Fibre–polymer composites for aerospace structures and engines
Composites are an important group of materials made from a mixture of metals, ceramics and/or polymers to give a combination of properties that cannot otherwise be achieved by each on their own. Composites are usually made by combining a stiff, strong but brittle material with a ductile material to create a two-phase material characterised by high stiffness, strength and ductility. Composites consist of a reinforcement phase and a matrix phase and, usually, the reinforcement is the stiffer, stronger material which is embedded in the more ductile matrix material.
The main purpose of combining materials to create a composite is to gain a synergistic effect from the properties of both the reinforcement and matrix. For example, carbon–epoxy composite used in aircraft structures is a mixture of carbon fibres (the reinforcement phase) embedded in epoxy resin (the matrix phase). The carbon fibre reinforcement provides the composite with high stiffness and strength while the epoxy matrix gives ductility. Used on their own, carbon fibres and epoxy are unsuitable as aircraft structural materials, the fibres being too brittle and the epoxy too weak, but when combined as a composite they create a high-performance material with many excellent properties.
This chapter examines fibre-reinforced polymer matrix composites used in aircraft. In chapter 14, the manufacture of composite materials, including the production of the fibre reinforcement and core materials, was described. This chapter examines the applications and properties of the manufactured composites. The application of fibre–polymer composites in aircraft and jet engines is described, including the benefits and problems with using these materials. The mechanical properties of composites are discussed, including the mechanics-based theories used to calculate the elastic and strength properties. Control of the mechanical properties of composites by the type, volume fraction and orientation of the fibre reinforcement as well as by the selection of the polymer matrix is explained. Other properties of polymer matrix composites are dealt with in following chapters, including their fracture properties (chapters 18 and 19), fatigue properties (chapter 20), creep properties (chapter 22) and methods for their recycling and disposal (chapter 24).
Composites are usually classified according to the material used for the matrix: metal matrix, ceramic matrix or polymer matrix. There is also a special type of composite called fibre–metal laminate that does not really fit into any of these three classes. The composites used in aircraft are almost exclusively polymer matrix materials, with the polymer often being a thermoset resin (e.g. epoxy, bismaleimide) and occasionally a high-performance thermoplastic (e.g. PEEK). Metal matrix composites, ceramic matrix composites and fibre–metal laminates are used in much smaller amounts, as described in chapter 16.
The matrix phase of composite material has several important functions, including binding the discrete reinforcement particles into a solid material; transmitting force applied to the composite to the stiff, strong fibres, which carry most of the stress; and protecting the fibres from environmental effects such as moisture and abrasion.
Composites can also be classified based on the shape and length of the reinforcing phase. The three main categories are particle-reinforced, whisker-reinforced and fibre-reinforced composites (Fig. 15.1). The reinforcement, regardless of its shape and size, is always dispersed within the continuous-matrix phase. The main functions of the reinforcement are to provide high stiffness, strength, fatigue resistance, creep performance and other mechanical properties. In some instances, the reinforcement may be used to alter the electrical conductivity, thermal conductivity or other nonmechanical properties of the composite. The stiffening and strengthening attained from the reinforcement is dependent on its shape, and increases in the order: particle, whisker and fibre. For this reason, the composite materials used in aircraft structures contain long, continuous fibres. Composites containing particles or whiskers are rarely used in structural components.
The size of the reinforcement used in composites can range from ultrafine particles in the nanometre size range to large particles up to 1 mm or more. The reinforcement used in aircraft composite materials is mostly in the micrometre size range, typically 5–15 μm, in diameter or slightly larger, as shown in Fig. 15.2. The composites used in aircraft structures are reinforced with continuous fibres rather than whiskers or particles. The mechanical properties in the fibre direction increase with the fibre length, as shown in Fig. 15.3. Therefore using continuous fibres that can extend the entire length of the structure maximises the structural efficiency.
There is growing interest in nanoparticle composites, such as polymers reinforced with carbon nanotubes, for use in aircraft structures due to their exceptional mechanical properties (see chapter 14). In some cases, hybrid composites consisting of two or more types of reinforcing fibres (e.g. carbon and glass or carbon and aramid) are used together to optimise the material properties. For example, aramid fibres used in combination with carbon fibres increase the ballistic resistance of carbon-fibre composite materials. Hybrid composites containing both glass and carbon fibres are used in helicopter rotor blades.
It is worthwhile at this point to distinguish between composites and metal alloys, which are also created by combining two or more materials. The constituent materials in a composite, which are the reinforcement and matrix, are combined but there is no dissolution of one phase into the other phase. In other words, the reinforcement and matrix remain physically discrete phases when combined into the composite. In contrast, in metal alloys the solute material (i.e. alloying elements) dissolves into the solvent (i.e. base metal) and, therefore, the solute does not retain its original physical condition. Only materials which retain the physical properties of the constituents are considered composites.
Fibre–polymer composites alongside aluminium alloys are the most used materials in aircraft structures. The use of composites in civil aircraft, military fighters and helicopters has increased rapidly since the 1990s, and it is now competing head-to-head with aluminium as the material of choice in many airframe structures (as described in chapter 2). The use of composites in gas turbine engines for both civil and military aircraft is also growing. The main reasons for using composites are to reduce weight, increase specific stiffness and strength, extend fatigue life, and minimise problems with corrosion.
Composites used in aircraft structures are in the form of laminates or sandwich materials. Laminates consist of continuous fibres in a polymer matrix, and the most common type used in aircraft is carbon fibre–epoxy. The fibres are stacked in layers, called plies or laminae, often in different orientations to support multidirectional loads. Laminates made of glass fibre–epoxy, carbon fibre–bismaleimide and carbon fibre–thermoplastic may also be used in aircraft, although in smaller amounts than carbon–epoxy. Laminates are used in the most heavily-loaded composite structures and are often supported by ribs, spars and frames. The thickness of the laminate can range from a few millimetres to around 25 mm, depending on the design load.
Sandwich composites are used in lightweight secondary structures requiring high buckling resistance and flexural rigidity, and are often constructed with thin carbon–epoxy face skins covering a lightweight core of polymer foam, Nomex or (in older aircraft types) aluminium honeycomb.
The use of composites in aircraft has been led by the military, particularly with fighter aircraft and helicopters. Carbon fibre–epoxy composites have been used in the primary structures of fighter aircraft for many years, including the wings and fuselage, to minimise weight and maximise structural efficiency. The amount of composite varies between different types of fighter aircraft, and the usage has increased greatly since the 1970s (see Fig. 2.12). Composite material used in fighter aircraft ranges from a small amount, as in the F-14 Tomcat (4% of the structural mass) and F-16 Fighting Falcon (5%), to greater amounts as in the F/A-18 E/F SuperHornet (20%), AV8B Harrier II (22%), Dassualt Rafale B/C (25%) and F-22 Raptor (25%). Composites are the most used structural material in late-generation fighters such as Eurofighter (40%) and F-35 Lightning II (35%). The structural applications of composites, mostly carbon fibre–epoxy and carbon fibre–bismaleimide, in the F-35 are shown in Fig. 15.4.
Composites have been used in single- and twin-aisle passenger aircraft since the late-1960s, with the first applications being in non-safety critical components such as fairings and undercarriage doors. The use of composites in the primary structures of passenger airliners has lagged behind fighter aircraft. The first primary composite structure was the carbon fibre–epoxy horizontal stabilizer for the Boeing 737, which was certified in 1982. The uses of composites in airframe structures developed gradually during the 1990s, and typical applications are shown in Fig. 15.5.
A major expansion in the use of composites occurred with the Airbus 380, which is a 280 tonne airliner (typical operating empty weight) built from 25% composite materials. Carbon-fibre composites are used in many major structural components including the vertical and horizontal tail planes, tail cone, centre wing box, and wing ribs (see Fig. 15.6). The aircraft is also constructed of fibre–metal laminate, which is another type of composite material and is discussed in the next chapter. The Boeing 787 represents the latest in composite applications to large airliners, with 50% of the structural mass consisting of carbon fibre–epoxy composite material. The Airbus A350 is constructed using a similar amount of composite material (about 50% of the structural mass). Table 15.1 gives a summary of the structural components in Airbus and Boeing aircraft made using composite material. Although the use of composites is now widespread in commercial aircraft, it is unusual for the entire airframe to be built with these materials. Apart from a limited number of Beech Starship aircraft, no all-composite commercial airliner has yet gone into production. Aluminium alloy and other metals remain important structural materials to be used in combination with composites for the foreseeable future.
Composites are often used in the fuselage and rotor blades of helicopters. Carbon, glass and aramid fibre composites are regularly used in the main body and tail boom of many commercial and military helicopters to reduce weight, vibration and corrosion as well as to increase structural performance. Composites are being used increasingly to replace aluminium in the main rotor blades to prolong the operating life by improving resistance against fatigue. Most metal blades must be replaced after between 2000 and 5000 h of service to ensure fatigue-induced failure does not occur, and the operating life can be extended to 20 000 h or more with a composite blade. Figure 15.7 presents a cross-section view of a composite rotor blade, which is a sandwich construction containing both carbon and glass fibres in the face skins.
Composites are used in gas turbine engine components including the fan blades, front fan case, nacelle, outlet guide vanes, bypass ducts, nose cone spinner and cowling (Fig. 15.8). The replacement of incumbent metal components with composites provides a saving in the vicinity of 20–30% of the total engine lifetime operation cost. The use of composites is restricted to engine parts required to operate at temperatures below about 150 °C to avoid softening and heat distortion. Carbon-fibre laminates with a high-temperature polymer matrix (such as bismaleimide) is often the composite material of choice for jet engines.
The three major reasons for the use of composites in engines are lower weight, improved structural performance and reduced operating/maintenance cost. The trend in gas turbine engines is towards larger (and consequently heavier) fan sections. The addition of 1 kg to the fan blade assembly needs a compensatory increase in the weight of the fan blade containment case of 1 kg. This 2 kg increase requires a compensatory weight increase of about 0.5 kg to the rotor as well as incremental increases to the weights of the wings and fuselage. Any weight savings gained by using light materials in the fan blade translates to significant savings in the weight of other components. For example, the use of carbon-fibre composite in the two GEnx engines on the Boeing 787 provides an overall weight saving of about 350 kg, which translates directly into fuel burn saving and greater aircraft range. The other benefit of using lightweight composite fan blades is reduced centrifugal force (owing to their light mass) thus increasing the fatigue life.
Polymer matrix composites have been used for many years in space structures, including truss elements, antennas, and parabolic reflectors. The main truss of the Hubble space telescope, for example, is made of carbon fibre–epoxy composite for lightness, high stiffness and low coefficient of thermal expansion. As another example, the main cargo doors of the space shuttle orbiter are made of sandwich composite material and the arm of the remote manipulator system is made of carbon-fibre composite (Fig. 15.9).
There are many benefits to be gained from using carbon–epoxy and other types of fibre–polymer composite materials instead of aluminium alloy in aircraft. The main advantages are summarised here and a comparison between carbon–epoxy composite and an aircraft-grade aluminium alloy is given in Table 15.2.
The lower density and superior mechanical properties of carbon-fibre composite compared with aluminium results in significant weight saving. With careful design, it is possible to a achieve weight savings of 10–20% using carbon fibre–epoxy composite. For instance, the Boeing 787 is about 20% lighter than an equivalent aircraft made entirely of aluminium which, together with other factors, translates to a fuel saving of about 20% and reduced seat-mile costs by around 10%.
Composite manufacturing processes are better suited than metal forming processes to produce one-piece integrated structures with fewer parts. Typically, assembly accounts for about 50% of the production cost of the airframe, and the possibility of producing integrated structures offers the opportunity to reduce the labour, part count, and number of fasteners. For example, one mid-sized aircraft fuselage barrel made using aluminium requires about 1500 sheets held together with nearly 50 000 fasteners. An equivalent barrel made of composite material only requires about 3000 fasteners because of the fewer number of parts. The F-35 Lightening II has just one-half the number of parts of the F-18 SuperHornet, one of the aircraft it may replace.
The mechanical properties of composites can be tailored by aligning the fibre reinforcement in the load direction, thereby providing high stiffness and strength where it is needed. As a result, the specific stiffness and strength properties of carbon-fibre composites are superior to aluminium alloy. Furthermore, the choice of high-stiffness or high-strength carbon fibres provides the aircraft engineer with greater flexibility in the design of structurally efficient components.
Composites have high fatigue resistance under cyclic stress loading, thereby reducing the maintenance cost and extending the operating life. The superior fatigue resistance of carbon-fibre composite compared with aluminium is a major reason for its use in aircraft structures, helicopter rotor blades, and fan blades for gas turbine engines.
Carbon-fibre composites are immune to corrosion, which is a major problem with aluminium. The high cost of inspecting and repairing metal structures damaged by corrosion is minimised with the use of composites. The corrosion resistance of composite material allows for a higher humidity level inside the cabin, which creates a more comfortable environment for crew and passengers. The humidity for an aluminium fuselage must be kept at a very low level (around 5%) to avoid corrosion, and this can be increased in a composite fuselage (15–20%) for greater comfort.
Using specialist designs, materials and manufacturing processes, composites can be made with high radar absorption properties. When radar absorbing composite material is used in stealth military aircraft they are difficult to detect using radar.
The thermal conductivity of composites is much lower than metals, which makes them a good heat insulator. This property is used in fighter aircraft to reduce heat conduction from the engines through the fuselage, thereby making the aircraft harder to detect using infrared devices.
Carbon-fibre composites have a very low coefficient of thermal expansion, which means they experience little or no expansion or contraction when heated or cooled. This property is utilised in structures that need high dimensional stability with changes in temperature. For example, the truss supporting the antenna and deep-space observation devices in the Hubble space telescope is made of composite material. When the temperature of the telescope changes the truss does not change shape, thereby providing a stable platform for the antenna and devices.
The cost of producing aircraft components from composite material is often more expensive than using aluminium. The higher cost of composite is caused by several factors, including the high cost of carbon fibres, the labour-intensive nature of many of the manufacturing processes, and high tooling costs.
The production of aircraft components using composites can be slower than with aluminium owing to the long time needed to lay-up the fabric or prepreg ply layers and the curing time of the polymer matrix. The use of automated lay-up processes and multilayer non-crimp fabrics reduce production time, although most manufacturing processes are not well suited to the rapid production of a large number of thermoset matrix parts.
Designing aircraft components made of composite materials is more challenging than for metal alloys owing to their anisotropic properties. Careful design and manufacture is required to ensure the fibres are aligned in the load direction, otherwise the composite may have inferior mechanical performance.
The stiffness, strength, damage tolerance and other mechanical properties of composites are low in the through-thickness direction owing to the absence of reinforcing fibres. Figure 15.10 shows typical values for the in-plane and through-thickness mechanical properties of carbon fibre–epoxy composite. The through-thickness property values are just a small fraction of the inplane values, and the application of through-thickness loads on to composites must be avoided.
Composites are susceptible to delamination cracking when impacted at low energies because of their low through-thickness strength and fracture toughness. Impact events such as bird strike, large hail stone impacts, and tools which are accidentally dropped during maintenance can damage composites more severely than occurs with a metal alloy. A related problem is impact damage, which can reduce the compressive strength and other mechanical properties of the composite.
The growth of damage (e.g. delamination cracks) in composite materials is difficult to control and predict. A large amount of damage growth can occur rapidly with little or no warning. For this reason, primary composite aircraft structures must be designed according to the so-called ‘ no growth’ damage tolerance philosophy, which means that pre-existing damage must not grow over a specified period of time of aircraft service (usually two or more inspection intervals). As a result, composite structures must be over-designed to ensure adequate damage tolerance, thus increasing their weight and cost.
The reduction in the failure strength of composites owing to notches (e.g. bolt holes, windows) can be greater than the loss suffered by metal alloys. Composite materials with highly anisotropic properties (i.e. a large percentage of fibres aligned in one direction) experience a high concentration of stress near notches when under load, resulting in a large knock-down in strength.
Composites soften and distort at lower temperatures than aluminium owing to the glassy-to-rubbery transformation of the polymer matrix. The maximum operating temperature of epoxy matrix composites is typically in the range 100–150 °C, which limits their use for high-temperature applications.
Composite materials are much poorer conductors of electricity than the metals used on aircraft. An electrically conductive material (such as copper mesh) must be incorporated into composite materials used on the external surface of aircraft to dissipate electricity in the event of lightning strike.
Understanding the mechanical behaviour of continuous-fibre–polymer composites is different to understanding the mechanical properties of metals and their alloys, which has been a major focus of this book. The properties of metals are controlled by vacancies, dislocations, grain boundaries, precipitates, solute (alloying and impurity) elements, and other microstructural features. These features have no role to play in the mechanical properties of composite materials. Instead, understanding the mechanical properties is based on the key concept of load sharing between the fibre reinforcement (which is stiff and strong) and the polymer matrix (which is compliant and weak). The properties of composites are determined by the fibre properties, matrix properties, interfacial properties between the fibre and matrix, and how effectively the load is shared between the fibres and matrix.
When an external load is applied to a composite, a certain proportion of that load is carried by the fibre reinforcement and the remainder by the polymer matrix. At the simplest level, the applied load is shared between the fibres and matrix based on their relative volume fractions. This is expressed mathematically as:
where P is the total applied load, PfVf is the proportion of load on the fibres, and PmVm is the proportion of the load on the matrix. Pf and Pm are the loads carried by the fibres and matrix, and Vf and Vm are the volume fractions of the fibres and matrix, respectively.
Provided the response of the composite to the applied load is elastic, then the distribution of load-sharing between the fibres and matrix is independent of the stress level. In other words, the relative distribution of load between the fibres and matrix remains the same for any stress up to the elastic stress limit of the composite, which quite often is the ultimate failure stress. Because the fibre reinforcement is much stiffer and stronger than the polymer matrix, it is important that as high a proportion of the applied load as possible is borne by the fibres. For this reason, the fibre content of aircraft composites is as high as practically possible. The fibre volume content is typically in the range of 55–65%, the maximum values that can be achieved using the manufacturing processes described in chapter 14. The relative proportions of the load borne by the fibres and the matrix is dependent not only on their relative volume fractions, but also on the shape and orientation of the fibres together with the elastic properties of the fibres and matrix.
The concept of load-sharing between the fibres and matrix is the basis for the theoretical understanding of the mechanical properties of composites as well as the practical aspects in the design and manufacture of composites with maximum mechanical performance. The concept of load-sharing is applied to the micromechanics of composites, which is the mathematical analysis of the mechanical properties based on the properties and interactions of the fibres and matrix. Analytical and finite element models are used to simulate the microstructure of composites, and from this the mechanical properties (such as elastic modulus or strength) are calculated in terms of the properties and volume fractions of the fibres and matrix. Figure 15.11 illustrates the basic premise of micromechanics, where the properties of the fibre reinforcement are combined with the properties of the polymer matrix to calculate the ‘average’ properties of the composite material based on volume-averagis of the two constituents.
The averaging approach to micromechanical analysis assumes that the fibres are arranged in a regular pattern. In actual composite materials, the fibres are randomly distributed throughout the matrix phase, with regions of higher than average volume content of fibres and other regions with lower than average fibre content. Micromechanical analysis assumes that the fibres are packed in a regular pattern. For example, Fig. 15.12 shows two regular fibre packing patterns: the square array and the hexagonal array. Either array can be viewed as a repetition of a single unit cell of the composite which contains the fibre and matrix. The unit cell represents the basic building block of the composite material, and is the simplest geometry for analysing using the volume averaging approach to micromechanics.
The simplest approach to micromechanical analysis of composite materials involves treating the fibre and resin separately within a unit cell, and assuming that there is no interaction between the two constituents. The property averaging concept is called ‘rule-of-mixtures’, and it is applied to derive simple-to-use equations for mechanical and other properties including:
Once the properties of the unit cell have been calculated using micromechanics, it is then possible to calculate the properties of composite structures using a hierarchical approach to modelling as shown in Fig. 15.13. Micromechanical modelling of the unit cell is the foundation for higher levels of modelling which follow the sequence of single-ply modelling, laminate (or ply-by-ply) modelling, and finally structural modelling. Single-ply modelling calculates the properties of one ply layer within a composite, in which the fibres and matrix are treated separately (as depicted in Fig. 15.11). All of the fibres within the single ply are assumed to be aligned in the same direction (i.e. unidirectional fibre laminate). Taking the properties calculated for a single ply, the next level of modelling involves analysing multiple-ply layers with different fibre angles to calculate the properties of a multidirectional composite. Laminate modelling treats each ply layer separately and the fibres and matrix within each ply are treated as a continuum. Structural (also called macroscopic) modelling analyses the composite as a single orthotropic material with the geometric features of the final component. Structural modelling is used to predict the stiffness, strength and other properties of the final composite structure, and this is performed using finite element and other numerical methods.
One of the main reasons for using continuous-fibre composites in aircraft structures is their high specific stiffness compared with many metal alloys. The elastic modulus properties of carbon-fibre composites are superior to aluminium, which is an important reason for the increasing use of composites and the corresponding decline in the application of aluminium in aircraft structures.
The elastic modulus of a unidirectional composite reinforced with straight, continuous fibres when loaded in the fibre direction can be calculated using rule-of-mixtures modelling. The unidirectional composite shown in Fig. 15.10 is loaded in tension along the fibre direction (which is also called the 1– or longitudinal direction). In this load state, the fibre and matrix are assumed to act in parallel and both constituents experience the same elastic strain (ε1). This iso-strain condition is true even though the elastic moduli of the fibres and matrix are different. The iso-strain condition is expressed as:
where εf and εm are the strain values of the fibre and matrix; σf and σm are the stresses carried by the fibre and matrix, and Ef and Em are the Young’s modulus of the fibres and matrix, respectively.
Because the elastic modulus of the matrix is much lower than the fibre reinforcement, the fibres contribute between 95 and 99% of the in-plane stiffness of a unidirectional composite containing carbon at the typical volume contents used in aerospace materials (Vf ~ 0.55–0.65). The type of polymer matrix (e.g. epoxy, bismaleimide, PEEK) does not have much effect, and the matrix is chosen for reasons other than longitudinal stiffness, such as cost, maximum operating temperature or durability.
Rule-of-mixtures analysis is remarkably accurate in the calculation of the longitudinal Young’s modulus of unidirectional composites. For instance, Fig. 15.14 compares the measured and calculated Young’s modulus values for a unidirectional composite over a range of fibre contents, and the excellent agreement demonstrates that the in-plane stiffness can be accurately predicted using this simple analysis. Rule-of-mixtures modelling can be used to determine the longitudinal modulus for various types of composite materials to identify which provides the highest stiffness, as shown in Fig. 15.15. Rule-of-mixtures is only accurate over the fibre volume content between about 0.2 and 0.7, which is within the range used in aerospace composites.
Composite materials must be designed to ensure that the external load is applied parallel to the fibres. The load should never be applied in the antifibre direction because of the low transverse stiffness and strength of the composite. The simplest model to calculate the transverse modulus assumes that the fibres and matrix act in series under an external load, and is expressed as:
Figure 15.16 shows the effects of fibre type and fibre volume content on the transverse Young’s modulus of different types of unidirectional composites. The fibre reinforcement has a much smaller effect on the transverse Young’s modulus compared with the longitudinal modulus, with virtually identical stiffness properties for composites containing fibres with vastly different modulus values.
The transverse modulus is not always accurately calculated using equation [15.4] because the strain is not uniform when a composite is loaded in the antifibre direction. Owing to the large difference between the elastic modulus of the fibres and matrix, the strain is distributed unevenly in the matrix under a transverse load. Figure 15.17 shows the heterogeneous strain field in a composite loaded in transverse tension, where the darkest fringe lines (located above and below the fibres) indicate regions where the matrix is highly strained. The heterogeneous strain affects the accuracy of the micromechanical analysis given in equation [15.4], with the model underpredicting the transverse modulus.
15.17 Photoelastic image of a composite-type material loaded normal to the fibre direction showing the heterogeneous strain distribution indicated by the fringe lines (from A. Puck, Zur Beanspruchen und Verformung von GFK-Mehrschichtenverbund-Bauelmenten, Kunststoffe, 57 (1967), 965–673.
Other models have been developed to account for the heterogeneous strain distribution through the polymer matrix. One such model is the Halpin–Tsai model, which states that the transverse modulus is calculated using:
Other elastic properties as well as the thermal properties of composites can be calculated using rule-of-mixtures analysis. Table 15.3 gives equations to predict the shear modulus, Poisson’s ratio, thermal conductivity, and specific heat capacity of unidirectional composites.
The Young’s modulus of a unidirectional composite is much higher when loaded in the longitudinal direction than in the transverse direction owing to the high stiffness provided by the fibres. For example, the longitudinal modulus for a high modulus carbon fibre–epoxy composite is more than 200 times higher than the transverse modulus. The longitudinal (or 0°) and transverse (90°) directions are the two extremes for loading of a unidirectional composite: parallel and perpendicular to the fibre direction. Between these two extremes the Young’s modulus changes continuously with the fibre angle. For instance, Fig. 15.18 shows the effect of fibre angle on the elastic modulus properties of a unidirectional carbon–epoxy composite. The Young’s modulus decreases with increasing fibre angle, particularly from 0° to 30° when the stiffness falls sharply by nearly 80%. The high sensitivity of the Young’s modulus to fibre angle occurs for any type of unidirectional composite material, and their fibres must be closely aligned to the load direction to achieve high stiffness.
The other elastic properties of unidirectional composites are also dependent on the fibre angle. Figure 15.17 also shows the variation in the shear modulus of carbon/epoxy with fibre angle, and this property is highest at 45° and lowest when the load is applied in the fibre (0°) and anti-fibre directions (90°). The change in the shear modulus with fibre angle can be determined using:
The highly orthotropic nature of the elastic properties of unidirectional composites means they should not be used in aircraft and other engineering structural applications. The loads applied to aircraft structures are rarely unidirectional. The majority of aircraft structures are subject to multidirectional loads which cannot be effectively carried using a composite material in which all the fibres are aligned in the same direction.
The composites used in aircraft structures are designed with the fibres aligned in two or more directions to support multidirectional loads. The two most common fibre patterns are cross-ply [0/90] where 50% of the fibres are aligned in the 0° direction and 50% in the 90° direction and quasi-isotropic [0/± 45/90] where 25% of the fibres are aligned in each of the 0°, + 45°, -45° and 90° directions. The quasi-isotropic fibre pattern is shown in Fig.14.18. The 0° and 90° fibres in the quasi-isotropic pattern is used to carry in-plane tension, compression and bending loads whereas the + 45° and -45° fibres carry in-plane shear loads. The cross-ply pattern is used when shear loads are absent. The effect of load angle on the Young’s modulus of quasi-isotropic and cross-ply composites is shown in Fig. 15.19, and the variation in stiffness with angle is much less extreme than the unidirectional condition. The Young’s modulus of a quasi-isotropic composite is relatively insensitive to the angle, and its in-plane elastic behaviour is approaching that of a fully isotropic material. The elastic properties of composite laminates with multidirectional fibre patterns can be calculated using laminate analysis.
Other fibre patterns are used occasionally in composite aircraft structures, although the most common is the quasi-isotropic pattern followed by the cross-ply pattern. An important consideration in the fibre arrangement of multidirectional composites is that the fibre pattern is symmetric about the mid-thickness point; that is, the fibre pattern in the upper half of the composite is a mirror image of the fibre pattern in the lower half. As mentioned in the previous chapter, symmetric fibre arrangement is essential to avoid warping and distortion of the composite component.
The longitudinal tensile strength of composite materials is determined mostly by the strength and volume content of the fibre reinforcement. The breaking strength of the fibres is much greater than the strength of the polymer matrix, and therefore the fibres determine the ultimate strength of the composite. Table 15.4 lists the tensile strength and failure strain values for several types of fibres and polymers used in aircraft composites. The fibre strengths are typically 50 to 100 times higher than the matrix and, consequently, the strength of the matrix has little influence on the in-plane tensile strength of composite materials.
The fibres used in aircraft composites have high stiffness and strength, but they have little or no ductility and as a result fail at low strain. Figure 15.20 shows tensile stress–strain curves for various types of fibres used in aircraft composite materials. Carbon fibres are brittle materials that sustain no plastic deformation before breaking at low strain (less than 0.5–2.0%). Similar behaviour occurs with glass fibres whereas aramid fibres experience a small amount of plastic flow before final failure. The brittle nature and low failure strain of carbon fibre has important consequences on the tensile behaviour of carbon-fibre composites used in aircraft structures and engines.
Brittle materials such as carbon and glass do not have a well-defined tensile strength, unlike metals or polymers. For example, the aircraft-grade 7075-T76 aluminium alloy has a fixed tensile yield strength close to 470 MPa. Likewise, epoxy resin has a constant strength of about 100 MPa when fully cured. The strength of carbon fibre, however, varies over a very wide range (1400 to 4800 MPa or more) despite being produced, like metals and polymers, under controlled conditions which should ensure a consistent strength value.
The strength of brittle fibres is extremely variable because the failure stress is highly sensitive to the presence of flaws and defects. Tiny cracks and voids are present in fibres, and these have a major influence on fibre strength. For example, a surface crack in carbon fibre as small as 0.3–0.4 μm reduces the breaking strength by more than 50%. Cracks are accidentally created at the fibre surface when the individual filaments are collimated into bundles after being produced. The fibres slide and rub against each other during winding and handling which introduces tiny scratches on the surface. Even when fibres are wound and handled with great care it is difficult to avoid surface damage. Lubricants within the size coating are applied to fibres to minimise sliding friction, but this does not completely eliminate surface damage. The small cracks and other surface damage caused to the fibres are not all the same size, but vary randomly along the fibre length as illustrated is Fig. 15.21. Fracture of the fibre always occurs at the largest flaw, which determines the tensile strength, and the other (smaller) flaws have no influence on fibre strength. Because the maximum flaw size can vary from the fibre to fibre, despite being manufactured and handled under the same conditions, the strength is different from fibre to fibre.
where Kc is the fracture toughness of the fibre material. (The concept of fracture toughness and the fracture properties of composites and other aerospace materials are described in (chapters 18 and 19). The fracture toughness of carbon is very low (about 1 MPa m1/2), which means its strength is very sensitive to crack size. Figure 15.22 shows the dependence of carbon fibre strength on the largest crack size, and the strength decreases rapidly with very small increases in the crack length. Owing to this high sensitivity of fibre strength to crack size, and the random nature of the maximum crack size between fibres, the tensile strength varies over a wide range. Figure 15.23 shows the strength distribution plot for carbon fibre and, unlike metals and polymers that have a single strength value, the fibre strength varies over a wide range.
The tensile strength of unidirectional composite materials such as carbon fibre–epoxy is much higher than many aerospace alloys, including the aluminium alloys used in aircraft structures. For example, Table 15.2 shows that the specific tensile strength of carbon fibre–epoxy is about twice that of 2024 aluminium alloy. Owing to the failure stress of the fibres used in aerospace composites being much higher than the polymer matrix, the longitudinal tensile strength is strongly influenced by the tensile strength and volume fraction of the fibres. For example, the fibres typically account for 95–98% of the tensile strength of a unidirectional carbon–epoxy composite at the volume contents used in aircraft structures (Vf ~ 0.55–0.65). The contribution of the polymer matrix to the strengthening of unidirectional composites is very small.
Unlike the calculation of the elastic properties, micromechanical modelling is usually not accurate in the calculation of the tensile strength of composite materials. The accurate determination of the strength properties requires tensile testing of the material. Modelling based on weighted averaging can be used to approximate the longitudinal strength, but should never be used as an accurate value without being validated by tensile testing.
The simplest micromechanical model for longitudinal tensile strength assumes two cases: (i) the fibre failure strain is lower than the matrix failure strain or (ii) the fibres fail at a higher strain than the matrix. The first case is represented by the stress–strain condition shown in Fig. 15.24, and it applies to carbon fibre–epoxy and most other composite materials used in aircraft.
As mentioned, the failure strength of brittle fibres is determined by the size of the largest flaw, which varies from fibre to fibre. As a result, fibres break over a range of stress values when a composite is loaded in tension. Those fibres with the lowest strength (because they contain the largest flaw) are the first to fail within the material. As the tension stress applied to the composite increases the fibres containing smaller flaws break until all the fibres are broken, and this event defines the ultimate tensile strength of the composite material. For example, Fig. 15.25 shows the effect of increasing tensile stress on the number of broken fibres within a unidirectional carbon-epoxy composite. The fibres begin to break at a threshold stress level, above which the total number of broken fibres rises at an increasing rate until the final failure stress. The design load limit for composite materials used in aircraft structures is well below this threshold limit to ensure that no fibres are broken during routine flight operations.
An important process in the tensile failure of fibres is shear lag, which is illustrated in Fig. 15.26. When a fibre breaks, its capacity to carry the applied tensile stress does not immediately drop to zero. The broken fibre can continue to carry load by transferring the stress across the break by shear lag. This is a process whereby tensile stress is transferred between the two ends of a broken fibre by shear flow of the surrounding polymer matrix. The efficacy of the shear lag process to transfer stress across a broken fibre decreases with increasing tensile strain because the two fibre ends become further apart. Therefore, the tensile load capacity of broken fibres decreases with increasing strain, and this has the result of more stress being exerted on the unbroken fibres. These fibres then become overstressed which causes them to break. As more fibres break the number of remaining intact fibres becomes fewer until eventually the composite fails. Without this shear lag process, the longitudinal tensile strength of composite materials would be lower than it actually is.
The transverse tensile strength of composite materials is much lower than their longitudinal strength owing to the absence of fibres aligned in the transverse direction. Figure 15.27 compares the longitudinal and transverse tensile strengths of a carbon–epoxy composite over a range of fibre contents, and the failure strength in the transverse direction is just a small fraction (typically less than 5–10%) of the longitudinal strength. This is one reason for the poor impact damage resistance of composite structures, with impact events such as bird strike causing significant damage owing to the low transverse strength. (The impact damage resistance of composites is described in chapter 19).
The transverse strength is controlled by the failure strength of the interface between the fibres and matrix. Failure under transverse tensile loading usually occurs by cracking along the fibre–matrix interface, as shown in Fig. 15.28. The use of sizing compounds on the fibre surface to promote strong adhesion bonding with the polymer matrix is one of the common ways to increase the transverse tensile strength. Like the transverse Young’s modulus, the transverse tensile strength is not influenced significantly by the type or properties of the fibres and, therefore, different types of composite materials have similar failure strengths (usually in the range of 30–80 MPa).
Analytical models have been developed to calculate the transverse tensile strength of composite materials. The simplest model treats the fibres as cylindrical holes in the polymer matrix that have no strength, as illustrated in Fig. 15.29. The transverse strength (YT) is then calculated by considering the reduction in the effective load-bearing area owing to the holes created by the fibres, which can be expressed as:
when the fibres are arranged in a square grid pattern. σm is the tensile strength of the polymer used for the matrix.Equation [15.10] gives an approximate estimate of the transverse tensile strength of composite materials, but like the longitudinal strength the most accurate method for determining the strength is by tensile testing.
The tensile strength of a unidirectional composite is dependent on the fibre angle in a similar way to the Young’s modulus. Figure 15.30 shows the effect of fibre angle on the tensile strengths of unidirectional carbon–epoxy and glass–epoxy materials. The strength drops rapidly with increasing angle owing to the loss in load-bearing capacity of the fibres, which are only highly effective when aligned close to the load direction (within 3–5°). This behaviour, along with the highly orthotopic modulus properties, is the reason why unidirectional composites should not be used in aircraft structures.
The effect of fibre angle on the tensile strength of a unidirectional composite can be calculated in several ways, although the easiest method is to treat the failure process using three arbitrary stress states. Failure is assumed to occur in one of three modes:
Failure of a unidirectional composite under off-axis tensile loading can be estimated using the maximum stress criterion. This criterion assumes that the stresses for the three mode conditions are related to the fibre angle as follows:
where XT, S12 and XY represent the failure strengths of the composite in axial tension (ϕ = 0°), in-plane shear (ϕ = 45°) and transverse tension (ϕ = 90°), respectively. These three equations are solved over the range of fibre angles between 0° and 90°, as shown in Fig. 15.31. The curve which has the lowest strength value over the range of fibre angles is then used to define the failure strength. It is shown in Fig. 15.31 that this approach provides a good estimation of the reduction to the tensile strength of a unidirectional composite with increasing fibre angle.
15.31 Calculation of the effect of fibre angle on the tensile strength of a unidirectional carbon–epoxy composite using the maximum stress criterion. The data points are experimental strength values. The full curves indicate the range of fibre angles when the equations for the maximum stress criterion are valid.
The failure strength of composite materials under compression loading is usually different than under tension owing to the different failure behaviour. The most common failure modes of composites under in-plane compression are microbuckling or kinking of the load-bearing fibres. Microbuckling involves the lateral (out-of-plane) buckling of fibres over a small region, as shown schematically in Fig. 15.32. Kinking involves the localised out-of-plane rotation and fracture of the fibres, as shown in Fig. 15.33, and occurs more often in carbon fibre–epoxy composites than microbuckling.
Kinking occurs from a local buckling instability, which develops in the load-bearing fibres. The instability arises from a defect (e.g. void, resin-rich region) or free edge which does not provide sufficient lateral restraint to the fibre under compression loading. The instability allows the fibre to rotate in a very small region towards the defect or free edge, with the amount of fibre rotation increasing with the compressive strain. As the small region of fibres is rotated, the surrounding polymer matrix is plastically deformed by shear stress generated by the rotation process. Eventually, the fibres rotate over a sufficiently large angle that they break over a small length (usually 100–200 μm long), which is called a kink band. Once the first fibres fail by kinking, the other fibres are overloaded and they also then fail by kinking. A kink band propagates rapidly through the composite material causing compressive failure. Although the kink band is short in length, once it has propagated through the load-bearing section of the composite the material can no longer support an applied compression load.
Brittle fibre composites, such as carbon or glass-reinforced materials, fracture along a well-defined kink plane when loaded to the compressive failure stress. The compressive strength of a unidirectional composite that fails by kinking is calculated using:
where τy and γy are the shear yield stress and shear yield strain, respectively, of the polymer matrix, and ϕ is the fibre misalignment angle. This equation shows that the kinking stress is dependent on the fibre angle, and maximum compressive strength is attained when the fibres are aligned with the direction of compression loading. The kinking stress decreases rapidly with increasing fibre angle in a manner similar to the tensile strength (Fig. 15.30). The kinking stress is also dependent on the shear properties of the polymer matrix, and one of the most effective ways of increasing the compressive strength of composites is to have a polymer matrix with high shear resistance. For example, Fig. 15.34 shows the large improvement to the compressive kinking stress that can be attained by increasing the matrix shear strength.
As mentioned, unidirectional composites are not used in aircraft structures owing to their highly orthotropic properties. These materials are only structurally efficient when the load is applied parallel with the fibre, which is difficult to ensure for aircraft structures. Instead, quasi-isotropic and cross-ply composite materials are commonly used, and the effect of load angle on their tensile and compressive strength properties are similar to that shown in Fig. 15.19 for the Young’s modulus. The strength properties are reasonably constant at the different load angles in the quasi-isotropic composite owing to the presence of an equal proportion of 0°, +45°, −45° and 90° fibres. The tensile and compressive strengths of the cross-ply composite are highest in the 0° and 90° directions and lowest at 45°.
The mechanical properties of composite materials can be adversely affected by defects which are inadvertently created using manufacture. Manufacturing defects include voids, dry spots, resin-rich regions, irregular fibre distribution, micro-cracks, shrinkage, and misaligned fibres. Controlled manufacturing conditions and stringent quality control checks are applied to ensure that the composites produced for aircraft structures have a defect content which is below an acceptable level. Composites having high defect content must be repaired (if economically feasible) or scrapped.
Of the many types of defects, the most common are voids which are created by air being trapped within the material during ply lay-up or by gas created as a by-product of the cure reaction of the polymer matrix. The processes used to manufacture aerospace composites, such as autoclave curing, resin transfer moulding and vacuum bag resin infusion, result in low void contents (typically under 1% by volume). In addition, the epoxy and other thermoset resins used as the matrix material are formulated to yield little or no gaseous volatiles during curing and, therefore, void formation is minimised. Despite these measures, voids can still occur, thus reducing the mechanical properties. Voids are more detrimental to the mechanical properties influenced by the matrix, the so-called matrix-dominated properties, than the fibre-dominated properties. Matrix-dominated properties include interlaminar shear strength, impact resistance and delamination fracture toughness. Fibre-dominated properties include the tensile modulus and strength, and these are not affected significantly by voids unless present at a high volume content.
Figure 15.35 shows the effect of volume content of voids in an aircraft-grade carbon–epoxy composite on the interlaminar shear strength (matrix-dominated property) and tensile strength (fibre-dominated property). The interlaminar shear strength decreases rapidly with increasing void content above about 1%, and this behaviour is typical for matrix-dominated properties, whereas the tensile strength is largely unaffected. For this reason, the volume content of voids (and other defect types) in aircraft composite materials must be kept to very low levels (under 0.5–1.0%) to avoid any significant loss in mechanical performance.
Sandwich composite materials consist of thin face skins bonded to a thick low-density core material. By the correct choice of materials for the skins and core, sandwich composites provide higher ratios of stiffness-to-weight and strength-to-weight than many metals and fibre–polymer laminates. In most cases, the main reason sandwich construction is used is to reduce weight, but it also provides other advantages including improved acoustic damping (noise reduction), thermal insulation, and impact absorption.
Sandwich composites are used almost exclusively in secondary aircraft structures such as the control surfaces (e.g. ailerons, flaps, spoilers, slats), engine nacelles, radomes, floor panels, and the vertical tailplane (Fig. 15.36). Rarely, if ever, are sandwich materials used in primary structures of conventional civil or military aircraft. An exception is the sandwich fuselage of the Beech Starship, which is an all composite aircraft (Fig. 15.37). Sandwich composites are also used in the fuselage and wings of sailplanes and other ultralight aircraft.
Sandwich construction is used most often for aircraft structures subjected to bending loads. The separation of the face skins, which are stiff and strong, using the lightweight core offers the possibility of making the resulting stresses act far from the neutral axis of the structure. Moving the skins away from the neutral axis greatly increases the bending stiffness and thereby makes sandwich structures resistant to buckling. The skins carry most of the in-plane tension, compression and bending loads applied to the sandwich structure. The function of the core is to carry shear stresses resulting from transverse loads, to support the face skins, and to maintain a separation distance between the skins. To fulfil these functions the core material must have sufficient shear stiffness. Table 15.5 compares the structural efficiency of a monolithic material (in this case a 0.8 mm sheet of aluminium) against two sandwich composite materials with different thickness. When loaded in bending, the stiffness and strength increases with the thickness of the sandwich material. At the highest thickness, the strength is increased over 9 times and the stiffness by 37 times, but with only a 6% increase in weight.
The face skins are made of a thin sheet of fibre–polymer laminate such as carbon–epoxy, although in earlier sandwich constructions aluminium alloy was used. There are four main groups of low-density core materials: honeycomb cores, corrugated cores, foam cores, and monolithic cores of homogeneous material. The first three types are used in aircraft sandwich composites, whereas the monolithic core materials (of which the most common is balsa wood) are rarely used.
Honeycomb core consists of very thin sheets attached in such a way that connected cells are formed. The cell structure closely resembles the honeycomb found in a beehive. The two most common honeycomb configurations are shown in Fig. 15.38: hexagonal and rectangular. Various types of material are used for the honeycomb, with aluminium alloy (usually 5052-H39, 5056-H39 or 2024-T3) and Nomex being the most common in aircraft sandwich components. Nomex is the tradename for a honeycomb material based on aramid fibre paper in a phenolic resin matrix. Aluminium and Nomex core materials have a lightweight cellular honeycomb structure which provides high shear stiffness.
The aerospace industry is increasingly using polymer foams instead of aluminium honeycomb and Nomex because of their superior durability and resistance to water absorption. The foams used are often closed cell, which means the cell voids are completely surrounded by a thin membrane of the solid polymer. Some types of polymer foams have an open cell structure where the cells are interconnected, although these are used rarely in aircraft in preference to closed-cell foams. Examples of closed cell foams are polyetherimide (PEI) and polymethacrylimide (PMI), and these materials have a low-density cellular structure which has high specific stiffness and strength (Fig. 15.39). New types of foam metals are emerging, such as aluminium foams, although these are not currently used in aircraft sandwich materials.
The mechanical properties of the face skins and core must be balanced to achieve high structural efficiency. The core must have sufficient shear and compressive strength to ensure effective support of the skins. If the core is too weak in shear, the skins bend independently and the bending stiffness of the sandwich material is low. The stiffness and strength properties of core materials increase linearly with their density, as shown in Fig. 15.40, and therefore increased mechanical performance comes with a weight penalty. The core materials used in aircraft structures commonly have density values in the range of 75 to 250 kg m−3, which provides a good balance between mechanical performance and light weight. The stiffness and strength of the skins must be sufficient to provide high tensile and compressive properties to the sandwich composite.
The properties of laminates and sandwich composite materials may be affected by the environmental operating conditions of the aircraft. Composites absorb moisture from the atmosphere and this can degrade the physical, chemical and mechanical properties over time so that measures have to be taken to minimise any deterioration by the environment. The combination of high humidity and warm temperature found in the tropics has the synergistic effect of increasing the rate of moisture absorption and accelerating the deterioration of the material. This effect is called hygrothermal (hot/wet) ageing, and it is the most common environmental condition for degrading the properties of composite materials. Aircraft components may also be affected by various fluids used on aircraft, including fuels, fuel additives and hydraulic fluids, as well as by ultraviolet radiation. The long-term durability of composite materials in the aviation environment is essential to maintain structural integrity and safety whilst also minimising maintenance costs and aircraft downtime for inspection and repair.
The absorption of moisture occurs mostly via the diffusion of water molecules through the polymer matrix. Moisture is also absorbed along the fibre–matrix interfacial region, which in some materials can provide a pathway for the rapid ingress of water. Carbon and glass fibres do not absorb moisture and therefore do not assist in the diffusion process. Organic-based fibres, on the other hand, can absorb significant amounts of moisture; for example, aramid fibres can absorb up to 6% of their own weight in water. Because most of the moisture is absorbed by the polymer matrix, the type of resin used in a composite has a large effect on environmental durability. Moisture absorption and its effect on properties can vary by an order of magnitude between different types of polymers. For example, some thermoplastics such as PEEK absorb less than 1% of their own weight in water whereas some thermosets (such as phenolics) can absorb more than 10%. Epoxy resins absorb between about 1% and 4% in water (depending on the type of epoxy), and the epoxies used in external aircraft structures exposed directly to the atmosphere are often selected for their low moisture absorption.
Exposure of composite materials to humid air and water (rain) allows moisture to diffuse through the outer plies of the composite towards the centre. Water molecules diffuse through the matrix via the ‘ free space’ between the polymer chains under a concentration gradient where the moisture content is highest at the surface and decreases towards the centre. After a period of time at constant humidity, an even distribution of moisture occurs through the material and this is the saturation limit of the polymer matrix.
The diffusion of moisture into aircraft composite materials can usually be described by Fickian diffusion behaviour. Fickian diffusion is characterised by a progressive increase in weight of the material owing to the uptake of water until an asymptotic value is reached at full saturation, as shown in Fig. 15.41. The rate of moisture absorption increases with the temperature, but is not strongly influenced by the relative humidity of the environment. In other words, the moisture uptake by a composite material occurs faster in warm than in cool conditions, but is not affected significantly by whether it is dry or moist air. The saturation limit, defined by the maximum, steady-state weight gain of the composite, increases with the humidity level of the atmosphere. The absorption of moisture by composites which exhibit Fickian diffusion behaviour is often reversible, and when the material is exposed to a dry environment the water can diffuse out into the atmosphere. Therefore, composite structures used in aircraft that operate in different environmental conditions may undergo cyclic absorption and loss of moisture.
The absorption of moisture into a composite material can be calculated using Fickian diffusion kinetics, which allows the time to saturation to be determined. Fickian behaviour assumes that the moisture concentration gradient through-the-thickness of a material can be expressed as a function of time:
where c is the moisture concentration, t is time, x is the distance below the material surface, and D is the diffusion coefficient of moisture through the material. The diffusion coefficient is calculated using an Arrhenius equation:
where Do is the diffusion coefficient at a known temperature (T), E is the activation energy for diffusion, and R is the universal gas constant. The diffusion coefficient is related exponentially to temperature, and therefore the rate of moisture absorption increases rapidly with temperature. Figure 15.42 shows the relationship between the diffusion coefficient for water in a carbon fibre–epoxy composite with temperature. An increase in temperature of 10 ° C typically doubles the diffusion rate.
where t is the time, M∞ is the mass change at saturation, and h is the thickness of the composite material. This expression can be used to predict the weight gain with time for any fixed environment condition (constant humidity and temperature) where the diffusion coefficient is known.
The diffusion coefficient for anisotropic materials such as a unidirectional composite depends on the fibre orientation, as shown in Fig. 15.43. The diffusion coefficient when moisture ingress is perpendicular to the fibres, which is the case for most composite materials exposed to a humid environment, is calculated using:
where it is assumed that the fibres are arranged in a regular square pattern. Diffusivity of water in the fibre direction in a carbon–epoxy composite can be up to 10 times higher than that perpendicular to the fibres. However, most composite structures are designed with the fibres parallel to the surface, and therefore only moisture ingress perpendicular to the fibres is considered.
Environmental degradation of the polymer matrix can occur in different ways, depending on the type of resin. Plasticisation, which is characterised by a loss in stiffness and strength, occurs with epoxy and other thermoset resins used in aircraft composites. For example, Fig. 15.44 shows the reduction in the open-hole compressive strength with increasing temperature for a carbon fibre–epoxy composite in a dry or saturated condition. The reduction in strength becomes more extreme in the saturated composite with increasing temperature owing to plasticisation of the matrix phase. Matrix-dominated properties are more sensitive to plasticisation than fibre-dominated properties. As a general rule, for a carbon fibre–epoxy composite cured at 180 °C, moisture reduces the matrix-dominated mechanical properties by up to 10% for subsonic aircraft and 25% for supersonic aircraft (owing to their higher skin temperatures). The absorption of water can also reduce the glass transition temperature and consequently the maximum operating temperature of the composite. Reductions in the glass transition temperature of 30 °C or more can be experienced, as discussed in chapter 13. Plasticisation is a reversible process, and the properties recover when moisture is driven out of the composite by drying.
More severe forms of degradation of the matrix include chemical changes by hydrolysis or chain scission reactions between the polymer and water. The polymer chains can be broken down by chemical reactions with the water. These changes are irreversible, and epoxies and other aerospace resins are formulated to avoid this type of damage. Swelling and volumetric stresses can occur when a large amount of moisture is absorbed, which causes warping and cracking of the composite.
Moisture absorption is a problem for sandwich composites with an open-cell core material, such as Nomex or aluminium honeycomb. Moisture diffuses through the laminate face skin and forms water droplets within the core, as shown in Fig. 15.45. Often after long-term exposure the water builds up at the bottom of the cells where it can soften Nomex and corrode aluminium honeycomb. This problem can be avoided by the use of closed cell polymer foam.
15.45 Moisture absorption into open-cell cores within sandwich composites owing to skin damage (reproduced from G. Kress, ‘Design Criteria’, in: ASM Handbook Volume 21: Composites, ASM International, Materials Park, OH, 2001).
Water trapped within composites can seriously affect the properties when it freezes. Water that collects within voids and cracks in composites and within the cells of core materials freezes when exposed to cold conditions at high altitude. Water expands when it freezes and has a higher bulk modulus than epoxy resin. Therefore, when water freezes it exerts tensile pressure on the surrounding composite material that can cause permanent damage, such as delamination cracking. The water freezes and melts each time the aircraft ascends or descends from high altitude, which causes cyclic stressing of the composite in a process called freeze/thaw.
The surface of composite materials can be degraded by ultraviolet (UV) radiation which breaks down the polymer matrix and, if present, organic fibres such as aramid. UV radiation destroys the chemical bonds in epoxy resins and many other types of polymers, and the degraded material is then removed from the surface by wind and rain. UV damage is often confined to the topmost layers of the material, and the bulk properties are unlikely to be affected owing to the slow rate of degradation. It is possible to dope polymers with UV-absorbing compounds to minimise the damage or, alternatively the composite surface can be protected using UV absorbent paint.
Composites consist of a reinforcement phase and a matrix phase. The reinforcement can be particles, whiskers or continuous fibres. Aerospace composites are reinforced with continuous fibres which provide higher stiffness and strength than particles or whiskers. The matrix phase can be metallic, ceramic or polymeric, although the matrix for most aerospace composites is a thermoset resin (e.g. epoxy, bismaleimide) or high-performance thermoplastic (e.g. PEEK). The functions of the polymer matrix are to bind the reinforcement fibres into a solid material; transmit force applied to the composite on to the fibres; and to protect the fibres from environmental damage.
There are many advantages as well as several problems with using carbon-fibre composites rather than aluminium in aircraft. The advantages include reduced weight, capability to manufacture integrated structures from fewer parts, higher structural efficiency (e.g. stiffness/weight and strength/weight), better resistance against fatigue and corrosion, radar absorption properties, good thermal insulation, and lower coefficient of thermal expansion. The disadvantages of composites include higher cost, slower manufacturing processes, anisotropic properties making design more difficult, low through-thickness mechanical properties and impact damage resistance, higher sensitivity to geometric stress raisers such as notches, lower temperature operating limit, and lower electrical conductivity.
The mechanical properties of composite aircraft structures are determined using a hierarchical approach to modelling that begins with micromechanical analysis of the unit cell followed by ply analysis then laminate analysis and, finally, structural analysis.
The elastic modulus and strength properties of unidirectional composites are anisotropic, with the mechanical properties decreasing rapidly with fibre angle from the longitudinal (0°) to the transverse (90°) direction. The longitudinal tensile properties are controlled mostly by the properties and volume content of the fibre reinforcement. The transverse properties are more dependent on the matrix properties. The anisotropic properties of composites is minimised by using cross-ply [0/90] or, in particular, quasi-isotropic [0/± 45/90] fibre patterns.
The carbon and glass fibres used in composite materials are brittle and fail at low strain. The fibre strength is determined by the largest flaw, which varies from fibre to fibre. As a result, the tensile strength of fibres varies over a wide range, and does not have a single strength value.
Failure of composites under compression loading occurs by microbuckling or, more often, kinking. Kinking involves the localised out-of-plane rotation and fracture of load-bearing fibres caused by a buckling instability. The compressive strength is determined by the applied stress required to cause kinking, which increases with the shear properties of the polymer matrix and the closer the fibres are aligned to the load direction.
Microstructural defects such as voids, microcracks, resin-rich regions and wavy fibres can reduce the Young’s modulus and, in particular, the strength properties of composite materials. Matrix-dominated properties (e.g. interlaminar shear strength, impact resistance) are affected more by defects than fibre-dominated properties (e.g. tensile strength).
Sandwich composites are used in aircraft structures that require high bending stiffness and buckling resistance. The core material must have sufficiently high shear stiffness to transfer load to the face skins. Different types of core materials are used, including honeycombs and closed-cell polymer foams.
Composites can absorb moisture and other fluids (e.g. fuel, hydraulic oil) which can adversely affect their physical and mechanical properties. The most common environmental durability issue for aerospace composites is moisture absorption from the atmosphere, particularly under hot/wet conditions. Thermoset resins generally absorb more moisture than highperformance thermoplastics, and epoxies can gain 1–4% of their original weight owing to water absorption. The composite materials used in aircraft structures often exhibit Fickian behaviour which is characterised by a steady uptake of water until the saturation limit is reached. The absorption rate increases rapidly with temperature whereas the maximum amount of moisture absorbed at the saturation limit is determined by the relative humidity of the environment. The absorption of moisture by epoxy matrix composites causes plasticisation, which is reversible when the material is dried. The absorption of moisture reduces the glass transition temperature and mechanical properties (particularly the matrix-dominated properties). Composites are also susceptible to environmental degradation by freeze/thaw and UV radiation.
Matrix-dominated property: Mechanical property which is strongly influenced by the properties of the matrix phase. Examples are interlaminar shear strength, transverse tensile strength, interlaminar fracture toughness and impact damage resistance.
Reinforcement: Constituent material of a composite that forms the discrete (discontinuous) phase. Reinforcement phase often used to increase structural properties such as stiffness and strength of the composite.