You are on page 1of 7

ME420 Composite Materials

Fall 2013
Dr. Surojit Gupta
12/10/13

Lab Assignment Part II


Composite Fabrication and Testing
University of North Dakota
Eric Wilderson

Summary of Results
The purpose of this laboratory exercise is to experimentally test a sample of composite beams,
evaluating the load and displacement at first ply failure, and then comparing the experimental results with
the mathematical calculations for the beams. Beams were constructed by hand using unidirectional and
bidirectional fiberglass (E-Glass) sheets and West Systems 105/206 epoxy resin and hardener. Two unique
layups were created using the supplied materials, with 8 test specimens created of each layup. After the
construction of the beams was completed, an analysis of the resulting loads at first ply failure and
deflection were completed using Matlab.
During the analysis stage, it was discovered that the beams constructed were not made per the provided
instructions, where the orientation of the ply(s) was the same between the 2 plates, have only either
bidirectional or unidirectional sheets used. Testing was attempted on the unidirectional samples, but could
not be tested to failure with the equipment available. The analysis estimated the first ply failure at a load
of ~145 lbs, which could not be accurately recreated. Comments on the results and details discovered
during testing are reported.

Methods and Procedures


Fabrication
The fabrication of the 2 composite plates was conducted using a composite materials practice kit supplied
by Aircraft Spruce & Specialty Co. The kit consisted of unidirectional and bidirectional cloths, epoxy
resin and hardener, mixing cups, measuring pumps and an assortment of other related materials. Each
layup would need to have a minimum of 10 plies, and needed to be symmetrical. Symmetrical layups
mean that the orientation of the plies are mirrored across the midplane of the layup. For these plates, both
the unidirectional and bidirectional plates would have the same symmetrical layup, consisting of 12 layers
with the orientation 0-90-0-90-0-90--90-0-90-0-90-0. With each cloth having a nominal thickness of .
008in, the resultant thickness should be ~.100 in thick, depending on the amount of resin and
compression used.
The fiberglass cloths were laid out flat and cut into ~12in squares using a rotary cutter on a sheet of
cardboard. The ply orientation is referenced using a tape strip on the edge of the cloth, and the cut
sections need to stay in the proper orientation after cutting to correctly place them in the layup. After
cutting the cloths into squares, plastic sheets are laid out, and a sheet of peel-ply is placed on top. A mold
release agent is sprayed onto the peel ply using squirt bottle until it is uniformly coated. The mold release
is allowed to dry before the first ply is laid down. Once the first play is laid down, the epoxy/resin is
poured on top, and spread generously over the sheet until the reflective shine of the fiberglass is no longer
visible, indicating that the epoxy has penetrated through. Another ply is then added in the proper
orientation (0 or 90 degrees from the previous ply) and the epoxy filling process is repeated. This process
is repeated, following the planned layup orientations until the final ply is coated in epoxy. The plates are
then covered with peel ply and plastic sheeting, also treaded with mold release in the same manner as the
bottom layer. The plates are then covered with a cardboard sheet, and compressed using 5 gallon paint
buckets filled with water. They are allowed to dry overnight (12+ Hours). It should be noted that the
temperature where the plates cured varied in temperature from 16 degrees to 50 degrees F, which is low to
outside of the recommended range for the West Systems 205 hardener (35+ Degrees F).
The finished plates were then cut into 1in X 10in test beams using a band saw. The thickness of the plates
was greater than estimated, where the average unidirectional beam thickness was 0.134 inches and the
bidirectional beams average thickness was 0.142 inches.
Analysis
The analysis of the beams was completed using Maximum Strain Criterion. A modified version of the
ME420 OrthoLaminateHT Matlab code was used to predict the outcome of the first ply failure load,
displacement and expected location of failure. The input parameters for the calculations, such as lamina
strengths were gathered from the Principles of Composite Material Mechanics 3 rd edition text book (Table
4.1). The Scotchply 1002 series E-Glass Epoxy values were used. This resulted in the estimated first ply
failure for 3 point bending at 194 Nmm/mm. This would translate is a load of ~149.2 lbf.
Testing
Due to the limited resources available to accurately test the beams, accurate testing was not able to be
completed. However limited 3 point bending tests were completed using 2 automotive jack stands, and
assorted weights. To support the beam under test, the two jack stands were positioned such that the
distance between the supports would be 8.0 inches. 25lb and 50 lb weights would be balanced on the
center of the beam to load the specimen for testing. The figures below show the test apparatus.

Figure 1: Testing Setup


As expected from the analysis, these loads were not able to cause any failure in the beam. To load the
beams additionally, the setup was placed on the floor, and a persons were allowed to step onto the center
of the beams. Their weights were ~ 130 lbs and 250 lbs. In each case, the beam did not experience failure.
To gather results for discussion, the beams were fractured by bending (in hand) or by
stepping/jumping/bouncing on the beams while over the test apparatus.
Discussion and Recommendations
While not able to accurately test the failure of the beams, interesting results were found during the testing.
It was determined during testing that the actual failure load of the beams was higher than calculated. This
conclusion is based upon the results of 2 persons standing on the beams without failure. However, the
person standing on the beam distributes the load over the beam more evenly and is not a point load, which
may have increased the load carrying ability of the test beams.
All of the beams were tested to failure, even though the load at the failure was not measured. The 2 types
of beams (unidirectional ply and bidirectional ply) did fail differently from each other. The unidirectional
ply beams would fail in a single position, always in the orientation perpendicular to the supported length
of the beam. This is likely due to the orientation of the first ply, where the fibers are parallel to the
fracture. Figure 2 below is a representation of 2 different single ply failures on a unidirectional beam.

Figure 2: Unidirectional Beam Ply Failure(s)


The failures seen on the bidirectional beams is significantly different than that of the unidirectional
beams. When the bidirectional beam begins to fail, the beams fibers begin to fail and crack/release across
the surface of the beam, before the fracture in the ply. This is due to the bidirectional weave in the cloth. It
is possible to see the hatching in the failed beams in Figure 3 that matches the pitch of the fibers in the
cloth.

Figure 3: Bidirectional Beam Ply Failure(s)

Overall, the experience of building the plates was an exciting introduction to composites. The fabrication
process developed a better understanding of the labor intensive process involved in the layup, which
proved to be very time consuming for producing just simple plates/beams. It was disappointing not being
able to accurately test the beams, and in retrospect, a better testing method could be constructed for the
distance students. Possibly supporting the beams between 2 tables, or stands, and using a spring scale and
various weights to apply the loads would have been produced more valuable data. Building beams with
fewer plies would have failed at a much lower applied load, and could have possibly been tested in the
manner described in this report. A hint to the distance students in this regard would have been worthwhile.
The amount of direction from the university on testing was lacking, but it should be expected that
students at this level in their coursework be resourceful enough to develop a testing method.

Attachment 1: Literature Review #1


Source: Botelhoa, E., Silvac, R., Pardinia L., Rezend, M. A Review on the Development and Properties of
Continuous Fiber/epoxy/aluminum Hybrid Composites for Aircraft Structures, Materials Research, Vol. 9,
No. 3, 247-256, 2006.
Reviewer: Eric Wilderson
This article presents the uses of newer/next generation Fiber/Metal Laminates (FML) in aircraft
structures, particularly Glare and Arall composites. FML composites such as Glare combine fiber/epoxy
composites with a metal layer, such as 7475 aluminum or 2024 aluminum alloys. The resultant product
has greater fatigue resistance, impact resistance and strength to weight ratios. Where many composites
have similar performance gains in the fatigue resistance and strength to weight ratios, FML composites
exhibit a higher impact resistance due to the metal layer.
The article specifically presents the uses of FML composites in aircraft structures, where the aerospace
industry has been developing FML type materials for structural applications. The Glare (trade name) for
FML composites was specifically targeted by ALCOA to be used in the aviation industry, which seems
clear in the abundant use of 2024 aluminum, which is one of the most common grades of aluminum used
in aircraft structures. 2024 aluminum is used in 8 out of the 9 grades of Glare FML. Based on the tables
provided, the different grades of Glare will use between 0.008 in and 0.020in thick aluminum layers in the
layup. The fiber/epoxy layers are designated as prepreg, most likely due to manufacturing reasons, as the
labor costs associated with hand layups would likely make the materials uneconomical for use.
An interesting perspective that is noted about the FML uses in the aircraft structures application is the
ability to form the material. Many aluminum parts used in aircraft structures are stamped or formed,
requiring the material to be yielded to form to the proper shape. This is one weakness that I see in the use
of these FML composites for aircraft structures. The costs associated with tooling and forming the
composite to the desired shape/contour would be a significant investment, however it is likely to be
within reason to traditional fiber matrix composites. Interestingly, the article does mention that studies
have estimated that FML materials are 5-10 times more expensive per kilogram than a traditional
aluminum alloy structure. This study also concluded that the weight savings from this material would be
in the range of 20%, which was an attractive position to the airplane builders in the study.
The article continues to cover the mechanical properties of the FML style composite. One except
describes the impact resistance of the FML as being greater than that of aluminum alone, and significantly
better than glass fiber composites, which is less than an aluminum alloy. The article describes an impact
in FML composites to leave a dent of the surface of the aluminum, which would lead one to believe that
the bond between the epoxy/fiber matrix to the aluminum sheet would be compromised, however this
issue of delamination between the aluminum and glass fiber is not directly addressed in the article from
the perspective of impact resistance. I found this odd, since this was a characteristic that portrayed the
material very positively.
Another interesting point that is described in the article is the environmental effects on the FML
composite. Water absorption is a sensitive issue for the FML, as the moisture absorption can affect the
bond between the fiber and epoxy matrix. Techniques are described that would combat eh moisture
intrusion, such as anodizing or painting the outer surface of the FML. Testing if Glare and Caral type
FMLs showed only .2% weight increase in high temperature 90% RH environments, where a
carbon/epoxy composite in the same environment absorbing 1.4% more moisture weight.

The article concludes by describing the development of new FML composites. The Glare composite from
ALCOA was introduced in the early 90s, and newer more advanced materials are becoming available as
of the time the article was published (2006). These new materials introduce Titanium and
Aluminum/Boron as newer more advanced FML configurations.

Attachment 2: Literature Review #2

Source: Grimshaw, M., Grant, C., Diaz, J. Advanced Technology Tape Laying For Affordable
Manufacturing Of Large Composite Structures.
Reviewer: Eric Wilderson
After reading the first article in FML composites in aerospace manufacturing, I researched composites
and prepreg uses in aerospace. This article was a good overview of the automated tape laying process, in
which a multiaxis CNC controlled gantry system is used to accurately dispense prepreg composite tape
onto surfaces. This was particularly interesting, as it gave a summary/history of the use of composites in
the commercial markets.
Automatic tape laying machines are a manufacturing tool developed in the 1970s to reduce the labor costs
of manually laying composite structure. While the development of the automated tape layers was initially
a military research project headed by General Dynamics, it soon worked into the commercial aircraft
industry. The scale of the automated tape layers is massive, many able to layup parts over 96in X 96in.
This is significant when compared to the size of traditional machining centers more commonly found in
manufacturing centers. While very large in size, the tape layers are geared toward flat, or very lightly
contoured parts. A traditional tape layer is an open bay gantry configuration, where the center of the
machine is open, so that a table, or tooling mold can be rolled into place and fixture to the floor for
manufacturing. The multiaxis head of the automated tape layer contains a spool of prepreg composite
tape. This tape is impregnated with an epoxy resin that need to be heated for application, which is done on
the head of the tape layer just before it is dispensed onto the mold. The tape laying machines have been
claimed to reduce the labor hours on layups by 70-85%, a significant cost savings of the generally labor
intensive layup of composite materials.
An interesting detail of the article is the tracking of installed automated tape laying machines around the
world. Because the market for these types of machines is so small, it is relatively simple to track the
installation and usage of these types of equipment. The article claims to know of only 45 machines
worldwide to be installed and in use as of the late 1990s. Of these 45 machines, 21 belong to the Boeing
Company. Most other companies, such as Bell Helicopter, Airbus, and NASA own 1-2 tape layers. It was
interesting to see companies such as Kawasaki and Fuji Heavy Industries own automatic tape layers.
The article then describes multiple products produced by the various aerospace manufacturers using the
automatic tape layers. The finished products are surprisingly well finished, as in my experience, most
composite layups require a significant amount of trimming and finishing after layup. An example of a
tape layer in process, and a finished part are shown in the images below.

You might also like