...

Anderson-Altan2012-PropertiesCompositedCylinders-+

by user

on
Category: Documents
2

views

Report

Comments

Transcript

Anderson-Altan2012-PropertiesCompositedCylinders-+
Properties of Composite Cylinders
Fabricated by Bladder Assisted
Composite Manufacturing
J. P. Anderson
e-mail: [email protected]
M. C. Altan
e-mail: [email protected]
School of Aerospace and Mechanical Engineering,
University of Oklahoma,
Norman, OK 73019
An innovative manufacturing method, bladder assisted composite
manufacturing (BACM), to fabricate geometrically complex, hollow
parts made of polymeric composite materials is presented. Unlike
the conventional bladder or diaphragm assisted curing processes,
BACM uses an internally heated bladder to provide the consolidation pressure at the required cure temperature. The feasibility of this
manufacturing method is demonstrated by fabricating laminated
composite cylinders using multiple cure pressures and number of
plies. The elastic moduli, failure strength, fiber volume fraction, and
void contents of the cylinders were found to be comparable to the
values obtained from flat laminates produced by hot plate molding
of the same material. Compared to conventional bladder manufacturing methods, the BACM process reduced the energy required to
cure the cylinders by almost 50% due to internal heating and insulated mold. [DOI: 10.1115/1.4007017]
Keywords: bladder molding, composite manufacturing, composite
tubes
Introduction
The manufacturing of high quality, geometrically complex, hollow parts made of composite materials has presented significant
challenges. Currently, pultrusion and filament winding are the two
most commonly used processes to create hollow, structural composite components that contain continuous fiber reinforcement. However, despite achieving good surface quality, low microstructural
defects, and excellent structural performance, composite parts fabricated using these processes are generally axisymmetric with limited
geometric complexity and often have uniform wall thicknesses.
To address the shortcomings of filament winding and pultrusion, the composites industry has used variations of bladder molding to fabricate hollow composite parts with increased geometric
complexity, nonaxisymmetric shapes, and varying wall thicknesses. In conventional bladder molding, an inflatable mandrel
(bladder) presses a composite material against a female metallic
mold, which is subsequently heated in an oven or autoclave to
process the part. This technique has been used with manufacturing
techniques, such as resin transfer molding (RTM) successfully
[1,2]. Lehmann and Michaeli [1] produced hollow composite
components using a braided tube preform and RTM/bladder molding and found several processing difficulties. Specifically, the
obtainable fiber volume fractions were limited by the permeability
of the preform, which was governed by bladder pressure. Hence,
the use of prepregs may be desirable to achieve higher fiber volume fraction in bladder molding, as resin infusion would not be
Contributed by the Materials Division of ASME for publication in the JOURNAL OF
ENGINEERING MATERIALS AND TECHNOLOGY. Manuscript received June 8, 2011; final
manuscript received May 29, 2012; published online August 9, 2012. Assoc. Editor:
Ashraf Bastawros.
necessary. Bladder molding has been used as well for both thermoplastic and thermoset prepregs [3,4]. Parts, such as “bubble”
golf shafts, composite bicycle frames, bicycle stems, and composite propeller blades, are examples of parts that have been produced
using conventional bladder molding [2,5–7].
Generally, in an oven/autoclave, heat is transferred by convection to a conductive mold as well as the tooling or bagging placed
around the part, which in turn, transfers the heat to the part to facilitate curing. As a result, a large amount of energy is expended in
heating both the air inside the oven/autoclave, tooling, and the part,
which increases operating costs. Because of the large capital investment required and the disadvantages discussed above, the use of an
autoclave based process generally results in higher part production
costs when compared to out-of-autoclave processes [8].
Finding a replacement for the autoclave curing process capable of
producing comparable part quality has been actively pursued due to
the disadvantages outlined above. The QuickstepTM process is to
date one of the most successful out-of-autoclave processes developed that produces parts with comparable quality to autoclave cured
laminates. The QuickstepTM process functions by curing a composite component between two fluid filled pressure membranes which
heat the mold/part [9]. Due to the way in which the part is heated,
the heating rate can be increased beyond the typical rate for autoclave processed parts. The QuickstepTM process solves the thermal
inertia and exothermic heating problems associated with the autoclave. However, the required capital investment remains high and a
large amount of energy is expended in heating the fluid chambers.
The bladder molding technique introduced herein, BACM eliminates the conventional external heating of the tooling/mold in an
oven or autoclave by moving the heating to the inside of the
mold/part. Due to the change in heating dynamics, the temperature of the composite part is controlled accurately by circulating
and venting air inside the bladder, thus improving the process
quality. Using BACM, geometrically complex, hollow composite
laminates can be fabricated with relative ease.
Materials and Experimental Procedure
Geometry and Materials Used. Cylindrical samples
(D ¼ 50.8 mm and L ¼ 101.6 mm) were made for this study using
NewportV 321/7781 prepreg. This prepreg is composed of a balanced, eight-harness woven E-glass fabric with a 121 C cure epoxy
system. The epoxy system used is toughened and has a glass transition temperature of 149 C [10]. Tomblin et al. [10] processed the
material in a vacuum bag with an absolute pressure of 76 kPa producing laminates with a nominal fiber volume fraction of 46% and
elastic modulus of 27.5 GPa. NewportV 321 epoxy is currently marketed for use in general aviation and aerospace applications and is
available with a variety of woven and unidirectional fibers [11].
R
R
Sample Fabrication Process. Two separate methods were
used to fabricate the cylinders in this study. The first was the conventional bladder molding technique, which used an oven to heat
an uninsulated aluminum mold externally. The air inside the bladder was not heated directly but was pressurized to provide the
consolidation pressure during curing. Four composite cylinders
were fabricated using this method while recording the energy consumption for each cycle. The goal was to obtain the average
energy consumption per cure cycle when curing is performed via
an external heating source, such as an oven or an autoclave. Later
the energy requirements of conventional bladder manufacturing
molding and the newly developed BACM process are compared.
The second curing method, the BACM process, utilized a bladder, heating/cooling module, and a thermally insulated mold made
of aluminum. Eager Plastics P50 room temperature vulcanizing
(RTV) silicon, which has a maximum operating temperature of
343 C, was used to produce the bladders used during curing of
the composite cylinders. The cylindrical bladders had an outer diameter of 47.6 mm and length of 203.2 mm. The heating/cooling
module was composed of a 450 W cartridge heating element
Journal of Engineering Materials and Technology
C 2012 by ASME
Copyright V
OCTOBER 2012, Vol. 134 / 044501-1
Downloaded 08 Oct 2012 to 129.15.98.220. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
inside the bladder and an air-pressure regulation system. This
heater can provide high heating rates during the BACM process,
thus enabling shorter cure times for fast cure resin systems or prepregs. The air pressure and flow inside the bladder was controlled
by an air regulator and an electronic air valve which intermittently
vented air to reduce the temperature gradient within the laminate.
The aluminum mold was insulated by incasing it in 20-mm of gypsum which has a thermal conductivity of 0.6 W/m C [12]. The
resulting mold assembly was composed of two layers: an outer minimally conductive gypsum casing and an inner aluminum shell
(6.4 mm thick). In Fig. 1, a schematic representation of the bladder,
heating/cooling module, and mold assembly is shown.
A number of 101.6 mm long composite cylinders with 2 -, 4 -,
and 6-plies were manufactured with bladder pressures of 207 kPa,
345 kPa, 483 kPa, and 621 kPa. The cylinders were laid-up by
wrapping a single piece of prepreg material around a mandrel,
while allowing for an overlap of 6.4 mm at the seam of the cylinders. The principal material axes were oriented along the length
and circumference of the cylinders. One thermocouple was placed
in the center of the laminates at each end of the cylinders during
layup for process control purposes. The cylinders were cured
using a cure cycle consisting of a 1 C/min ramp from 25 C to
121 C where they were held for 2 h before being cooled to 60 C
and removed from the mold as recommended by the supplier.
Two cylinders were fabricated for each of the pressure/ply combinations resulting in the production of 24 cylinders. A power meter
measured the total energy consumed during the manufacturing of
each of the cylinders. During the cure cycle, the average temperature deviation from the set-point was 3 C with a maximum deviation of 5 C, which indicates a thermally stable cure process.
In addition to the BACM process, 4-ply flat composite plates
were fabricated using the same cure cycle in a hot press. In this
case, compressive forces resulted in the laminates experiencing
consolidation pressures of 207 kPa, 345 kPa, 483 kPa, and
621 kPa, which are equivalent to the bladder pressures used in
fabricating the cylinders. The properties of these flat laminates
establish baseline values, which could be used to compare the
properties of the cylinders produced by BACM. These comparisons are expected to help assess the quality of the composite cylinders and thus validate the viability of the BACM process.
The bladder pressure maintained during the cure cycle and the
cure pressure applied during hot pressing will be referred to as
processing pressure while comparing the properties of the composite samples.
Void Analysis Procedure. Comparing void contents would be
useful in assessing the quality of the cylinders and flat laminates
produced. The density of the bulk composite specimens was determined by suspending the samples in heavy liquids diluted by
water (Cargill Labs 2.49 g/cm3 heavy liquid). The weight fraction
of both the fiber reinforcement and matrix was determined by
removing the matrix using the burn-off technique recommended
by ASTM D2584-08. Using the weight fractions and density
measurements, the void content of the specimens was calculated
as recommended by ASTM D2734-09. By using the outlined procedure, the void content was resolved within 6 0.22%. This level
of accuracy indicates the suspension/burn-off method is useful
even for composite laminates with void contents below 1%. It
should be noted that when applicable all experimental data was
reported with a 95% confidence interval.
Mechanical Testing. To demonstrate the quality of the hollow
parts made by the BACM process, flexural modulus and strength
of the composite cylinders and hot pressed laminates were determined. Six 12.7 mm (0.500 ) wide ring specimens from the center of
each cylinder were obtained resulting in 12 ring specimens for
each of the pressure/ply combinations. Three rings from each tube
were quartered producing 24 ring segments for each of the pressure/ply combinations. In addition to the cylinder specimens, ten
50.8 mm 12.7 mm (200 0.500 ) specimens were cut from each of
the four hot pressed laminates.
The elastic moduli of the rings were found by laterally compressing them between two flat platens using a Com-TenV 705TN
testing system. For quasi-static loading as in Ref. [13], the rings
were compressed to 10% of their original diameter between two
parallel platens at a rate of 2 mm/min. Six specimens were tested
for each of the pressure/ply combinations using this procedure.
The elastic moduli of the rings, E, were determined as recommended by ASTM D2412-10 a at a deflection of 5.08 mm using
Eq. (1)
R
E¼
0:149PR3
d 3
1þ
4R
Id
(1)
Fig. 1 BACM assembly illustrating the placement of the heating element, RTV bladder, and the composite prepreg within the cylindrical aluminum shell mold
044501-2 / Vol. 134, OCTOBER 2012
Transactions of the ASME
Downloaded 08 Oct 2012 to 129.15.98.220. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
where P is the applied load at deflection d, R is the mean radius,
and I is the moment of inertia. The flexural strength of the rings
was determined by testing ring segments in three-point bending as
described in Ref. [14]. The support span for all tests was set such
that a 16:1 or greater span to depth ratio was maintained and the
specimens were placed with the concave side down to reduce the
effect of geometric nonlinearities [15]. Twelve ring segment
specimens from each of the pressure/ply combinations were
tested.
The flexural properties of the flat laminates were also determined by testing specimens in three-point bending. The elastic
modulus and stress at failure were found using Eqs. (2) and (3),
respectively, as recommended by ASTM D790-10
E¼
L3 m
4bd3
(2)
rf ¼
3PL
2bd2
(3)
where L is the support span, m is the slope of load–deflection
curve, b and d are specimen width and depth, respectively.
Results and Discussion
Energy Consumption. The energy consumed during the curing of cylinders was reduced by approximately 50% by the
BACM process. Cylinders cured in an oven consumed on average
1.09 6 0.1 kW-h, whereas those cured by the BACM process consumed only 0.56 6 0.03 kW-h at 95% confidence interval. The
BACM process could be considered a cost-effective manufacturing process if the laminate quality and mechanical properties
could be shown as equivalent to a commonly used manufacturing
process, such as hot press molding.
Processing Pressure Effect. If a composite part is cured at
higher pressure, the fiber volume fraction is expected to increase,
as more resin is likely to be squeezed out during cure. Obviously,
if the resin flow is restricted during cure, the increase in fiber volume fraction may not materialize as expected. Figure 2 illustrates
the effect of pressure on the fiber volume fractions obtained from
both BACM and hot press cured parts. From the figure, the dependence of fiber volume fraction on the applied pressure is
clearly observed. In addition, the number of plies present in the
laminate was not found to affect the final fiber volume fraction
significantly, particularly at higher bladder pressures. Also note
that as the processing pressure is increased the fiber volume fractions of both the BACM and hot pressed parts increase at approximately the same rate until 621 kPa. At 621 kPa, the fiber volume
fraction of the BACM parts began to reduce while the hot pressed
parts continued to increase. This indicates that, as processing pressure is increased, a critical point is reached in which the RTV silicon bladder begins to restrict resin flow and eventually seal both
ends of the cylindrical part. The seal point will vary with composite thickness and appears to be located between 483 and 621 kPa
for the cylinders produced in this study. The sealing of the part by
the bladder would most likely result in an increase in the hydrostatic pressure of the resin and thus would compress and reduce
the size of the microvoids. Several studies have examined the
effect of resin pressure on void fractions and have concluded that
increases in resin pressure will serve to decrease the void fractions
of parts produced [16–18]. It should be pointed out that for the
BACM process, some of the excess resin is always squeezed out
between the mold halves. Hence, despite the sealing of the ends of
the composite cylinders, the resin flow is not totally eliminated
even at higher bladder pressures. The thorough pursuit of this concept for the BACM process is outside of the scope of this study
but further investigation is warranted.
Laminate Quality. It is important to consider void volume
fraction (i.e., often referred to as void content) when assessing the
quality of composite laminates. In industrial applications, the
overall void content of a laminate is often times utilized to determine if that component is acceptable for a particular application.
The component acceptance threshold usually varies depending on
intended use. For example, aerospace structures generally contain
less than 1% voids but for other less demanding applications, void
contents of up to 5% are acceptable [19–22]. Boey and Lye [21]
examined the effects of voids on the flexural strength and modulus
of composite laminates fabricated in an autoclave. They found
flexural strength to degrade by approximately 5% and the modulus
by 2% for every 1% of voids present. The applicability of the
BACM process for producing high performance components is
therefore reliant on its ability to produce components with low
void content at high fiber fractions.
The presence of voids will decrease the density of a composite
when compared to one with negligible voids and comparable fiber
volume fraction. As a result, the relative void fractions of composite specimens can be assessed from a specimen density versus
fiber fraction plot. In Fig. 3, the average densities of six tested
specimens for each of the pressure/ply combinations are shown
plotted against their average fiber volume fraction. In addition, the
Fig. 2 Effect of processing pressure on fiber volume fraction. Note that the
increase in marker size is used to illustrate increase in pressure applied during
processing.
Journal of Engineering Materials and Technology
OCTOBER 2012, Vol. 134 / 044501-3
Downloaded 08 Oct 2012 to 129.15.98.220. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Fig. 3 Density of specimen as the fiber volume % is increased. The average
void content is shown to be below 5% for all specimens. Note that the increase
in marker size is used to illustrate increase in pressure applied during
processing.
lines of theoretical density of the material with 0% and 5% voids
are shown. All the pressure/ply combinations fabricated using
both BACM and hot press manufacturing processes were found to
contain void fractions below 5%. Interestingly, the void fractions
seem to vary little for each sample group regardless of fiber volume fraction and processing pressure. A close examination of the
figure also seems to indicate that void fraction is slightly increasing with ply number for the BACM specimens. This may indicate
that as plies are added, air or moisture is increasingly trapped in
the prepared laminates. Kardos et al. [17] developed a void
growth model which showed that the application of an absolute
pressure greater than 35.5 kPa would prevent the formation of
voids by moisture. Specimens in this study were processed with
pressures well above this limiting pressure. Therefore, the voids
present in the processed specimens are less likely to be environmentally driven and are most likely caused by the entrapment of
air mechanically.
The void content in the processed specimens normalized with
ply number is shown in Fig. 4. The normalized void content of the
BACM specimens is shown to exhibit similar trends as processing
pressure is increased. The results for the 4-ply and 6-ply specimens appear to be identical but the 2-ply specimens depict a
slightly higher void content per ply. The lay-up process itself may
be the cause of this increase. Considering that wrapping the prepreg around the mandrel may be serving as a debulking process,
the addition of plies could cause greater normalized compaction
(i.e., lower per ply thickness), which would in turn reduce the volume of trapped air per ply [18]. It is also interesting to note that
above 207 kPa the void content/ply for the BACM specimens
remain nearly constant. This seems to indicate that the resin and
not the voids are removed as processing pressure is increased.
Cross-sectional micrographs of a select number of specimens
were taken to observe the morphology of the voids present and to
determine if evidence of void mobility by resin flow could be
found. A select number of these micrographs are shown in Fig. 5.
It is evident that as processing pressure is increased the morphology of the voids is changing. At 345 kPa, the voids present are
nearly circular and as pressure is increased to 621 kPa, the voids
tend to elongate as they become mobile or are pressed against
fiber bundles. This change in void morphology may be a result of
higher resin flow rates inside the laminate as processing pressure
is increased, as Hamidi and Altan [23] showed. Additionally, the
movement of the void agglomerates is expected to be restricted by
the fibers present in the material. Blackmore et al. [24] showed
experimentally that in order for bubbles/voids to detach in a slit
microchannel, the fluid drag force must overcome the bubbles
Fig. 4 Void % as processing pressure is increased normalized by number of
plies present
044501-4 / Vol. 134, OCTOBER 2012
Transactions of the ASME
Downloaded 08 Oct 2012 to 129.15.98.220. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Fig. 5 Representative void morphologies in composite cylinders fabricated by BACM method observed at 100 magnification.
Bladder pressure: (a) 345 kPa and (b) 621 kPa.
adhesion force. Barraza et al. [25] showed this adhesion force to
be reliant on bubble/void contact diameter. In general, as the contact perimeter of a void increases, it’s associated adhesion force
increases as well, resulting in a reduction in its ability to move.
Figure 5 and above discussion support the assertion that at higher
pressures the voids are becoming deformed and mobile with the
resin flow, but the elongated void morphology is hindering their
removal. The removal of the voids can further be facilitated by
using a bleeder cloth between the bladder and the prepreg. The
use of a porous bleeder and its effect on the fiber volume fraction
and the void content will be explored in a subsequent study.
Flexural Properties. The flexural properties of composite
components produced by different manufacturing process can be
used to assess the effectiveness of these processes in manufacturing high quality, structural components. By comparing flexural
properties, most process-induced effects, such as compaction,
volume fraction variations, voids, degree of cure, etc., are considered. As a result, the subtleties of process-induced properties
become evident. The effects of processing technique on the elastic moduli of the various specimens are shown in Fig. 6. As
expected, the elastic moduli of the samples show a clear uptrend
as the fiber volume fraction is increased. Moreover, it is shown
that the elastic moduli of the BACM processed specimens are
comparable to the hot pressed laminates and greater than that
reported by Tomblin et al. [10]. The improvement in moduli is
not surprising, as the mechanical properties reported were
obtained from laminates with a fiber volume fraction of 46%
[10]. However, the trend line fit to the BACM results (i.e.,
excluding the results of Tomblin et al.) depicts a good correlation with the stiffness reported by Tomblin et al. [10]. The moduli of the BACM specimens showed no discernable dependence
on ply number and were found to increase following the same
trend as the hot pressed specimens. The presence of voids did
not appear to have a significant effect on the elastic moduli of
the specimens. This is believed to be a result of the high fiber
volume fractions and fairly low amount of voids present in the
tested specimens. It is also well known that the elastic modulus
is primarily dependent on the fiber property [19,26–28].
It is worth noting that the effect of voids may not be significant
if the void content can be maintained below a certain level. For
example, Muller de Almeida and Santos Nogueira Neto [22]
developed a fracture model which showed that the presence of
voids in a composite laminate would not significantly affect its
strength until a critical void fraction is surpassed. This critical
fraction has been observed experimentally in Refs. [29,30] to be
between 2% and 3% and is dependent on laminate toughness, the
ratio of constituent shear moduli and Poisson’s ratio, and finally
the strength of a laminate with no voids present. The specimens in
this study were found to contain void fractions from a minimum
of 0.2% to a maximum of 3.4% with an average of 1.9%. Because
of this, the failure strengths of the tested specimens were placed
into two groups. The first consisted of specimens with void fractions below 2.0% and the second for those with void fractions
Fig. 6 Effect of processing technique and fiber content on measured elastic
modulus. The dashed line shown is a trend line fit to all data points except
Tomblin et al.
Journal of Engineering Materials and Technology
OCTOBER 2012, Vol. 134 / 044501-5
Downloaded 08 Oct 2012 to 129.15.98.220. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Fig. 7 Void effects on the flexural strength of composite specimens. The
dashed markers indicate specimen with void fractions above 2.0% threshold.
None of the hot pressed specimen exceeded the threshold value of 2.0%.
above 2.0%. The average void content of the first and the second
group was 1.19% and 2.27%, respectively. The failure strength of
the specimens tested and the trends of the associated void groupings are shown in Fig. 7. As expected, the failure strength of both
groups shows an upward trend as fiber volume fraction is
increased. The slopes of the low and high void groupings were
found to be 12.3 and 11.6 MPa/% fiber volume fraction, respectively. Hence, the rate of strength increase is not believed to be
affected by the presence of voids due to the very similar values of
these slopes. However, the presence of voids has caused the trend
lines of the two groups to be offset by 5% as void contents
increase above 2.0%. For example, the flexural strength of a sample that contains 50% fibers is estimated to be 695 MPa for the
group that has less than 2.0% voids. This value decreases to
652 MPa when the void fraction surpasses 2.0%.
Conclusions
A new, cost-effective manufacturing process, BACM has been
presented. The BACM process can fabricate hollow, geometrically complex structural composite components using a variety of
prepreg material. Unlike conventional oven or autoclave curing,
the heating of the composite to be cured is provided by hot air circulated inside of a pressurized bladder. The proposed process was
shown to produce composite cylinders: (i) in a more energy efficient manner when compared to traditional cure processes with an
external heat source; (ii) with void fractions and flexural properties comparable to hot pressed components. It was also shown that
the properties of the cylinders fabricated were not dependent on
number of plies. As a result, the BACM process can be considered
as a viable alternative to filament winding, pultrusion, or autoclave curing for the production of medium to large, structural
composite components.
Acknowledgment
The authors would like to acknowledge the donation of the prepreg material by Newport Adhesives and Composites Inc., a division of Mitsubishi Rayon Company Ltd. This work is partially
funded by a grant to SEAM Center at the University of Oklahoma
by the Oklahoma EDGE program.
References
[1] Lehmann, U., and Michaeli, W., 1998, “Cores Lead to an Automated Production of Hollow Composite Parts in Resin Transfer Moulding,” Composites, Part
A, 29A, pp. 803–810.
044501-6 / Vol. 134, OCTOBER 2012
[2] Ghiasi, H., Lessard, L., Pasini, D., and Thouin, M., 2010, “Optimum Structural
and Manufacturing Design of a Braided Hollow Composite Part,” Appl. Compos. Mater., 17(2), pp. 159–173.
[3] Salomi, A., Greco, A., Felline, F., Manni, O., and Maffezzoli, A., 2007, “A Preliminary Study on Bladder-Assisted Rotomolding of Thermoplastic Polymer
Composites,” Adv. Polym. Technol., 26(1), pp. 21–32.
[4] Visconti, I. C., and Langella, A., 1992, “Analytical Modelling of Pressure Bag
Technology,” Compos. Manuf., 3(1), pp. 3–6.
[5] Trimble, B. J., 1992, “Method of Making Composite Cycle Frame
Components,” U.S. Patent No. 5,158,733.
[6] Rebard, D., 2004, “Bladder Molding With Latex in the Recreational Industry
Lessons Learned,” Proceeding of 49th International SAMPE Symposium and
Exhibition, SAMPE, Long Beach, CA, pp. 79–82.
[7] McCarthy, R. F. J., Haines, G. H., and Newley, R. A., 1994, “Polymer Composite Applications to Aerospace Equipment,” Compos. Manuf., 5(2), pp.
83–93.
[8] Bader, M. G., 2002, “Selection of Composite Materials and Manufacturing
Routes for Cost-Effective Performance,” Composites, Part A, 33, pp. 913–934.
[9] Hodgkin, H. H., and Rabu, N., 2000, “A New Development in High-Speed
Composite Fabrication for Aerospace, Automotive, and Marine Applications,”
Proceedings of 45th International SAMPE Symposium and Exhibition,
SAMPE, Long Beach, CA, pp. 2274–2282.
[10] Tomblin, J. S., McKenna, J., Ng, Y. C., and Raju, K. S., 2001, “B-Basis Design
Allowables for Epoxy-Based Prepreg Newport E-Glass Fabric 7781/NB-321,”
AGATE-WP3.3-033051-097, NASA, Washington, DC.
[11] Raju, K. S., Dandayudhapani, S., and Thorbole, C. K., 2008, “Characterization
of In-Plane Shear Properties of Laminated Composites at Medium Strain
Rates,” J. Aircr., 45(2), pp. 493–497.
[12] Karni, J., and Karni, E. Y., 1995, “Gypsum in Construction: Origin and Properties,” Mater. Struct., 28, pp. 92–100.
[13] Gupta, N. K., and Abbas, H., 2000, “Lateral Collapse of Composite Cylindrical
Tubes Between Flat Platens,” Int. J. Impact Eng., 24, pp. 329–345.
[14] Prosen, S. P., Karpe, S., Kinna, M. A., Mueller, C., Perry, H. A., and Barnet, F.
R., 1963, “Compression, Fatigue, and Stress Studies on NOL Ring Specimens,”
Symposium for Filament Wound Reinforced Plastics, ASTM International, Naval Ordnance Laboratory White Oak, Silver Spring, MD, pp. 105–122.
[15] Guedes, R. M., and Sa’, A., 2008, “Numerical Analysis of Singly Curved Shallow Composite Panels Under Three-Point Bend Load,” Compos. Struct., 83, pp.
212–220.
[16] Hamidi, Y. K., Aktas, L., and Altan, M. C., 2005, “Effect of Packing on Void
Morphology in Resin Transfer Molded E-Glass/Epoxy Composites,” Polym.
Compos., 26(5), pp. 614–627.
[17] Kardos, J. L., Dave, R., and Dudukovic, M. P., 1988, “Voids in Composites,”
Manufacturing Science of Composites, T. G. Gutowski, ed., ASME, Atlanta,
GA, pp. 41–48.
[18] Browning, C. E., 1986, “Processing Science of Graphite/Epoxy Composites,”
Chem. Eng. Prog., 82(6), pp. 41–44.
[19] Liu, L., Zhang, B.-M., Wang, D.-F., and Wu, Z.-J., 2006, “Effects of Cure
Cycles on Void Content and Mechanical Properties of Composite Laminates,”
Compos. Struct., 73, pp. 303–309.
[20] Ghiorse, S. R., 1993, “Effect of Void Content on the Mechanical Properties of
Carbon/Epoxy Laminates,” SAMPE Q., 24(2), pp. 54–59.
[21] Boey, F. Y. C., and Lye, S. W., 1992, “Void Reduction in Autoclave Processing
of Thermoset Composites Part 1: High Pressure Effects on Void Reduction,”
Composites, 23(4), pp. 261–265.
[22] Muller de Almeida, S. F., and dos Santos Nogueira Neto, Z., 1994, “Effect of
Void Content on the Strength of Composite Laminates,” Compos. Struct., 28,
pp. 139–148.
Transactions of the ASME
Downloaded 08 Oct 2012 to 129.15.98.220. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
[23] Hamidi, Y. K., and Altan, M. C., 2003, “Spatial Variation of Void Morphology
in Resin Transfer Molded E-Glass/Epoxy Composites,” J. Mater. Sci. Lett., 22,
pp. 1813–1816.
[24] Blackmore, B., Li, D., and Gao, J., 2001, “Detachment of Bubbles in Slit Microchannels by Shearing Flow,” J. Colloid Interface Sci., 241(2), pp. 514–520.
[25] Barraza, H. J., Hamidi, Y. K., Aktas, L., O’Rear, E. A., and Altan, M. C., 2004,
“Porosity Reduction in the High-Speed Processing of Glass-Fiber Composites
by Resin Transfer Molding (RTM),” J. Compos. Mater., 38(3), pp. 195–226.
[26] Hagstrand, P. O., Bonjour, F., and Månson, J. A. E., 2005, “The Influence of Void
Content on the Structural Flexural Performance of Unidirectional Glass Fibre Reinforced Polypropylene Composites,” Composites, Part A, 36(5), pp. 705–714.
Journal of Engineering Materials and Technology
[27] Hancox, N. L., 1975, “The Compression Strength of Unidirectional Carbon
Fibre Reinforced Plastic,” J. Mater. Sci., 10, pp. 234–242.
[28] Lenoe, E. M., 1970, “Effect of Voids on Mechanical Properties of Graphite
Fiber Composites,” AVSD-0166-71-RR, U.S.DOT Navy, U.S. Naval Air Systems Command, Washington DC.
[29] Tang, J. M., Lee, I. W., and Springer, G. S., 1987, “Effects of Cure Pressure on
Resin Flow, Voids and Mechancial Properties,” J. Compos. Mater., 21, pp.
421–440.
[30] Stringer, L. G., 1989, “Optimization of the Wet Lay-Up/Vacuum Bag Process
for the Fabrication of Carbon Fibre Epoxy Composites With High Fibre Fraction and Low Void Content,” Composites, 20(5), pp. 441–452.
OCTOBER 2012, Vol. 134 / 044501-7
Downloaded 08 Oct 2012 to 129.15.98.220. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
Fly UP