1

A study of Composite materials and

Structures for Aircraft applications

with Damage Analysis in

1 Fighter + 1 Transport Aircrafts.

A REPORT BY:

MEENAL DUTT

Under the guidance of Gp. Capt. Praveen Khanna

A report submitted in

partial fulfillment of the requirements

of MINOR PROJECT.

16th JUNE 2011

2

Contents

ACKNOWLEDGEMENTS ............................................................................................................................ 4

1. Introduction........................................................................................................................................ 5

2. Carbon-fiber-reinforced polymer (CFRP) ............................................................................................. 7

3. Glass-reinforced plastic (GRP) ............................................................................................................. 8

4. The Use of Composites in Aircraft Design ............................................................................................ 9

4.1 Popular Examples of composite use ............................................................................................. 10

5. Basic principles (micromechanics) of fiber composite materials ........................................................ 11

6. Moulding Methods ............................................................................................................................ 12

6.1 Vacuum bag moulding ................................................................................................................. 12

6.2 Pressure bag moulding ................................................................................................................. 13

6.3 Autoclave moulding ..................................................................................................................... 14

6.4 Resin transfer moulding (RTM) ..................................................................................................... 14

6.5 Other ........................................................................................................................................... 15

7. Types of Composite Materials ............................................................................................................ 16

7.1 Matrices ...................................................................................................................................... 16

7.2 Polymers ...................................................................................................................................... 16

7.3 Metals ......................................................................................................................................... 18

7.4 Ceramics ...................................................................................................................................... 19

7.5 Polymer Matrix Composites ......................................................................................................... 21

7.6 Non-polymeric Composite Systems .............................................................................................. 21

7.6.1 Metal-Matrix Composites ...................................................................................................... 21

7.6.2 Particulate MMCs.................................................................................................................. 23

7.6.3 Ceramic-Matrix Composites .................................................................................................. 24

7.7 Hybrid Metal/PMC Composites .................................................................................................... 25

8. Application and Operating Experience .............................................................................................. 27

8.1 US. Air Force (T. Reinhart) ............................................................................................................ 27

8.2 US. Navy (D. Mulville) .................................................................................................................. 27

8.3 U.S. Army (P. Haselbauer) ............................................................................................................ 28

9. Turbine inlet guide vanes ................................................................................................................... 29

3

10. Failure Analysis of Composite Structures in Aircraft Accidents ......................................................... 31

10.1 Examination of Failed Metallic Structures .................................................................................. 32

10.1.1 Tension ............................................................................................................................... 36

10.1.2 Compression ....................................................................................................................... 37

10.1.3 Bending ............................................................................................................................... 38

10.1.4 Impact ................................................................................................................................. 39

10.1.5 Fatigue ................................................................................................................................ 40

11. Testing ............................................................................................................................................. 42

11.1 Composite FEA ......................................................................................................................... 42

11.2 Ultrasonic Testing ..................................................................................................................... 43

11.3 Thermography ........................................................................................................................... 43

11.4 Shearography ............................................................................................................................. 44

11.5 X ray Radiography ..................................................................................................................... 45

12. Composite Repair ............................................................................................................................ 46

12.2 Typical Damage .......................................................................................................................... 49

12.3 Laminates and Sandwich Panels ................................................................................................. 49

12.3.1 Patch repair ......................................................................................................................... 49

12.3.2 Taper sanded or scarf repair................................................................................................. 50

12.3.3 Step sanded repair ............................................................................................................... 50

12.4 Typical Laminate Repairs ............................................................................................................ 51

12.5 Typical Sandwich Panel Repairs .................................................................................................. 52

12.6 Repair Sequence for Double Sided Repair .................................................................................. 53

12.7 Equipment and Ancillaries for Repairs ........................................................................................ 54

12.8 Repair Process ........................................................................................................................... 56

13. Use of composites in the Advanced Light Helicopter (ALH) ............................................................... 57

14. Conclusion ...................................................................................................................................... 59

BIBLIOGRAPHY ...................................................................................................................................... 60

4

ACKNOWLEDGEMENTS

I would like to thank my guide Gp. Capt. Praveen Khanna, for initiating the work on this

project. His guidance has made me work efficiently and diligently. This would not have been

possible without the constant encouragement and motivation from my guide.

I am obliged to the director of AIAERS, O.P. Varshney for keeping the course Minor Project

available to the students.

I extend my thanks to everyone who helped me during the preparation of this project.

Name: Meenal Dutt

Enrolment number: A3705507015

5

1. Introduction

The unrelenting passion of the aerospace industry to enhance the performance of commercial and

military aircraft is constantly driving the development of improved high performance structural

materials. Composite materials are one such class of materials that play a significant role in

current and future aerospace components. Composite materials are particularly attractive to

aviation and aerospace applications because of their exceptional strength and stiffness-to-density

ratios and superior physical properties.

A composite material typically consists of relatively strong, stiff fibres in a tough resin matrix.

Wood and bone are natural composite materials: wood consists of cellulose fibres in a lignin

matrix and bone consists of hydroxyapatite particles in a collagen matrix. Better known man-

made composite materials, used in the aerospace and other industries, are carbon- and glass-

fibre-reinforced plastic (CFRP and GFRP respectively) which consist of carbon and glass fibres,

both of which are stiff and strong (for their density), but brittle, in a polymer matrix, which is

tough but neither particularly stiff nor strong. Very simplistically, by combining materials with

complementary properties in this way, a composite material with most or all of the benefits (high

strength, stiffness, toughness and low density) is obtained with few or none of the weaknesses of

the individual component materials.

CFRP and GFRP are fibrous composite materials; another category of composite materials is

particulate composites. Metal matrix composites (MMC) that are currently being developed and

used by the aviation and aerospace industry are examples of particulate composites and consist,

usually, of non-metallic particles in a metallic matrix; for instance silicon carbide particles

combined with aluminium alloy. Probably the single most important difference between fibrous

and particulate composites, and indeed between fibrous composites and conventional metallic

materials, relates to directionality of properties. Particulate composites and conventional metallic

materials are, nominally at least, isotropic, i.e. their properties (strength, stiffness, etc.) are the

same in all directions, fibrous composites are anisotropic, i.e. their properties vary depending on

the direction of the load with respect to the orientation of the fibres. Imagine a small sheet of

balsa wood: it is much easier to bend (and break) it along a line parallel to the fibres than

perpendicular to the fibres. This anisotropy is overcome by stacking layers, each often only

fractions of a millimetre thick, on top of one another with the fibres oriented at different angles

to form a laminate. Except in very special cases, the laminate will still be anisotropic, but the

variation in properties with respect to direction will be less extreme. In most aerospace

applications, this approach is taken a stage further and the differently-oriented layers (anything

from a very few to several hundred in number) are stacked in a specific sequence to tailor the

properties of the laminate to withstand best the loads to which it will be subjected. This way,

material, and therefore weight, can be saved, which is a factor of prime importance in the

aviation and aerospace industry.

Another advantage of composite materials is that, generally speaking, they can be formed into

more complex shapes than their metallic counterparts. This not only reduces the number of parts

making up a given component, but also reduces the need for fasteners and joints, the advantages

of which are twofold: fasteners and joints may be the weak points of a component a rivet needs a hole which is a stress concentration and therefore a potential crack-initiation site, and

fewer fasteners and joints can mean a shorter assembly time.

6

Shorter assembly times, however, need to be offset against the greater time likely to be needed to

fabricate the component in the first place. To produce a composite component, the individual

layers, which are often pre-impregnated (pre-preg) with the resin matrix, are cut to their required shapes, which are all likely to be different to a greater or lesser extent, and then stacked

in the specified sequence over a former (the former is a solid or framed structure used to keep the

uncured layers in the required shape prior to, and during, the curing process). This assembly is

then subjected to a sequence of temperatures and pressures to cure the material. The product is then checked thoroughly to ensure both that dimensional tolerances are met and that the curing

process has been successful (bubbles or voids in the laminate might have been formed as a result

of contamination of the raw materials, for example).

7

2. Carbon-fiber-reinforced polymer (CFRP)

Carbon-fiber-reinforced polymer or carbon-fiber-reinforced plastic (CFRP or CRP), is a very

strong and light fiber-reinforced polymer which contains carbon fibers. The polymer is most

often epoxy, but other polymers, such as polyester, vinyl ester or nylon, are sometimes used. The

composite may contain other fibers such as Kevlar, aluminium, glass fibers as well as carbon

fiber.

Although it can be relatively expensive, it has many applications in aerospace and automotive

fields, as well as in sailboats, and notably finds use in modern bicycles and motorcycles, where

its high strength-to-weight ratio and good rigidity is of importance. Improved manufacturing

techniques are reducing the costs and time to manufacture, making it increasingly common in

small consumer goods as well, such as laptops, tripods, fishing rods, paintball equipment,

archery equipment, racquet frames, stringed instrument bodies, classical guitar strings, drum

shells, golf clubs, and pool/billiards/snooker cues.

8

3. Glass-reinforced plastic (GRP)

Glass-reinforced plastic (GRP), also known as glass fiber-reinforced plastic (GFRP), is a fiber

reinforced polymer made of a plastic matrix reinforced by fine fibers of glass. It is also known

as GFK (for Glasfaserverstrkter Kunststoff), or simply by the name of the reinforcing fibers

themselves: fiberglass.

GRP is a lightweight, extremely strong, and robust material. Although strength properties are

somewhat lower than carbon fiber and it is less stiff, the material is typically far less brittle, and

the raw materials are much less expensive. Its bulk strength and weight properties are also very

favorable when compared to metals, and it can be easily formed using molding processes.

The plastic matrix may be epoxy, a thermosetting plastic (most often polyester or vinyl ester)

or thermoplastic.

Common uses of fiber glass including boats, automobiles, baths, hot tubs, water tanks, roofing,

pipes, cladding and external door skins.

9

4. The Use of Composites in Aircraft Design

Among the first uses of modern composite materials was about 40 years ago when boron-

reinforced epoxy composite was used for the skins of the empennages of the U.S. F14 and F15

fighters.

Initially, composite materials were used only in secondary structure, but as knowledge and

development of the materials has improved, their use in primary structure such as wings and

fuselages has increased. The following table lists some aircraft in which significant amounts of

composite materials are used in the airframe. Initially, the percentage by structural weight of

composites used in manufacturing was very small, at around two percent in the F15, for example.

However, the percentage has grown considerably, through 19 percent in the F18 up to 24 percent

in the F22. Composite materials are used extensively in the Euro fighter: the wing skins, forward

fuselage, flaperons and rudder all make use of composites. Toughened epoxy skins constitute

about 75 percent of the exterior area. In total, about 40 percent of the structural weight of the

Euro fighter is carbon-fibre reinforced composite material. Other European fighters typically

feature between about 20 and 25 percent composites by weight: 26 percent for Dassaults Rafael and 20 to 25 percent for the Saab Gripen and the EADS Mako.

The B2 stealth bomber is an interesting case. The requirement for stealth means that radar-

absorbing material must be added to the exterior of the aircraft with a concomitant weight

penalty. Composite materials are therefore used in the primary structure to offset this penalty.

The use of composite materials in commercial transport aircraft is attractive because reduced

airframe weight enables better fuel economy and therefore lowers operating costs. The first

significant use of composite material in a commercial aircraft was by Airbus in 1983 in the

rudder of the A300 and A310, and then in 1985 in the vertical tail fin. In the latter case, the 2,000

parts (excluding fasteners) of the metal fin were reduced to fewer than 100 for the composite fin,

lowering its weight and production cost. Later, a honeycomb core with CFRP faceplates was

used for the elevator of the A310. Following these successes, composite materials were used for

the entire tail structure of the A320, which also featured composite fuselage belly skins,

fin/fuselage fairings, fixed leading-and trailing-edge bottom access panels and deflectors,

trailing-edge flaps and flap-track fairings, spoilers, ailerons, wheel doors, main gear leg fairing

doors, and nacelles. In addition, the floor panels were made of GFRP. In total, composites

constitute 28 percent of the weight of the A320 airframe.

The A340-500 and 600 feature additional composite structures, including the rear pressure

bulkhead, the keel beam, and some of the fixed leading edge of the wing. The last is particularly

significant, as it constitutes the first large-scale use of a thermoplastic matrix composite

component on a commercial transport aircraft. The use of composites enabled a 20 percent

saving in weight along with a lower production time and improved damage tolerance.

The A380 is about 20-22 percent composites by weight and also makes extensive use of GLARE

(glass-fibre reinforced aluminium alloy), which features in the front fairing, upper fuselage

shells, crown and side panels, and the upper sections of the forward and aft upper fuselage.

GLARE laminates are made up of four or more 0.38 mm (0.015 in) thick sheets of aluminium

alloy and glass fibre resin bond film.

10

4.1 Popular Examples of composite use

Fighter Aircraft

U.S. Europe Russia

AV-8B, F16, F14, F18, YF23, F22, JSF,

UCAV Harrier GR7, Gripen JAS39,

Mirage 2000, Rafael, Euro fighter, Lavi,

EADS Mako MIG29, Su Series

Bomber

U.S

B2

Transport

U.S. Europe

KC135, C17, 777, 767, MD1 1 A320,

A340, A380, Tu204. ATR42, Falcon 900,

A300-600

General Aviation

Piaggio, Starship, Premier 1, Boeing 787

Rotary Aircraft

V22, Eurocopter, Comanche, RAH66,

BA609, EH101, Super Lynx 300, S92

Figure 1.1 Growth of use of advanced composites in air frame structures

11

5. Basic principles (micromechanics) of fiber composite materials

As an example, to a good first approximation, the stiffness under loading in the fiber direct ion

(unidirectional fibers) may be determined by the simple law of mixtures. This is simply a sum of

the volume (or area) fraction of the fibers and the matrix multiplied by the elastic modulus. The

strength estimation is similar (for a reasonably high fiber-volume fraction) but with each elastic

modulus multiplied by the breaking strain of the first-failing component. In the case of carbon

fiber/epoxy composites, this is generally the fiber-breaking strain. If, however, the lowest failure

strain is that of the matrix, the first failure event may be the development of extensive matrix

cracking, rather than total fracture. This damage may or may not be defined as failure of the

composite. However, toughness is usually much more than the sum of the toughness of each of

the components because it depends also on the properties of the fiber/matrix interface. Therefore,

brittle materials such as glass fibers and polyester resin, when combined, produce a tough, strong

composite, most familiarly known as fiberglass, used in a wide range of structural applications.

Control of the strength of the fiber/matrix interface is of paramount importance for toughness,

particularly when both the fiber and the matrix are brittle. If the interface is too strong, a crack in

the matrix can propagate directly through fibers in its path. Thus it is important that the interface

is able to disband.

12

6. Moulding Methods

In general, the reinforcing and matrix materials are combined, compacted and processed to

undergo a melding event. After the melding event, the part shape is essentially set, although it

can deform under certain process conditions. For a thermoset polymeric matrix material, the

melding event is a curing reaction that is initiated by the application of additional heat or

chemical reactivity such as organic peroxide. For a thermoplastic polymeric matrix material, the

melding event is solidification from the melted state. For a metal matrix material such as

titanium foil, the melding event is a fusing at high pressure and a temperature near the melt

point.

For many moulding methods, it is convenient to refer to one mould piece as a "lower" mould and

another mould piece as an "upper" mould. Lower and upper refer to the different faces of the

moulded panel, not the mould's configuration in space. In this convention, there is always a

lower mould, and sometimes an upper mould. Part construction begins by applying materials to

the lower mould. Lower mould and upper mould are more generalized descriptors than more

common and specific terms such as male side, female side, a-side, b-side, tool side, bowl, hat,

mandrel, etc. Continuous manufacturing processes use a different nomenclature.

6.1 Vacuum bag moulding

A process using a two-sided mould set that shapes both surfaces of the panel. On the lower side

is a rigid mould and on the upper side is a flexible membrane or vacuum bag. The flexible

membrane can be a reusable silicone material or an extruded polymer film. Then, vacuum is

applied to the mould cavity. This process can be performed at either ambient or elevated

temperature with ambient atmospheric pressure acting upon the vacuum bag. Most economical

way is using a venturi vacuum and air compressor or a vacuum pump. A vacuum bag is a bag

made of strong rubber-coated fabric or a polymer film used to bond or laminate materials. In

some applications the bag encloses the entire material, or in other applications a mold is used to

form one face of the laminate with the bag being single sided to seal the outer face of the

laminate to the mold. The open end is sealed and the air is drawn out of the bag through a nipple

13

using a vacuum pump. As a result, uniform pressure approaching one atmosphere is applied to

the surfaces of the object inside the bag, holding parts together while the adhesive cures. The

entire bag may be placed in a temperature-controlled oven, oil bath or water bath and gently

heated to accelerate curing.

In commercial woodworking facilities vacuum bags are used to laminate curved and irregular

shaped workpieces. Vacuum bagging is widely used in the composites industry as well. Carbon

fiber fabric and fiberglass, along with resins and epoxies are common materials laminated

together with a vacuum bag operation. Typically, polyurethane or vinyl materials are used to

make the bag, which is commonly open at both ends. This gives access to the piece, or pieces to

be glued. A plastic rod is laid onto the bag, which is then folded over the rod. A plastic sleeve

with an opening in it, is then snapped over the rod. This procedure forms a seal at both ends of

the bag, when the vacuum is ready to be drawn.

A "platen" is used inside the bag for the piece being glued to lay on. The platen has a series of

small slots cut into it, to allow the air under it to be evacuated. The platen must have rounded

edges and corners to prevent the vacuum from tearing the bag. When a curved part is to be glued

in a vacuum bag, it is important that the pieces being glued be placed over a solidly built form, or

have an air bladder placed under the form. This air bladder has access to "free air" outside the

bag. It is used to create an equal pressure under the form, preventing it from being crushed.

6.2 Pressure bag moulding

This process is related to vacuum bag moulding in exactly the same way as it sounds. A solid

female mould is used along with a flexible male mould. The reinforcement is placed inside the

female mould with just enough resin to allow the fabric to stick in place (wet layup). A measured

amount of resin is then liberally brushed indiscriminately into the mould and the mould is then

clamped to a machine that contains the male flexible mould. The flexible male membrane is then

inflated with heated compressed air or possibly steam. The female mould can also be heated.

Excess resin is forced out along with trapped air. This process is extensively used in the

production of composite helmets due to the lower cost of unskilled labor. Cycle times for a

14

helmet bag moulding machine vary from 20 to 45 minutes, but the finished shells require no

further curing if the moulds are heated.

6.3 Autoclave moulding

A process using a two-sided mould set that forms both surfaces of the panel. On the lower side is

a rigid mould and on the upper side is a flexible membrane made from silicone or an extruded

polymer film such as nylon. Reinforcement materials can be placed manually or robotically.

They include continuous fibre forms fashioned into textile constructions. Most often, they are

pre-impregnated with the resin in the form of prepreg fabrics or unidirectional tapes. In some

instances, a resin film is placed upon the lower mould and dry reinforcement is placed above.

The upper mould is installed and vacuum is applied to the mould cavity. The assembly is placed

into an autoclave. This process is generally performed at both elevated pressure and elevated

temperature. The use of elevated pressure facilitates a high fibre volume fraction and low void

content for maximum structural efficiency.

6.4 Resin transfer moulding (RTM)

A process using a two-sided mould set that forms both surfaces of the panel. The lower side is a

rigid mould. The upper side can be a rigid or flexible mould. Flexible moulds can be made from

composite materials, silicone or extruded polymer films such as nylon. The two sides fit together

to produce a mould cavity. The distinguishing feature of resin transfer moulding is that the

reinforcement materials are placed into this cavity and the mould set is closed prior to the

introduction of matrix material. Resin transfer moulding includes numerous varieties which

differ in the mechanics of how the resin is introduced to the reinforcement in the mould cavity.

These variations include everything from vacuum infusion (for resin infusion see also boat

building) to vacuum assisted resin transfer moulding (VARTM). This process can be performed

at either ambient or elevated temperature.

15

6.5 Other

Other types of moulding include press moulding,transfer moulding,pultrusion moulding, filament

winding, casting, centrifugal casting and continuous casting. There are also forming capabilities

including CNC filament winding, vacuum infusion, wet lay-up, compression moulding,

and thermoplastic moulding, to name a few. The use of curing ovens and paint booths is also

needed for some projects.

16

7. Types of Composite Materials

7.1 Matrices

The matrix, which may be a polymer, metal, or ceramic, forms the shape of the component and

serves the following additional functions:

1) Transfers load into and out of the fibers.

2) Separates the fibers to prevent failure of adjacent fibers when one fails.

3) Protects the fiber from the environment.

The strength of the fiber/matrix interfacial bond is crucial in determining toughness of the

composite. The interface, known as the interphase, is regarded as the third phase in the

composite because the matrix structure is modified close to the fiber surface. The interface is

even more complex in some fibers, notably glass fibers, which are pre-coated with a sizing agent

to improve bond strength, to improve environmental durability, or simply to reduce handling

damage.

Properties of the composite that are significantly affected by the properties of the matrix (matrix-

dominated properties) include:

1) Temperature and environmental resistance.

2) Longitudinal compression strength.

3) Transverse tensile strength

4) Shear strength.

The matrix may be brittle or tough. Figure 1.4 shows the inherent toughness of some candidate

materials. Economic production requires that the techniques used for matrix introduction allow

simple low-cost formation of the composite without damaging or misaligning the fibers. The

simplest method is to infiltrate an aligned fiber bed with a low-viscosity liquid that is then

converted by chemical reaction or by cooling to form a continuous solid matrix with the desired

properties. Alternatively, single fibers, tows of fibers, or sheets of aligned fibers may be coated

or intermingled with solid matrix or matrix precursor and the continuous matrix formed by

flowing the coatings together (and curing if required) under heat and pressure.

7.2 Polymers

Thermosetting or thermoplastic polymers that are used for the matrix of polymer composites.

Thermosetting polymers are long-chain molecules that cure by cross-linking to form a fully

three-dimensional network and cannot be melted and reformed. They have the great advantage

that they allow fabrication of composites at relatively low temperatures and pressures since they

pass through a low-viscosity stage before polymerization and cross- linking. The processes used

to manufacture components from thermosetting polymer composites include:

17

Impregnating a fiber preform with liquid resin, which is then cured (resin transfer molding; RTM). This process requires the resin to transition through a period of low viscosity (similar to

light oil).

Infusing a melted resin film into a fiber preform under pressure and then curing (resin-film infusion; RFI).

Pre-impregnating fiber sheet bundles or tows with a "staged" liquid resin (prepreg) for subsequent arrangement (stacking) followed by consolidation and cure under temperature and

pressure.

Epoxies have excellent mechanical properties, low shrinkage and form adequate bonds to the

fibers. Importantly, they pass through a low-viscosity stage during the cure, so allow the use of

liquid resin-forming techniques such as RTM. Epoxy systems curing at 120 C and 180 C have

respectively upper service temperatures of 100C and 130-150C. Bismaleimide resins (BMIs)

have excellent formability and mechanical properties similar to epoxies and can operate at higher

temperatures; however, they are more costly. BMI systems curing at about 200C have upper

service temperatures above 180 C. High-temperature thermosetting polymers such as

polyimides, curing at around 270C, allow increases up to 300C. However, they are even more

expensive and much more difficult to process. Thermosetting materials generally have relatively

low failure strains. This results in poor resistance to through-thickness stresses and mechanical

impact damage that can cause delaminations (ply separations) in laminated composites. They

also absorb atmospheric moisture, resulting in reduced matrix-dominated properties in the

composite, such as elevated temperature shear and compressive strength. Recent developments

have resulted in much tougher thermoset systems, some with improved moisture resistance,

through modifications in resin chemistry or alloying with tougher polymeric systems, including

rubbers and thermoplastics. Thermoplastic polymers, linear (none-cross-linked) polymers that

can be melted and reformed, are also suitable for use as matrices. High-performance

thermoplastics suitable for aircraft applications include polymers such as polyetheretherketone

(PEEK), application approximately to 120C; polyetherketone (PEK), to 145C; and polyimide

(thermoplastic type), to 270C. Thermoplastic polymers have much higher strains to failure

because they can undergo extensive plastic deformations resulting in significantly improved

impact resistance. Because these polymers are already polymerized, they form very high

viscosity liquids when melted. Thus fabrication techniques are based on processes such as resin-

film (or resin-fiber) infusion and pre-preg techniques. The main approach is to coat the fibers

with the resin (from a solvent solution) and then consolidate the part under high temperature and

pressure. Alternatively, sheets of thermoplastic film can be layered between sheets of dry fiber or

fibers of thermoplastic can be woven through the fibers and the composite consolidated by hot

pressing. Because thermoplastics absorb little moisture, they have better hot/wet property

retention than thermosetting composites. However, they are generally more expensive and are

more costly to fabricate because they require elevated temperature processing. In addition, with

improvements in thermosets, even the toughness advantage is being eroded. There is little doubt

that thermoplastics will be used extensively in the future for aircraft structures, particularly in

areas subject to mechanical damage.

18

7.3 Metals

The light metals, magnesium, aluminum, and titanium alloys (including titanium aluminides), are

used to form high-performance metal-matrix composites. These materials offer the possibility of

higher temperature service capabilities--approximately 150C, 300C, 500C, and >700C,

respectively--and have several other advantages, as discussed later, over polymer-matrix

composites. However, these advantages are offset by more costly, complex, and limited

fabrication techniques. Metals often react chemically with and weaken fibers during manufacture

or in service at elevated temperatures, so translation of fiber properties is often poor. The

tendency for a metal to react with the fiber is termed fiber/matrix compatibility. Generally,

because of compatibility problems, ceramic fibers such SiC, A1203, and Borsic (boron fibers

coated with silicon carbide) are most suited for reinforcing metals. However, carbon fibers may

be used with aluminum or magnesium matrices, provided that exposure to high temperature is

minimized. Methods based on infiltration liquid metal have many advantages for aluminum,

provided damaging chemical interaction between the metal and fibers does not occur and the

metal is able (or is forced under pressure) to wet the fibers. The process of squeeze casting is

attractive because time in contact with liquid metal is limited, minimizing chemical interaction,

and the high pressure overcomes wetting difficulties. Another major advantage of this process is

that alloys other than casting alloys can be employed.

19

7.4 Ceramics

For much higher temperatures than can be achieved with polymer or metal matrices, the options

are to employ a silica-based glass; a ceramic such as silicon carbide, silicon nitride, or alumina; 6

or a carbon matrix. These are called ceramic matrix composites (CMCs). In the case of the high-

modulus ceramic matrices, the fibers provide little stiffening; their purpose is to increase

toughness. This is achieved mainly by blunting and deflecting cracks in the matrix and

contributing to increased fracture energy through the various energy-absorbing mechanisms,

such as crack bridging and fiber pull-out. Several techniques are used to form composites with

ceramic matrices. These include infiltration of aligned fibers by

1) CVD

2) impregnation of fibers with a fine powder and consolidating

3) impregnation of fibers with a liquid

Ceramic precursor, generally a polymer, and converting to ceramic at elevated temperature. The

powders may be added to the aligned fibers or fiber preforms by injection molding or by sol-gel

techniques. Densification of powder coatings may be achieved by hot-pressing, sintering, hot

isostatic pressing, or superplastic forging. In most respects, the precursor route is the most

promising for ceramics because dense matrices can be produced at low temperatures without

causing fiber damage, and complex components can be formed directly. Glass and glass-ceramic

matrices are readily formed by consolidation of fiber preforms impregnated with fine powders

applied from a dispersion or gel. The glass melts easily and flows between the fibers to form a

continuous pore-free matrix. The procedure is similar to that adopted for thermoplastic matrix

composites. In glassceramic matrices, the matrix may subsequently be crystallized by heat

treatment, greatly enhancing performance at elevated temperatures. Carbon matrices may also be

formed by CVD of carbon from high-carbon content gases, such as methane, propane, and

benzene into a fiber preform. They can also be formed by liquid phase impregnation of fibers

followed by pyrolytic decomposition of a precursor with a high carbon content. Suitable

precursors include phenolic resin, pitch, and tar-based materials, all of which can have over 40%

yield of carbon on pyrolysis. The fibers are generally carbon and the composite called

carbon/carbon. Silicon carbide fibers are also used in some applications as an alternative to

carbon, particularly where improved resistance to oxidation is required. With the resin-based

route, standard polymer-matrix composite manufacturing processes, such as filament winding or

braiding, can be used before pyrolysis. The precursor route is the most efficient for making

carbon matrix composites; however, multiple impregnations and pyrolysis steps are required to

produce a matrix with an acceptably low porosity level. This is a slow process resulting in high

component costs. The CVD process is even slower, therefore it is mainly used to fill-in fine

interconnected near-surface voids in composites produced by pyrolysis. The CVD is, however,

suited to manufacture of thin-wall components. PMCs are extensively used in aerospace

structures; however, carbon/epoxy is by far the most exploited so is the main focus of this book.

Some current airframe applications are described in Chapter 12. Based on the drivers set out in

Table 1.1.

20

21

7.5 Polymer Matrix Composites

The nomenclature used in the U.S. identifies the composite in the format fiber/ matrix. For

example, the main composites discussed in this book are carbon fibers in an epoxy resin matrix

and are referred to as carbon/epoxy or graphite/epoxy (also c/ep and gr/ep). Other common

composite systems are carbon/BMI, carbon/polyimide, glass/epoxy, aramid/epoxy, and

boron/epoxy. This notation can readily be expanded to specific composite systems; for example,

a well-known commercial composite system, Hercules AS fibers in a 3501-6 epoxy resin matrix,

is AS/3501-6. In the U.K. the terminology for carbon/epoxy is carbon fiber reinforced epoxy, or

more usually, carbon fiber reinforced plastic (CFRP).

7.6 Non-polymeric Composite Systems

In this section, some of the important non-polymeric composite systems are briefly discussed.

7.6.1 Metal-Matrix Composites

Metal-matrix composites (MMCs), 4'7'8 with continuous or discontinuous fiber reinforcement

have been under development for well over 30 years, but have yet to be widely exploited.

The main MMCs based on continuous fibers, and their advantages and disadvantages compared

with PMCs. Potential aircraft applications of the MMCs include engine components, such as fan

and compressor blades, shafts, and possibly discs, airframe components, such as spars and skins,

and undercarriage components, such as tubes and struts. Carbon/aluminum alloy and

carbon/magnesium alloy composites are particularly attractive for satellite applications,

including aerials and general structures. These MMCs combine the high specific properties and

low, thermal expansion coefficients exhibited by the PMCs together with the advantages

indicated in Table 1.5. For example, high conductivity serves to minimize thermal gradients, and

therefore distortion, when a space structure is subjected to directional solar heating. However,

MMCs based on carbon fibers, although potentially low-cost, suffer several drawbacks for non-

space applications. These include oxidation of carbon fibers from their exposed ends at elevated

temperature and corrosion of the metalmatrix in wet environments due to galvanic action with

exposed fibers. Other potential non-structural applications of carbon/metal composites include

1) carbon/lead and carbon/copper-tin alloys for bearings,

2) carbon/copper for high-strength conductors and marine applications

3) carbon/lead for battery electrodes.

The earliest developed and probably still the most exploited aluminum matrix MMC is

boron/aluminum, based on CVD boron filaments. This MMC is used in the tubular structure in

the Space Shuttle. In the future, boron/aluminum may be superseded by CVD silicon

22

carbide/aluminum (or silicon carbide coated boron), which has the advantage of much greater

resistance to attack by liquid aluminum. The increased resistance simplifies composite

fabrication and improves fiber/ matrix compatibility at elevated temperature.

A typical microstructure of a boron/aluminum composite is shown in Figure 1.6a, whereas, for

comparison, Figure 1.6b shows the microstructure of a typical carbon/aluminum composite.

Current aluminum matrix MMCs do not offer a significant increased temperature capability over

PMCs based on high-temperature matrices such as BMIs and polyimides. Thus, unless some

other properties are required, such as thermal conductivity, aluminum MMCs generally have no

major advantage over PMCs and are far more expensive. In contrast, titanium alloy and titanium

aluminide MMCs, based on CVD silicon-carbide-fiber reinforcement, have a large margin on

temperature capability over PMCs. They also have excellent mechanical properties compared

with conventional titanium alloys (100% increase in stiffness and 50% increase in strength);

however, they cannot match PMCs in terms of moderate temperature properties and are much

more expensive. Titanium-based MMCs are damage tolerant, and so in addition to high

temperature applications in high-speed transport and gas-turbine engines, they are also being

evaluated as a replacement for steel undercarriage components where they could prove to be

cost-effective. Titanium MMCs lend themselves very well to selective reinforcement (where

reinforcement is applied only in high-stress areas), as titanium is readily diffusion bonded. For

example, layers of titanium/silicon carbide can be used to reinforce a high-temperature

compressor disk 9 with a 70% weight saving. The large weight saving results from the

elimination of much of the inner material of the disk. The resulting construction is a titanium

MMC reinforced ring. If the disk has integral blades, it is called a bling. Blings provide marked

improvements in the performance of military gas-turbine engines. Titanium MMCs can also be

used to reinforce titanium-skinned fan blades or for the face skins of a sandwich panel with a

super-plastically formed core. MMCs capable of operation to temperatures over 800C are also

keenly sought for gas-turbine applications. Unfortunately, the use of available high-performance

carbon or ceramic fibers is not feasible with high-temperature alloy matrices because of severe

compatibility problems. Attempts to use barrier layers on fibers, such as metal oxides or

23

carbides, to prevent chemical reaction have been unsuccessful. In addition, due to the high

temperatures and mismatch in coefficients of thermal expansion, thermal fatigue would be a

serious problem with these composites. A practical, but not very attractive solution because of

the poor specific properties, is the use of refractory metal wire as the reinforcement. This

approach has the potential to produce turbine blade materials with an additional 100C capability

over conventional superalloys. A promising composite is based on tungsten alloy wires (W-l%

ThO2 or W-Hf-C type) in an iron-based (Fe-Cr-A1-Y) matrix. This alloy has relatively high

ductility and excellent oxidation resistance requiring no protective coating. However, a coating

such as TiC or TiN may be needed on the fibers to avoid attack by the matrix. Costs of the

continuous fiber MMCs are (and almost certainly will continue to be) very high compared with

PMCs, and the range of sizes and shapes that can be produced is much more limited. As

mentioned previously, MMCs based on aluminum alloy matrices will be strongly challenged for

most elevated temperature applications by current and emerging PMCs. An alternative to the use

of "artificial" fiber reinforcement to produce high temperature MMCs is to use directionally

solidified eutectics. Here the reinforcing phase, produced by eutectic (or eutectoid)

decomposition, is in the form of aligned platelets or fibers. These "natural" composites have a

great advantage in that the matrix and reinforcement are in chemical equilibrium. However,

surface energetics can cause the fibers or laminates to form spherical particles over long periods

at elevated temperature, destroying the reinforcing effect. In addition, thermal fatigue can cause

internal cracking as well as accelerating spheroidizing of the microstructure. Promising systems

studied in the past include Co-Ta-C and Ni-Ta-C.

7.6.2 Particulate MMCs

Particulate MMCs should be mentioned in this overview because they may have extensive

aerospace applications 1 as structural materials. In these composites, aluminum or titanium

alloy-matrices are reinforced with ceramic particles, generally silicon carbide or alumina in the

micron range. Because reinforcement is not directional as with fiber-reinforced MMCs,

properties are essentially isotropic. The specific stiffness of aluminum silicon-carbide particulate

MMCs (A1/SiCp, where the subcript p refers to particulate) can exceed conventional aluminum

alloys by around 50% at a 20% particle volume fraction. For comparison, an MMC with

inclusion of silicon-carbide fibers at a similar volume fraction will increase its specific stiffness

increased by around 100%.

The primary fabrication techniques are rapid-liquid-metal processes such as squeeze casting or

solid-state powder processes based on hot-pressing. Particulate MMCs also have the

considerable cost advantage of being formable by conventional metal-working techniques and

possibly super-plastic forming and diffusion bonding in the case of titanium-matrix systems.

However, because of their high wear resistance, special tools such as diamond-coated drills and

Diamond-impregnated grinding wheels are required for machining. When fabricated using clean

high-grade particles with low porosity and moderate particulate volume fraction, particulate

MMCs have high strength, acceptable fracture toughness, and good resistance to fatigue crack

propagation. The MMCs also have high stiffness and wear resistance compared with

24

conventional alloys. They are therefore suited to small components requiring high stiffness

combined with fatigue and wear resistance.

7.6.3 Ceramic-Matrix Composites

Ceramic-matrix composites (CMCs) offer the main long-term promise for high-temperature

applications in gas turbine engines and for high-temperature airframe structures, although there

are formidable problems to be overcome. The main requirement is for lightweight blades able to

operate uncooled in environments around 1400C.

The main limitation is the unavailability of fibers with high-elastic moduli and strength, chemical

stability, and oxidation resistance at elevated temperatures. For suitable reinforcement of ceramic

matrices (such as alumina and silicon carbide or silicon nitride), the fiber must have high

oxidation resistance at high temperature because microcracking of the ceramic allows contact

between the fibers and the external environment. The fiber must also be chemically compatible

with the matrix and must closely match it in its coefficient of thermal expansion. Thus, the use of

similar materials for both components appears to offer the most promise, for example, silicon-

carbide-fibers/silicon-carbidematrix or alumina fibers/alumina matrix. Unfortunately, available

fibers eitherdo not maintain strength at high enough temperatures or (in the case of carbon fibers,

for example) have adequate oxidation resistance to provide anywhere near the full exploitation of

the potential benefits. CMCs are sometimes based on three-dimentional fiber architectures

because in many (but not all) applications, the fibers are required to provide toughness, including

through-thickness toughness, rather than stiffness as required in other classes of composites.

Thus, for some CMCs, the relatively low fiber volume fraction resulting from this form of

construction is not a major limitation. Glass and glass-ceramic matrices are promising for

applications at temperatures around 500C because of their excellent mechanical properties

and relative ease of fabrication. In contrast to CMCs based on conventional ceramics, such as

silicon carbide, the low modulus matrix can be effectively stiffened by suitable fibers and

relatively high toughness achieved (typically, an increase of over 30 times the matrix glass

alone). Because the matrix does not microcrack at relatively modest strain levels and

temperatures, carbon fibers can be used. However, for higher-temperature applications more

oxidation resistant fibers such as silicon carbide fibers must be used. Carbon/carbon composites

12 have no significant chemical or thermal expansion compatibility problems. However, unless

protected, they are also prone to rapid attack at elevated temperature in an oxidizing

environment. Even where oxidation is a problem, the composites can be used where short

exposures to severe applications at temperatures over 2000C are experienced, for example, in

rocket nose-cones, nozzles, and leading edges on hypersonic wings. In the presence of oxygen-

reducing conditions, for example with a hypersonic engine running slightly rich on hydrogen

fuel, operations for prolonged periods can be maintained. Carbon/carbon composites could be

used for prolonged periods at elevated temperature, above 1600 C, if effective oxidation-

preventative barrier coatings were available.

25

7.7 Hybrid Metal/PMC Composites

Structural metals, such as aluminum alloys and composites, including carbon/ epoxy, have a

variety of advantages and disadvantages for airframe applications. For example, metals are prone

to fatigue cracking but PMCs are not; PMCs are easily damaged by low-energy mechanical

impacts but metals are not. Thus, the potential exists to combine these materials in such a way as

to get the best of both materials. One such approach is the aluminum/fiber composite hybrid

laminate, 13 which consists of thin sheets of aluminum alloy bonded with a fiber-reinforced

adhesive. When a crack grows through the metal, the fibers, which are highly resistant to fatigue

damage, are left spanning or bridging the crack in its wake (Fig. 1.7). The result is a reduction in

crack growth rate by approximately one order of magnitude and an insensitivity to crack length.

However, the fibers have little influence on crack initiation and, indeed, because the hybrid

composite has relatively low modulus, the increased strain in the aluminum alloy can encourage

earlier crack initiation. The fibers also significantly increase the postyield strength compared

with unreinforced aluminum alloy, and the composite has a much higher damping capacity.

Disadvantages of these materials include sensitivity to blunt notches due to the inability of the

fibers to withstand very high strain levels. Thus, the notch insensitivity of metals is not retained

in the hybrid. Also, depending on the reinforcement used, the elastic modulus of the hybrid is

generally lower than aluminum alloys, however, this is compensated for by a reduction of

specificgravity of between 10-15%. Another problem is cost, which is typically 7-10 times that

of standard aerospace-grade aluminum alloys. The aluminum alloy is generally either 2024 T3 or

7475 T761, 0.2-0.4 mm thick. The composite is aramid (Kevlar) or glass fibers in an epoxy

nitrile adhesive, around 0.2 mm thick for unidirectional reinforcement, or 0.25-0.35 mm thick for

(glass reinforcement only) cross-ply. With aramid reinforcement, the laminate is called ARALL

(aramid reinforced aluminum laminate), and with glass fiber, GLARE. Because of the sensitivity

of aramid fibers to compressive

sheets

26

bonded with an epoxy film adhesive reinforced with glass or aramid fibers. The fibers are left

spanning or bridging fatigue cracks if they develop in the aluminum sheets, vastly reducing the

rate of crack growth. Stresses and the favorable residual strength that is produced, ARALL may

be pre-stretched. This also overcomes, at a cost, the adverse residual stresses arising from the

differences in thermal expansion coefficient between aramid, or glass, and aluminum. GLARE

does not require pre-stretching as the high-strain glass fiber used is less susceptible to

compressive stresses. Consequently, the glass fibers can be cross-plied to give crack growth

resistance in two orthogonal directions as may be required for a fuselage structure. Although

GLARE has a lower modulus than conventional aluminum alloys, with a reduction of around

20% (particularly with cross-plied fibers), it has the best resistance to fatigue crack growth.

Significant weight savings--20% or so---can be achieved in fatigue-prone regions such as

pressurized fuselage skins and stiffeners and lower wing skins by the use of these materials. The

hybrid composites are also suited to high-impact regions such as leading edges and inboard flaps

and to components subject to mishandling, such as doors. For applications requiring higher

stiffness and strength, as well as a higher temperature, capability studies have been conducted 13

on hybrid laminates made of thin sheets of titanium alloy (Ti-6A1-4V) and a low-modulus

carbon fibercomposite. The matrix for the composite and adhesive is a thermoplastic (PEEK).

This laminate is reported to have excellent resistance to fatigue crack growth as well as good

blunt-notch strength.

27

8. Application and Operating Experience

8.1 US. Air Force (T. Reinhart)

The Air Force has had 8 to 10 years of operational experience with composites, with good

success; composites are being used on rotor blades and other parts of helicopters and for

secondary structures on other types of aircraft. Plans indicate that some 40 to 60 percent of the

structural weight of new aircraft will be composites. Operational problems include cracking and

corrosion of meta1 honeycomb, incidents of maintenance damage, quality control in

manufacture, paint removal, and repair.

The operational environment appears to have had no adverse effects on composite material and

structural characteristics other than corrosion of associated metal parts because of inappropriate

composite structure design. In summary, operational experience is good while maintenance

experience is poor. Needs include improved damage tolerance, large-area inspection capability,

understanding of failure mechanism, reliable joints and attachments, and designs that can

efficiently handle the transfer of large loads. For transport class aircraft, there is a need for more

design data for highly loaded parts.

8.2 US. Navy (D. Mulville)

The Navy has had extensive experience with both fixed- and rotary-wing aircraft composite

applications. The Navy is favorably impressed with its application of composites including the

use of composite load-carrying wing skins and enginecasings (replacing titanium). Problems

found are related to operations, maintenance, and repair of battle damage. Care needs to be taken

in design where high temperatures can impinge on composite structures (e.g., hot duct blowouts).

The AV8-B aircraft primary structure is about 26 percent composites by weight. The JVX/V-22

structure is expected to be 70 percent composites by weight. An A-6 composite wing-box

program is under development, as are studies of composite control surfaces. Field repairs of

composite structures are a major concern. A substantial program is in progress with emphasis on

the minimization or elimination of the need for special repair equipment. Generally there has

been little use of thermoplastics, except for repair. Problems are related to damage during

maintenance, erosion/abrasion, and wear around holes. Moisture intrusion and its impact on

metal components is a long-term problem. Fuel leakage and lightning strikes are other areas

requiring special attention in design and manufacture. In summary, experience with composites

has been good. Operational support and repairs is an area requiring and getting attention in the

Navy program.

28

8.3 U.S. Army (P. Haselbauer)

The major composites experience has been with rotorcraft rotor blades (AH-15, CH-47D, UH-

GOA, and OH-58D) and some secondary airframe components. The OH-58D production articles

will have composite main rotor yokes. This yoke has been through qualification testing. The

service is moving toward greater use of composites in its future rotorcraft. The types of problems

encountered include: rough skins, skin/core voids, fit tolerances, moisture retention, retention of

blade-tip weights, and the sealing of fuel in composite structures. Correction of these design and

operating problems requires detailed attention to design, manufacturing processes including

quality control, and knowledge of the operating environment. In summary, the Army has found

that composites are viable for its aircraft structures; trade studies that consider costs, weight,

performance, and support dictate the use of composites; three-dimensional stress analyses are

important for design; and the use of composite structures for the containment of fuel should be

avoided.

29

9. Turbine inlet guide vanes

Turbine inlet guide vanes have been fabricated from composites of silicon carbide fibers in

silicon carbide matrices. A unique design for a cloth made from SiC fibers makes it possible to

realize the geometric features necessary to form these vanes in the same airfoil shapes as those of

prior metal vanes.

The fiber component of each of these vanes was made from SiC-fiber cloth coated with boron

nitride. The matrix was formed by chemical-vapor infiltration with SiC, then slurry-casting of

SiC, followed by melt infiltration with silicon.

These SiC/SiC vanes were found to be capable of withstanding temperatures 400 F (222 C)

greater than those that can be withstood by nickel-base-superalloy turbine airfoils now in

common use in gas turbine engines. The higher temperature capability of SiC/SiC parts is

expected to make it possible to use them with significantly less cooling than is used for metallic

parts, thereby enabling engines to operate more efficiently while emitting smaller amounts of

NOx and CO.

The SiC/SiC composite vanes were fabricated in two different configurations. Each vane of one

of the configurations has two internal cavities formed by a web between the suction and the

pressure sides of the vane. Each vane of the other configuration has no web (see Figure 1).

It is difficult to fabricate components having small radii, like those of the trailing edges of these

vanes, by use of stiff stoichiometric SiC fibers currently preferred for SiC/SiC composites. To

satisfy the severe geometric and structural requirements for these vanes, the aforement ioned

unique cloth design, denoted by the term Y-cloth, was conceived (see Figure 2). In the regions away from the trailing edge, the Y-cloth features a fiber architecture that had been well

characterized and successfully demonstrated in combustor liners. To form a sharp trailing edge

(having a radius of 0.3 mm), the cloth was split into two planes during the weaving process. The

fiber tows forming the trailing-edge section were interlocked, thereby enhancing through-

thickness strength of the resulting composite material.

30

For vanes of the webless configuration, each made from a layup of six plies of Ycloth, the length

of each Y-cloth layer was cut so that the two strips corresponding to the aforementioned two

planes would wrap around the perimeter of a graphite vane preform tool with a 10-mm overlap.

The overlap was used to join the two strips in a fringe splice. To make the external sixth ply, a

standard woven cloth was cut to the required final length and a fringe splice joined the two ends

of the cloth at the trailing edge. The cloth was then prepregged. The entire assembly was then

placed into an aluminum compaction tool designed to form the outer net shape of the vane. After

the prepreg material was allowed to dry, the preform was removed from the aluminum tooling

and placed into an external graphite tool before being shipped to a vendor for matrix infiltration.

To make the SiC fiber preform for a vane having an internal web, a slightly different initial

approach was followed.

31

10. Failure Analysis of Composite Structures in Aircraft Accidents

The impending generation of aircraft, represented by the Boeing 787, the Airbus 380, the newly

emerging very light jets, and the new generation of military fighters, marks a shift in airframe

technology in which primary structural components that have been traditionally constructed of

metal are being constructed of composites. This advancement creates the possibility of aircraft

accidents involving composite failures. Despite this possibility, the current body of knowledge

and experience regarding aircraft accidents is largely dependent on metallic aircraft. The basic

concepts involved in analyzing failed composites under a variety of fundamental loading

conditions such as tension, compression, bending, impact, and fatigue. These concepts are

demonstrated by discussing the analysis by the NTSB of the failed composite vertical stabilizer

involved in American Airlines flight 587.

32

10.1 Examination of Failed Metallic Structures

The science and art of analyzing failed metallic structures has matured in part as a result of the

analysis of accidents involving metal aircraft. Employing knowledge accrued during this period

of time, investigators often rely heavily on their ability to analyze failed structures in an effort to

determine the cause and events of an accident. Some investigators have emphasized the role of

such analysis: The bent metal speaks The story is written in the wreckage You have to learn

how to read the bent metal For the purposes of this paper, the evidence contained within the

wreckage will be referred to in two categories macrostructural evidence and microstructural

evidence. Macrostructural evidence refers to the overall deformation of failed structures a

buckled fuselage panel, a twisted propeller blade, a dented leading edge. Figure 3 shows an

example of macrostructural evidence, a collection of dents on the leading edges of an aircraft.

The value of macrostructural evidence in failed metal structures is enhanced by the fact that

typical aircraft metals, such as aluminum, are ductile, which means they undergo significant

deformation prior to final failure. Ductility allows for the permanent bending, twisting, and

denting of structures, essentially recording evidence of events in the accident. The evidence

33

contained in Figure 3 immediately identifies impact as a factor in this accident. Moreover, the

evidence identifies the possible size, shape, and energy associated with the impactor or

impactors. According to the NTSB, this aircraft impacted a set of power lines on approach

(NTSB, 1995). Ductility in metals provides macrostructural evidence in a variety of ways. One

method for determining whether a jet engine was powered at the time the aircraft impacted the

ground is to examine the fan blades. Metallic fan blades of a powered turbofan will generally

bend upon impact in a direction opposite the direction of rotation. This deformation can reveal

whether the engine was powered at the time of the accident. Another example is the deformation

produced by an explosion occurring inside a metallic fuselage. The bulging of fuselage panels,

the curling of ruptured edges away from the explosion, and the stretching and unzipping of

panels along rivet lines, all indicate the role of an explosion in an accident. Typical aircraft

composites are not ductile; they are brittle, which means they undergo relatively minor

permanent deformation prior to final failure. Without ductility, the macrostructural evidence

from an accident, such as the examples discussed above, will likely change. What evidence

would be produced by a failed composite structure? What evidence would be produced by a

GEnx engine, with its composite fan blades, impacting the ground? What evidence would be

produced by an explosion inside a 787 composite fuselage? With changes in macrostructural

evidence associated with the change from ductile to brittle structural materials, the analysis of

microstructural evidence becomes paramount. Microstructural evidence refers to relatively local

deformation and changes in the structure, such as fracture surfaces, that typically require close

visual or microscopic analysis. To interpret microstructural evidence in failed metallic structures,

investigators rely upon a well established and widely used body of knowledge, which has, in the

past, often provided rapid and insightful results. One example is the recent crash of Chalks

Ocean Airways flight 101 in December 2005 off the coast of Miami, Florida. Initial evidence

indicated that the right wing had separated in flight. Within days, the NTSB had identified

fatigue damage in metallic structural components in the right wing (Figure 4), with

corresponding damage in the structure of the left wing. As shown in the figure, 5 Such events

typically produce microstructural evidence as well. An unaided visual inspection of the wing

spar cap reveals beach marks, which is evidence widely accepted to be indicative of fatigue

failure. As a result of this established analysis, the microstructural evidence, supported by an

accrued body of knowledge regarding the interpretation of fracture surfaces in metals, rapidly

established the wing spar cap as a critical component to consider in this investigation. The

analysis of failed composite structures cannot rely solely on the knowledge and experience

accrued for metallic structures. The analysis of failed composite structures involves terms such

as fiber pullout, delamination, and interfacial failure. These terms do not even exist in the

analysis of failed metallic structures. These and other rudimentary components of knowledge

must be understood by accident investigators in order to analyze failed composite structures.

Examination of Failed Composite Structures Transitioning from failed metallic structures to

failed composite structures requires, in many ways, a new mindset. Although composites are

often considered to be materials and are generally classified as engineered materials, composites

34

are actually structures, made of multiple materials. Typical aircraft composites are made of two

materials, long fibers that are stiff and strong (typically carbon or glass) and a matrix, essentially

hardened plastic glue, that holds the fibers together. The glued fibers are typically assembled

layer-bylayer, called plies. The fibers in each ply typically run parallel to each other or are

woven together in the manner of a textile. Ply-wise variations in fiber orientation and other

variables often exist in a composite. In contrast to typical aircraft metals, the physical properties

of composites vary from location to location, and their response to loads usually varies with the

direction in which the load is applied. Composites can respond to loads in ways aircraft metals

cannot. A simple tensile load, for example, can cause a composite to twist; a simple twisting load

can cause a composite to bend. While designers know of, understand, and can predict these

phenomena, accident investigators must be able to recognize and reconstruct them. Composites

have design variables that are not available in metals. Some of these variables are fiber

orientation, fiber-to-matrix volume ratio, ply thickness, and ply stacking sequence, among others.

With new variables come new opportunities for manufacturing errors or imperfections. Some of

these imperfections are fiber waviness, poor adhesion between fibers and matrix, poor adhesion

between plies, excessive voids in the matrix, and an improperly cured matrix, among others.

Changes in design variables and accumulated imperfections directly affect the failure of a

composite.

35

For example, Figure 5 shows twenty failed composite specimens, four groups of five specimens,

representing four different ply-wise fiber orientations. Each specimen was subjected to simple

tensile loading. Despite the similarity in loading, the failure in each specimen looks unique.

Some of the failed specimens have a shredded appearance with a very rough fracture surface;

some of the specimens have a smoother, angular appearance. Some specimens even broke into

three pieces, rather than two. The differences in the appearance of these failures are a result of

two primary sources of variation among the specimens. The first source of variation is the

intentional variation in design variables, in this case, fiber orientation. The second source of

variation is the accumulation of imperfections, as discussed above. The result is that these

composites, all of which failed in tension, appear very different from each other. This is one of

the challenges of analyzing failed composites. In many cases, this challenge can be addressed by

performing a microscopic analysis of the failure surfaces to identify common features that

indicate failure in tension.

36

10.1.1 Tension

Regardless of the macroscopic variation of the fractures discussed above, tensile fractures of

fibrous composites typically exhibit some common characteristics that can help identify failure

under tensile loads. One characteristic is that the fracture surface generally has a rough

appearance, as can be seen in the failed specimens in Figure 5. Figure 6 shows a microscopic

view of a fracture surface of a composite that failed under tensile load, with the fibers aligned

with the direction of the load. One clear characteristic of the fracture surface is that fractured

fibers are sticking out of the fractured matrix, contributing to the rough appearance of the

fracture surface. Called fiber pullout, this characteristic is a typical indication of tension failure

in a composite. Fiber pullout is the result of a fiber breaking and being extracted from the matrix.

Close inspection of Figure 6 reveals, in addition to pulled-out fibers, holes in the matrix that

were created by other pulled-out fibers. In some cases of tensile failure, the fibers do not

37

completely fracture and only the matrix completely fractures. The fibers then span the matrix

fracture in a phenomenon called fiber bridging. In either case, the investigator can use the pulled

out fibers to identify tensile loading, and in the case of stacked laminates, identify those plies that

have been loaded in tension.

The length of the pulled out fibers can provide perspective on important fundamental conditions

present in the composite at the time of fracture, such as temperature, exposure to moisture, and

rate of loading. As long thin members, the fibers are designed to carry tensile loads and

composites are nominally designed such that the fibers run parallel to the tensile loads. However,

in the common case of composites with ply-wise variations in fiber orientation, tension loads do

not run parallel to the fibers and failure can occur in the matrix. Common matrix failures

associated with such loading conditions are tension failures between fibers, particularly at the

fiber-matrix interface, and shear failures in the matrix-rich region between plies, typically

associated with rough features on the fracture surface called hackles. Such inter-ply shear

failures can also be produced under compression.

10.1.2 Compression

Under compression, the fibers are relatively less effective. One common chara

cteristic of the compressive failure of fibrous composites is the formation of kink bands, as

shown in Figure 7. Kink bands are a result of structural instability, much like a person standing

38

on and eventually crushing a soda can. The fibers buckle as the compressive load approaches a

critical level, which is a function of material, geometric, and environmental factors. Fiber

buckling can also be identified by examination of the fiber ends. As shown in Figure 8, chop

marks indicate fibers that have buckled and have bent to failure. The chop marks coincide with

the neutral axis of the fiber in bending, separating the tension side of the fiber from the

compression side of the fiber. Often associated with kink bands is matrix splitting, which can be

seen in Figure 7 as gaps in the matrix. Matrix splitting occurs at weak points in the matrix or at

areas of high stress concentration, such as at the fiber-matrix interface and the interface between

plies. Matrix splitting at the interface between plies is referred to as delamination and is

discussed further in the paragraphs below regarding impact.

10.1.3 Bending

The difference between tensile and compressive fracture surfaces is readily demonstrated in

composites that have failed in bending. Figure 9 shows a specimen that has failed in bending.

Divided by a neutral bending axis, one part of the fracture surface contains pulled-out fibers and

the other part is relatively flat. This is a result of the fact that, in bending, one part of the cross-

section is in tension and the other part is in compression. These characteristics can readily

translate to a macroscopic level. Figure 10 shows a composite aircraft wing that has reportedly

39

failed in bending (Stumpff, 2001). The bottom surface of the wing, which was subjected to

tension in bending, has a very fibrous texture relative to the top side of the wing, which was

subjected to compression in

bending.

10.1.4 Impact

As discussed above, typical aircraft composites are brittle rather than ductile. Ductile metal

structures undergo relatively high levels of permanent deformation prior to final failure and this

deformation provides information regarding the events preceding structural failure. As brittle

structures, composites exhibit relatively little permanent deformation prior to final failure. The

metallic aircraft discussed above and shown in Figure 3 provides a clear indication of impact by

a foreign object. Impact evidence may not be as readily observed in a composite structure.

In fact, impact loading can cause damage to a composite without any visible evidence on the

surface. Consider an aircraft mechanic dropping a wrench on the top surface of a wing. If the

wing is made of aluminum, the impact may leave a dent, essentially recording the impact and

providing some rudimentary indication of the significance of the resultant damage. If the wing is

a brittle composite, the impact of the wrench may produce local crushing of the fibers and matrix

or it may not produce any damage on the surface at all. In either case, the level of damage below

the surface of a composite can be much more extensive than that indicated on the surface. One

common type of sub-surface damage from impact is delamination. A delamination is a split

between plies in a composite. The split can propagate along the interface at which neighboring

plies were joined during manufacturing, or it can propagate along the fiber-matrix interface.

Figure 11 shows a couple views of the cross-section of a composite plate after impact. As

40

indicated in the figures, the impact caused extensive delamination among multiple plies. Such

damage can dramatically degrade the load bearing capability of the composite even though the

fibers may remain intact. Moreover, the damage, if unnoticed, can continue to propagate upon

further loading of the composite. Without visible evidence on the surface, delaminations must be

identified by cross-sectioning the composite in the location of the delamination or by employing

non-destructive techniques such as ultrasonics or x-ray tomography. If destructive techniques are

employed, delaminations may be identified visually. In graphite-epoxy composites,

delaminations can be identified by a dull, whitish appearance, relative to the shiny, black

appearance of neighboring areas free from delamination.

10.1.5 Fatigue

One of the attractive qualities of composites is that they generally have better fatigue

performance than typical aircraft metals such as aluminum. Despite this fact, composites can fail

under fatigue loading and such failures result in particular failure features. Fatigue failure in

metals can be readily identified, in many cases, by an unassisted visual inspection. A typical

fatigue failure in metals will produce a fracture surface with beach marks. An example of beach

marks was already discussed above and shown in Figure 4. Fatigue fracture surfaces in

composites, on the other hand, do not typically have visible beach marks. In fact, fatigue

fractures in composites typically do not appear any different from a corresponding overload

failure. While fatigue fractures lack macroscopic evidence, some evidence may be identified

microscopically. Figure 12 shows striations at the fiber-matrix interface of a composite. One

striation typically corresponds to one load cycle. Although these striations indicate fatigue

failure, areas containing striations are typically small in size, few in number, and may be

dispersed over multiple locations in the composite. In addition, the striations are often

identifiable only under high magnification and oblique lighting (Figure 12 was captured under a

magnification of 2000x). In short, the identification of fatigue failure in composites can be very

challenging. One macroscopic feature that can provide evidence of fatigue is abrasion between

mating fracture surfaces. With repeated loading, the growing fracture surfaces may rub against

each other and leave abrasive marks on the ends of broken fibers and in the matrix.

41

42

11. Testing

To aid in predicting and preventing failures, composites are tested before and after construction.

Pre-construction testing may use finite element analysis (FEA) for ply-by-ply analysis of curved

surfaces and predicting wrinkling, crimping and dimpling of composites. Materials may be tested

after construction through several nondestructive methods including ultrasonics, thermography,

shearography and X-ray radiography.

11.1 Composite FEA

Composite Finite Element Analysis - can help creating Robust Composite Designs

using"Composite Design Principles" - which is by far the most critical competency that a

composite manufacturer can have. Only a panoramic understanding of the various design aspects

of composites can lead to a realistic Finite Element Analysis result and an Optimal Composite

Design - that performs as intended.

Ironically these design aspects are numerous and many of them defy conventional engineering

design logic. For example, as design engineers we are taught to "ignore" shear forces, in say a

"beam" made in a Structural Steel Rectangular Box Section - because as compared to the flexural /

bending forces, the shear forces are negligible (Actually the shear forces are not negligible - its

just that the shear modulus and shear strength of isotropic materials like steel etc. is relatively high

- thus the shear forces pale in comparison).

If one were to follow this logic and design a composite rectangular box section - then one is

setting oneself up for a rude awakening. In Advanced Composite Structures shear forces cause

significant deflections and in most poorly designed cases - it is the primary cause for failure.

Over the past few years we have developed sufficient expertise in using "Composite Design

Principles" or "Composite Science" to design optimal composite components - relying heavily on

the Finite Element Analysis Method.

Composite Finite Element Analysis (FEA) Capabilities

Our Composite Finite Element Analysis Capabilities can lead to the following tailor-made

43

properties of advanced composites,

Composite Mechanical Properties - Tensile, Compression, Deflection etc.

Composite Thermal Properties - Thermal Expansion, Thermal Conductivity etc.

Composite Fatigue Properties

Composite Electric Properties - Electric Insulation, Electric Conductivity etc.

Composite Radio Transparency - Electromagnetic Radiation, Diagnostic and Treatment

Radiation like X-Rays, Gamma Radiation etc.

We use special software, tailor-made for composite finite element analysis and optimization for

designing composite products. Dedicated high-end workstations for product modeling as well as

for finite element analysis support the composite design process.

11.2 Ultrasonic Testing

Nondestructive testing (NDT) methods play an important role in physical characterization of new

composite materials and in assessment of their quality and serviceability in structures. Ultrasonic

NDT methods are an effective instrument for evaluation of elastic modules, strength, stiffness,

and other essential parameters which are vital for analysis