4/4/2014

Fiber-Reinforced Polymer
Composite
Characterization

Nick
1

Abstract
Composites are of great interest in the automotive and aerospace industries because of
their load bearing capabilities coupled with their lightweight design. Style 4180 Glass Fiber
Reinforced Epoxy (Cycom

919) was tested using Dynamic Mechanical Analysis (DMA) to
determine its glass transition temperature (T
g
), which predicts when the composite changes
form and loses its effectiveness. This limit was determined graphically to be T
g
= 125.11C. T
g
is
crucial when choosing materials for specific applications that include high temperatures. Also,
resin burn-off was utilized to determine the fiber volume fraction (FVF) which roughly
correlates to how much strength is gained in the composite when the load carrying fibers are
present. The mean FVF was found to be 0.34543 with a standard of 0.00374.





















2

Contents
Abstract 1
Introduction 3
Methods 5
Results 6
Discussion 8
Conclusion 11
References 11
Appendix A 12



















3

Introduction
Fiber-reinforced polymer composites play a big role in load bearing structural
applications in the automotive and aerospace industries. Their high specific strength, the ability
to shape their structure to produce more aerodynamically efficient configurations, and
lightweight designs makes them far superior to conventional materials. [1] Composites
maintain structural integrity and load-bearing capability at elevated temperatures, which is
particularly important in structures located near heat sources such as engines and electronic
equipment [2]. Also, exterior aircraft and spacecraft structures experience high temperatures
due to aerodynamic heating and other phenomena. The high strength-to-weight ratio and
ability to withstand these temperatures make these composites ideal. [3]

Composites refer to a material composed of two or more distinct phases. In the case of
fiber-reinforced composites, which was the subject of this experiment, the two phases are
represented by a fiber and a polymer matrix. The composite specimen under analysis here was
Style 4180 Glass Fiber Reinforced Epoxy (Cycom

919).

The fiber reinforcement is the more crucial component because it provides the strength
and stiffness in the composite. The fibers carry the in-plane loads and also reduce the
coefficient of thermal expansion for the material. On the other hand, the polymer matrix holds
the filaments in place and transfers the load between fibers. The matrix is woven through the
fibers similar to the weaving of clothing. The matrix also protects the fibers from damage,
provides transverse strength (compression), and increases toughness and durability [4]. A few
common types of fibers include glass, carbon (graphite), and aramid (Kevlar).

An important parameter to consider for a fiber-reinforced polymer composite is the
glass transition temperature, T
g.
This temperature can be thought of as the temperature at
which polymer chains can partially move, known as local segmental motion. Below this
temperature, the bulk polymer behaves as a glass which is more or less brittle, and rigid. Above
the glass transition temperature, the material transitions to a more rubbery state and
behaves like a viscous fluid. This temperature is crucial because once this temperature is
exceeded, the composites load-carrying capability is drastically reduced, basically rendering it
useless. To determine the glass transition temperature, the storage modulus of the material is
analyzed as it is subjected to cyclic deformation as temperature is increased.


Additionally, there are a few important terms to consider. The storage modulus
measures the ability of a material to store energy. On the contrary, the loss modulus measures
the ability of a material to dissipate energy lost to viscous flow. Also, tan delta refers to the
4

(1)
(2)
(3)
dissipation of energy in a material under cyclic loading. Tan delta is shown in this lab as the
phase difference between strain and stress when the material is stressed cyclically.

Another important parameter of a composite is the fiber-volume fraction, which relates
the relative volume of the fibers to the overall volume of the composite material. The volume
fraction is given by:

where

is the volume of the

constituent, and

is the volume of the composite.

Where f, m, and v represent the fibers, matrix and voids respectively.

The volume of fibers relative to the matrix is dependent on a number of factors,
including fabrication process, cure cycle, initial volume of fiber and matrix, and fiber packing
arrangement. Generally, the composite failure stress increases linearly with increasing fiber
volume fraction. Also, the impact strength and shear strength of the composite both increase
with increasing fiber volume fraction mainly because the fiber strength is much greater than
that of the matrix. Ideally, the void content should be less than 5% and the fiber volume
fraction should be around 40-70% to ensure the polymer composite maximizes its strength
potential.


Once volume fractions are known, the upper limit of the modulus of elasticity of the
composite can be found. Shown in Equation (3), the rule of mixtures is shown as follows:

where E represents the elastic modulus, is the volume fraction, f or m represents the fiber
or matrix, and 1 refers to the composite material as a whole in the direction of the fiber. The
rule of mixtures is a basic attempt to determine the overall properties of a composite based on
the properties of its constituents. Equation (3) is only valid in the longitudinal direction of
continuous fibers, however the material in this experiment is balanced so direction is not
important.
The objective was to find the two important parameters, fiber volume fraction and glass
transition temperature, of the specified fiber-reinforced composite. The fiber volume fraction
was found by a method known as resin-burn off and was expected to be around the 40-70%
5

mark, considering this material is used in aerospace applications. After measuring the weight of
the composite samples, the matrix was burned off and only the fibers were left, which allowed
for the determination of the fiber weight and the subsequent byproducts. The glass transition
temperature was found by the use of dynamic mechanical analysis (DMA). This value was
expected to be reasonably high and was used to determine suitable applications for the
material. Also, a literature value was found and used as a reference to determine the accuracy
of the experiment.



Method

10 samples of Style 4180 Glass Fiber Reinforced Epoxy
(Cycom

919) were used in the two parts of this lab. One was
used in the DMA to determine the glass transition temperature
of the material and the nine remaining were used to determine
the fiber volume fraction by the process of resin burn-off

The DMA was done by use of a Q800 DMA made by TA
Instruments shown by Figure 1 to the right. This machine utilized
multi-frequency iso-strain to subject the specimen to a cyclic
load at an increasing temperature. The frequency was 1Hz and
the amplitude of strain was 15 micrometers. Also, the
temperature was started at 25C and increased to 160C at a
ramping step at 5C/min. As the process progressed, the data
was collected through a computer program that allowed for
a plot of the data for analysis.

The resin burn-off method was used to
determine the fiber volume fraction. Initially, the
nine samples were measured by length, width,
and thickness to find the volume and then
weighed. These values can be seen in Table 1 in
Appendix A and the samples and precise
micrometers used to take the measurements are
shown in Figure 2 to the left. As shown, the samples were not too large but nine were used and
averages were taken to ensure the results were accurate as possible and were not skewed by
any random occurrences.
Figure 1: DMA machine used to find T
g

Figure 2: Composite Samples and Tools
6

Figure 4: Determination of Glass Transition Temperature
Next, they were placed in an oven at an initial temperature of
321C and then ramped up to 800C after about 45 minutes to
complete the resin burn-off. This process successfully removed the
polymer matrix because of the differing melting points of the
constituents. The fiber has such a high melting point that it can
withstand the 800C temperature while the matrix cannot. The fibers
were the only things that remained when removed from the oven and
they were subsequently weighed. A Mettler Toledo digital scale was
used for measuring the weights throughout the lab and is shown in
Figure 3 to the right. This scale was extremely precise and was
sensitive to the smallest vibrations so extreme care was exercised.

From the data collected, the fiber volume fraction and
matrix volume fraction were able to be determined using
Equation (1). Additionally, the mean, standard deviation,
variance, and coefficient of variation was found for the fiber
volume fraction.

Results
The DMA produced a large set of data that included variables such as time,
temperature, storage modulus, loss modulus, stress, and tan delta among others. The variables
of interest were plotted below and consist of temperature vs. storage modulus, loss modulus,
and tan delta.

Figure 3: Digital Scale Used in Lab
7

Figure 4 represents the change in storage modulus and loss modulus as temperature
was increased. As shown by the black lines, tangent lines were drawn to the two linear portions
of the storage modulus line. The intersection of these two tangent lines is the glass transition
temperature, which was found to be T
g
= 125.11C. Also of importance was the change in tan
delta as temperature increased, which is shown in Figure 5 below:













The second part of the lab, done to determine the Fiber Volume Fraction and the Matrix
Volume Fraction, was completed by resin burn-off and a series of weighing the samples with an
extremely precise digital scale. To begin, the 9 samples were weighed and measured in all three
dimensions to determine their volumes. These values are shown in Table 1 in Appendix A. Next,
the samples underwent resin burn-off to remove the matrix polymers, which left only the
fibers. These fibers were weighed and the weights are shown in Table 2 as follows:
Table 2
Sample 1 2 3 4 5 6 7 8 9
Fiber
weights(g) 0.2590 0.2640 0.1871 0.2588 0.2728 0.2572 0.2454 0.2520 0.2671



Figure 5: Increase of Tan Delta as Temperature Increased
8

Using the density of the fiber, 2.48g/cc, the fiber volumes were easily calculated using
Equation (1) and are shown in Table 3 in Appendix A. More importantly, the fiber volume
fractions were found and are listed in Table 4 as follows:
Table 4
Sample 1 2 3 4 5 6 7 8 9
Fiber Volume
Fraction 0.3513 0.3463 0.3380 0.3424 0.3462 0.3440 0.3462 0.3480 0.3467

Along with these values, some statistical information was calculated and is shown in Table 5:
Table5
Fiber Volume Fraction Stats
Mean 0.34543
Standard Deviation 0.00374
Variance 0.00001
Coefficient of Variation 0.01083

Based on the weights of the fibers, the weights and volumes of the matrix could be
determined by subtracting the fiber weights from the overall sample weights and using the
density of matrix polymers at 1.31 g/cc. These values are shown in Table 6 in Appendix A. Also,
the matrix volume fractions were found and are listed in Table 7 below:
Table 7
Sample 1 2 3 4 5 6 7 8 9
Matrix Volume
Fraction 0.4990 0.5133 0.5003 0.4916 0.5069 0.4985 0.5119 0.5074 0.4914


Discussion
The objectives of this lab were carried out in two parts, first with DMA and then by resin
burn-off. T
g
, the glass transition temperature, and the combination of Fiber Volume Fraction
and Matrix Volume Fraction were found in these two sections respectively. These parameters
reveal a lot of information about the material of interest, Style 4180 Glass Fiber Reinforced
Epoxy (Cycom

919). Suitable applications and situations of certain failure can be determined by
these critical parameters.

In part one of the lab, T
g
was found graphically to be 125.11C. This was done by
analyzing Figure 4 and using tangent lines drawn from the linear portions of the storage
9

modulus curve. Although a literature value of T
g
for the exact epoxy used in this lab could not
be found, values of around T
g
=127C were found for glass/epoxy fiber-reinforced composites
[5]. Therefore, the graphically determined value of 125.11C can be assumed to be fairly
accurate. However, this method does have the potential to include some systematic error. The
error stems from the tangent lines that are drawn from the linear portions of the curve,
which leads to imperfect results since these locations are left up to the user to decide. The
intersection of these lines should be reasonably close for most tangent lines, but an error of
less magnitude will most likely be incorporated.

Nevertheless, Figure 4&5 shown in the results depict the data that was gathered during
DMA. By definition, the storage modulus and loss modulus curves are opposites of each other.
Similarly, tan delta is basically the loss modulus divided by the storage modulus, so it follows
the same basic path of the loss modulus itself. All three of these curves changed exponentially
when the glass transition temperature was approached because of the nature of the
experiment. The storage modulus represents the elastic portion which is why it starts out high,
and the loss modulus represents the viscous portion, which is why it starts out low. As the test
progressed, the composite moved closer and closer to the viscous portion which is why the loss
modulus increased and the opposite for the storage modulus. As it reached T
g
, the curves
dramatically change and the composite eventually enter the fully viscous portion and is
rendered useless in load bearing situations.

The second process was done mainly to determine the Fiber Volume Fraction of the
material. The average FVF was found to be 35.453% which is less than what was expected,
considering most composites have a FVF of around 40-70%. Also, the matrix volume fraction
was found based on the burned off weight of the matrix and had an average of 50.23%. Based
on Equation (2), this would lead to the determination of a void volume fraction of about 15%,
which is certainly too high. The void content is important to consider because it refers to the
spacing in between the fibers and matrix. Fatigue test results indicate that voids have a large
detrimental effect on the fatigue life of composite structures if the void content surpasses a critical
value. [6] Although it was not found in this lab, most composites have a void volume fraction of
under 5%

The void volume fraction was initially assumed to be around 2.5% to begin with, so
somewhere in the process the data can assumed to have been skewed. The potential sources of
error that would have altered this value could have come from the weighing of the samples.
One observation that was made during the experiment was the difference in fiber weights that
was found based on the two possible methods. The fiber weights were found after resin burn-
off by simply weighing what was left of the sample after it was heated. However, the fiber
10

weights could have been found by simply subtracting the crucible weight from the
crucible+fiber weight which was also weighed after heating. The two methods would
theoretically produce identical results, however an average difference of 0.0056g was found
between the two methods. This difference is intriguing, but may be explained by the
assumption that the lower magnitude of weight of the fibers was not accurately distinguished
when compared to the weight of the crucible. The direct weighing of the fibers is a better
approach that does not include this error.

The rule of mixtures, shown by Equation (3), can be used to determine the overall
properties of a composite based on the properties of the fibers and the matrix. It can be a good
rule of thumb, but it is not always precisely valid. The composite properties were not explicitly
calculated in this lab because of the many shortcomings of this rule, which are explained as
follows. First, the matrix and fiber structures could be altered in the fabrication of the
composite which may lead to initial residual stress in the material [7]. More importantly, the
differential contraction from cooling that occurs after fabrication may cause the matrix to begin
plastic deformation, which would dramatically change the overall properties of the composite
[8]. In short, the rule of mixtures should be used as a guideline but never an exact
measurement because the changes that may occur during fabrication will cause the properties
of the combined constituents to differ from the properties of the isolated fibers and polymers.

As mentioned, there may have been a few sources of error pertaining to the fiber
volume matrix and its byproducts. Averages were taken in most cases to diminish the effect of
human error in weighing and measuring the samples, but it is the nature of the experiment to
experience some error. Nevertheless, the DMA, done by a highly accurate machine, did
produce a glass transition temperature of 125.11C that was close to the literature value. This
T
g
value is more critical in determining applications for the composite anyway. If one was to
decide whether or not to choose Style 4180 Glass Fiber Reinforced Epoxy (Cycom

919), the
glass transition temperature should be the one of the first properties to consider. To be safe, if
the environment did reach around 100C, it would be unwise to choose this material
considering the consequences that could result if an outlier temperature was experienced near
T
g
. The composite would lose almost all effectiveness if it shifted into a rubbery state and
considering the load bearing applications it is used in, this may lead to disaster.




11

Conclusion
The analysis of Style 4180 Glass Fiber Reinforced Epoxy (Cycom

919) resulted in a
deeper understanding of composites and how their properties may or may not be suitable for
certain applications. Composites contain specific fibers that greatly increase the overall
strength of the material which makes them perfect for load bearing structures in the
automotive and aerospace industries. The fiber volume fraction is a crucial parameter that
roughly determines the increase in strength that the composite experiences. The FVF of the
material of interest was calculated to have a mean of 0.34543 with a standard of 0.00374. Also,
the glass transition temperature is the key parameter that predicts when the composite will
shift from a rigid, load bearing material to a fluid like state with no real effectiveness. This was
determined graphically by analysis of the temperature vs. storage modulus plot and found to
have a value of T
g
=125.11C.





References
[1] Progress in Aerospace Sciences. Soutis,C. Feb. 2005.
http://www.sciencedirect.com/science/article/pii/S0376042105000448
[2] http://en.wikipedia.org/wiki/Carbon-fiber-reinforced_polymer
[3] https://www.princeton.edu/~ota/disk2/1988/8801/880106.PDF
[4] Composite Material Characterization Lab Powerpoint
[5] Composites Science and Technology. Moll, Jericho L. April 2013.
http://www.sciencedirect.com/science/article/pii/S0266353813000717
[6] Composite Structures. Muller, Sergio Frascino. 1994
http://www.sciencedirect.com/science/article/pii/0263822394900442
[7] http://www.mech.utah.edu/~rusmeeha/labNotes/composites.html
[8] On the Applicability of the "Rule-of-Mixtures" to the Strength Properties of Metal-
Matrix Composites. Chawla, K. K.
1974http://www.sbfisica.org.br/bjp/download/v04/v04a25.pdf



12

Appendix A
Table 1
Sample 1 2 3 4 5 6 7 8 9
Sample weight (g) 0.4533 0.4707 0.3334 0.4551 0.4838 0.4541 0.4371 0.4461 0.4671
Sample Volume (cc) 0.2973 0.3074 0.2232 0.3048 0.3178 0.3015 0.2858 0.2920 0.3107

Table 3
Sample 1 2 3 4 5 6 7 8 9
Volume of Fibers
(cc) 0.1044 0.1065 0.0754 0.1044 0.1100 0.1037 0.0990 0.1016 0.1077

Table 6
Sample 1 2 3 4 5 6 7 8 9
Matrix
Weight (g) 0.1943 0.2067 0.1463 0.1963 0.2110 0.1969 0.1917 0.1941 0.2000
Matrix
Volume (cc) 0.1483 0.1578 0.1117 0.1498 0.1611 0.1503 0.1463 0.1482 0.1527