Zhu, J., Chandrashekhara, K., Flanigan, V., and Kapila, S

Composites: Part A 35 (2004) 95–101
www.elsevier.com/locate/compositesa
Manufacturing and mechanical properties of
soy-based composites using pultrusion
J. Zhua, K. Chandrashekharaa,*, V. Flanigana, S. Kapilab
a
Department of Mechanical and Aerospace Engineering and Engineering Mechanics, University of Missouri-Rolla,
101 Mechanical Engineering Building, Rolla, MO 65409 1350, USA
b
Department of Chemistry, Center for Environmental Science and Technology, University of Missouri-Rolla, Rolla, MO 65409, USA
Received 19 April 2003; revised 8 August 2003; accepted 20 August 2003
Abstract
Two primary cost driving factors for the composites industry are raw materials and labor. Inexpensive alternative epoxy resin systems
based on epoxidized soyate resins are developed for fiber reinforced composite applications. This research investigated on the manufacturing
and mechanical characterization of fiber/epoxy composites using chemically modified soy-based epoxy resins. Co-resin systems with up to
30 wt% soyate resins were used to manufacture composites through pultrusion. Mechanical tests show that the pultruded composites with soy
based co-resin systems possess comparable or improved structural performance characteristics such as flexural strength, modulus, and impact
resistance. Maximum mechanical properties enhancement is demonstrated by the enhanced epoxidized allyl soyate (EAS) formulation.
Further property improvement is obtained through using a two-step prepolymer process. The EAS holds great potential as partial supplement
for polymer and composites applications from renewable resources.
q 2004 Elsevier Ltd. All rights reserved.
Keywords: A. Glass fibres; A. Resins; B. Mechanical properties; E. Pultrusion
1. Introduction
Fiber reinforced polymer (FRP) composite materials
possess superior properties such as high strength and
stiffness to weight ratio, resistance to environmental
deterioration, high electrical insulation and low assembly
cost. Over the years, FRP composites have found
increasingly wide applications in a variety of industries
including the aerospace, transportation, sporting goods,
chemical engineering, construction and electrical industries. More recently, FRP composites are gaining
increasing market share in civil infrastructure applications due to their unique advantages over traditional
steel and concrete materials. Examples include the
external bonding of FRP plates on concrete buildings
for strengthening, and external confinement by wrapping
FRP shells on highway bridge columns to increase both
strength and durability. Emerging new construction
applications include FRP reinforcement for concrete
* Corresponding author. Tel.: þ 1-573-341-4587; fax: þ1-573-341-6899.
E-mail address: chandra@umr.edu (K. Chandrashekhara).
1359-835X/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compositesa.2003.08.007
structures, bridge decks, concrete-filled FRP tubes, and
FRP reinforcing tendons, etc.
Although this fast growing industry consumes over 20%
of total nationwide composite shipments, composite
materials represent only a very small percentage of the
entire civil infrastructure market. Widespread application in
this construction market is still limited by the cost of raw
materials like fiber and resin. With growing opportunities to
use pultruded composites in civil structural applications, the
development of increasingly cost effective raw materials is
of great interest. Innovative combination of FRP with other
inexpensive materials is becoming an active research
subject.
Triglyceride vegetable oils such as epoxidized soybean
oil (ESO) possess functional epoxy groups, which can react
with suitable reagents. They can be polymerized to form
elastomeric networks and are attractive raw material
resources for polymer synthesis [1]. Low cost soy-based
resins would offer a significant cost reduction through the
partial replacement of epoxy resin for composite applications. Furthermore, vegetable oil is biodegradable and,
therefore, environmentally friendly. With the depleting of
96
J. Zhu et al. / Composites: Part A 35 (2004) 95–101
the limited petroleum resources, use of this renewable
resource for industrial applications is attracting great
interest. Epoxidized vegetable oils therefore provide
potentially inexpensive alternatives to other traditional
petroleum-based resins. Another major concern for FRP
composites is that typical thermoset resin composites have
low impact damage resistance tolerance because of brittleness or low toughness of resins such as epoxy [2]. The
composite materials are susceptible to impact damage either
during assembly or in service. This restricts composite
applications to load levels far below their real capabilities
when the safety factor is considered. Modification of epoxy
resins to improve these properties has been the subject of
intense research interest. In the research at Eastern
Michigan University, prepolymerized ESO has been used
to toughen epoxy resins [3]. The ESO is comprised of long
chain fatty acid, which make the material more flexible
(Fig. 1). The mixture of ESO into epoxy resin resulted a
more ductile resin behavior. However, it also resulted in
some degradation on the mechanical strengths when used as
a toughening agent.
The oxyran functionality of ESO allows it to participate
in cross-linking with a suitable curing agent, therefore
allowing it to participate in cross-linking reaction along
with epoxy resin, as a result, offers opportunity to be
chemically incorporated into cross-linked matrix structure
rather than just act as plasticizer. The curing characteristic
and mechanical properties have been investigated using
modified epoxidized soyate resins as partial supplements in
conventional epoxy systems [4]. Several researchers have
studied the feasibility of manufacturing FRP composites
from soybean oil. Crivello et al. [5] reported the fabrication
and mechanical characterization of glass fiber reinforced
composites from ESO. Photopolymerization by ultra-violet
initiation was used to fabricate glass fiber reinforced multilayer composite laminates. A group of researchers from the
University of Delaware [6] investigated the fabrication of
natural fiber-reinforced composite from soybean oil derived
polyester. They were looking for a way to modify the
soybean oil chemical structure, to add more functional
hydrogen group into soybean oil by genetic engineering. Lu
and Stoffer [7] investigated the curing of commercial ESO
with various amine curing agents. A glass fiber reinforced
composite was fabricated using ESO as matrix and
mechanical properties were evaluated.
The feasibility of using ESO in the pultrusion
process was previously investigated at the University of
Fig. 1. Structure of fully epoxidized soybean oil.
Missouri-Rolla [8]. Addition of commercial ESO manufactured by Witco to the EPON epoxy resin showed
comparable properties over the base resin at a low
content of ESO. The objective of this research was to
explore the utilization of the enhanced soyate resins,
namely epoxidized allyl soyate (EAS) resin, and
epoxidized methyl soyate (EMS) as additives for the
composite materials manufactured through pultrusion
process. Co-resin systems with a higher content of
soyate resins up to 30 wt% were successfully used for
fabricating composites. Manufacturing characterization
and the effect of soyate resins on the mechanical
properties of final composites products were evaluated.
2. Experiment
2.1. Materials
Glass fiber reinforced composite specimens were fabricated through a pultrusion process with a varied of soyate
resin formulations. The base epoxy resin system was
comprised of Shell EPONq9500/EPI-CUREq9550. This
is a modified bisphenol epoxy resin specially designed for
pultrusion. Different weight ratios of ESO from Witco
Corporation, EAS and EMS, which were synthesized at the
University of Missouri-Rolla, were added to the base EPON
9500 system to form the soyate co-resin systems. Internal
lubricant Mold WIZ INT-PUL 24 from Axel Plastics and
clay filler ASP400P from Engelhard Corporation were
added to improve the process of pultrusion. The reinforcement for the unidirectional pultrusion was E-glass fiber
roving with 113 yields from Owen’s Corning. The resin
mixtures by weight are shown in Table 1.
2.2. Manufacturing of fiberglass composites by pultrusion
Pultrusion is the fastest and the most cost-effective
composite manufacturing processes, and is well suited to
high volume production for structural applications. Pultrusion technology also improves composite properties compared to other methods because the fibers are under tension
as the resin cures and are tightly bonded to each other [9].
When the composite material is put under a mechanical
Table 1
Formulation of resin system for pultrusion
Materials
Parts by weight (pbw)
Epoxy resin mixture
MOLD WIZ INT-PUL 24
ASP-400P
Air release BYK 55
EPI-CURE curing agent
100
1.0
10
0.2
Refer to Table 2
J. Zhu et al. / Composites: Part A 35 (2004) 95–101
97
Fig. 2. The schematic representation of the pultrusion process.
load, the load is taken immediately by the high strength
fibers.
FRP composite specimens were manufactured using a
Durapul 6000 Labstar pultrusion machine at the University
of Missouri-Rolla. The 36 in. length die has a cross-section
profile of 2 in. £ 0.125 in. (Fig. 2). Glass reinforced
composites were made with fiberglass rovings of 113
yield that provides a fiber content of 63% by volume. The
EPON 9500/9550 will cure at a die temperature ranging
from 170 to 190 8C. Maximum safe processing temperature
should be controlled under 200 8C to prevent the resin
premature at the die entrance area. The pulling speed was
12 in./min and varied slightly throughout the process. After
pultrusion, all test specimens were post cured for 90 min at
160 8C in an oven.
the trigger valve is released. By varying the pressure in
the system, the velocity of the ball and the kinetic energy of
impact are regulated. The timer uses two lasers signals to
record the time duration as the ball crosses the two signals.
This is used to calculate the velocity of the ball. All samples
were impacted with the same energy level of 37 J, which
produced propagated crack damage that was easily seen. For
unidirectional fiber oriented composite by pultrusion, the
crack initiated and grew along the fiber direction. The crack
lengths were measured and used for preliminary evaluation
on the impact properties.
2.3. Mechanical testing
The rectangular composite test specimens with dimensions of 10 in. £ 0.5 in. £ 0.125 in. were cut from pultruded
panels with a high speed router and edges were polished
with very fine aluminum oxide sandpaper. Tensile testing of
pultruded fiberglass composites specimens was performed
using an INSTRON 4204 machine. Each specimen was
tested to failure following the procedure in accordance with
ASTM D3039-97. Flexural strength and modulus were
tested in the same INSTRON machine following ASTM D790 standard. The specimens dimension is 5 in. £ 0.5
in. £ 0.125 in., the span was 2 in., and cross-head speed
was 0.1 in./min.
Impact testing is also one part of the overall program to
characterize new polymer composites by their physical and
mechanical properties [10]. In this study, a low velocity
impact test was used to compare the effect of the different
resin formulations on impact resistance of composites. A
custom-built gas gun ballistic apparatus (Fig. 3) was used to
fire a 3/8 in. diameter spherical steel ball on to the
composite panels. It consists of a pressurized nitrogen gas
tank, a pressure barrel, the gun barrel, a high-precision laser
timer, and supporting fixtures mounted at the end of the gun
barrel. The composite panel was mounted on the fixture and
covered by a projectile containing box. To fire the balls, the
chamber is pressurized to a prescribed level and then
Fig. 3. Layout of the gun barrel (a) with cross-section shown in (b).
98
J. Zhu et al. / Composites: Part A 35 (2004) 95–101
3. Results and discussion
3.1. Process feasibility for pultrusion
Many types of resins have been used for the pultrusion
process, including the thermoset resins (unsaturated polyester, vinylester, epoxy, phenolic, etc) and thermoplastic resin.
In general, a resin mixture for pultrusion must possess suitable
properties such as low viscosity (ranging from 500 to
2000 cps), high reactivity and long pot life in the impregnation
bath [11]. After adding epoxidized soyate resins, the properties of the co-resin mixtures change. In general, the viscosity
of the co-resin formulation drops as the proportion of soyate
resin additive is increased. The pure EPON 9500 epoxy has a
viscosity range of 4000– 9000 cps, whereas partially ESO has
a much lower viscosity of 94 –543 cps at room temperature
[12]. The reduced viscosity allowed resin to penetrate into the
fiber reinforcement easily and resulted in good wet-out.
Soyate resin prepolymer had similar viscosity to the EPON
epoxy resin and did not reduce the viscosity of co-resin
systems. Gel time and the pot life of the resin mixture were also
not noticeably affected by the addition of soyate resin
materials. Overall, the soyate co-resin systems with various
formulations performed well for the pultrusion process.
3.2. Mechanical characterization of pultruded composites
3.2.1. Using one-step blending co-resin systems
In the first stage of this research, the co-resin systems for
pultrusion were prepared through a one-step direct blending
process. The varied weight ratios from 10 to 30% of the
soyate resins ESO, EAS, and EMS were added in base
EPON epoxy. The effect of varied contents of soyate resins
on the mechanical properties of the FRP composites were
investigated.
3.2.1.1. Tensile and flexural properties. While the high
tensile strength and elastic modulus of the fiber is primarily
responsible for the strength and stiffness of a fiber reinforced
composite structure, the resin matrix provides the load
transfer between fibers, maintains fiber alignment, and
stabilizes the fibers against buckling. Although the mechanical properties of a composite part is dominated by the
amount and orientation of the fiber reinforcements, the resin
must be strong and stiff enough to adequately transfer the
load to the fibers for the structure to benefit from the
presence of the fibers. The main factor that affects the resin
properties is the cross-linking density. In addition, long
molecules with high molecular weight increase the intertwining and entangling of the molecular chains when large
and stiff molecular groups are attached to the backbone.
ESO possesses flexible nature due to long-chain molecular
structure. It contains functional epoxide rings, which are
capable of chemically participating in the cross-linking
reaction together with existing epoxy resins. The addition of
flexible ESO into epoxy resins would increase failure strain
of the resin.
The results of the mechanical testing of the pultruded
composites are shown in Figs. 6 and 7. The average ultimate
strength of specimens were almost the same as the neat
EPON epoxy resin at very low level 10 wt% of soyate resins
but began to decrease with the increasing content of the
soyate resins. Because soy-based resins have lower reactivity
and less functional epoxy group density, the cross-linking
density of the soyate co-resin systems will decrease when the
proportions of soyate resins increase. However, Young’s
modulus of the composites using soyate co-resins did not
show much change compared with the base epoxy resin when
low content soyate resins less than 20% were added.
Unlike tensile properties, the flexural properties of a
unidirectional composite structure are much more dependent on the nature of the resin because the fibers are
generally not oriented in the direction of the applied force.
Therefore, the resin makes the greater contribution to the
flexural properties than to the tensile properties. The
movement of molecules relative to each other is a major
factor in influencing the flexural strength and stiffness of the
polymer. In fact, higher proportions of soyate resins in the
new co-resin systems yielded a more flexible final product.
They may allow for further alignment of the fibers and
inhibit initial matrix crack formation and subsequent
propagation. As a result, the ESO and EAS co-resin systems
showed some improvement in flexural properties, especially
on the modulus at low level of soyate resin additives.
Compared to the commercial Witco ESO, the modified EAS
systems showed superior mechanical properties. These
experimental results demonstrated the greatest potential
for the EAS resin as a supplement to epoxy resin.
3.2.1.2. Impact properties. The impact response in fiber
reinforced epoxy composites reflects a failure process
involving crack initiation and growth in the resin matrix,
fiber breakage and pullout, delamination and disbonding. The
energy absorbing capability of the composites during impact
is therefore strongly dependent on the tensile strain capacity
of the resin and the interface between fiber and resin [13]. The
flexible resins derived from soybean oil have high strain-tofailure properties. The increased flexibility and reduced yield
strength could improve impact resistance of the composites
using the co-resins with the soyate resin additives.
In the impact testing, the damage was visually inspected.
The damage included a circular area radiating out from the
point of impact and cracks propagated along the unidirectional roving direction. It was found that the various
resin formulations produced different crack lengths. The
crack length can be used to qualitatively indicate the impact
resistance of each specimen. Fig. 4 gives the crack length
comparison for different resin formulations. The results
showed that specimens with soy-based resins ESO and EAS
had reductions in crack length of about 20% as compared to
the neat EPON formulations. The flexible ESO and EAS
J. Zhu et al. / Composites: Part A 35 (2004) 95–101
Fig. 4. Impact crack length of composites with various resin formulations.
resin systems improved the impact damage resistance of
composites. But the EMS formulation conversely gave high
impact crack length and showed the decreased impact
properties. This was because the EMS with low degree of
epoxidation made the matrix even more brittle than pure
EPON epoxy resin.
3.2.1.3. Effect of soy-based resin on pultrusion pulling
forces. Pulling force provides a relative measure of resistance
to pulling parts out of the die through pultrusion [14]. Under
normal operating conditions, overall forces can be affected
by numerous factors such as the viscosity of the resin, the
relative percentage of glass and resin, the coefficient of
thermal expansion of the gel and solid phase of the materials,
the length of the die, the cross-sectional area of the cavity,
and most importantly the degree of shrinkage of the resins.
Typically, epoxy resin is much more difficult to pultrude than
polyester and vinylester resins because epoxy shrinks very
little upon curing and is prone to stick to the die. A high
content of mold release additive is normally needed to
perform the epoxy composites pultrusion successfully [15].
In this research, an extra benefit using soyate co-resin system
was found, i.e. the pulling force during pultrusion process
was significantly reduced and epoxy composites pultrusion
process becomes easier. The pulling force values are
compared for various resin formulations as shown in Fig. 5.
This benefit is due to the good lubricity properties from the
oily fraction of soy-based resins.
3.2.2. Using two-step prepolymer co-resin systems
Due to the low reactivity of epoxidized soy-based resins,
the co-resin systems from one-step direct blending did not
achieve a high density cross-linking due to short curing time
during the pultrusion process. The utilization of an
epoxidized soy-based resin supplement was limited to low
levels below 20 wt% for pultruded composite without
sacrificing mechanical properties. In order to increase the
utilization of ESO, a two-step process was developed.
Epoxidized soyate prepolymers were prepared using amine
99
Fig. 5. Pulling forces of pultrusion with various resin formulations.
curing agent PACM and then blended into base EPON
epoxy [4]. The weight ratio 10/4 of all soyate resins to
PACM was used. The corresponding amounts of EPI-CURE
9550 were reduced based on the EPON epoxy amount minus
the portion that is assumed to react with unreacted
secondary amine of PACM. The resin formulation is listed
in Table 2. When preparing the prepolymer, the heating time
was controlled so that prepolymer viscosity was adequate
and pot life was long enough for pultrusion.
Figs. 6 –9 give the tensile and flexural properties of
pultruded composite specimens using the two different
processes. Considerable improvements on mechanical
properties are observed using the two-step mixture compared to the one-step mixture. The tensile and flexural
properties of the composites with 10 –20 wt% levels of EAS
and ESO were comparable with that of a base epoxy resin
system. At the higher content soyate resin up to 30 wt%, the
reduction in tensile and flexural properties was also much
less pronounced. For example, tensile strength only
decreased 8.1% from 1020 to 937 MPa for 30 wt% EAS
formulation. The flexural strength only decreased 5.7%
from 1448 to 1365 MPa. In comparison, the tensile strength
Table 2
Co-resin system formulations by weight ratio for two-step process
Resin
formulation
EPON
9500
(parts)
Epoxidized
soyate resins
(parts)
PACM (parts)
EPI-CURE
9550 (parts)
Neat EPON epoxy
10 wt% ESO
20 wt% ESO
30 wt% ESO
10 wt% EAS
20 wt% EAS
30 wt% EAS
10 wt% EMS
20 wt% EMS
30 wt% EMS
100
90
80
70
90
80
70
90
80
70
0
10
20
30
10
20
30
10
20
30
0
4
8
12
4
8
12
4
8
12
32.6
28.0
23.7
18.9
28.0
23.7
18.9
28.0
23.7
18.9
ESO
ESO
ESO
EAS
EAS
EAS
EMS
EMS
EMS
ESO, epoxidized soybean oil; EAS, epoxidized allyl soyate resin; EMS,
epoxidized methyl soyate resin.
100
J. Zhu et al. / Composites: Part A 35 (2004) 95–101
Fig. 6. Variation of tensile strength with soyate resin content by two
methods.
Fig. 9. Variation of flexural modulus with soyate resin content by two
methods.
functionalities, resulting in the higher cross-linking density.
In comparison, the EMS had the least epoxide functionalities
and thereby resulted in the lowest performance. Overall, the
EAS showed the greatest potential as an epoxy supplement
even with a lower degree of epoxidation than that of the
commercial ESO. It is expected that a higher content of EAS
resin over 30 wt% can be used as a epoxy resin supplement
for the composites, and even better mechanical properties can
be obtained once a high degree of epoxidation is achieved.
Fig. 7. Variation of flexural strength with soyate resin content by two
methods.
was reduced by 14.7% for 20 wt% EAS formulation when
the one-step mixtures were used. All tensile and flexural
modulus increased at low content of soy-based resins. These
strength improvements were attributed to the increased
incorporation of soy-based resins into the cross-linked
epoxy network by the prepolymerizing process, and therefore, resulting a stronger matrix.
Among the three types of soy-based resins, the EAS
formulation co-resin system consistently gave the highest
strengths for the pultruded composites. This was because
the EAS had the highest reactivity and more epoxide
Fig. 8. Variation of tensile modulus with soyate resin content by two
methods.
4. Conclusion
This research demonstrated a new method for utilization
of ESO in structural composites manufacturing by pultrusion. The alternative inexpensive soyate resins will lower
the composite material cost without limiting the mechanical
properties of the final products. The lubricity of soybean oil
significantly reduces the pull force. This program proved the
process feasibility of pultruding the composites with the
soy-based resin systems.
The test results indicated that pultrusion epoxy resin
formulation could contain up to 30 wt% soyate resins with
similar or improved mechanical properties. Furthermore,
the soyate resin additives increased impact resistance of the
pultruded composites because of the benefits of the added
elongation and ductility provided by the epoxidized soyate
resins. More mechanical properties enhancement was
demonstrated for the modified EAS formulation when
compared to a commercial ESO. Further property improvement was obtained through using a two-step prepolymer
method. Higher content of soy-based resins can be used as
partial supplements for conventional petrochemical-based
epoxy resins once higher degree of epoxidation is achieved.
The modified soyate resins from the renewable resources
showed considerable promise for widespread use in
the polymer and composite industries. For future work,
optimization of curing agent formulation will be conducted
to further improve the properties of soybased composites.
Also, environmental effects like moisture and UV exposures
will be investigated.
J. Zhu et al. / Composites: Part A 35 (2004) 95–101
Acknowledgements
This project was sponsored by Missouri Soybean
Merchandising Council and US Department of Agriculture.
References
[1] Gunstone FD. Chemistry and biochemistry of fatty acids and their
glycerides. London: Chapman and Hall; 1967.
[2] Strong AB. Molecular design advances in resin chemistry. Compos
Fabricat***ion 1998;November/December:1218–24.
[3] Frischinger I, Dirlikov S. Two-phase epoxy thermosets that
contain epoxidized triglycerides oil. Polym Mater Sci Engng 1994;
70:1–8.
[4] Zhu J. Development and characterization of soy-based epoxy resins
and pultruded FRP composites, PhD Dissertation, University of
Missouri-Rolla; 2002.
[5] Crivello JV, Narayan R, Sternstein SS. Fabrication and mechanical
characterization of glass fiber reinforced UV-cured composites from
epoxidized vegetable oils. J Appl Polym Sci 1997;64:2073–87.
[6] Williams GI, Wool RP. Composite from natural fibers and soy oil
resins. Appl Compos Mater 2000;7:421–32.
101
[7] Lu PP, Stoffer OJ. Composite prepared from epoxidized soybean oil.
Poly Prepr 2000;41:1340–1.
[8] Chandrashekhara K. Flanigan VJ, Berring NS, Unser JF. Pultrudable
resin from soybean oil, 44th International SAMPE Symposium; 2002.
p. 1857–65.
[9] Mallick PK. Fiber-reinforced composites: materials, manufacturing,
and design. New York: Marcel Dekker; 1993.
[10] Sjoblom PO, Hartness JT, Cordell TM. On low velocity impact testing
of composite materials. J Compos Mater 1988;22:30–52.
[11] Meyer RW. Handbook of pultrusion technology. New York: Chapman
and Hall; 1985.
[12] Muturi P, Wang D, Dirlikov S. Epoxidized vegetable oils as reactive
diluents. I. Comparison of vernonia, epoxidized soybean and
epoxidized linseed oils. J Polym Mater 1992;67:79– 80.
[13] Richardson M, Whisheart MJ. Review of low-velocity impact
properties of composite materials. Compos Part A 1996;27:
1123– 31.
[14] Price HL, Cupschalk SG. Pulling force and its variation
in composite materials pultrusion, polymers blends and composites in multiphase system. Washington, DC: C.D. Han; 1984.
p. 301 –22.
[15] Lackey LE, Vaughan JG. An analysis of factors affecting pull force for
the pultrusion of graphite/epoxy composites. J Reinf Plast Compos
1994;13:188 –98.