Kutz M. Handbook of materials selection
Подождите немного. Документ загружается.
REFERENCES 1341
69. S. M. Spearing, ‘‘Materials Issues in Microelectromechanical Systems (MEMS),’’ Acta Metr., 48,
179–196 (2000).
70. A. M. Flynn, et al., J. Microelectromech. Syst., 1, 44 (1992).
71. S. N. Wang, et al., Sensor Mater., 10, 375 (1998).
72. D. J. Sadler, L. T. M., and A. C. H., ‘‘A Universal Electromagnetic Microactuator Using Magnetic
Interconnection Concepts,’’ J. Microelectromech. Syst., 9(4), 460–468 (2000).
73. C. L. Shih, et al., ‘‘A Robust Co-sputtering Fabrication Procedure for TiNi Shape Memory Alloys
for MEMS,’’ J. Microelectromechan. Syst. 10(1), 69–79 (2001).
74. B. J. Feder, ‘‘A Big Step Forward in Tiny Technology,’’ New York Times, 2000, p. C1.
75. C. R. Giles, D. J. Bishop, and V. Aksyuk, ‘‘MEMS for Light-wave Networks,’’ MRS Bull., 26(4),
328–329 (2001).
76. T. Laverghetta, T., Microwave Materials Fabrication and Techniques, 2 ed., Norwood, Artech
House, Norwood, MA, 1991.
77. J. S. Goela, et al., ‘‘Chemical-Vapor-Deposited Materials for High Thermal Conductivity Appli-
cations,’’ Mat. Res. Soc. Bull., 26(6), 458 (2001).
78. D. L. Wilcox, R. F. Huang, and R. Kommrusch, ‘‘The Multilayer Ceramic Integrated Circuit
(MCIC) Technology: An Enabler for the Integration of Wireless Radio Functions,’’ Adv. Mi-
croelect., (July / August) (1999).
79. P. K. Gallagher, et al., ‘‘Preparation of Mn-Zn Ferrites Having a Low Temperature Coefficient
of Permeability,’’ J. Am. Cer. Soc., 66(7), C110 (1983).
80. J. L. Vossen, and W. Kern (eds.), Thin Film Process, Academic, New York, 1978.
81. M. Ohring, The Materials Science of Thin Films, Academic, New York, 1992, p. 704.
82. H. Ehrenreinch, Fundamentals of Amorphous Semiconductors, National Academy of Sciences,
Washington, DC, 1972.
83. R. Glang, ‘‘Vacuum Evaporation,’’ in Handbook of Thin Film Technology, L. I. Maissel and R.
Glang (eds.), McGraw-Hill, New York, 1970.
84. R. C. Jaeger, Introduction to Microelectronic Fabrication, Vol. 5, Modular Series on Solid State
Devices, G. W. Neudeck and R. F. Pierret (eds.), Addison-Wesley, New York, 1988.
85. S. M. Sze, Semiconductor Devices: Physics and Technology,Wiley, New York, 1985.
86. G. H. B. Thompson, Physics of Semiconductor Laser Devices, Wiley, New York, 1980.
87. A. Y. Cho, ‘‘Recent Developments in Molecular Beam Epitaxy,’’ J. Vac. Sci. Tech., 16, 275
(1979).
88. M. A. Herman, and H. Sitter, Molecular Beam Epitaxy-Fundamentals and Current Status,
Springer, Berlin, 1989.
89. M. B. Panish, and H. Temkin, Ann. Rev. Mater. Sci., 19, 209 (1989).
90. R. A. Smith, Semiconductors, 2nd ed., Cambridge University Press, Cambridge, 1978.
91. M. Grayson, (ed.), Encyclopedia of Semiconductor Technology. Encyclopedia Reprint Series,
Wiley, New York, 1984.
92. C. G. Barrett, W. D. Nix, and A. S. Tetelman, The Principles of Engineering Materials, 1973,
Prentice-Hall, Englewood Cliffs, NJ, 1973.
93. R. A. Laudise, The Growth of Single Crystals, Prentice-Hall, Engelwood, NJ, 1970.
94. J. W. Mayer, and S. S. Lau, Electronic Materials Science: For Integrated Circuits in Si and
GaAs, Macmillan, New York, 1990.
95. G. O. Mallory, and J. B. Hajdu (eds), Electroless Plating, American Electroplaters and Surface
Finishers Society, Orlando, FL, 1990.
96. W. Riedel, Electroless Nickel Plating, ASM International, Metals Park, OH, 1991.
97. R. W. Filas, ‘‘Metallization of Silica Fibers,’’ Mat. Res. Soc. Symp. Proc., 531, 263–272 (1988).
98. R. Comerford, ‘‘Handhelds Duke It Our for the Internet,’’ IEEE Spectrum, August, 35 (2000).
99. V. K. Varadan, K. A. Jose, and V. V. Varadan, ‘‘Design and Development of Electronically
Tunable Microstrip Antennas,’’ Smart Mater. Struct., 8, 238 (1999).
100. S. C. Jordan, ‘‘Packaging Automation Drives WDM Component Assembly,’’ Laser Focus World,
37(2), 45 (2001).
1343
CHAPTER 44
USING COMPOSITES
Hans J. Borstell
Milledgeville, Georgia
1 INTRODUCTION 1343
2 EVALUATING POTENTIAL
PRODUCTS 1344
3 DIFFERENCES BETWEEN
COMPOSITES AND METALS 1348
4 MANUFACTURING 1348
5 THE DESIGN, MANUFACTURING,
AND QUALITY CONTROL
INTERFACE 1349
6 SELECTION OF MATERIAL
AND MANUFACTURING
CONCEPT 1350
7 DETAIL DESIGN 1356
8 PRODUCIBILITY CHECKLIST 1363
9 ETERNAL QUALITY CONTROL
QUESTION 1365
10 ENVIRONMENTAL CONCERNS 1367
1 INTRODUCTION
When I started working with fiberglass in 1962, many products such as small
boats were already in production. Most parts were built by hand labor using
rudimentary processes and nonoptimum resin systems.
When boron and carbon fibers became available, they offered the opportunity
to compete with aluminum for lightweight stuctures. I was fortunate to be part
of an integrated team dedicated to demonstrating the structural capabilities of
these fibers by the design, fabrication, and testing of representative aircraft com-
ponents.
The success of the group led to the opportunity to teach short engineering
courses in the United States and Europe. From these experiences I learned that
there was a lack of understanding of the interaction between engineering design,
materials selection, and manufacturing processes. There were many people who
knew what they wanted to achieve but had no understanding how to accomplish
their desires.
This chapter is intended to guide the reader through the complex process of
conceiving a suitable application for composites and then generating a produc-
ible and inspectable design that can be manufactured at a profit. If there are any
applicable catch phrases, they would be:
HandbookofMaterialsSelection.EditedbyMyerKutz
Copyright Ó 2002 John Wiley & Sons, Inc., NewYork.
1344 USING COMPOSITES
●
To surpass what is presently available, Imagine what is possible!
●
Cooperate within the product team to spread the burden of achieving goals
of quality and cost.
●
Perfection is not common; define what level of ‘‘bad’’ is acceptable.
●
Consider failure a learning experience rather than a mistake requiring
blame.
●
Try to fill a vacuum when there is no one providing your intended product
or service. (For example, a recent list of qualified contractors for the
restoration of concrete structures with composite patches had no listings
for Maine, New Hampshire, and Vermont. Surely there is a need for that
service in those states.)
Because it is impossible to predict the specific needs of every reader, I take
a broad-brush approach to describe how to convert a designer’s idea into a
profitable reality. This chapter discusses materials, processes, and applications
to enhance the reader’s understanding of the factors impacting composite design.
The composites industry is divided by product lines, which determine which
particular materials and processes are used in a particular plant. Available com-
binations of resin and fiber are increased by the choice of fiber form: continuous,
chopped, the many types of woven fabrics or braided preforms. Processes range
from hand labor to sophisticated computer-controlled mechanical marvels. Al-
though the combinations of materials and processes are numerous, much spe-
cialization exists because the technologies are too complex for one person, or
even one company, to master them all. However, the philosophy underlying a
successful composite application is always the same.
2 EVALUATING POTENTIAL PRODUCTS
Composites are truly wonder materials capable of being made into excellent
products, but only if they are utilized properly. That is a big ‘‘if.’’ Some notable
failures have occurred, such as attempts to use expensive aerospace materials
and techniques to replace thin-gauge aluminum sheet metal structures on utility
aircraft.
History offers interesting examples of successful composite applications. The
ancient Egyptians added straw to the mud used for bricks so that taller watch
towers could be built to provide earlier warning of attacks. Assyrian archers
increased the range of their arrows by switching from vines and leather for
bowstrings to tendons from the legs of antelopes, which are composites of fi-
brous cells and other cells that provide binding. Tendons have higher strength
and stiffness than either vines or leather, which are also composites. And as
electricity became more widely used, laminates of cotton cloth and phenolic
resins were used as insulators because the laminates were cheap, effective, and
easy to make.
The commercial introduction of fiberglass yarns in 1937 led to the establish-
ment of a huge industry with many applications based on the attractive attributes
of these materials:
2 EVALUATING POTENTIAL PRODUCTS 1345
●
Pleasure boats and tanks and pipes in corrosive environments based on
resistance to water and certain chemicals.
●
Printed circuit boards, radomes, and antenna dishes based on low electrical
conductivity.
●
Filament-wound tanks for firefighters’ breathing apparatus based on high
ratios of strength and stiffness to weight
●
Ducts supplying cooling air to the bays containing electronic systems on
military aircraft based on restrictions on part shape. The ducts are the last
thing fitted into these cramped spaces and thus must have complex shapes.
There is no shape limitation because they are molded on removable cast
ceramic tooling. Artificial mountains used in movies are made by draping
fiberglass cloth over wooden pegs, spraying on resins, and painting as
needed.
Composites offer a wide range of advantages:
●
Rather than having to make assemblies from individually made component
parts, an automobile manufacturer can mold an entire assembly, such as
a press-molded grill panel, which has tabs for attaching lamps and is
shaped to match a car’s contour.
●
Large structures can be made from composites, such as yachts longer than
100 ft. Using continuous processing and tubular fiberglass cloth, liners for
leaking water mains have been produced in lengths up to 1400 ft on site.
●
Processes exist to produce either a few prototype parts or a large produc-
tion run. A 40-ft-diameter orange, featured in the Florida building at the
1964 New York World’s Fair, was produced by spraying chopped glass
fiber and pigmented resin on an inflated war surplus weather balloon,
which served as a mold. Corrugated panels used for patio roofs are pro-
duced by a continuous pultrusion process in which dry mat and liquid
resin are converted to a stack of panels cut to required lengths by an
automated system.
●
An exotic application is the development of ablative plastics capable of
operating at 3000
⬚F for rocket motors. These materials slowly decompose
to form an insulating char that contains the gases that provide the thrust
for liftoff. High-temperature metals would have been too heavy and re-
quire cooling by the evaporation of excess fuel to prevent melting.
The accidental rediscovery in the early 1960s of boron fiber, originally dis-
covered in the 1870s, made possible composites that have significantly higher
strength and stiffness-to-weight ratios than metals, especially aluminum, which
was important for the aerospace industry. The excellent performance of the
F-14 Tomcat was due in part to the large weight reduction in the horizontal
stabilizer accruing from the use of boron skins. A low-volume application was
the development of expensive rods for dry fly trout fishing. Purchasers can
choose from a series of rods with different stiffness and dynamics characteristics,
thereby essentially tailoring the rod they select to their casting technique.
1346 USING COMPOSITES
In the late 1960s, some 30 companies around the world were attempting to
produce graphite fibers with properties similar to the boron fiber. Only 5 sup-
pliers in the United States, the United Kingdom, and Japan successfully pro-
duced reproducible graphite fibers. Subsequently, stiffer and stronger graphite
fibers were introduced. Recently, inexpensive graphite fibers for commercial ap-
plications have reached production levels of several million pounds a year.
Graphite fiber applications were based initially on testing of demonstration
aircraft and missile structures by the U.S. military, notably the Air Force Lab-
oratories in Dayton, Ohio. Since 1975, the composite content of military aircraft
and helicopters, as well as that of commercial jetliners, has increased steadily.
Applications ranging from fighter wings to interior shelves are based on the
following material characteristics: (1) ease of tailoring parts to load require-
ments, (2) high strength-to-weight ratios, (3) no fatigue degradation, (4) no cor-
rosion in service, and (5) high radar absorption.
Initial commercial applications of graphite epoxy fibers used off-grade aero-
space fibers with lower stiffness. Applications included golf club shafts, fishing
poles, tennis rackets, and drive shafts for pickup trucks. Variation in fiber price
negatively affected all these markets, however. The drive shafts performed well,
for example, but manufacturers went back to metal based on price. With the
introduction of low-cost commercial fibers, these sporting goods applications
have rebounded to be a multimillion dollar market.
Newer applications and the reasons for using graphite, now more commonly
called carbon, fibers include:
●
High-performance race cars (stiffness-to-weight ratio, crash worthiness)
●
Rods to stiffen elevated walkways (high modulus)
●
Picker sticks to remove drills from oil wells (tensile strength-to-weight
ratio)
●
Reinforcing corroded bridge columns to bring them back to design
strength (stiffness)
●
Bows for stringed instruments (stiffness)
Aramid fibers (Kevlar) are lighter than glass and have good tensile properties.
Applications include lightly loaded helicopter structures. In a ballistic applica-
tion, the kinetic energy of a bullet is expended in causing delamination of the
armor.
Spectra polypropylene fibers are used in chemical environments where glass
fibers are attacked, and in lightweight armor, notably helmets.
An effective nonstructural application uses pads of brittle, high modulus,
graphitized pitch fiber in heat sinks to cool densely packed airborne electronic
systems. The thermal conductivity to weight ratio of these fibers is five times
that of aluminum, and the fibers are 25% lighter.
Reasons why composites are substituted for other materials include:
2 EVALUATING POTENTIAL PRODUCTS 1347
●
To improve a product currently made from other materials by using the
unique properties of composites.
●
Fiberglass boats eliminate the need to scrape, caulk, and paint wooden
boats every 2 years.
●
To provide solutions for problem applications where other materials do
not work well or do not work at all.
●
An all-fiberglass shovel was developed to remove snow from electri-
fied railroad tracks.
●
To produce luxury items with performance superior to the competition.
●
The Ferrari F-40 had a carbon fiber body produced using techniques
applicable to a next-generation fighter plane. It went a little faster, cost
$1 million, and met demand by several years production at a low rate
to maintain exclusivity.
But once a wonderfully inventive product is conceptualized, reality sets in.
The designer must tear himself away from his CAD/NASTRAN computer and
join the project team in addressing the constraints in design implementation.
These include:
●
Production rate
●
Limits on product cost
●
Availability of processes and facilities
●
Capital required for mechanization/automation/enhanced process control
systems/quality control facilities
●
Environmental issues arising from materials selection
The best planned programs can still have problems that must be resolved by
cooperation between team members. Mr. Murphy is usually lurking nearby and
can strike at any time. Because we do not know everything about composites,
good ideas can turn into unsatisfactory products. For example, radio-frequency
(RF) energy is sent from the transmitter of a radar set to an antenna by means
of square tubes that must remain precisely the right size over a wide range of
temperature. Composites offered significant weight savings over the heavy metal
parts made from Invar, but composite waveguides were a failure because their
low coefficient of thermal expansion (CTE) was not low enough. However, the
CTE data from this effort showed that the CTE of certain graphite epoxy lam-
inates precisely matches that of mirrors of space telescopes. Now graphite/epoxy
egg crates support these mirrors in the precise orientation needed despite large
temperature swings. Considerable weight savings accrue because massive tita-
nium mounts and thermal strain isolators have been eliminated. Also, Invar tool-
ing is now widely used to produce composite parts to precise dimensions because
the tool expands very little as the resin is cooled at elevated temperatures. So
new applications sometimes arise from previous failures by understanding what
went wrong and using the acquired data in a more suitable application.
1348 USING COMPOSITES
3 DIFFERENCES BETWEEN COMPOSITES AND METALS
Composites and metals are different! Designers take care!
Alloy and heat treatment identify metals. Mechanical properties are well de-
fined and maintained during normal processing. In contrast, fiberglass/epoxy
can refer to several different types of materials with vastly different properties
based on the fiber length and form and the quantity of fibers parallel to the load
path. Laminate properties can be varied to match the applied loads by controlling
the fiber orientation of tape and fabric plies. Thus, the designer must match the
material and process used for his design allowables to that actually used in the
product. The required control is obtained by material and process specifications.
Composites are not ductile and do not yield before reaching their breaking
strength. Failing strains are much lower than those of metals. The design strain
for crossplied graphite/epoxy tape panels is 3000
in. per inch. Thus, the design
must eliminate high point loads such as those introduced by forcing mating parts
to fit. Load eccentricities must be avoided to prevent internal ply delamination.
Also, there is a significant interaction between the lay-up process, which is
geared to a certain fiber form, and part shape. Fiber form also impacts the
mechanical properties used in the design. For example, a small cylindrical tank
is best made by filament winding of continuous glass/carbon fibers because high
mechanical properties are combined with low cost. Irregularly shaped water lily
ponds used by gardeners are buried in earth so the weight of the water is sup-
ported. These items are best produced by spraying a chopped glass fiber/poly-
ester resin mix onto the mold. The low cost process is well suited to irregular
shapes, and the ponds do not require high mechanical properties.
Lastly, thickness variations of composite parts, typically
Ⳳ5%, may exceed
those of metal parts, especially as part thickness increases. Also, each part can
have areas that are either above or below nominal thickness. These variations
are due to variations in the weight of resin and fiber per unit area in the materials,
and resin movement during processing. Sometimes an extra ply is required to
carry the load when minimum part thickness is inadequate. Thickness variations
complicate assembly by fastening or bonding because composite parts cannot
be forced into shape without damage.
4 MANUFACTURING
As a preview to the major discussion of this chapter, a brief description of the
manufacturing operation is desirable. The steps are:
●
Clean the mold and apply a release agent to prevent the part from sticking.
●
Prepare plies.
●
Put the plies on the mold in the specified order (to reduce warpage and
obtain optimum mechanical properties).
●
Cure the resin by chemical reaction; the resin is transformed from a liquid
or paste to a brittle solid.
●
Remove the part from the mold.
●
Trim, assemble, and paint.
5 THE DESIGN, MANUFACTURING, AND QUALITY CONTROL INTERFACE 1349
Accommodation
• Manufacturing accepts that
there is a problem when QC
reports it
• QC clearly explains what the
defect/discrepancies are without
hyperbole
• QC expedites inspection of
experimental parts to accelerate
problem solving
Requirements
• Optimal processes/tooling
• Optimize factory operation
• Conform to QC plan
• Project
Secret Wish
• Make it simple and
just like what we
know how to do
and have always
done
Secret Wish
• This has to work even if manufacturing
screws it up
Secret Wish
• We are not going to
ship anything which
isn’t perfect.
Scrappage forces
manufacturing to
improve
Requirements
• Implement QC plan
• Document defects
and repairs
• Highlight recurring
problems
Requirements
• Sturctural integrity
• Popular styling
• Producibility/design cost
• Low cost materials which
process reproducibly
Accommodation
• Design still meets requirements and is
easier to make
• Design revisions to overcome
persistant manufacturing problems
Accommodation
• Confirm inspectability before design release
• Provide allowable defect criteria
• Respond to need for redesign to resolve
repetitive defects
Engineering
Quality ControlManufacturing
Program
Objectives
Met
Fig. 1 Design, manufacturing, and quality control interfaces.
There are three differences from metallic manufacturing:
●
Metal parts usually start with constant thickness stock and material is
removed in low-stress areas. Composites add plies in high-stress areas to
a constant thickness lay-up designed to withstand minimum loads on the
part.
●
Except for metallic diffusion bonding, it is difficult to produce an assem-
bly in one operation from sheet stock. Composites can be molded easily
into assemblies.
●
Sandwich construction, especially with skins having several thicknesses,
is difficult and expensive to produce in metals. The parts must fit together
to obtain a structural bond because only the film adhesive can deform
during adhesive curing. However, composite sandwich panels using a va-
riety of core materials are produced in huge quantities by the cure process
where the adhesive and skin materials are forced to fit snugly on the core
before the resins solidify by curing.
5 THE DESIGN, MANUFACTURING, AND QUALITY CONTROL
INTERFACE
A product development team consists of representatives from four departments
whose functions are well defined. Management has the responsibility for sched-
ule, cost, and quality and functions as a referee among the other three groups
whose objectives and wishes are shown in Fig. 1. Engineering, manufacturing,
and quality control usually are strong organizations that are protective of their
1350 USING COMPOSITES
function and are provincial in outlook. Their objectives are not always compat-
ible with the needs of the program.
For the success of the program, there must be a sense of give and take to
enable each department to achieve its objectives more easily. As an example, a
senior vice president at my employer believed in high-performance composites
when they were in their infancy in 1966. He directed that an integrated team be
formed with all the skills required so that we would work together with these
materials under his prodding. Whether it was his directive or the thrill of break-
ing new ground, each department performed its tasks with a high level of co-
operation and concern for other departments’ problems. Perhaps that is why we
built more types of prototype structures for the first time than anyone else, and
they all met the design requirements in full-scale structural and flight tests.
Subsequent production programs got off to a rocky start, however, because par-
ticipants were more interested in protecting their vested interests than following
the method of operation used in the development phase.
6 SELECTION OF MATERIAL AND MANUFACTURING CONCEPT
There are four classes of fiber, eight types of matrices, several fiber forms, and
eight general manufacturing processes. At first glance, this array of options is
bewildering. Often, however, choices are limited. The selection of material and
manufacturing process is controlled by design factors such as loading, weight
limitations, operating temperature, and perceived value. In addition, customer,
government, and industry specifications may apply. Special requirements, such
as the increasing use of phenolic materials in trains and aircraft to reduce fire
and smoke in case of accident, may apply.
Part shape significantly influences process selection. Fiberglass channels for
ladders are a perfect application for pultrusion because they have a constant
cross section. Skins for wind turbine blades must have a smooth aerodynamic
outer surface, and inner surfaces must fit to the substructure within glue line
tolerances. The skins are made by resin injection molding in a closed mold to
obtain two controlled surfaces. Resin injection tooling/facilities are less expen-
sive than the alternative of press molding. Parts of complex shape are either
vacuum bag or autoclave molded because only one tooled surface is needed for
these processes.
Despite these instructions, a designer should take a ‘‘clean-sheet’’ approach
and evaluate the factors shown in Fig. 2, which includes both engineering and
manufacturing requirements. To assist in material selection, it should be noted
that higher properties are obtained from purchased compounds or prepregs.
These materials contain matrices that cannot be prepared on the shop floor. If
no other data are available, vendor data sheets for these products are a good
source of material properties.
At the outset, the intrinsic nature of the various fiber/resin combinations elim-
inate many possibilities. Service temperature is often the dominant factor. In a
process similar to the conversion of ice into slush, the matrix softens due to a
thermally induced loss of internal bonding. The softened matrix reduces the
ability of the resin to transfer load from one fiber to another. Typically, a very
significant decline in laminate properties occurs within a small range of tem-
6 SELECTION OF MATERIAL AND MANUFACTURING CONCEPT 1351
Fig. 2 Factors affecting the selection of materials and processes.
Table 1 Typical Service Temperatures for Resins
Matrix Service Temperature (⬚F)
Polyester and vinyl ester 150
Epoxy on carbon fibers 260
a
Epoxy on glass fibers 350
BMI (BisMaleIrnide) 400
Phenolic 500
Addition polyamide 550
Condensation polyamide 600
Silicone laminating resins 600
a
Service temperature reduced to compensate for moisture
pickup.
peratures, when the resin is no longer able to function. Typical service temper-
atures for resins are shown in Table 1.
When service temperature exceeds 150
⬚F for parts having significant applied
loads, polyester/vinyl ester type resins are not an option. Epoxy resins have
better thermal stability and improve laminate properties. Epoxy resin mixes are
of three types: liquid mixes prepared on the shop floor, liquid compounds with
better properties prepared by a specialized supplier, and prepreg or molding
compounds.
Liquid epoxy resins with enhanced thermal stability are available for filament
winding and resin injection parts. In this process, a dry fibrous preform is en-
closed in a mold and liquid resin is pumped into the mold to fill the cavity
completely. Except for these two cases, all the higher temperature materials use
preimpregnated fiber/resin combinations. Difficult-to-handle ingredients can be
blended and uniformly applied to the reinforcement to produce unidirectional
tape or woven fabric prepregs. For example, a widely used matrix for graphite