that the testing
instrument requires high force capacity and a very stiff design. If the load frame lacks
sufficient stiffness, it will absorb too much energy as the specimen is loaded. Then, as
the specimen begins to fail, energy will be transferred from the load frame back to the
specimen, causing premature failure and giving an erroneous maximum strength value
[3].
1.2.2 Stress Concentration
Some of the most common tests performed on composite materials are designed
to determine tensile properties. However, preparation of tensile specimens is
particularly challenging. Because most composites have a high ratio of tensile strength
to shear strength, the tab area of dog bone geometry tends to shear off. If a straight-
sided specimen is used, stress concentrations from gripping generally cause failure.
Therefore, many composite specimens require double tabs to be bonded to both sides
at the ends. The tabs distribute gripping stresses and prevent specimen failure caused
by grip jaws damaging the specimen's surface [16]. Specimen bending during tensile
4
tests can arise from misalignment of the grip or the specimen. The testing machine also
must have good repeatability of grip alignment from test to test to ensure that stresses
are uniform across the specimen's width. Considering that the material is strong, the
necessary clamping forces can be significant so that the specimen will not slip through
the grips of the testing machine. The problem then is how to reduce these required
clamping forces, and maintain them from dropping the precise tensile strength of the
specimen. One solution to this problem is to build the specimen as thin as theoretically
possible while maintaining it as representative of the material specification. Specimen
width is not an issue in this case because both gripping area and applied force fluctuate
in direct proportion to the specimen width. By forcing these restrictions, it is always
beneficial to get rid of, or at least minimize the stress concentrations introduced by the
grips. Generally the end tabs used are of composite materials only due to the ability of
machining the tabs and test material at the same time. The most recent version of
ASTM D 3039 [1] has revised the use of end tabs with reinforcement at ± 45°. The
reason for specifying the ± 45° material is that it imposes less constraint on the
specimen in both the longitudinal and lateral directions [27].
1.2.3 Alignment Precision
Many composites are designed to be much stronger in one orientation than
another. Precise alignment in the test fixture is vital, because even a slight angle can
produce an unexpected failure mode and invalidate the test. For example, a
misalignment might produce a transverse failure rather than the intended axial failure,
because the transverse strength of the material might be extremely low relative to its
5
axial strength. A load frame with high lateral stiffness provides better alignment and
reduces stress concentrations [3].
1.2.4 Strain Measurement
Strain measurement of composites is more difficult than with conventional
materials. Because composites release a large amount of energy at failure,
conventional clip-on extensometers can be damaged. Bonded strain gauges are often
required to measure bending or determine Poisson's Ratio. A test coupon should
include a region of uniform stress over which strain measurements can be made i.e.
gage length and in that region the expected failure should occur [3].
CHAPTER 2
6
BACKGROUND AND LITERATURE REVIEW
2.1 General Terms and Purpose of Tensile testing
Tensile tests are performed for several reasons. The results of these tests are
used for selecting the proper materials for engineering applications. Tensile properties
are measured during development of new materials so that new materials can be
developed and compared. Tensile properties of composites are often measured to
predict the material behavior under loading conditions other than uni-axial tension.
Strength of the materials is always primary concern in tensile tests. Strength can be
interpreted as the stress necessary to cause plastic deformation or the maximum stress
that materials can withstand [4].
Tensile testing of materials include following main factors as primary concerns:
Test specimens and testing equipment.
Stress-Strain curves and corresponding modulus.
Test methodology and data analysis.
High strain rate tensile testing is necessary to understand the material response
under dynamic loading conditions. Strain rates ranges between 100 s -1 to 104 s-1 can
occur in many practical events like foreign object impact damage, blast loadings,
structural impacts etc. The behavior of materials under high strain rate tensile loads
may be considerably different than that of static loadings. Strain rate sensitivity also
depends upon whether engineering or true strain formulations are used because local
stress concentrations such as necking are normally not present in high rate loadings [4].
2.2 Literature Review
7
One of the aspects of structural engineering that is and will be always appealing
is destructive testing. At different high stroke rates the performance of the fiber-
reinforced composite materials has been experimentally calculated in many conditions
mainly under tensile and compressive loads. These tests are carried out with Split
Hopkinson Pressure Bar (SHPB) method, as it depicts the properties of the materials
over a wide variety of strain rates [2, 11, and 21]. The experimental method used to
determine the dynamic properties depends on the range of strain rate required. The
easiest system is the Charpy pendulum with special fixtures for tensile tests. This
system is used in many laboratories but the drawback behind this system is limited
range of strain rates. The Falling Weight method can also produce accurate results but
this technique can also become complicated in data interpretation due to the stress
wave reflections in the fixtures and the specimen. The most widely used technique for
obtaining direct determination of material properties at higher strain rates, the split
Hopkinson bar, was introduced by Herbert Kolsky [2]. Table 1 summarizes the various
instrument types and corresponding strain rate ranges for high rate tests of composite
materials.
TABLE 1
8
SUMMARY OF SYSTEMS USED FOR DYNAMIC LOADING
Strain Rate Range

Test System Test Type
(sˉ¹)
Charpy Pendulum Tensile <100
Falling Weight Tensile <10
Hydraulic Machines Tensile, Shear 10ˉ³ to 50
Hopkinson Bar Tensile, Shear & Compression 100 to 3000
In this technique a short material specimen was deformed at a high rate while it
is essentially under a constant state of stress till it breaks. This technique was not
widely used for tensile tests in its early days because in the simple form of this test set
up the specimen could be used to get loaded only in compression.
In later research in this area researchers found out a separate test set up of
Kolsky bar (SHPB) technique for tensile testing of materials. In attempt to determine
the mechanical properties of composite materials under dynamic tensile loads, a review
of techniques was given by Harding and Welsh [6]. Difficulties encountered in the
design of a satisfactory tensile impact testing machine for composite materials were
discussed and a new method using modified version of SHPB was suggested. In the
standard tensile version of Kolsky bar apparatus the input loading bar became the
weigh-bar tube within which the output bar slides freely. Dynamic stress strain curves
for unidirectionally reinforced carbon epoxy composite in which failure occurs in less
than 30 µs at a mean strain rate of about 400s-1 were presented and their validity was
established by the authors. They also did an extension of the technique to allow the
testing of woven reinforced glass/epoxy composites and dynamic stress-strain curves
obtained for which the time to failure approach 100 µs and the average strain rate was
9
around 1000s-1. Comparative stress-strain curves at low and intermediate rates of strain
were obtained and compared.
Many researchers use and thus show that SHPB technique is the most popular.
Although, use of split bars for material testing needs very intricate study of wave
propagation in composite materials. These factors have to be taken into account to
interpret the correct result values of tests at range of any stroke rate.
Use of servo hydraulic machines can minimize the issues of wave propagation,
which is a great help in producing better and very much repetitive results [7, 16, and
26]. Pardo and Baptiste [16] carried out tests of unidirectional E-Glass/polyester
composite specimens on a Schenck high strain rate hydraulic test machine to explore
the effect of strain rate on tensile properties. Different fiber orientations of 0°, 90° and
5% weft fibers combined with unidirectional fibers were tested from velocities of quasi-
static up to 20 m/s. Difficulties faced due to high velocities were solved by modifying the
specimen geometry and reducing the shock at the grip engagement link, in order to
obtain reliable material properties. 5% volume weft fibers were present in the
specimens and these were the source of the damage development. 0° and 90°
specimens with and without weft fibers were tested. The failure behavior of the pure
unidirectional fibers was linear and brittle. Also the rate effect on 90° composite was
significant on maximum stress but the mechanical characteristics of the 0° composite
evolved weakly.
Hayes and Adams [5] conducted various tests at various tests speeds and load
levels to characterize the tensile impact behavior and rate sensitive materials properties
of unidirectional glass/epoxy and graphite/epoxy composites. An instrumented tensile
10
impact test system with partial loading capabilities has been developed and tested. A
standard Pendulum type impact testing machine was modified for these tests. The
various test speeds used in this research were 2.7, 3.4, 3.8, 4.5, and 4.9 m/s. The
micromechanics of fracture within the composite caused by impact were studied, using
partially loaded impact test specimens and scanning electron microscopy. The
glass/epoxy material exhibited a considerable increase in the strength and modulus as
the strain rates were increased but in case of the graphite/epoxy material system the
results were opposite to that of glass/epoxy. There was decrease in the strength of
graphite/epoxy test coupons.
Peterson and Pantano [7] studied the mechanical response of discontinuous fiber
reinforced styrene-maleic anhydride (S/MA) polymer and have been characterized at
static and high strain rates. Five different materials were tested and ultimate strength,
failure strains and effective moduli for each material were investigated as function of
strain rate under dry and wet test conditions. The authors got a different type of results
here in terms of failure strains for the materials tested. Results of the S/MA tests
showed 60% increase in the strength but there was reduction in failure strains at higher
speed tests. All these different materials showed twice the strength and 2.5 times the
stiffness and less than a tenth of the strain to failure compared to the unreinforced
S/MA. There was significant increment in ultimate strength in the materials in dry test
conditions than that of wet test environments.
Barre’ and Chotard [8] studied the tensile dynamic behavior of glass fiber
reinforced polyester and phenolic resins in order to find the effects of strain rate on the
mechanical properties of composite materials produced by resin transfer molding (RTM)
11
and pultrusion processes. They created a new specimen design and validated using
drop-weight dynamic tests. The results showed that the dynamic elastic modulus and
strength increased by a ratio for majority of the materials studied. The shear modulus
measured with off-axis and ±45 coupons produce different effects as a function of strain
rate.
The effects of strain rate on the mechanical behavior of Scotchply type 1002
glass/epoxy angle-ply laminates were investigated by Staab and Gilat [9]. Tests were
conducted at high strain rates of approximately 1000/s using direct tension split
Hopkinson bar apparatus and quasi-static tests of strain rate approximately 0.0001 s-1
using servo-hydraulic testing machine. Results showed that maximum normal stress
experienced by glass/epoxy laminates is higher for dynamic than for quasi-static loading
conditions. Authors described that both fibers and matrix are sensitive to strain rates but
fibers dominate the laminate properties in case of high rate loadings. Authors also
mentioned that the failure patterns change with the fiber orientations.
On the basis of loading and unloading tests, the transient temperature rise
measurement and energy analysis of unidirectional fiber reinforced epoxy under tensile
impact, Yuanming and Xing [10] proposed a coated fiber bundle model. According to
this model, a one dimensional constitutive equation for glass fiber reinforced epoxy in
tensile impact tests from 300 to 2000 s-1 was derived from statistical analysis of the test
data. Yuanming and Xing also proposed that the constitutive equation for coated fiber
bundle model has general meaning and it can be used for other unidirectional brittle
fiber-reinforced resins.
12
Lifshitz and Leber [11] investigated the interlaminar tensile strength and modulus
of two material systems namely PW E-glass/epoxy and Unidirectional Carbon fiber
epoxy of 30-32 mm thick plates at high strain rates with SHPB. Results for CFRP
specimens were too scattered and also showed that both strength and modulus were
rate sensitive and increase with the loading rates.
Hou and Ruiz [12] tested CFRP T300/914 laminates at different strain rates
ranging from 10-4 s-1 to 103 s-1. Specimens were waisted symmetrically in the thickness
direction and there were difficulties in testing of ±45 tension test specimens. So the
specimens for ±45 were used as rectangular in cross section. In tension tests,
specimens remained virtually linear elastic up to failure. Tensile modulus and strength in
0° direction were rate dependent and tests on ±45 specimens gave non-linear stress
strain curves. Authors have discussed one more important issue that due to brittle
failure pattern of the specimen’s, plasticity should not be considered as a factor in
damage predictions of the composite materials.
In later research work with Zhou and Xia [13] carried out tensile impact tests on
T300/Al and M40J/Al (metal matrix) composite specimens with the help of a custom-
designed Split Hopkinson Pressure Bar (SHPB) apparatus. Quasi-static tensile tests
were performed on the MTS-810 apparatus to compare the data with dynamic test data.
Specimens were tested in the range of strain rates of 0.001 to 1300 s -1. From the
experimental results it appears that both the materials show rate sensitive behavior. The
elastic-plastic coated fiber bundle model and one dimensional damage constitutive
equation for the composite wires has been put forth by the authors based on these
experimental results. The statistical analysis of the experimental data was performed
13
using Weibull distribution, which is a continuous probability distribution. The relationship
between the mechanical parameters in the constitutive equation of the composites wires
and strain rate was derived from this statistical study.
Srikanth and Sun [14] observed the rate dependent non-linear characteristics of
the unidirectional S2 glass/8553-40 and the woven 7781/F155 E-glass fabric. The
motive was to find out a rate dependent failure model. Various tension tests were
conducted on different off-axis test coupons to acquire stress/strain data for various
strain rates. The test speeds used were in the limits of 0.0001 to 1/s. Srikanth and Sun
thus reached on a conclusion of a three parameter constitutive model to fit the test data.
Wang and Xia [15] proposed a one dimensional elastic brittle damage
constitutive equation for GRP on the basis of Double Weibull distribution function to
interpret the stress distribution. This equation was on the basis of stress-strain curves of
unidirectional GRP at a strain rate of 300/s and the coated fiber bundle model. Further-
more the tensile impact experiments on unidirectional KFRP were performed at strain
rates up to 1500/s. Experimental results showed that the mechanical properties of
KFRP were also rate dependent. Consistency between simulated results and the
experimental data confirmed that the coated fiber bundle model and the modified
constitutive equation are valid.
Strain rate dependent behavior of IM7/977-2 carbon/epoxy matrix composite in
tension is studied by testing various laminate configurations at different strain rates up
to 400-600 s-1 by Gilat and Goldberg [17]. High rate tests were conducted with SHPB
and low rate tests were conducted on conventional hydraulic testing machine. To avoid
the uncertainties in the data interpretation due to different test set ups, authors kept the
14
specimen geometry same for all specimens. Strain rates showed significant effect on
material response. General observation showed that the response at higher strain rates
is stiffer and stress-strain curve increased more rapidly with increased strain rates i.e.
higher maximum stresses were associated with deformation at higher rates.
Strain rate dependent behavior was observed for continuous filament random
mat glass polyester material when testing both in tension and compression by Fernie
and Warrior [18]. In their test data they noticed an increase in ultimate strength of 115%
in tension and corresponding 43% increase in modulus. In this work, in order to facilitate
the testing of materials with a relatively large unit cell, an instrumented falling weight
testing machine had been modified with appropriate loading rigs to allow the use of
larger specimens to test in compression, tension and shear. Till date research was
focused on various methods of reducing the test system noise and vibration from the
load data collected. These include extreme data reduction, data processing methods
and using the principle of impedance mismatching in order to absorb stress waves. But
the work had focused on eliminating noise and vibration through the design of the
system and impactor itself.
To characterize the high strain rate response of composite materials, Tsai and
Sun [19] developed a constitutive model using non-zero axis composite specimens.
Based on the experimental data, a viscoplasticity model was developed for strain rates
up to 1/s and verified with data obtained from high strain rate experiments conducted on
a SHPB using off-axis specimens. Authors used the rate dependent constitutive model
based on low strain rate tension tests on off-axis coupon specimens to predict the
15
dynamic laminate response observed in the SHPB test. Assumption made here was
that, material behavior was same in tension and compression.
At NASA Glenn Research Center Goldberg and Roberts [20, 25] used an
approach to modify the state variable constitutive equations originally developed for
metals in order to model the nonlinear, strain rate dependent deformation of polymeric
materials. One objective behind this study was to see the effects of hydrostatic stresses
on strength predictions which can be significant in polymers composites. To get the
detailed view in this direction, authors modified the equations of inelastic strain rate
tensor using classical plasticity theory definitions of effective stress and effective
inelastic strain. The results from the constitutive equations associated well with the
experiments. These constitutive equations then implemented within strength of
materials based micromechanics approach to predict the nonlinear strain rate
dependent deformation of polymer matrix composites. The mechanics approach was
then verified for different fiber angles and for different strain rates. Also in the later
research work of these authors [25], as there is not any specific material model for
predicting High Strain rate behavior using FE codes, the developed constitutive
equations were implemented in a user defined sub-routine material card to an explicit
transient dynamic finite element code LS-Dyna to get better idea of dynamic material
behavior.
Majzoobi and Saniee [21] used the high rate tensile testing apparatus called
“flying wedge” for testing of R2000 Glass/Epoxy. Specimens were tested at low rates of
10-3 s-1 using conventional Instron testing machine and up to 850 s-1 with the help of
flying wedge apparatus. Observations showed that there is significant increase in failure
16
strength and reduction in failure strain in dynamic tests compared to the static tests.
Also the rate of the increase of stress versus strain slowed down as ply angle
increased. With the help of Electron microscopy authors also showed that failure
mechanisms do not change much at high strain rates compared to low strain rates.
Fitoussi et al. [22] dealt with development and optimization of an experimental
methodology devoted to the microscopic and macroscopic characterization of
composites mechanical behavior under high speed loadings. The applied experimental
procedure has been optimized in order to isolate the inertial disturbances attributed to
the system. The optimization aims at minimizing the amplitude of noise in
measurements in order to obtain homogeneous stress-strain regions within the tested
specimen. Using a servo-hydraulic machine, monotonic and interrupted tensile tests
were performed at different strain rates. Strain rates up to 200s-1 have been applied to
SMC-R26 and woven carbon-epoxy laminates. Authors concluded that due to time
dependent (high strain rate) damage behavior the strength was increased. In extension
to this research work, currently authors are trying to develop a multi-scale FE model to
implement as a simulation tool to obtain material’s micro-structural effects in high rate
loadings.
Rio et al. [24] performed dynamic tensile tests on different carbon/epoxy
laminates of configurations 0º, 90º and quasi-isotropic [±45/0/90] s. All these tests were
carried out at two temperatures 20ºc and -600c. The work here mainly validated the Split
Hopkinson bar theory using one-dimensional wave propagation theory and analyzed the
effect of low temperature on elastic wave propagation along the bars of the and the
subsequent wave dispersion phenomena. The results of the dynamic tests showed little
17
effect of temperature and strain rate on the tensile strength of a unidirectional laminate
loaded in the fiber direction but in the other case strength increases evidently in the
transverse direction at low temperature and high strain rate. The test results presented
here by the authors indicate the non-sensitivity of carbon fibers to the temperature but
strain rate plays important role in ultimate strength values.
Shenoi and Makarov [26] dealt with experimental investigation of the high strain
rate behavior of unidirectional glass fiber-reinforced composite materials. High speed
hydraulic test machine is used to dig out the properties of unidirectional E-Glass/Epoxy
materials. Specimens were tested for static and high speed conditions separately. The
conclusion from the results here was also the same that there was significant increase
in ultimate strength and modulus of the material.
The overall issues of testing and summary for various strain rate ranges are
discussed here. These strain rates are generally covered by universal test instruments,
servo-hydraulic machines, and specialized drop towers. Strain rates above 500s -1
require other testing solutions such as SHPB/Kolsky bar. Slack grip system is used on
servo-hydraulic machines is used to ramp to the needed test speed prior to engaging
the test specimen. Specimen geometry details, Force and strain measurement details
and data collection and data reduction issues are elaborated.
Recent regulations by the government regarding automobile/aircraft safety issues
[28] have served as motivation for the development of high strain rate test data to
observe the changes in the material properties in case of a crash event. These test
results and be used as inputs in the computational softwares to build accurate material
models. By doing this expensive test methods and testing set up can be eliminated
18
which is a major step in today’s design process. Also, this will be helpful in efficient and
multiple design checks by making minor changes in the models which will lead to best
designs in the industry.
Generating high stroke rate test data is a great deal than obtaining static test
data. In a high speed tensile test, the gage section (actual material area) of a coupon
will not be under a consistent stress or strain state. In a particular test set up for high
strain rate material testing, the mechanical links should be having lowest weight
possbile while maintaining the strongest material properties. Due to this stress waves
propagate along the specimen and travel through all of the mechanical links which can
affect the load readings by load cell as the load cell will is calibrated and fixed in
between the top portion of the links. By making small test coupons this effect can be
diminished and average stress state can be simulated. Additionally, a very precise
engineering study and understanding is required when filtering high strain rate data.
Because of this, the test data recorded from the high strain rate tests should be
compared only with data obtained from tests having comparable experimental setups
and data analysis methods.
At present no specific method exists for performing high strain rate tests or for
obtaining the data generated from these tests. Major different methods in high strain
rate test and analysis techniques exist between researchers. Due to these reasons, a
standard methodology that specifically points out and resolve issues related to high
speed testing and data analysis is needed.
19
CHAPTER 3
PROBLEM STATEMENT AND SOLUTION APPROACH
3.1 Introduction
The requirement of high strain rate tensile tests is to obtain the data containing
the basic material properties like Ultimate Strength, Strain Rate distribution over the test
time and Failure Strain. This test data can be compared with the static tests data for the
same material configurations and changes in material properties can be observed with
respect to the increase in the strain rates.
The modified specimen geometry for the tests conducted is as shown in Figure 2.
The specimen geometry as per the ASTM 3039 standard is different in dimensions. But
as discussed earlier for reduction in problems of stress propagation, the specimen gage
length is reduced to 2 inches. Also, the number of laminas and width of the specimen
are reduced to trim down the strength so that it should be within the load carrying
capacity of the High Rate test machine used. (Facility provided by National Institute for
Aviation Research). Also the tabs used according to ASTM 3039 are tapered at ≥5º
angle to reduce the stress concentrations in the tabbed region. But in this present work
as the specimen geometry is modified as per requirements, the specimen thickness is
considerably thin than that of specified in ASTM standard. So the glass fiber laminate
used for tabbing is also of just 2 plies. As the tabbing laminate is very thin there is no
possibility of providing a taper for the tabs.
20
90°
0°
0.5
1.25
3.25
4.50
Figure 2. Modified specimen geometry (All dimensions in inches).
The goal of this work is to discover the high speed tensile testing factors and
analysis technique that may have a noteworthy effect on the resulting data and to
recommend a procedure that specifically addresses the problems related to high strain
rate tensile testing of composite materials.
3.2 Statement of Problem
Making it simple, the purpose of tensile testing is to determine ultimate tensile
stress and tensile modulus of the material. With the crashworthiness perspective, the
following aspects of the composite laminated coupons in static and high strain rate
tensile loadings were to be studied:
1. The ultimate strength of the material.
2. Failure pattern of the test coupons according to the strain rates.
3. Nature of the strain rate with respect to the time and strain throughout the
test.
4. Change in ultimate strength of the materials corresponding to the fiber
orientation and strain rates.
21
3.3 Solution Approach
Using MTS test systems, the specimens are tested in static and high stroke rate
tensile loadings. A 22 Kip MTS hydraulic machine is used for all static tests and a
special purpose high rate MTS hydraulic machine is used for all high rate tests. Tests
are conducted at quasi-static rate of 0.05 in/min and at five different higher stroke rates
of 1 in/s, 10 in/s, 100 in/s, 250 in/s and 500 in/s. For each stroke rate including the static
test 3 specimens are tested of each material and each lay-up combination.
The test data obtained in each test is analyzed for iso-strain rate behavior,
ultimate tensile strength and failure strain.
22
CHAPTER 4
MATERIAL SPECIFICATIONS AND TEST SET-UP
4.1 Material Specifications
Using vacuum bag forming, both quasi-static and dynamic test specimens with
identical physical and mechanical properties are manufactured. Glass and carbon fiber
pre-preg with epoxy matrix cured at the specified cure cycle in autoclave.
The materials and their lay-up combinations used to study in this present work
are as follows:
Materials used,
1. Newport NB321/7781 Satin Weave Fiberglass.
2. Newport NB321/3K70 PWCF.
Fiber orientations,
1. [0]4
2. [+15/-15]2s
3. [+30/-30]2s
4. [+45/-45]2s
The tabs are fabricated using NEWPORT NB321/7781 FIBERGLASS prepreg.
The tabs consist of two laminas (plies) oriented at 45° to the loading direction. Hysol-
9394 is used as adhesive for bonding of the tabs to the specimens.
23
4.2 Static Testing
At constant actuator velocity of 0.05 in/min, quasi-static tests were carried out
using a servo-hydraulic MTS testing machine. An MTS Load cell 100KN/22kip
specification measured the tensile load. The longitudinal strain of the specimen was
measured by uni-axial tension strain gage (Vishay micro measurements-CEA-06-
250UN-120). The M-BOND-200 adhesive is used for bonding stain gauges. Three
specimens of each combination were tested to failure at quasi-static stroke rate to get
the strain values and maximum load carrying capacity. A specimen set-up for static test
is as shown in Figure 3.
Stationary Top
Fixture
(crosshead)
Specimen
Moving
Bottom Fixture
Figure 3. Static test set-up.
24
All the materials specified above are tested in static rate for all lay-up combinations to
get the ultimate strength data in static stroke rate.
4.3 High Speed Testing
An MTS 810 high rate testing machine is used to perform dynamic tests on the
composite material specimens. This machine is designed to function at dynamic loads
of maximum of 9 Kip and actuator velocity of 500 in/s. By providing the slack-inducer
mechanism it is assured that the actuator plunger reaches the required velocity before
starting to apply load to the test coupon. The plunger displacement and recorded loads
by the load cell are calibrated to read the voltage readings in inches and pounds
respectively. Figure 4 shows the design of grips and slack inducer mechanism
(Courtesy: National Institute for Aviation Research). In this design, specimen is
manually tightened in the top and bottom grips with high strength steel bolts. In between
the bottom grip and the hydraulic actuator shaft a slack inducer mechanism is attached
so that the desired speed can be achieved before the specimen starts taking actual load
at that speed. A crushable foam ring is placed in between the bottom grip and the slack
inducer to absorb the shock due to the high speeds. The supply pressure chargers of
the hydraulic actuator must be fully charged before the start of the actual test. These
supply the force to achieve the “high speeds” which will be dissipated during the test. By
providing the slack inducer mechanism it is ensured the required constant velocity of the
plunger is achieved before the grips are engaged to start applying tensile load on the
specimen. The specimen is loaded in tension until it fails completely and the load and
strain data is simultaneously recorded by high rate data acquisition system (National
Instruments data acquisition system).
25
Load Cell
Top and Bottom

Grips
Note that
thesis
and
dissertati
on work
is
protected
by Foam
copyright, Ring
with all
rights
reserved.
Only the
author
has the
legal right Slack
Inducer
to
publish,
produce,
sell, or
distribute
this work.
Author
permissio
n is
Actuator
needed
Shaft
for others
to directly
Figure quote
4. Set-up of mechanical grips and slack inducer mechanism.
significan
t amounts
of
informatio
n in their
own
work.
Limited 26
amounts
of
informatio
The primary aim of this study is to keep the same specimen design and strain
gages for static and high speed tests. This is done to ensure that specimen geometry,
size and end fixing conditions do not affect the tensile load and strain readings. Typical
signals from the compact-size piezoelectric PCB PIEZOTRONICSINC load cell are
obtained for getting the load values in terms of voltages and then the calibration sheet
provided by the manufacturer is used to convert the voltage values in load values in
pounds. The details about the load cell are as follows;
Model: 206M33 Serial No: 1051
Type: ICP Force Sensor Bias: 10.67 VDC
The load cell is calibrated in compression mode and with a customer supplied
bolt as per the requirements of mechanical fittings available on the test fixture. Load cell
is preloaded to 26000 lbs prior to calibration. So the bias value of 10.67 V corresponds
to the preload value of 26000 lbs. This means if the load cell records a value of 10.67 V
in a test, then the load carried by the specimen in terms of pounds is 26000 lbs. So thus
the conversion factor is used to convert the obtained load data in terms of voltage to the
load values in lbs.
The Vishay 2210 signal conditioning amplifier system is used to obtain the strain
gage signals. As the high speed data acquisition system is set to record the data in
terms of 10 V calibration value, the strain gage signal conditioning amplifier is also set
for 10 V excitation voltage and respective gain settings. The maximum strain value the
specified strain gage can read is 50000 micro-strains. So the 50000 micro-strain value
corresponds to 10 V value in terms of voltage. Thus conversion factor is found out and
strain values are obtained in terms of in/in strains.
27
The high speed data acquisition system used is BNC-2090, a rack-mountable
breakout box designed for use with National Instruments data acquisition cards. The
BNC-2090 has 16 BNC inputs on the front panel, labeled ACH 0 to ACH 15 (ACH =
Analog Channel). Consequently, the BNC-2090 will accommodate 8 or 16 differential
channels of analog data inputs. In the test set-up we are interested in only four data
inputs which are actuator displacement, Load, Strain and Test time. These data signals
are then processed through the associated software provided by National Instruments
to convert the analog input signals into the desired output values.
28
CHAPTER 5
EXPERIMENTAL RESULTS
5.1 Test Results of Newport PWCF 0º Specimens
Considering the number of materials tested and the lay-up sequences for each
material, the task to present and explain the test data in detail for every material in this
chapter is pretty much vast. For this reason the data presented in this chapter limits for
one material and all of its lay-up combinations tested for each stroke rate.
The material taken into consideration in this chapter is Newport PWCF. Following
table describes the test results for each stroke rate for 0º lay-up combination.
TABLE 2
SUMMARY OF TEST DATA FOR NEWPORT PWCF MATERIAL (0º)
Stroke Rate Failure Stress Failure Strain Average Strain

Orientation
(in/s) (lbf/in²) (in/in) Rate (1/s)
0º (4 ply) 0.000833 107934.926 0.0146 0.000415
0º (4 ply) 1 115104.151 0.0124 0.1460
0º (4 ply) 10 118085.333 0.0120 1.7225
0º (4 ply) 100 179144.346 0.0121 57.0393
0º (4 ply) 250 87679.101 0.0121 92.2193
0º (4 ply) 500 118885.87 0.012 98.482
From the above data presented it is observed that even if the failure stress is
changing with respect to the stroke rate, the failure strain is not changing and it is
almost constant. The average strain rate mentioned in the table above is the mean
value taken for the strain rate values recorded at specific equal time intervals. From the
29
data presented above we can observe that the material strength is increasing with
respect to the stroke rate.
5.2 Test Results of Newport PWCF 15º, 30º and 45º Specimens
Now the following table describes the test data for 15º, 30º and 45º fiber angle
test coupons tested at each stroke rate.
TABLE 3
SUMMARY OF THE TEST DATA FOR NEWPORT PWCF MATERIAL (15º, 30º, 45º)
Stroke Rate Failure Stress Failure Strain Average Strain

Orientation
(in/s) (lbf /in²) (in/in) Rate (1/s)
15º (8 ply) 0.000833 79462.971 0.0124 0.000415

15º (8 ply) 1 82006.097 0.0112 0.1438
15º (8 ply) 10 88179.962 0.0112 0.5106
15º (8 ply) 100 178670.676 0.0114 48.8590
15º (8 ply) 250 87675.169 0.0117 85.2219
15º (8 ply) 500 90780.18 0.0116 92.255
30º (8 ply) 0.000833 56177.947 0.0184 0.000415
30º (8 ply) 1 64003.513 0.0170 0.1834
30º (8 ply) 10 65995.159 0.0177 1.5366
30º (8 ply) 100 151521.068 0.0172 68.1032
30º (8 ply) 250 98391.463 0.0195 95.2602
30º (8 ply) 500 82042.131 0.01823 98.574
45º (8 ply) 0.000833 31956.908 0.0604 0.000415
45º (8 ply) 1 37098.937 0.0415 0.2482
45º (8 ply) 10 38542.971 0.0414 1.9253
45º (8 ply) 100 81960.809 0.0410 58.7862
45º (8 ply) 250 94892.736 0.0411 131.1468
45º (8 ply) 500 63717.575 0.0344 170.58
30
The test results for the angled fiber orientation specimens show that the material
strength decreases as the fiber angle increases. But still taking one angle into
consideration, the failure strength is directly proportional to the stroke rate.
The Strength comparison chart for 0º fiber orientation specimens is presented
below from which one can get clear understanding of the material behavior in the
different stroke rate tensile tests.
Figure 5. Strength comparison for Newport PWCF (0º) specimens on the basis of
stroke rates.
In the Figure 5 we can clearly understand that the material strength is increasing
as the stroke rate is increasing but it is true till the 100 in/s tests and after that as the
stroke rate increases the strength is reduced than that of the 100 in/s tests.
31
Now considering for one stroke rate and different fiber orientation angles, we can
compare the strength distribution over the range of fiber angles as presented in Figure 6
below,
Figure 6. Strength distribution over the range of fiber orientation angles for Newport
PWCF in quasi-static stroke rate.
Here in we can clearly see that the tensile strength of the material clearly
diminish as the fiber angle of the test coupon increases. So the 0º fibers are the
strongest in a tensile test along the longitudinal direction. Same as the quasi-static
stroke rate, the material strength distribution holds true for all different stroke rates,
which means as the fiber orientation angle increases tensile strength of the material
decreases irrespective of the test rate.
32
5.3 Stress-Strain Plots for Newport PWCF (0º, 15º, 30º, 45º) Specimens in Static
Loading
To study the nature of the stress-strain graph for Newport PWCF material, the
following figure describes the brittle behavior of the material in 0º orientation. The
specimen does not elongate but breaks suddenly after a particular load range.
Figure 7. Stress-Strain behavior in quasi-static test for Newport PWCF (0º) material.
In case of higher orientation angles the nature of the stress-strain graphs goes
on changing and the specimen tends to go into the plastic deformation region as the
fiber orientation angle increases. Figure 8 describes the nature of the stress-strain
graph for Newport PWCF 15º orientation. Here we can observe that the failure strain
values are increasing and the strength of the material is diminishing as the fiber angle
increases.
33
34
We can observe from the above graphs that as the fiber angle is increasing,
more specimen elongation is occurring and there is reduction in ultimate strength. This
suggests that as the fiber angle increases, the material shows plastic deformation
region before failure.
5.4 Stress-Strain Behavior for Newport PWCF (0º) Specimens in Higher Stroke
Rates
Considering the various types and combinations of fiber orientations and stroke
rates, in this chapter we will observe only stress-strain nature of 0º fiber orientation with
all different stroke rates. The following figures describe the stress-strain nature of the 0º
orientation for different stroke rates.
35
Figure 11. Stress-Strain behavior of Newport PWCF (0º) material in 1 in/s stroke rate.
Figure 12. Stress-Strain behavior of Newport PWCF (0º) material in 10 in/s stroke rate.
36
Figure 13. Stress-Strain behavior of Newport PWCF (0º) material in 100 in/s stroke
rate.
Figure 14. Stress-Strain behavior of Newport PWCF (0º) in 250 in/s stroke rate.
37
5.5 Failure Patterns of Newport PWCF (0º) Specimens in different Stroke Rates
Following figures show the failure mode patterns of Newport PWCF (0º) material
in different stroke rate tests
Figure 16. Failure mode of Newport PWCF (0º) in static stroke rate.
38
Figure 17. Failure mode of Newport PWCF (0º) in 1 in/s stroke rate.
39
The figures above suggest that there is no change in failure pattern of the specimens
according to the stroke rates. In each stroke rate test the specimen is failing in the gage
section and no change can be seen in the mode of failure.
40
CHAPTER 6
METHODOLOGY FOR ISO-STRAIN RATE CHARACTERIZATION
6.1 Introduction
Up till now many researchers have done extensive work in the area of tensile
testing of polymer composites. Unlike metals there are no specified material properties
are mentioned anywhere for polymers as they exhibit anisotropic behavior. As the fiber
orientation changes, these materials display different properties. Till now many
scientists suggested many methods to determine the specific material properties so that
one can use those to implement in a Finite Element code to reduce the extensive work
of testing for individual material configuration each time.
Unlike metals, there are no specific material property cards provided in any Finite
Element Analysis (FEA) code for polymer composites. So with the data obtained from
actual testing of these materials, one needs to develop user defined material cards to
input the data for a FEA code. This is one of the ways to interpret the test data and get
an analysis model for a particular material combination so that further testing for that
combination can be eliminated and analysis software can replace the actual testing.
From the tensile testing of a material, one can get fundamental properties like
Young’s Modulus, Failure Stress and Failure Strain. To obtain the Young’s modulus,
stress-strain curves are needed. Generally to obtain a stress strain curve from a tensile
test data, the test should be at constant strain rate. In case of metals, as they show
isotropic properties and ductile in nature, one can easily achieve a constant strain rate
test requirements. But in case of polymers, these materials exhibit brittleness and non
41
ductile properties. When brittleness comes into picture, these materials can not show a
constant strain rate elongation in case of a tensile test. So throughout the test period the
specimen shows a variable strain rate elongation as the fiber breakage takes place
randomly.
To validate the real life test scenario in FEA software, one needs to input the
stress-strain curves in the material model. For this reason we need to input the stress-
strain curves obtained from a constant strain rate test. In this present study a unique
approach can be used to obtain the stress-strain curves at constant strain rates and to
input into the user defined material model.
6.2 Variable Strain Rate Behavior
From the data obtained from various stroke rate tests, the strain rates are found
out. In this present study, to obtain the strain rate throughout the test strain-time curves
are plotted for each stroke rate test. This strain-time data then differentiated with
respect to the each time instance and strain rate data is obtained. The following graph
in figure 22 shows the nature of the strain-time data. Fitting a best fit polynomial to the
strain-time curve and differentiating with respect to the time instances can get us to the
strain rate at each time point. The red line is the best fit polynomial for the strain-time
curve for a 100 in/s test. The reason to fit the polynomial is to get rid of some noise from
the data obtained from actual test. As one can obtain the strain rate at each data point
from the original data also but it gives some erratic values of strain rate due to the
noise. So fitting the polynomial is to get rid of those noisy data points and obtain a
smooth curve.
42
Figure 22. Strain-Time data for Newport PWCF (0º) in 100 in/s test.
The reason behind fitting a polynomial curve to find out the strain rate is that the
individual derivative at each time instance gives some erratic values for strain rates.
After differentiating the polynomial equation displayed above with respect to time, we
can get the strain rate data for the overall strain-time data.
Following figure 23 describes the nature of the strain rate vs. strain data. In this
graph we can clearly see that the strain rate is not constant for the total test time and is
continuously changing.
43
Figure 23. Strain Rate-Strain data for Newport PWCF (0º) in 100 in/s test.
As the strain rate is not constant in nature, we need to extract the stress strain
data from the available test data for a particular constant strain rate. As one can see
that the strain rate is starting from zero value at the start of the test and it reaches
approximately to the value of 70 s-1 at the end of the test. Like this in every stroke rate
test, the strain rate starts initially from a zero value and reaches a certain value at the
end of the test.
For extracting the constant strain rate data for a particular strain rate, the strain
rate at which the isostrain-rate stress/strain curves must be obtained are chosen. Here
the chosen strain rates are 1s-1, 10s-1, 30s-1, 50s-1, and 100s-1. To get the family of
stress-strain curves for mentioned iso-strain rates, first the total strain interval is divided
into smaller divisions. In case of Newport PWCF (0º) specimens the failure strain is
0.012 in/in. So this total strain interval of 0.012 in/in is divided into 12 strain intervals.
44
For each strain interval value Stress vs. Strain Rate values are obtained. Stress vs.
Strain rate curves are plotted from these values and equations for stress as a function
of strain rate are obtained for the same. Figure 24 describes the Stress vs. Strain rate
behavior for random strain intervals of 0.001 in/in, 0.004 in/in and 0.008 in/in.
Figure 24. Family of Stress-Strain Rate curves for different strain intervals.
For each Stress vs. Strain rate curve, one equation is obtained by fitting a trend
line to the data as shown in Figure 25. The equation for each line represents an
equation for stress in terms of strain rate at each of different strain levels.
45
1000000
y = 86953x0.0759
100000
y = 38769x0.0695
y = 8124.5x0.0828
Stress (psi)
10000
0.001 in/in
1000
0.004 in/in
0.008 in/in
100 Power (0.008 in/in)
Power (0.004 in/in)
Power (0.001 in/in)
10
0.0001 0.001 0.01 0.1 1 10 100
Strain Rate
Figure 25. Family of Stress-Strain Rate curves for different strain intervals.
Now to get an iso-strain rate Stress-Strain curve for a particular strain rate we
can input the desired constant strain rate value in each of these equations and for each
strain interval we will get a stress value. The family of Stress-Strain curves obtained
from different equations of Stress vs. Strain rate data for Newport PWCF (0º) material
are as shown in Figure 26.
Figure 26 suggests the iso-strain rate nature of the family of the stress-strain
curves for different strain rates. These curves for different strain rates can be input to
the FEA code to develop the material model for particular configuration.
46
Figure 26. Family of Stress-Strain curves for Newport PWCF (0º) material at
different strain rates.
In the figure above we can notice that the failure stress is increasing in case of higher
strain rates than that of Quasi-static tests.
Similarly, following the same procedure to obtain the family of stress-strain
curves, figure 27 shows the Iso-strain rate Stress-Strain curves for Newport Glass fiber
(0º) material. Comparing the stress-strain data for Newport PWCF (0º) and Newport
Glass fiber (0º) materials we can observe that there is some plastic deformation in case
of glass fiber specimens. Also the important observation can be seen that the failure
strength of Newport PWCF material is higher than that of glass fiber material.
47
Figure 27. Family of Stress-Strain curves for Newport Glass Fiber (0º) material at
different strain rates.
48
CHAPTER 7
CONCLUSIONS AND FUTURE RECOMMENDATIONS
7.1 Conclusions
In-plane tensile tests are carried out on Glass and Carbon fiber materials of
various lay-up configurations for different strain rates. All the tests are carried out on the
MTS static and high stroke rate hydraulic machines.
Following conclusions can be drawn from the static and high strain rate experimental
data:
1. In case of Glass fiber materials, a region of plastic strain hardening can be
seen.
2. In case of 0º fiber orientation of Newport PWCF materials, the failure modes
are quite brittle in nature so the plastic strain hardening can’t be seen.
3. In case of higher fiber orientation angles, there can be seen a region of plastic
strain hardening as the fibers get separate from each other and load is carried
by the matrix material.
4. As the ply orientation angle increases in lay-up schedules, the failure strength
of the material decreases.
5. The failure strength of both the materials increases proportionately to the
strain rate.
49
7.2 Future Recommendations
Research activities, aimed to expand the applications in composite industry, must
be addressed to improve manufacturing composite technology, through a better
integration of product and process design; to develop new constituent materials with
better performances and for the development of new processes and new manufacturing
technologies. The future developments in this present research work can be as follows;
1. In case of the test set-up, there should be enough damping provided in
between the slack adapter and the engaging pin so that to reduce the
vibrations and data scatter in higher stroke rate tests.
2. These iso-strain rate curves can be introduced in FEA packages to simulate
the presented experiments here and can be used to modify the existing
material models for response prediction of composite materials in different
strain rates.
3. This data can be used to determine the parameters of the material models
presented by other researchers.
4. Future focus can be on the development of the test standards for evaluation
of dynamic material properties for composites.
50
REFERENCES
51
REFERENCES
[1]. Anonymous, “Standard Test Method for Tensile Properties of Polymer Matrix
Composite Materials,” ASTM Standards, Designation: D 3039/D 3039M, Dec.
2002.
[2]. Kolsky, H., “An investigation of the mechanical properties of materials at very
high rates of loading,” Proceedings of Physics Society, London, 62-B: 676-700,
1949.
[3]. Hodgkinson, J.M., “Mechanical Testing of Advanced Fiber Composites,”

Woodhead Publishing Limited, 2000.
[4]. Davis, J.R., & Associates, “Tensile Testing” Second edition, ASM International,
2004.
[5]. Hayes, S. V., and Adams, D.F., “Rate sensitive tensile impact properties of fully
and partially loaded unidirectional composites,” Journal of Testing and
Evaluation, Vol. 10, No.2, pp 61-68, 1982.
[6]. Harding, J., and Welsh, L.M., “A tensile testing technique for fiber reinforced
composites at impact rates of strain,” Journal of Material Science, Vol. 18, pp
1810-1826, 1983.
[7]. Peterson, B.L., and Pantano, C.G., “Static and high strain rate response of a
Glass Fiber reinforced thermoplastic,” Journal of Composite Materials, Vol. 25,
pp 887-905, 1991.
[8]. Barre’, S., and Chotard, T., “Comparative study of strain rate effects on
mechanical properties of glass fiber-reinforced thermoset matrix composites,”
Composites Part A, Vol. 27A, pp 1169-1181, 1991.
[9]. George, H.S., and Gilat, A., “High strain rate response of angle-ply Glass/Epoxy
Laminates,” Journal of Composite Materials, Vol. 29, No.10, 1995.
[10]. Yuanming, X., and Wang, X., “Constitutive equation for unidirectional composites
under tensile impact,” Composites Science and Technology, Vol. 56, pp 155-160,
1996.
[11]. Lifshitz, J.M., and Leber, H., “Response of fiber-reinforced polymers to high
strain-rate loading in interlaminar tension and combined tension/shear,”
Composites Science and Technology, Vol. 58, pp 987-996, 1998.
[12]. Hou, J.P., and Ruiz, C., “Measurement of the properties of woven CFRP
T300/914 at different strain rates,” Composites Science and Technology, Vol. 60,
pp 2829-2834, 2000.
52
[13]. Yuanxin, Z., and Yuanming, X.,”Experimental study of the rate sensitivity of
unidirectional-fiber-reinforced metal-matrix composite wires and the
establishment of a dynamic constitutive equation,” Composites Science and
Technology, Vol. 61, pp 2025-2031, 2001.
[14]. Srikanth, V., and Sun, C.T., “Models for the strain-rate-dependent behavior of
polymer composites,” Composites Science and Technology, Vol. 61, pp 1-12,
2001.
[15]. Wang, Y., and Yuanming, X., “A modified constitutive equation for unidirectional
composites under tensile impact and the dynamic tensile properties of KFRP,”
Composites Science and Technology, Vol. 60, pp 591-596, 2000.
[16]. Pardo, S., Baptise, D., and Fitoussi, J., “Tensile dynamic behavior of a quasi-
unidirectional E-Glass polyester composite,” Composites Science and
Technology, Vol. 62, pp 579-584, 2002.
[17]. Gilat, A., and Goldberg, R.K., “Experimental study of strain rate dependent
behavior of Carbon/Epoxy composite,” Composites Science and Technology,
Vol. 62, pp 1469-1476, 2002.
[18]. Fernie, R., and Warrior, N.A., “Impact test rig for high strain rate tensile and
compressive testing of composite materials,” Strain, Blackwell Science Ltd, Vol.
38, pp 69-73, 2002.
[19]. Tsai, J. and Sun, C.T., “Constitutive model for high strain rate response of
polymeric composites,” Composites Science and Technology, Vol. 62, pp 1289-
1297, 2002.
[20]. Goldberg, R.K., Roberts, G., and Gilat, A., “Implementation of an Associative
Flow rule including Hydrostatic stress effects into the high strain rate deformation
analysis of polymer matrix composites,” NASA Technical Publication: NASA/TM-
2003-212382.
[21]. Majzoobi, G.H., and Sainee, F., “A tensile impact apparatus for characterization
of fibrous composites at high strain rates,” Journal of Material Processing
Technology, Vol. 162-163, pp 76-82, 2005.
[22]. Fitoussi, J., Meraghni, F., and Hug, G., “Experimental methodology for high strain
rates tensile behavior analysis of polymer matrix composites,” Composites
Science and Technology, Vol. 65, pp 2174-2188, 2005.
[23]. Testing Brief, Anonymous, “Testing at High Strain Rates,” Axel physical testing
services, July 2005.
53
[24]. Barbero, E., Zaera, R., and Navarro, C., “Dynamic tensile behavior at low
temperature of CFRP using a Split Hopkinson Pressure Bar,” Composites
Science and Technology, Vol. 65, pp 61-71, 2005.
[25]. Zheng, X., and Goldberg, R.K., “Rate dependent shell element composite
material model implementation in LS Dyna,” 46th AIAA/ASME/ASCE/AHS/ASC
Structures, Structural Dynamics & Materials Conference, 2005.
[26]. Makarov, G., and Shenoi, R.A., “Deformation and fracture of unidirectional GFRP
composites at High strain rate tension,” School of Engineering Sciences,
University of Southampton, UK, COMP Test 2004, [2] BS EN ISO 527-4:1997, [3]
BS EN ISO 14129:1998.
[27]. “Unidirectional composite axial tensile specimens,” High Performance

Composites, Magazine, Jan 2006.
[28]. http://www.sae.org/standardsdev, June 2006.
54
APPENDIX
55
APPENDIX
As discussed earlier in chapter 5, following are the Stress-Strain graphs of Newport
PWCF material for 15º, 30º and 45º fiber orientations tested at high stroke rates.
56
57
58
59
60
61
62
63
Failure Patterns of Newport PWCF (15º) Specimens in different Stroke Rates
64
The figures 43 to 48 suggest that there is no change in failure pattern of the
specimens according to the stroke rates. In each stroke rate test the specimen is failing
in the gage section and no change can be seen in the mode of failure.
65
Failure Patterns of Newport PWCF (30º) Specimens in Different Stroke Rates
66
67
Failure Patterns of Newport PWCF (45º) Specimens in different Stroke Rates
68
69
As discussed earlier in Chapter 5, following is the test data of Newport Glass-Fiber
material for 0º, 15º, 30º and 45º fiber orientations tested at quasi-static and high stroke
rates.
TABLE 4
SUMMARY OF THE TEST DATA - NEWPORT GLASS-FIBER (0º, 15º, 30º, 45º)
Stroke Rate Failure Stress Failure Average Strain

Orientation
(in/s) (lbf /in²) Strain (in/in) Rate (1/s)
0º (4 ply) 0.000833 63413.975 0.0258 0.000415

0º (4 ply) 1 86062.082 0.0252 0.1311
0º (4 ply) 10 100937.341 0.0285 2.0346
0º (4 ply) 100 160730.954 0.035 58.480
0º (4 ply) 250 206450.554 0.0325 109.420
0º (4 ply) 500 181191.315 0.0295 142.967
15º (8 ply) 0.000833 33550.218 0.055 0.000415
15º (8 ply) 1 40217.102 0.0387 0.2499
15º (8 ply) 10 42177.69 0.0488 1.724
15º (8 ply) 100 77445.547 0.0401 67.506
15º (8 ply) 250 56840.376 0.0458 150.071
15º (8 ply) 500 54834.214 0.0475 184.212
30º (8 ply) 0.000833 61678.547 0.0299 0.000415
30º (8 ply) 1 73402.535 0.0303 0.1947
30º (8 ply) 10 74627.877 0.0294 1.1332
30º (8 ply) 100 133332.347 0.0357 55.963
30º (8 ply) 250 114574.476 0.0274 101.937
30º (8 ply) 500 116228.774 0.0394 123.266
45º (8 ply) 0.000833 45242.599 0.0418 0.000415
45º (8 ply) 1 52579.231 0.0418 0.239
45º (8 ply) 10 54925.751 0.0357 1.4758
45º (8 ply) 100 108352.22 0.0420 63.468
45º (8 ply) 250 87113.034 0.029 115.008
45º (8 ply) 500 89379.554 0.037 154.384
70
Stress-Strain behavior of Newport Glass-fiber in Quasi-static Loading
Following graphs describe the stress-strain nature of the Newport Glass-fiber test
coupons in quasi-static loading.
Figure 61. Stress-Strain behavior of Newport Glass-Fiber (0º) in Quasi-static test.
71
72
73
Stress-Strain behavior of Newport Glass-fiber (0º) in High Stroke Rate Loading
Following graphs describe the stress-strain nature of the Newport Glass-fiber (0º) test
coupons in high stroke rate loading.
Figure 65. Stress-Strain behavior of Newport Glass-fiber (0º) in 1 in/s stroke rate.
74
75
76
77
78
79
80
81
82
83
84
85
Failure Patterns of Newport Glass Fiber (0º) Specimens in different Stroke Rates
Figure 85. Failure mode of Newport Glass Fiber (0º) in static stroke rate.
Figure 86. Failure mode of Newport Glass Fiber (0º) in 1 in/s stroke rate.
86
87
88
89
90
91
92
93
Family of Stress-Strain Curves at Iso Strain Rates for Newport PWCF and
Newport Glass fiber (30º and 45º) Materials:
Figure 109. Family of Stress-Strain Curves for Newport PWCF (30º) material.
Figure 110. Family of Stress-Strain Curves for Newport Glass Fiber (30º) material.
94
Figure 111. Family of Stress-Strain Curves for Newport PWCF (45º) material.
Figure 112. Family of Stress-Strain Curves for Newport Glass Fiber (45º) material.
95
The figures 109 and 110 suggest that the strength of Newport PWCF (30º)
material is higher than that of Newport Glass Fiber (30º) material. But as in case of
higher fiber angles the matrix plays more important role than that of fibers in strength
comparison. So in case of 45º fiber angle (Figures 111 and 112), Glass Fiber
specimens are depicting higher failure strength values than that of Newport PWCF
specimens in high strain rates as the matrix material is more stronger in strength in case
of Newport Glass Fiber material.
96

Characterization of CFRP and GFRP Composite Materials at High Strain Rate Tensile Loading PDFDrive (SFILE

Uploaded by

Copyright:

Available Formats

Characterization of CFRP and GFRP Composite Materials at High Strain Rate Tensile Loading PDFDrive (SFILE

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Characterization of CFRP and GFRP Composite Materials at High Strain Rate Tensile Loading PDFDrive (SFILE

Uploaded by

Copyright:

Available Formats

CHARACTERIZATION OF CFRP AND GFRP COMPOSITE MATERIALS AT

HIGH STRAIN RATE TENSILE LOADING

Bachelor of Engineering, Dr. B.A.M. University, India, 2002

Submitted to the Department of Mechanical Engineering

All Rights Reserved

Hamid M. Lankarani, Committee Chair

K. Suresh Raju, Co-Chair and Committee Member

Bob Minaie, Committee Member

To my parents and friends

I would like to present my sincere gratitude towards my advisor Dr. Hamid

from him in my entire journey of MS.

It is difficult to overstate my gratitude to my co-advisor Dr. K. Suresh Raju, who

towards him in words.

guidance and support.

I owe the deepest appreciation to my colleagues Vivek Mariyanna, Satish

and helping with various applications.

I am indebted to my many colleagues and friends for providing a stimulating and

entertainment and caring.

of encouragement and inspiration throughout my life.

complete this thesis.

material starts playing role in case of higher fiber angles.

code to develop a material model for computational simulations.

1.1 Simple Tensile Test ................................................................................... 2

2. BACKGROUND AND LITERATURE REVIEW ..................................................... 7

2.1 General Terms and Purpose of Tensile Testing ........................................ 7

3. PROBLEM STATEMENT AND SOLUTION APPROACH .................................. 20

3.1 Introduction .............................................................................................. 20

4. MATERIAL SPECIFICATION AND TEST SET-UP ............................................ 23

4.1 Material Specifications ............................................................................. 23

5. EXPERIMENTAL RESULTS .............................................................................. 29

5.1 Test Results of Newport PWCF 0º specimens ......................................... 29

6. METHODOLOGY FOR ISO STRAIN RATE CHARACTERIZATION.................. 41

6.1 Introduction .............................................................................................. 41

7. CONCLUSIONS AND FUTURE RECOMMENDATIONS ................................... 49

7.1 Conclusions ............................................................................................. 49

1. Summary of Systems Used for Dynamic Loading ................................................ 9

2. Summary of Test Data for Newport PWCF Material (0º) .................................... 29

1. General specimen geometry................................................................................. 3

2. Modified specimen geometry .............................................................................. 21

3. Static Test Set-up ............................................................................................... 24

4. Set-up of mechanical grips and slack inducer mechanism ................................. 26

5. Strength comparison for Newport PWCF (0º) specimens on the ........................ 31

basis of stroke rates

Newport PWCF material in quasi-static stroke rates

8. Stress-Strain behavior in quasi-static test for Newport (15º) material ................. 34

12. Stress-Strain behavior of 0º Newport PWCF in 10 in/s stroke rate ..................... 36

different strain rates

different strain rates

61. Stress-Strain behavior of 0º Newport Glass-Fiber in Quasi-static test ................ 71

62. Stress-Strain behavior of 15º Newport Glass-Fiber in Quasi-static test .............. 72

63. Stress-Strain behavior of 30º Newport Glass-Fiber in Quasi-static test .............. 72

64. Stress-Strain behavior of 45º Newport Glass-Fiber in Quasi-static test .............. 73

A composite material is formed by the combination of two or more different

composite material is wood consisting of cellulose fibers in a lignin matrix. When

because of one or more of the following advantages:

2. Strength and Stiffness:

High strength-to-weight ratio

Directional strength and/or stiffness

Tailored surface finish