Chapter 14: Manufacturing of fibre–polymer composite materials – Introduction to Aerospace Materials


Manufacturing of fibre–polymer composite materials

14.1 Introduction

The rapid growth in the use of fibre–polymer composites in aircraft and helicopters since the mid-1990s has led to it now sitting side-by-side with aluminium as the most commonly used structural materials. A critical factor underpinning the growth in composites has been the development of improved, cheaper manufacturing processes. Similarly to the casting and forming processes used to manufacture metal structures, the processes used to produce composite components have a major impact on the cost, quality and material properties. The choice of manufacturing process impinges on the design of the composite component; some processes are suited to flat or slightly curved structures whereas others are better for complex, highly curved parts. The choice of process also determines the mechanical properties of the composite by affecting the volume fraction and orientation of the reinforcing fibres.

The aerospace industry has invested heavily in the development of manufacturing processes capable of producing many types of composite components for aircraft structures and engines. The industry has also developed processes to reduce the manufacturing cost; which typically accounts for over 60–70% of the total production cost (with the remainder being materials, nondestructive inspection and other process-related costs). Figure 14.1 shows the approximate reduction in the cost of manufacturing composite components for civilian and military aircraft since 1980, and costs have fallen considerably owing to advances in automation, rapid processing and other manufacturing technologies.

14.1 Approximate reductions in the production costs of aerospace composites owing to improvements in manufacturing technologies.

Composites for aircraft applications are manufactured in two basic material forms: laminate and sandwich composites (Fig. 14.2). Laminates consist of multiple layers of fibre and resin (called plies) which are bonded together into a solid material. The fibres are oriented along the in-plane directions of principal loading to provide high stiffness, strength and fatigue resistance and the polymer matrix binds the fibres into the material. Laminates made of carbon fibre–epoxy resin composite are used in the most heavily-loaded aircraft structures. Sandwich composites consist of thin face skins (usually carbon–epoxy laminate) bonded (often with adhesive film) to a thick, lightweight core material. Sandwich structures are often used in lightly loaded structures that require high resistance to bending and buckling.

14.2 Two material forms of composites: (a) laminate and (b) sandwich composite.

The cost of manufacturing aircraft parts is usually higher with composites than with aluminium. This is because carbon–epoxy is more expensive than aluminium, the tooling costs are higher, and the production processes are often slower and more labour intensive. However, there are numerous benefits in the manufacture of composite components. The number of assemblies, parts and fasteners needed in the construction of composite structures is usually much less than the same structure made using metal. For instance, the fin box of the vertical tailplane of a single aisle aircraft contains over 700 parts and about 40 000 fasteners when made with aluminium, but this falls to only 230 parts and around 10 000 fasteners with composite. As another example, one metal fuselage barrel for a mid-sized airliner requires around 1500 sheets of aluminium held together by about 50 000 rivets. With composite construction, the barrel can be made as a single-piece and the number of fasteners and rivets drops by about 80%.

In this chapter the processes used in the manufacture of composite components for aircraft are studied. Firstly, the processes used to produce carbon, glass and aramid fibres, which are the main types of reinforcement used in aircraft composites, are examined. The processing methods used to produce fibres have a major impact on their mechanical properties and subsequently the structural properties of the composite part. Therefore, the manufacture of these fibres is important to fully understand. Composite manufacturing processes can be roughly divided into two categories depending on how the polymer is combined with the fibres: prepreg-based processes and resin-infusion processes. Several prepreg-based processes are described: autoclave curing, automated tape lay-up and automated fibre placement. A variety of resin infusion processes are explained, including resin transfer moulding, resin film infusion, vacuum-bag resin infusion, and filament winding. Most manufacturing processes for composites are capable of producing parts to the near-net shape, and a significant amount of machining is not required (unlike many metal products). After manufacturing, most composites only require a small amount of trimming and hole drilling, and the machining processes used to remove material are briefly explained.

14.2 Fibre reinforcements for composites

14.2.1 Carbon fibres: production, structure and properties

Carbon (graphite) fibres are very stiff, strong and light filaments used in polymer (usually epoxy) matrix composites for aircraft structures and jet engine parts. Polymer composites containing carbon fibres are up to five times stronger than mild steel for structural parts, yet are five times lighter. Carbon fibre composites have better fatigue properties and corrosion resistance than virtually every type of metal, including the high-strength alloys used in airframe and engine components. Carbon fibres do not soften or melt and, when used in reinforced carbon–carbon composites, have exceptional heat resistance for high-temperature applications such as thermal insulation tiles, aircraft brake pads and rocket nozzles (as described in chapter 16). Although carbon fibres do not soften/melt, they do ultimately lose strength at high temperature via oxidation.

Applications for carbon-fibre composites have expanded dramatically since their first use in the 1960s and, alongside aluminium, is the most used aircraft structural material. Important factors behind the greater use of composites have been improvement to the properties and reduction in the cost of carbon fibres. Figure 14.3 shows the large increase in the use and reduction in the average price of carbon fibres since their first application in aircraft. It is important to note that large variations exist in the price depending on the fibre properties, and therefore the figure provides an indicative trend.

14.3 Changes to the consumption and average price of carbon fibre owing to increasing use in both aerospace and non-aerospace applications.

The properties of carbon fibre that make it an important aerospace material (stiffness, strength and fatigue resistance), are determined by the process methods used in its production. Carbon fibre is extracted via heat treatment from a carbon-rich precursor material, the most common of which are polyacrylonitrile (PAN), pitch and rayon. Carbon fibres produced for aerospace applications are made using PAN, which is an organic polymer. Other precursor materials such as pitch and rayon are rarely used in the production of aircraft-grade carbon fibres because of their lower mechanical performance, but they are used to produce fibres for non-aerospace applications such as sporting equipment, civil infrastructure and marine craft.

Before the production of the carbon fibre, the PAN precursor material is stretched into long, thin filaments. This stretching causes the PAN molecules to align along the filament axis, which subsequently increases the stiffness of the carbon fibre after processing. The greater the stretch applied to the PAN, the higher the preferred orientation of the molecules along the filament, resulting in stiffer carbon fibre. PAN filaments are heat treated in a multistage process while under tension to produce carbon fibres. The process begins by stretching and heating the PAN filaments to 200–300 °C in air. This treatment oxidises and crosslinks the PAN into thermally stable carbon-rich fibres. The fibres are stretched during heating to prevent them contracting during oxidation. The PAN is then pyrolysed at 1500–2000 °C in a furnace having an inert atmosphere (e.g. argon gas), which stops further oxidation of the carbon. This heat treatment is called carbonisation because it removes non-carbon atoms from the PAN molecules (e.g. N, O, H) leaving a carbon-rich fibre with a purity content of 93–95%. The heat treatment changes the molecular bond structure into graphite. Carbon fibre consists of closely packed layers of graphite sheets in which the carbon atoms are arranged in a two-dimensional hexagonal ring pattern that looks like a sheet of chicken wire. The graphite sheets are stacked parallel to one another in a regular pattern (as represented in Fig. 14.4) and the sheets are aligned along the fibre axis.

14.4 Sheet structure of graphite. Each sphere represents a carbon atom arranged in a hexagonal pattern in sheets.

The mechanical properties of carbon fibre vary over a large range depending on the temperature of the final heat treatment. There are two general categories of carbon fibre produced depending on the final temperature: high-modulus or high-strength (Table 14.1). The stiffness and strength of carbon fibre is determined by the pyrolysis temperature for the final processing stage, as shown in Fig. 14.5. Peak strength is reached by heating the fibre at 1500–1600 °C whereas the elastic modulus increases with temperature. The production of high-strength carbon fibres involves carbonisation of the PAN filaments at 1500–1600 °C. An additional heat treatment step called graphitisation is performed after carbonisation to produce high modulus fibres, which are heated in an inert atmosphere at 2000–3000 °C. The higher temperature used for graphitisation increases the preferred orientation of the graphite sheets along the fibre axis and thus raises the elastic modulus.

Table 14.1

Properties of PAN-based carbon fibres

Property High-modulus carbon fibres High-strength carbon fibres
Density (g cm− 3) 1.9 1.8
Carbon content (%) + 99 95
Tensile modulus (GPa) 350–450 220–300
Tensile strength (MPa) 3500–5500 3500–6200
Elongation-to-failure (%) 0.1–1.0 1.3–2.2

14.5 (a) Production process for carbon fibres; (b) effect of final processing temperature on the tensile modulus and strength of carbon fibre.

By appropriate choice of the final process temperature it is possible to control the elastic modulus and strength of carbon fibres for specific structural applications. For instance, carbon fibres used in composite structures which require a high strength-to-weight ratio, such as the wing box, are pyrolysed at 1500–1600 °C to yield high strengths of 3500–6000 MPa. Structures needing a high stiffness-to-weight ratio such as the control surfaces may contain high stiffness fibres heated at 2500–3000 °C which have an elastic modulus of 350–450 GPa. Structures that require both high stiffness and strength such as the fuselage and wings may contain intermediate modulus fibres.

The properties of carbon fibres are determined by the internal arrangement of the graphite sheets, which is controlled by the final temperature. A single carbon filament is a long rod with a diameter of 7–8 μm. Packed within PAN-based fibres are tiny ribbon-like carbon crystallites which are called turbostratic graphite. In turbostratic graphite, the graphite sheets are haphazardly folded or crumpled together as shown in Fig. 14.6. The intertwined sheets are orientated more or less parallel with the fibre axis, and this crystal alignment makes the fibres very strong along their axis. This arrangement also makes the fibres highly anisotropic; the interatomic forces are much stronger along than between the sheets. For instance, the stiffness along the fibre axis is typically 15 to 30 times higher than across the fibre.

14.6 Three-dimensional representation of the turbostratic graphite structure of PAN-based carbon fibre (from S. C. Bennett and D. J. Johnston, Structural heterogeneity in carbon fibers, Proceedings of the 5th London carbon and graphite conference, Vol. 1, Society for Chemical Industries, London, 1978, pp. 377–386).

The stiffness of carbon fibre is dependent on the alignment of the crystalline ribbons and the degree of perfection of the graphite structure. The length and straightness of the graphite ribbons determines the fibre stiffness. The order between the ribbons becomes better with increasing temperature (up to 3000 °C), which is the reason for the steady improvement in elastic modulus with the final process temperature.

Fibre strength is determined by the number and size of flaws, which are usually tiny surface cracks. During processing, the fibres are damaged by fine-scale abrasion events when they slide against each other after leaving the furnace and during collimation into bundles. The rubbing action and other damaging events introduces surface cracks that are usually smaller than 100 nm. Although tiny, these cracks have a large impact on fibre strength.

After final heat treatment, carbon fibres are surface treated with various chemicals called the size which serves several functions. A thin film of chemicals is applied to the fibre surface to increase its bond strength to the polymer matrix. Carbon does not adhere strongly to most polymers, including epoxies, and it is necessary to coat the fibre surface with a thin sizing film to promote strong bonding. Size agents are also used to reduce friction damage between fibres. Shortly after the final heat treatment, the individual carbon fibres are collimated into bundles (which are also called tows) for ease of handling.

Section 14.11 at the end of this chapter presents a case study of carbon nanotubes in composites.

14.2.2 Glass fibres: production, structure and properties

Glass fibre composites are used sparingly in aircraft structures owing to their low stiffness. Glass fibre has a low elastic modulus (between 3 to 6 times lower than carbon fibre) which gives glass-reinforced composites a comparatively low stiffness-to-weight ratio. Glass fibre composites are only used in structures where specific stiffness is not a design factor, which is mainly secondary components such as aircraft fairings and helicopter structures such as the cabin shell. Glass fibre composites have low dielectric properties and, therefore, are used when transparency to electromagnetic radiation (i.e. radar waves) is important, such as radomes and aerial covers.

Despite the restricted use of glass-reinforced composites in components where stiffness is a critical property, whenever possible they are used instead of carbon-fibre materials because of their lower cost. Glass fibres are anywhere from 10 to 100 times cheaper than carbon fibres. The greatest use of glass-reinforced composite is inside the aircraft cabin. The majority of cabin fittings, including overhead luggage storage containers and partitions, are fabricated using glass fibre-phenolic resin composite that is inexpensive and lightweight with good flammability resistance.

Glass fibre is a generic name similar to carbon fibre or steel, and as for these materials there are various types having different properties. Glass fibres are based on silica (SiO2) with additions of oxides of calcium, boron, iron and aluminium. It is the different concentrations of metal oxides that allow different glass types to be produced. There are two types of glass fibres used for aircraft applications: E-glass (short for electrical grade) and S-glass (structural grade). E-glass is cheaper and lower in strength than S-glass and, therefore, is used mostly in composites for aircraft cabin fittings that do not require high structural properties. The higher strength S-glass composite is used in structural components. The composition and engineering properties of E- and S-glass fibres are given in Table 14.2.

Table 14.2

Composition and properties of E-glass and S-glass fibres

The internal structure of glass fibre is different from carbon fibre. Glass consists of a silica network structure containing metal oxides, as shown in Fig. 14.7. The network is a three-dimensional structure and, therefore, the fibre properties are isotropic, thus (unlike carbon) the elastic modulus is the same parallel and transverse to the fibre axis.

14.7 Internal molecular structure of glass fibre. The metal oxides in glass are not shown.

Glass fibre is manufactured using a viscous drawing process whereby silica and metal oxide powders are initially blended and melted together in a furnace at about 1400–1500 °C. The molten glass flows from the furnace via bushings or spinnerets containing a large number of tiny holes. The glass solidifies into thin continuous fibres as it passes through the holes. The fibre diameter is around 12 μm. Upon leaving the furnace the fibres are cooled using a water spray mist and then coated with a thin layer of size. The size consists of several functional components including chemicals to improve adhesion with the polymer matrix and lubricants to minimise surface abrasion during handling.

14.2.3 Aramid (Kevlar) fibres: production, structure and properties

Synthetic organic fibres are used in polymer composites for specific aerospace applications. Organic fibres are crystalline polymers with their molecular chains aligned along the fibre axis for high strength. Examples of organic fibres are Dyneema® and Spectra®, which are both ultra-high-molecular-weight polyethylene filaments with high-strength properties. Of the many types of organic fibres, the most important for aerospace is aramid, whose name is a shortened form of aromatic polyamide (poly-p-phenylene terephthalamide). Aramid is also called Kevlar which is produced by the chemical company Du Pont.

The most common aerospace application for aramid fibre composites is for components that require impact resistance against high-speed projectiles. Aramid fibres absorb a large amount of energy during fracture, thus providing high perforation resistance when hit by a fast projectile. For this reason, aramid fibre composites are used for ballistic protection on military aircraft and helicopters. They are also used for containment rings in jet engines in the event of blade failure. Aramid composites, similarly to glass-reinforced composites, have good dielectric properties, making them suitable for radomes. Aramid composites also have good vibration damping properties, and therefore are used in components such as helicopter engine casings to prevent vibrations from the main rotor blades reaching the cabin. Figure 14.8 shows the vibration damping loss factor for several aerospace materials; aramid–epoxy has ten times the loss decrement of carbon–epoxy and nearly 200 times higher than aluminium.

14.8 Vibration loss damping factor of aramid–epoxy composite compared with other aerospace materials.

The process of producing aramid fibres begins by dissolving the polymer in strong acid to produce a liquid chemical blend. The blend is extruded through a spinneret at about 100 °C which causes randomly oriented liquid crystal domains to develop and align in the flow direction. Fibres form during the extrusion process into highly crystalline, rod-like polymer chains with near perfect molecular orientation in the forming direction. The molecular chains are grouped into distinct domains called fibrils. The fibre essentially consists of bundles of fibrils that are stiff and strong along their axis but weakly bonded together, as shown in Fig. 14.9. As for carbon fibres, aramid fibres are highly anisotropic with their modulus and strength along the fibre axis being much greater than in the transverse direction. The properties of two common types of aramid fibre are given in Table 14.3, and they exceed the specific stiffness and strength of glass fibres. Aramid fibre composites are lightweight with high stiffness and strength in tension. However, these materials have poor compression strength (which is only about 10–20% the tensile strength) owing to low micro-buckling resistance of the aramid fibrils. Therefore, aramid composites should not be used in aircraft components subject to compression loads. Aramid fibres can absorb large amounts of water and are damaged by long-term exposure to ultraviolet radiation. Therefore, the surface of aramid composites must be protected to avoid environmental degradation.

Table 14.3

Comparison of average properties of aramid fibres against glass and carbon fibres

14.9 Polymer chains aligned in the fibre direction in aramid.

14.3 Production of prepregs and fabrics

Composite aircraft components contain anywhere from tens of thousands to many millions of fibres. For example, an average-sized wing panel made using carbon–epoxy composite comprises of the order of 20 000 million fibres. Fibres are too fine to easily handle and process into composite parts, with carbon fibres being only 1/10 to 1/20 the width of human hair (Fig. 14.10). Therefore, upon leaving the furnace the fibres are bundled together into tows or yarns. Aerospace fibres are collimated into tows that usually contain about 3 K (3000), 12 K (12 000) or 24 K (24 000) filaments (Fig. 14.11). These bundles are then used to make prepreg (resin pre-impregnated) material or dry fabric for the production of composite parts.

14.10 Comparison of the width of carbon fibre and human hair.

14.11 (a) Fibres bundled into a tow. (b) Internal view of a tow showing many individual fibres.

14.3.1 Prepregs

Major aircraft manufacturers use prepreg tape in the production of composite structural components. The term prepreg is short form for pre-impregnated fibres. Prepreg is a two-part sheet material consisting of fibres (e.g. carbon) and partially cured resin (e.g. epoxy). The most common prepreg used in aircraft is carbon–epoxy, although many other types are used including carbon–bismaleimide, S-glass–epoxy and aramid-epoxy. The benefits of using prepreg include accurate control of the fibre volume content and the ability to achieve high fibre content, thus allowing high-quality composite components to be produced with high mechanical properties.

Several methods are used to produce prepreg, the most common being the solution dip, solution spray, and hot-melt techniques. The solution dip method involves dissolving the resin in solvent to a solids concentration of 40–50%. Fibres are then passed through the solution so that they pick up an amount of resin solids. The fibres leave the solution coated with a thin film of resin, and they are then pressed into thin sheets of prepreg. The solution spray method simply involves spraying liquid resin onto the fibres whereas the hot-melt technique involves direct coating of the fibres with a low viscosity resin.

After the prepreg is produced, the resin is partially cured to a condition where it is semisolid; it is too hard to flow like a liquid but soft enough to be pliable and flexible. The resin needs to be hard enough so it does not leak out from between the fibres during the cutting and laying of the prepreg plies. The resin also needs to be soft enough to allow the prepreg to be easily deformed to the shape of the composite component. The resin must be tacky enough (sticky to touch) in order to bond the prepreg layers together during the lay-up of the composite. Epoxy resin used in carbon-fibre prepreg is slightly cured to a semisolid condition, whereby the crosslinking between the polymer chains is about 15–30% complete. Prepreg in this condition is called B-stage cured. The prepreg must be stored at low temperature (about − 20 °C) inside a freezer to avoid further curing and crosslinking of the resin matrix which occurs at room temperature.

B-stage cure prepreg is produced as a thin sheet (usually 0.1–0.4 mm thick) with a fibre content of 58–64%. The prepreg is protected on both sides with easily removable separators called backing paper. Backing paper stops the prepreg sheets from sticking to each other before the lay-up of the composite part. Prepreg is used in the manufacture of composite parts by simply cutting to shape and size, removing the backing paper, laying the prepreg sheets in a stack in the preferred fibre orientation, and then consolidating and curing the final composite material using an autoclave as described in 14.5.1.

14.3.2 Dry fabrics

The aerospace industry is increasingly using dry carbon fabric instead of carbon-fibre prepreg to manufacture aircraft structures. There are several advantages gained by using fabric rather than prepreg, including lower material cost, infinite storage life, no need for storage in a freezer, and better formability into complex shapes. The most common fabric is woven fabric produced on weaving looms, and the main styles are plain, twill and satin weaves as shown in Fig. 14.12. Woven fabrics contain fibre tows aligned in the warp direction (which is the weaving direction) and the weft direction, which is perpendicular to the warp. In plain woven fabric, each warp tow alternately crosses over and under each intersecting weft tow. Twill and satin fabrics are woven such that the tows go over and under multiple tows. When observed from the side, weaving causes periodic out-of-plane waviness of the fibres. The waviness results in significant loss in stiffness and strength of the composite because maximum structural performance is achieved when the fibres are absolutely straight and in-plane. Twill and satin weaves have lower degrees of fibre waviness than plain weave, and therefore their composites have higher in-plane mechanical properties. For this reason, twill and satin woven fabrics are preferred over plain woven fabric in the fabrication of aerospace composite components.

14.12 Common weave architectures: (a) plain, (b) twill and (c) satin.

Other types of fabric used in aircraft composite parts include non-crimp, braided and knitted fabrics. Non-crimp fabric consists of multiple layers of straight tows oriented at different angles, which are bound together by through-thickness stitches. The tows in non-crimp fabric are not forced to interlace with other tows to produce a weave, and therefore the fibres are straight and in-plane for high structural performance. Braided and knitted fabrics are used when high impact collision resistance and high conformed shapes in composite components are required, respectively, although they have low in-plane mechanical properties.

A family of fabrics containing fibre tows aligned in the in-plane direction together with tows running in the through-thickness direction are available for damage tolerant aircraft composite components. The fabrics have a three-dimensional array of fibres, as shown schematically in Fig. 14.13, which allows in-plane and through-thickness loads to be carried by the composite. Composite laminates made using woven fabrics (such as those shown in Fig. 14.13) do not have through-thickness fibres. Such composites have low through-thickness strength and damage resistance. Composites containing a three-dimensional fibre structure have superior damage tolerance because the through-thickness fibres can carry out-of-plane loads. There are various types of three-dimensional fabrics that are produced by orthogonal weaving, stitching, tufting and other specialist techniques. The application for these types of fabrics in aircraft structures is currently limited, although their use is expected to grow.

14.13 Three-dimensional fibre tow structure of fabrics.

14.4 Core materials for sandwich composites

Many aircraft structural components are required to carry only light or moderate loads. When a small load is applied on a composite laminate, the structure is designed with a thin skin and few ribs, frames and spars to reduce weight. When the skin becomes too thin and the stiffeners too few then the structure loses its resistance against buckling and, therefore, some additional form of stiffening is required. A common solution is to build the laminate skins as a sandwich by inserting a lightweight filler or core layer. Under bending, the skins carry in-plane tension and compression loads whereas the core is subjected to shear, as shown in Fig. 14.14. The skin-core construction greatly increases the bending stiffness and thereby makes sandwich structures more resistant to buckling, with only a small increase in weight. Examples of aircraft components made using sandwich composites are control surfaces (e.g. ailerons, flaps) and vertical tailplanes.

14.14 Basic loading of sandwich composite.

Various types of materials are used for the core, with aluminium honeycomb and Nomex being the most common in aircraft sandwich components. Nomex is the tradename for a honeycomb material based on aramid fibres in a phenolic-resin matrix. Aluminium and Nomex core materials have a lightweight cellular honeycomb structure, which provides high shear stiffness (Fig. 14.15). The aerospace industry is increasingly using polymer foams instead of aluminium honeycomb and Nomex because of their superior durability and high-temperature properties. Examples of polymer foams are polyetherimide (PEI) and polymethacrylimide (PMI), and these materials have a low density cellular structure which has high specific stiffness and strength (Fig. 14.16).

14.15 Honeycomb structure of aluminium.

14.16 Cellular microstructure of polymer foams.

14.5 Composites manufacturing using prepreg

14.5.1 Manual lay-up

Composites have been fabricated using prepreg since the 1970s, and the original fabrication method involved the manual lay-up of the prepreg plies into the orientation and shape of the final component. Manual lay-up involves cutting the prepreg to size, removing the backing paper, and then stacking the prepreg plies by hand onto the tool surface. This process is slow, labour-intensive and inconsistent because of human error in the accurate positioning of the plies. For these reasons, manual hand-lay is rarely used in the construction of large composite components, although it is still used for making small numbers of parts when it is not economically viable to automate the process.

The prepreg is laid-up by hand directly onto the tool, which has the shape of the final component. The prepreg plies are oriented with their fibres aligned in the main loading directions acting on the component when used in service. The most common pattern used in the lay-up of plies for aircraft components is [0/+ 45/–45/90], which is shown in Fig. 14.17. A composite with this lay-up is called quasi-isotropic because the stiffness and strength properties are roughly equal when loaded along any in-plane direction. The 0° and 90° fibres carry the in-plane tension, compression and bending loads whereas the 45° fibres support shear loads. Another common ply pattern is [0/90], known as cross-ply, which is used for composite components subject to in-plane tension or compression loads in service, but not shear loads.

14.17 Ply orientations for quasi-isotropic cross-ply composite reproduced with permission from D. R. Askeland, The science and engineering of materials, Stanley Thornes (Publishers) Ltd., 1998.

It is important that the orientation of the plies is symmetric around the mid-plane to ensure the material is balanced. That is, the plies in the upper half of the prepreg stack must be arranged as a mirror image of the plies in the lower half. If this does not occur, then the ply pattern is asymmetric and it is possible that the composite may distort or warp after the prepreg is cured. The symmetric lay-up of plies is required for all composite fabrication processes, and not just for prepreg.

After the prepreg is stacked to the correct ply orientation and thickness, it is then compacted by vacuum bagging to remove air from between the ply layers. The prepreg stack is sealed within a plastic bag from which air is removed using a vacuum pump, as shown in Fig. 14.18. The bag is a flexible membrane that conforms to the shape of the prepreg and the underlying tool to ensure good dimensional tolerance. Release film and bleeder layer cloth are placed over the prepreg stack within the vacuum bag. Release film, which is a nonstick flexible sheet containing tiny holes, is used to stop the prepreg from sticking to the bleeder cloth. This cloth is used to absorb excess resin, which is squeezed from the prepreg during consolidation. The prepreg is consolidated and cured within an autoclave, and excess resin flows from the prepreg through the fine holes in the release film into the bleeder layer where it is absorbed.

14.18 Schematic of vacuum bagging operation (from

The final stage in the fabrication of prepreg composite involves consolidation and curing inside an autoclave (Fig. 14.19). The prepreg, which is resting on the tool and enclosed within the vacuum bag, is placed inside an autoclave, which is a closed pressure chamber in which consolidation and cure processes occur under the simultaneous application of pressure and high temperature. An autoclave is a large pressure cooker, in which the prepreg is compacted using pressurised nitrogen and carbon dioxide to over 700 kPa. At the same time, the autoclave is heated to cure the polymer matrix, which for a carbon–epoxy prepreg requires temperatures of 120–180 °C. The combined action of pressure and heat consolidates the composite, removes trapped air, and cures the polymer matrix. The autoclave process produces high-quality composites with high fibre contents (60–65% by volume) which are suitable for primary and secondary components for aircraft and helicopters.

14.19 Autoclave used for consolidation and curing of prepreg composite.

The production of composite components using the autoclave process is practised throughout the aircraft industry. However, the industry is also using other manufacturing processes, as described in the following sections, that avoid the need for an autoclave. Autoclaves are expensive and can only produce components of limited size (less than about 15 m long and 4 m wide).

14.5.2 Automated tape lay-up

Automated tape lay-up (ATL) is an automated process used to lay-up prepreg tape in the fabrication of composite aircraft structures. The process is used in the manufacture of carbon–epoxy prepreg components for both military and commercial aircraft. Examples of military parts made using ATL are the wing skins of the F-22 Raptor, tilt rotor wing skins of the V-22 Osprey, and skins of the wing and stabiliser of the B-1 Lancer and B-2 Spirit bombers. ATL is used in the production of empennage parts (e.g. spars, ribs, I-beam stiffeners) for the B777, A340-500/600 and A380 airliners.

ATL is used instead of manual lay-up to reduce the time (by more than 70–85%) and cost spent in the lay-up of prepreg tape on the tool. For instance, the lay-up rate in the production of the primary wing spar to the Airbus A400M is increased from 1 to 1.5 kg h−1 of prepreg with manual lay-up to about 18 kg h−1 with ATL. The first wing spars to be produced using manual lay-up took about 180 h whereas the lay-up time using the ATL process is only 1.5 h. Other benefits of the ATL process include less material scrap and better repeatability and consistency of the manufactured parts.

The key component of the ATL process is a computer numerically controlled tape-laying head that deposits prepreg onto the tool at a fast rate with great accuracy (Fig. 14.20). The head device is suspended via a multi-axis gantry above the tool surface. A roll of prepreg (75–300 mm wide) in the tape-laying head is deposited on the tool in the desired orientations according to a programmed routine. As the tape is laid down, the head removes the backing paper and applies a compaction force onto the prepreg. The head can also apply moderate heat to the prepreg to improve tackiness and formability. The head is programmed to follow the exact contour of the tool at a speed of about 50 m min−1 for the rapid deposition of prepreg (4–45 kg h−1). When the tape-laying head reaches the location where the ply terminates then the tape is automatically cut by blades within the head, the head changes direction, and the tape laying process continues. The head can lay any number of prepreg plies on top of each other in a series of numerically controlled steps with a high degree of accuracy, thereby assuring consistent shape, thickness and quality for the part. After the automated tape lay-up is complete, the prepreg is consolidated and cured. The ATL process is capable of producing composite parts with a flat or slightly curved profile; it is not suited for highly contoured parts which are better made using the automated fibre placement process.

14.20 ATL process in the manufacture of aircraft panels.

14.5.3 Automated fibre placement

Automated fibre placement (AFP) is used in the automated production of large aircraft structures from prepreg (Fig. 14.21). The process involves the lay-up of individual prepreg tows onto a mandrel using a numerically controlled fibre-placement machine. In the AFP process, prepreg tows or narrow strips of prepreg tape are pulled off holding spools and fed into the fibre placement head. Within the head, the tows, which contain about 12 000 fibres and are approximately 3 mm wide, are collimated into bundles of 12, 24 or 32 tows to produce a narrow band of prepreg material that is deposited onto the mandrel which has the shape of the final component. The placement head is computer controlled via a gantry system suspended above the mandrel. During fibre placement, the mandrel is rotated so the prepreg is wound into the shape of the component. The head moves along the rotating mandrel to steer the fibres so they follow the applied stresses acting on the finished component in service. The head is able to stop, cut the prepreg, change direction and then re-start the lay-down of prepreg tows until the material is built up to the required thickness. Each tow is dispensed from the head at a controlled speed to allow it conform to the mandrel surface, thereby allowing highly curved components to be produced, not possible with ATL. After the lay-up process is complete, the prepreg is cured in the same way as composites manufactured by manual lay-up or ATL.

14.21 (a) Schematic of the automated fibre placement process. (b) ATL process in the construction of fuselage barrels for the B787 aircraft. Photograph by B. Nettles, Charleston Post and Courier.

The AFP process is used by the aerospace industry to fabricate large-circumference and highly contoured structures such as fuselage barrels, ducts, cowls, nozzle cones, spars and pressure tanks. The process is used in the construction of carbon–epoxy fuselage sections to the B787 Dreamliner, V-22 Osprey, and the Premier 1 and Hawker Horizon business jets. For the Premier 1, the AFP process is used to construct one-piece carbon–epoxy fuselage barrels measuring 4.5 m long and 2 m at the widest point. AFP is used to fabricate inlet ducts, side skins and covers to the F-18 E/F Superhornet.

14.6 Composites manufacturing by resin infusion

The aerospace industry is increasingly using manufacturing processes that do not rely on prepreg to produce composite components. Composites made using prepreg are high-quality materials with excellent mechanical properties owing to their high fibre content. However, prepreg is expensive, must be stored in a freezer, has limited shelf-life (usually 1–2 years), and must be cured in an autoclave which is slow and expensive. The aerospace industry is also keen to use ‘out of autoclave’ processes because of the faster manufacturing times and lower cost.

The industry is producing composite components using dry fabric which is infused with liquid resin and then immediately consolidated and cured. The types of fabrics used include the plain, twill, satin, knitted and non-crimped materials. These fabrics are infused with resin using one of several manufacturing processes, including resin transfer moulding, vacuum bag resin infusion, resin film infusion, filament winding and pultrusion. Many aerostructure companies also use their own proprietary manufacturing process that is developed in-house. Not all of the many processes are described here; instead just a few are outlined to illustrate the diversity of the processes.

14.6.1 Resin transfer moulding (RTM)

Resin transfer moulding (RTM) is used to fabricate composite components of moderate size (typically under 3 m), such as the fan blades of F135 engines and the spars and ribs for the mid-section fuselage and empennage of the F-35 Lightning II fighter. RTM is a closed-mould process that is illustrated in Fig. 14.22. Fabric is placed inside the cavity between two matched moulds with their inner surfaces having the shape of the final component. Fabric plies are stacked to the required orientation and thickness inside the mould, which is then sealed and clamped. Liquid resin is injected into the mould by means of a pump. The resin flows through the open spaces of the fabric until the mould is completely filled. The resin viscosity must be low enough for easy flow through the tiny gaps between the fibres and tows of the fabric. Aerospace-grade resins with low viscosity have been specifically developed for the RTM process. After injection, the mould is heated in order to gel and cure the polymer matrix to form a solid composite part. After curing, the mould is opened and the part removed for edge trimming and final finishing.

14.22 Schematics of the (a) resin transfer moulding and (b) vacuum assisted resin transfer moulding processes.

The RTM process can produce composites with high fibre volume content (up to 65%), making them suitable for primary aircraft structures that require high stiffness, strength and fatigue performance. However, it can be difficult to completely infuse some types of fabric with resin which leaves dry spots or voids in the cured composite. Furthermore, the fibre architecture can be disturbed by the high flow pressures needed to force the resin through some fabrics. To minimise these problems, a variant of the RTM process called vacuum-assisted resin transfer moulding (VARTM) is used.

The VARTM process is shown in Fig. 14.22b, and is different to conventional RTM in that a vacuum-pump system is used to evacuate air from the mould and draw resin through the fabric. After the mould containing the fabric is closed and sealed, a vacuum pump is used to extract air from the cavity placing it in a state of low pressure. Rather than resin being injected into the mould cavity under pressure as in conventional RTM, with VARTM the resin is drawn into the mould under the pressure differential created by the vacuum. Resin percolates between the fibres and tows of the fabric until the mould is filled, at which stage the infusion process stops and the part is cured at elevated temperature.

14.6.2 Vacuum bag resin infusion (VBRI)

The vacuum bag resin infusion (VBRI) process and similar processes are used to fabricate various types of carbon–epoxy structural components. Figure 14.23 illustrates the basic configuration of the VBRI process, which uses an open mould rather than the two-piece closed mould of the RTM and VARTM processes. This results in VBRI having lower tooling costs, which can be a large capital cost. Fabric plies are stacked on the tool with the top layer being a resin distribution fabric. The fabric stack is enclosed and sealed within a flexible plastic bag that is connected at one point to a liquid resin source and at another point to a vacuum-pump system. Air is removed by the vacuum pump which causes the bag to squeeze the fabric layers to the shape of the mould surface. This part of the process is similar to the vacuum bagging of prepreg before autoclave curing. In the VBRI process, liquid resin flows into the bag under the pressure differential created by the vacuum pump. Resin is drawn through the tightly consolidated fabric as well as along the top resin distribution fabric, from where it seeps downwards into the fabric. After the fabric is completely infused with resin, the material is cured at high temperature within an oven.

14.23 Schematic of the vacuum bag resin infusion process.

14.6.3 Resin film infusion (RFI)

The resin film infusion (RFI) process is suited for making relatively large structures such as stiffened skins and rib-type structures (Fig. 14.24). The process uses an open mould upon which layers of dry fabric and solid resin film are stacked. The film is a B-stage cured resin similar to the cure condition of the resin matrix in prepreg. Film is placed at the bottom, top or between the layers of fabric. The materials are sealed within a vacuum bag and then air is removed using a vacuum pump. The entire assembly is placed into an autoclave and subjected to pressure and heat. The temperature is increased to reduce the resin viscosity to a level when it is fluid enough to flow into the fabric layers under the applied pressure. Once the infusion is complete the pressure and temperature are raised to consolidate and fully cure the component.

14.24 Schematic of the resin film infusion process.

14.6.4 Filament winding

Filament winding is a manufacturing process where cylindrical components are made by winding continuous fibre tows over a rotating or stationary mandrel, as shown in Fig. 14.25. There are two types of filament winding process: wet winding and prepreg winding. Wet winding involves passing continuous tows through a resin bath before reaching a feed head which deposits them onto a cylindrical mandrel. Prepreg winding involves depositing thin strips of prepreg onto the mandrel in a similar manner to the ATP process. The feed head is rotated around the stationary mandrel or, more often, the mandrel is rotated while the feed head passes backwards and forwards along its length. Successive layers of tows are laid down at a constant or varying angle until the desired thickness is reached. Winding angles can range between 25° and 80°, although most winding processes are performed around 45° to provide the part with high hoop stiffness and strength. After winding is complete, the composite is cured at room temperature or at elevated temperature inside an oven or autoclave. The mandrel is then pulled away from the cured composite.

14.25 (a) Schematic and (b) photograph of the filament winding process.

The filament-winding process is used to produce cylindrical composite components. Examples of aerospace components produced using this process include motor cases for Titan IV, Atlas and Delta rockets, pressure vessels, missile launch tubes and drive shafts. Some of these components, such as rocket motor cases, are very large (exceeding 4 m in diameter).

14.6.5 Pultrusion

Pultrusion is an automated, continuous process used to manufacture composite components with constant cross-section profiles. Figure 14.26 illustrates the pultrusion process which is a linear operation starting from the right-hand side of the diagram. Continuous fibres (tows) are pulled off storage spools and drawn through a liquid resin bath. The resin-impregnated fibres exit the bath and are pulled through a series of wipers that remove excess polymer. After this, the fibre-resin bundles pass through a collimator before entering a heated die which has the shape of the final component. As the material passes through the die it is formed to shape while the resin is cured. Heated dies are often about 1 m long, and it is essential that, as the material travels through the die, there is sufficient time to fully cure the resin. When the pull-through speed is too fast the composite exits the die in a partially cured condition and when the speed is too slow then the production rate of the process is too slow. The pulling speed for epoxy-based composites is in the range of 10 to 200 cm min−1. The cured composite leaves the die and is cut by a flying saw to a fixed length. Unlike most other manufacturing processes, pultrusion is a continuous process with the material being pulled through the die by a set of mechanically or hydraulically driven grippers.

14.26 (a) Schematic of the pultrusion process and (b) I-beam demonstrator for aircraft produced by pultrusion photograph courtesy of the Cooperative Research Centre for Advanced Composite Structures.

There are few examples of pultruded composite components used in aircraft or helicopters. In part, this is because the process produces components having a constant cross-sectional shape with no bends or tapers. Few aerospace components are flat with a constant cross-section. Another problem is that the process is designed for large production runs whereas the production runs for aircraft are usually measured in the hundreds spread over several years. Nevertheless, the pultrusion process has potential for the production of high-strength floor beams and other selected parts for aircraft.

14.7 Machining of composites

The majority of processes used to manufacture composites produce components to the near-net shape. This is one of the advantages of manufacturing with composites rather than metals, which often require extensive milling and machining to remove large amounts of material to produce the final component. Most machining operations for composites simply involve trimming to remove excess material from the edges and hole drilling for fasteners. Trimming can be performed using high-speed saws and routers, although care is required to avoid edge splitting (delamination damage). The preferred method of trimming carbon–epoxy composites is water jet cutting because of high cut accuracy with little edge damage. Water jet cutting is a process involving the use of a high-pressure stream of water containing hard, tiny particles that cut through the material by erosion. Most water jet systems used by the aerospace industry are high-pressure units that use garnet or aluminium oxide particles as the abrasive.

Hole drilling of composites requires the use of specialist drill bits, several of which are shown in Figure 14.27. The aerospace industry often uses flat two-flute and four-flute dagger drills for carbon–epoxy. Drilling must be performed using a sharp bit at the correct force and feed-rate otherwise the material surrounding the hole is damaged. The application of excessive force causes push-down damage involving delamination cracking ahead of the drill bit (Fig. 14.28). Drilling at a high feed rate can also generate high friction temperatures at the hole, thereby overheating the polymer matrix. Aramid fibre composites are particularly difficult to drill without the correct bit because the fibres have a tendency of fuzz and fray. The aramid drill contains a ‘C’ type cutting edge that grips the fibres on the outside of the hole and keeps them in tension during the cutting process, thereby avoiding fraying.

14.27 Drill bits used for carbon–epoxy (two and four flutes daggers) and aramid fibre composites reproduced from F. C. Campbell, Manufacturing Technology for Aerospace Structural Materials, Elsevier Science & Technology, 2006.

14.28 (a) Peel-up and (b) push down damage to composite materials caused by incorrect drilling.

14.8 Summary

The three types of fibre reinforcement most often used in aircraft composite materials are carbon, glass and aramid. Carbon fibre is used in primary and secondary structures for aircraft, helicopters and spacecraft. Glass fibre is used in the radomes and components when stiffness is not a critical design property, and are used extensively inside cabins for fittings and furnishing. Aramid fibre is often used in structures requiring high vibration damping or impact resistance.

The production of aircraft components using composites instead of metals often results in significantly fewer parts and fasteners because the manufacturing processes for composites have better capability for making integrated parts.

The mechanical properties of carbon fibres are determined by the processing temperature. High stiffness fibres are produced at the highest treatment temperature (2500–3000 °C) whereas high-strength fibres are treated at about 1500 °C. The properties of glass fibres are determined mainly by the addition of metal oxides to the base silica material.

Carbon–epoxy prepreg is commonly used in primary and secondary aircraft structures. Various processing methods are used to produce composite components using prepreg, including manual hand lay-up, automated tape lay-up (for flat and slightly curved sections) and automated tow placement (for highly curved sections). The prepreg is usually consolidated and cured inside an autoclave.

Composites fabricated using prepreg have high stiffness, strength and other structural properties, although they are also expensive owing to the high cost of prepreg material, the need to store prepreg in a freezer, and the need to cure using an autoclave. Aerospace composites are increasingly being fabricated from dry fabric using out-of-autoclave processes. Various types of carbon fabrics are used, including twill and satin weaves as well as non-crimp fabrics. Fabrics are infused with resin using using several processes, including vacuum-assisted resin transfer moulding, vacuum-based resin infusion, and resin film infusion.

Composites are susceptible to delamination cracking and other damage during trimming, drilling and other material removal processes. Specialist trimming methods, such as water jet cutting, and drilling bits are required for composites.

14.9 Terminology

Carbonisation: Process used for the conversion of an organic substance into carbon or a carbon-containing residue through pyrolysis or other method.

Fibril: Very fine filament which forms part of a larger fibre.

Liquid crystal domains: Randomly oriented domains of highly crystalline polymer chains within a liquid polymer melt.

Polyacrylonitrile (PAN): Organic resin used as the chemical precursor for high-quality carbon fibre.

Pitch: Viscous carbon-rich resin usually formed from petroleum that is used in the production of non-aerospace grade carbon fibre.

Rayon: Organic substance manufactured from regenerated cellulose fibre (extracted from plants or produced artifically) that is used in the production of non-aerospace grade carbon fibre.

Size: Thin coating on fibres containing chemical agents which are used for various functions, including increased adhesion with polymer matrix, antistatic properties, binding and lubrication.

Tow: Untwisted bundle of continuous filaments. Tows are designated by the number of fibres they contain, e.g., 12 K tow contains about 12 000 fibres.

Turbostratic graphite: Basal sheets of graphite that are haphazardly folded or crumpled together.

Yarn: A continuous strand of twisted fibres used in weaving or knitting.

14.10 Further reading and research

Baker A., Dutton S., Kelly D., eds. Composite materials for aircraft structures. Reston, VA: American Institute of Aeronautics and Astronautics, Inc, 2004.

Campbell, F.C. Manufacturing processes for advanced composites. Amsterdam: Elsevier; 2004.

Harris, C.E., Starnes, J.H., Shuart, M.J. Design and manufacturing of aerospace composite structures state-of-the-art assessment. Journal of Aircraft. 2002; 39:545–560.

Sheikh-Admad, J.Y. Machining of polymer composites. Springer; 2009.

Strong, A.B., Fundamentals of composite manufacturing. Society of Manufacturing Engineers. 2008.

14.11 Case study: carbon nanotubes in composites

Carbon nanotubes were discovered in the early 1990s and, since then, have caused great excitement in the scientific community and attracted the interest of the aerospace industry because of their extraordinary properties. Strictly speaking, any tubes with nanometer dimensions are called ‘nanotubes’, but the term is generally used to refer to carbon nanotubes. Graphene is a tessellation of hexagonal rings of carbon that look like a sheet of chicken wire and a carbon nanotube is basically a graphene sheet rolled up to make a seamless hollow cylinder. Nanotubes can have a hemispherical cap of graphene at each end of the cylinder. Tubes typically have an internal diameter of 5 nm and external diameter of 10 nm. The tubes occur as a single cylinder, known as a single-wall nanotube, or two or more overlaid cylinders, called a multiwall nanotube, as shown in Fig. 14.29.

14.29 Images of (a) single-wall and (b) multiwall carbon nanotubes.

The mechanical properties of carbon nanotubes are extraordinary compared with conventional engineering materials. The Young’s modulus of single-walled nanotubes is 1000 to 1300 GPa, which is much higher than conventional carbon fibres (200–400 GPa) and many times greater than metallic alloys used in aircraft structures. The tensile strength of single-walled nanotubes is close to 200 GPa, which again is much higher than carbon fibres. Carbon nanotubes can be mixed with polymers to create nanocomposites which have good mechanical properties and improved flammability resistance. There is strong interest in the possible use of carbon nanotubes in nano-mechanical and nano-electronic devices, which may be used in future aircraft. It is also possible that carbon nanotubes may be used to strengthen and toughen polymer composites used in aircraft structures and gas turbine engines. However, the future of polymer nanocomposites is uncertain because several technical challenges exist in making high-performance composites containing carbon nanotubes, and currently their future in aircraft structures is uncertain.