1. Aerodynamic lift loads and wind-tunnel pressure tests to determine the principal structural loads that can generally be validated through static and cyclic fatigue testing.

2. Dynamic loads such as due to flutter, turbulence, and engine vibration that are used to size locations where these loads are most prevalent, such as wing-to-body joins, engine pylons, and flight-control surfaces.

3. Takeoff and landing loads are handled mainly by the landing gear, which is largely a metallic structure. The way that these loads are transferred into the airframe will guide the design of structures that interface with the landing gear.

4. Impact loads such as from bird strikes, hail, dust, sand, and runway debris must also be considered. These will be discussed in detail in Section 8.3.3.

Bird strike

The US Department of Agriculture compiles a database of all reported bird/wildlife strikes to United States civil aircraft and to foreign carriers experiencing strikes in the United States through an interagency agreement with the FAA. Between 1990 and 2008, 89,727 animals were struck with the vast majority (97.4%) being birds [17]. Because the frequency of impacts due to animals other than birds is small, only bird strikes will be discussed here. The number of reported bird strikes has been increasing and this is believed to be a combination of increased awareness and reporting as well as increases in both bird populations and flight frequency. Most bird strikes occur during descent, approach, or climb, and not during cruise.

Some countermeasures are used to minimize the likelihood of bird strikes, but notwithstanding the fact that a low percentage of flights actually suffer from high-velocity bird strikes, it must be assumed that a structure will sustain a bird strike in designing aircraft structures to ensure vehicle and passenger safety. Indeed, the FAA requires such performance in FAR Part 25, Sub-part 25.571 [1] ‘Damage-tolerance and fatigue evaluation of structure’, where it is stated that ‘The airplane must be capable of successfully completing a flight during which likely structural damage occurs as a result of impact with a 4-pound [1.8-kg] bird when the velocity of the airplane relative to the bird along the airplane’s flight path is equal to Vc [design cruise speed] at sea level or 0.85 Vc at 8000 ft [~2440 m], whichever is more critical.’

With increasing use of composites in aircraft like the Boeing 787, the need for robust design tools has become even more necessary. Indeed, Boeing has developed [18] some simulation techniques for characterizing bird strikes in order to certify composite structures for airworthiness. In this work, finite-element models were developed to simulate a bird strike and the models were corroborated with test data to prove their validity.

Lightning strike

The FAA regulations [1] provide the guidelines for lightning protection for airframes and these requirements are identical for all vehicle categories. Specifically, the requirements appear in Section 25.581 of the regulations ‘Lightning Protection of Structure’, which requires that aircraft be able to withstand a lightning strike without suffering catastrophic damage. Additionally, guidelines for military aircraft are contained in MIL-STD-464 [19].

Thus, it is the goal of designers using composites to prevent lightning strike from causing catastrophic structural damage and to prevent hazardous electrical shocks from reaching vehicle occupants and flight-critical electrical systems. Damage to electrical systems may cause a loss of aircraft flight-control capability. Lastly, it is imperative to prevent any lightning strike from causing sparking in fuel tanks that may lead to the ignition of fuel vapors. Composite materials present a unique challenge as they are, for the most part, non-conducting. Although carbon fibers may have some electrical conductivity along their length, they have far lower conductivity transverse to the principal fiber direction. Additionally, polymer matrix resins used for aerospace are electrical insulators. For these reasons, add-on solutions are used to prevent significant damage from affecting the underlying composite structure.

For non-metallic components such as composites, meeting the requirements may be accomplished by appropriate design considerations. This is especially important as polymeric composites have electrical conductivities roughly three orders of magnitude lower than aluminum, which is the standard material for nearly all metallic aircraft. Thus, it is necessary to incorporate lightning strike protection into composite design.

In order to provide protection from lightning strikes to composites, metallic materials like those shown in Fig. 8.4 are typically utilized to divert the electrical current from a lightning strike so as not to endanger the airplane. Different parts of the structure may have different lightning strike requirements as shown in the schematic in Figure 8.5, which shows zones susceptible to different likelihoods and severities of lightning strike. Methods for meeting the requirements for lightning strike protection are not rigidly specified to allow manufacturers to adopt different certification strategies. Zone 1 regions are likely to experience initial lightning strikes and first-return strokes. Zone 2 regions are likely to experience subsequent swept strokes, or re-strikes. Except for a few protruding areas of the airframe, such as the wing tips and nose, which comprise Zone 1, the vast majority of the airframe belongs to Zone 2. In addition to the zoning information provided by SAE International as shown in Fig. 8.5 [20], SAE also provides aerospace recommended practices (ARPs, available at http://www.sae.org) that can be utilized to aid designers in meeting these requirements.

The metallic materials mentioned above for providing lightning strike protection on composite structures are used near the outer surface of the composite for several reasons. First, it is essential to keep the high energy of a lightning strike away from any combustible materials, especially fuel that may ignite if exposed to a sufficiently large spark. Second, keeping the energy on the surface serves to mitigate the amount of damage that might be otherwise caused by a lightning strike. Figure 8.6 shows a comparison of test panels subjected to a simulated Zone 1A strike with and without lightning strike protection. Severe localized delamination and heat damage is observed without lightning strike protection. In addition to providing the high surface electrical conductivity necessary for diverting the energy of a lightning strike, designers must also complete the electrical circuit by establishing a current return or grounding plane for the electrical energy. Metallic aircraft, by contrast, can utilize the entire airframe as a grounding plane.

The necessity of efficient, lightweight lightning-strike protection for aircraft has become much more important recently with the introduction of the Boeing 787 Dreamliner and the Airbus A350XWB, which are the first two large commercial transports to use composite materials for the majority of their primary structure. Because of these new, significant applications of composites in aircraft structures and the high likelihood that this trend will continue with other new commercial transports, there is active work in developing newer and improved approaches to handling the problem. Some of this work is related to modeling lightning strike (e.g. [21] and [22]) to be able to design full systems that incorporate lightning-strike protection instead of the current state of using an add-on solution like metal meshes. One example of a systems-level approach is the creation of a Faraday cage within the structure as outlined by Hawley [23].

Moisture: rain and ice

Composite materials all pick up some amount of moisture depending on the chemical composition of the matrix resin used. Additionally, stress can enhance moisture absorption in composites [24] so that the effects in service may be more severe than simple tests may suggest. Once inside a polymer, moisture causes the modulus to decrease, especially at elevated temperatures. Furthermore, moisture depresses the glass transition temperature (T_g) of polymers, thereby limiting the upper use temperature. In all specifications for aerospace composites, testing is performed on composites that contain moisture to measure the property reductions at elevated temperatures. In these tests, moisture is introduced into the composites by either immersion in water at an elevated temperature for a period generally not less than 14 days or by exposure of the composite specimens to high humidity for a similar period of time. A general rule of thumb is a 25–30 °C reduction in the upper use temperature of a composite after significant exposure to moisture.

In addition to moisture absorption within laminates, composite structures, in particular sandwich structures and especially honeycomb-core sandwich structures, are subject to moisture ingress into the core material [25]. Although moisture that accumulates within a foam or honeycomb is a concern as it adds weight to an aircraft, the potential for the moisture to suddenly vaporize, which causes a large pressure increase inside the sandwich structure that may cause the skins to delaminate from the core, is the main source of concern. Such a failure was purported to be the cause of the failure of the X-33 cryogenic tank [26] which was a honeycomb-core sandwich structure.

It is well known that moisture acts as a plasticizer on polymer matrices and there is evidence of microcrack formation due to cyclic moisture exposure [27, 28] and to combined moisture and thermal cycling [29]. An example of microcracking caused by combined moisture and thermal cycling is shown in Fig. 8.7. It is also sometimes believed that microcracks such as those caused by moisture and thermal cycling can be repositories for moisture accumulation that could create further problems due to ice formation. However, this can only occur if the size of the microcracks is sufficient to allow the formation of ice crystals [30] and, if this is not the case, then the liquid present in these cracks becomes a vitreous solid [31] instead of ice. Because the water does not form ice if the cracks are sufficiently small, it does not experience the large volume expansion caused by the phase change from liquid water to ice. Because the density of liquid water and vitreous water remains essentially the same from 0 °C to −175 °C [32], no such volume expansion occurs.

Ice is a factor when it accumulates on aerodynamic surfaces but as this does not cause direct damage to the underlying composite structure, it will not be considered in this chapter. Ice is also a concern when it impacts the structure in the form of hail. It is a variant of impact damage as previously discussed above, but some details particular to hail will be discussed below.

Moisture inside cracks is of concern in a similar manner to that noted for honeycomb-sandwich structures above when the temperature rises sufficiently and rapidly under a phenomenon called ‘thermal spike’, which may cause this trapped water to suddenly turn to steam and cause damage within a composite structure [33, 34]. It has been observed that thermal spikes cause internal damage and these are a particular concern in both supersonic vehicles and exhaust-washed structures. McKague et al. [33] performed simulated flight cycles involving a subsonic mission, 54 °C for 90 minutes, and for a flight involving a subsonic segment with a supersonic dash, which caused rapid heating to a peak of 149 °C followed by rapid cooling. The simulated thermal spikes caused permanent changes in the subsequent moisture–diffusion behavior of the carbon-epoxy such that both the amount and rate of moisture absorption were significantly increased. In contrast, the subsonic exposures caused no detectable change in diffusion behavior. Similar results were observed by Adamson [35] who also observed irreversible damage in composites due to thermal spiking.

Moisture is also a problem when it freezes and becomes hail. Hail impact at the high velocities of flight can cause significant damage to aircraft, as evidenced by the large amount of damage seen in Fig. 8.8, which shows a metal aircraft. Composite structures will, naturally, have different failure modes than metallic ones with laminates experiencing similar damage patterns as those from other impacts like tool drop or bird strike. In addition to the size of the impact from hail, it is the potentially large number of impacts within a short time period that is of most concern and, because these types of events are significantly enough likely, structures must be designed to be able to carry design loads after such impacts.

For honeycomb-sandwich structures, hail impact damage was modeled by Horrigan and Aitken [36] based on observed damage modes that strongly resemble the quasi-static indentation damage shown in Fig. 8.9 [37] where damage is due to localized, near-surface cracking in non-metallic honeycomb and near-surface buckling in metallic core. Higher impact levels will cause skin failure and direct impact on to the underlying honeycomb structure [38].

Simulated testing of hail strikes has been performed [39, 40] to better understand the phenomena associated with these impacts. The experiments of Kim et al. [39] show a linear relationship between the peak measured force and the projectile kinetic energy, regardless of projectile size, and it was observed that, in composite panels, there is a linear relationship between the kinetic energy at which failure initiates and the thickness of the panel. At the failure initiation threshold, multiple localized failure modes may be observed and non-orthogonal impact energy levels may be accurately estimated using simple trigonometric scaling. Mahinfalah and Skordahl [40] showed how most of the impact energy from hail impact goes into disintegrating the hail such that little internal damage was observed, especially in comparison to similar mass aluminum projectiles, which were much more likely to cause internal damage to composite structures, though, below a certain threshold size and mass, no internal damage was seen. Subsequent fatigue testing was consistent with these observations as no significant changes in behavior were observed. More recent work [41, 42] has focused on trying to model hail impacts to enable predictive models to be used to improve composite designs.

Fluids

Aerospace composites must be able to withstand a large number of different fluids. Not least of these is moisture, which, in addition to the problems outlined above in honeycomb-sandwich structures, diffuses into polymer matrices. A general overview of the effects in composites is given by Weitsman [43] who notes that, although there are good means for characterizing fluid sorption, how these relate to durability and strength is much more problematic and requires that a sufficiently robust database be generated to properly account for degradation and damage due to fluids.

In addition to ever-present moisture, composites are also exposed to hydraulic fluids (e.g. Skydrol), fuel (e.g. JP-4), cleaning solvents (e.g. methyl ethyl ketone, methylene chloride, etc.), and a wide variety of other fluids used in servicing and maintaining aircraft. As fluid exposure/conditioning can induce strong matrix plasticization, it is possible that significant reductions in static and fatigue performance may result from the presence of fluids as compared to unconditioned laminates. Sala [44] summarizes the effects of various fluids on composites, noting that impact performance may also be degraded substantially due to fluid exposure.

For composites made with fiberglass, aramid (e.g. Kevlar®, Twaron®), or other organic fibers (e.g. Vectran®, Zylon®, Dyneema®, etc.), the susceptibility of the fibers to fluid degradation must also be taken into account. In cases where the fibers are not sufficiently compatible with likely fluids to which they would be exposed, then they may not be used for those structures. In these cases the use of carbon fibers is especially useful as it is generally chemically inert under normal usage in aerospace applications in addition to its high specific strength, which is the main reason for its ubiquitous use in aerospace composites.

Because carbon fibers are predominantly used for aerospace applications, it is the matrix susceptibility to degradation that is of most importance. In all specifications for aerospace composites, there are specific requirements for fluid resistance. Thus, there are built-in safeguards to prevent the use of unsuitable materials that will be exposed to harsh fluids.

Temperature

As mentioned previously, polymers are affected by temperature. In general, this concerns the change in mechanical properties that results from the softening of a matrix resin at elevated temperatures, as shown in Fig. 8.10. Cyclic temperature exposure is also important and is analogous to cyclic fatigue where, instead of direct mechanical loading, stresses are induced from the differences in the coefficients of thermal expansion between the fiber and matrix. Indeed, Li et al. [29] note that most microcracking in cyclic thermal fatigue occurs at low temperatures and also that increases in the ratio of high temperature to low temperature increase the density of cracking that is observed. The increased susceptibility to microcracking at low temperatures is also noted by Tsotsis [30] who cites prior research [45]–[47] on the characterization of polymeric composites at cryogenic temperatures.

Another issue of concern is long-term exposure to elevated temperatures. There is a large body of work on long-term thermo-oxidative degradation of polymer composites and this is summarized by Tsotsis [48] who notes that, in the absence of highly accurate, reliable modeling techniques to characterize long-term aging of polymeric composites, it is necessary to use materials that will remain essentially unchanged in their performance and properties during their desired lifetime. In fact, this is already the de facto approach used with polymeric composites, though no formal guidelines exist stating this.

All aerospace composites are used well below their glass transition temperature (and especially their wet glass-transition temperature) such that thermo-oxidative effects are largely not present. The only exceptions are for composites used for space applications where the exposure times are sufficiently short that it is highly unlikely that any degradation of any significance will occur during the lifetime of the structure (generally only a few hundred hours of exposure at maximum use temperature). For composites based on higher-temperature matrix resins like bismaleimides and polyimides, the desired lifetime at temperature will determine the allowable upper use temperature in service.

High-speed airflow

For supersonic aircraft temperature also can become a key design limiting parameter. A National Materials Advisory Board publication [49] summarizes the requirements for supersonic aircraft quite well based on prior studies, many related to the Concorde, which are also cited in this reference. At Mach 2.0, the skin temperature remains well below 100 °C but rises to 120 °C at Mach 2.2 and further to 150 °C at Mach 2.4. Usually, at Mach 2.0 or below, thermo-oxidative degradation is not a concern. However, as the skin temperature rises to 120 °C or above, thermo-oxidative degradation and its potential consequences can either prevent composites from being used at all or require them to be protected – usually through insulative layers – such that they operate at more benign temperatures. In many cases, the use of insulation is not permissible (such as on aerodynamic surfaces or parts exposed to high-speed airflow), but, recently [50, 51], some development on thin thermal-barrier coatings has been done to try to address this problem. The goal of thermal barrier coatings is to reduce the temperature of the substrate such that it operates at a more benign temperature where it is stable. As mentioned above, composite materials considered for use at elevated temperatures must be chemically stable during their desired lifetime or other solutions must be found.

In subsonic flow, turbulence, gusts, and flutter, rather than temperature, can affect the design of aircraft structures (note that these loads also persist in supersonic flow). These dynamic loads are a result of instabilities in the atmosphere as they encounter aerodynamic surfaces, such as a wing. Because these loads are generally much larger than general flight loads, they must be considered in the proper sizing and design of aerodynamic surfaces. Furthermore, they must also be taken into account when simulating flightload spectra both experimentally and analytically.

In order to certify aircraft structures, it is essential to eliminate flutter entirely from the flight spectrum. In order to do this, the analytical techniques differ somewhat from those for isotropic, metallic aircraft because composite materials involve non-classical effects such as transverse shear, warping inhibition/restraint, non-uniformity of contour-wise shear stiffness, and other 3D effects not present in metals. Shokrieh and Behrooz [52] demonstrate some finite-element techniques that may be used to eliminate flutter from a composite-based design and Kim and Hwang [53] showed techniques for dealing with gust loads for composite wings. Qin et al. [54] modeled composite wings using anisotropic thin-walled beams to represent the composite structure and showed how lay-up tailoring could be used to meet the loading requirements from gusts and how elastic coupling could be used to eliminate flutter. Some more general discussions of analytical methods to characterize flutter are given in Refs [55]–[57].

Fire safety

As for structural loads, there are strict requirements for fire safety on aircraft. These are outlined in the FAA guidelines for flammability [1] under FAR 25.856, which gives smoke-density and heat-release requirements, among other properties, and materials must be tested according to the FAA Aircraft Materials Fire Test Handbook [58], meeting these safety regulations in order to be certified for use on commercial aircraft structures. Most safety regulations pertain to the interiors of aircraft, where fires are most likely to start. However, because these materials are not primary load-bearing materials for the aircraft structure, they will not be considered any further here.

The growing use of composites over the past several decades has led the FAA [59] and other aviation-based international governing bodies to design requirements and certification procedures for composite-based aircraft structures. Fires involving composite structures can be quite difficult to extinguish, as shown in Fig. 8.11, and the main requirement, as for aircraft interiors, is to allow sufficient time for passengers to exit safely from a burning aircraft before heat and smoke adversely affect them. In this case, external fires, usually beginning as an engine fire, such as shown in Fig. 8.12 [60], are the primary concern in structural fire safety. Such fires have been characterized by the FAA [61] to help develop certification requirements and accompanying test methods for certification. In order to simulate fuselage burnthrough, the FAA specifies a modified oil burner to simulate the effects of an external fuel fire on an aircraft fuselage and interior components (Fig. 8.13) [62]. The burner used in this test is a typical home heating-oil burner, but using Jet A jet fuel in the burner.

The requirements for the fire safety of structural composites are no different than those for aluminum aircraft, and standards for compliance are listed under the FAA Flame Propagation test, FAR 25.856. The ability of a fuselage to resist penetration from an external fuel fire is directly related to the thickness and material of the skin. However, differing local structural load and other design requirements will necessarily create considerable variations in local skin thickness meaning that different airplane types will have different burnthrough resistances. Other factors internal to the airplane can also affect penetration of an external fire into the fuselage interior. For example, differences in aircraft ducting that affect air return within the cabin can influence the time required for an external fire to penetrate through the fuselage.

According to FAR 25.856, a fuselage must be able to resist flame penetration for at least four minutes. It is not essential that the fuselage structure itself be able to withstand this, but that the aircraft (the combination of the fuselage, thermal/acoustic insulation, fire barrier materials, and sidewalls plus any active fire prevention equipment/systems) be able to meet the requirement. For example, Mueller et al. [63] describe a conceptual design wherein the cargo floor comprises a fire barrier to minimize the amount of the addition of non-structural burnthrough protective materials. Other concepts based on the use of intumescents, such as by Heitkamp [64], or using inorganic materials to act as fiber barriers [65] are also known.

Because the fuselage must be dealt with as a system, the type of composite material used to fabricate a fuselage may not be the prime determining factor in the fire-safety performance of the fuselage. However, composite materials that have better flammability properties than others should, in principle, reduce the need for other non-structural materials to meet the mandated fire-safety requirements.

Ground-based damage due to human error

Ground-based damage is principally caused by maintenance crews impacting airframe structure, tool drop during maintenance, or from errors in ground-traffic control. The first of these is generally related to impacting the airframe with ground vehicles such as forklifts, refueling trucks, food-servicing trucks, etc. For the most part, these types of impacts are at levels much greater than can be practically designed for. In these cases, it is more important to consider how to repair such structures than how to prevent this type of damage.

The National Business Aviation Association (NBAA) has developed some best practices for preventing ground damage [66]. In this study, it was noted that nearly half of ground-based damage in business jets occurs during the towing of aircraft and approximately a quarter attributable to both ramp movements and ground-service equipment. Although these guidelines were writing specifically for business jets, similar types of damage can be inflicted on large commercial aircraft – as seen in Fig. 8.14, which shows landing gear that was damaged by towing an aircraft with its brakes locked – and on military aircraft [67] where struts and landing gear are among some of the damage caused during ground-based maintenance.

Wenner and Drury [68] evaluated some of the reasons for the human errors that cause ground damage to large commercial aircraft. The trends in types of damage differ slightly from the NBAA results mentioned above with over 60% of the damage occurring when the aircraft is parked. It is interesting to note that nearly 72% of all damage results from equipment, vehicles, or hangar doors impacting the vehicle. As mentioned previously, it is not weight-efficient to design aircraft to withstand such large impact levels, so such damage will necessitate repair to the damaged structure. In contrast, tool drops represent only 3% of all ground-based damage. Tool drops have impact energies similar to hail or debris (e.g. rocks and stones) and, combined, these impacts occur in sufficiently high probability that they must be included in materials specifications and structural designs. The impact energy imparted by a tool drop is used to specify the levels in tests like compression after impact (CAI) [11]. Means for dealing with impact damage will be discussed below.

Stitching

Toughening approaches with preforms have been developed previously. Most notable was the use of transverse or z-direction stitching during Boeing’s Advanced Subsonic Technology Composite Wing Program (NAS1-20546) [75], which showed significant improvements in impact damage tolerance as a result of stitching. Other studies such as those by Larsson [76], Jain et al. [77], and Sharma and Sankar [78] also report significant increases in damage tolerance and fracture toughness due to stitching. Although stitching provides significant out-of-plane property improvements and drastically reduces the amount of fasteners required if used to fabricate integrated preforms, it does so at a cost. This cost includes the amount of time required to provide for field stitching (i.e. stitching not used solely for assembly or attachment) as well as a reduction in in-plane properties due to the multiple penetrations of the stitching needle through the preform, as noted in [75] and by Mouritz et al. [79] and Reeder [80].

The result is that it is desirable to use stitching only where it provides the most benefit and limit the amount of field stitching required. This last point is the key driver for searching for other preform-based toughening methods that would provide the same benefits in toughness as stitching, but at a lower cost.

Tackifiers

One such candidate method is the use of toughening tackifiers. These have been used for some time and are formulated in two different ways. One is to use a partially solvated version of the matrix resin (the solvent provides greater tack and flow than the unsolvated resin) and another is the use of tacky, rubbery compounds. In both cases, the tackifiers are either sprayed or brushed on. In certain instances, heat may be used to apply the tackifiers, but this is more the exception than the rule. Hillermeier and coworkers [81, 82] describe some approaches and benefits of tackifier tougheners.

Tackifiers have several potential problems. First, there is a limited outlife or handling time on the preform. Solvent-modified tackifiers will lose their solvent to the atmosphere if exposed for too long a time, thereby negating their tack. In the case of systems that comprise a version of the matrix resin, the catalyzed resin may advance if exposed for too long a time, also eliminating the tack. Second, tackifiers may inhibit resin flow during infusion [83]. Third, in low-pressure processing, tackifiers will prevent full consolidation of a preform resulting in too high resin contents and corresponding low fiber volumes as shown in Fig. 8.17. Fourth, the presence of uncatalyzed, chemically incompatible, and/or unevenly distributed tackifier reduces the mechanical properties of molded composites [84].

Interlayers

Examples of attempts to use interlayers in fabrics are available in the patent literature. For example, Mitsubishi (EP 1 175 998) made laminated products formed of reinforcing fibers in which thermoplastic resin layers were provided between layers of the reinforcement fiber. The thermoplastic resin layer was described in the form of a porous film, fiber, network structure, knitted loop, etc. The laminated product used a thermoplastic layer of sufficient permeability between the layers of reinforcing fibers so as not to inhibit liquid resin flow during a liquid-molding process. However, the processes described in EP 1 175 998 produce preforms that have poor handling characteristics and, thus, are not very stable during resin infusion. In another example, under EP 1 125 728, Toray uses short fiber, non-woven fabrics that have a portion of the fibers of the non-woven fabric pass through the main reinforcing fiber to integrate the non-woven fiber to the fiber reinforcing material. Claims are also given for the use of pressure-sensitive adhesives instead of fiber entanglement for attaching the non-woven to the fiber-reinforcing material. Lastly, it also claimed by Toray to use a non-woven material that comprises both low-(5–50%) and high-melt materials such that the fabric is attached to the fiber reinforcing material by melt-bonding the low-melt material.

Recently, Hogg [85] discussed the use of thermoplastic fibers to improve the toughness of composites made with brittle resins, such as untoughened epoxies. In this work thermoplastic fibers were introduced either by commingling or by placing veils between layers of structural fibers. Both methods showed improved impact resistance compared with unmodified laminates using otherwise identical materials and lay-ups.

A successful implementation of non-woven interlayers on the Boeing 787 is based on work by Tsotsis [86]. In this work it was demonstrated that the use of lightweight, non-woven veils using out-of-the-autoclave resin infusion and dry-fiber preforms can be used to match the performance of the state-of-the-art prepreg composite materials. It was also noted that it is particularly important to select an interlayer material that is compatible with the matrix resin or, rather than improving impact resistance, it may actually reduce it. The incorporation of non-woven veils is compatible with existing fabric manufacturing technologies and test data showed that other properties like open-hole compression were not compromised by the veil, when selected and applied properly. Indeed, results suggest that, if the veil is melt-bonded to the carbon fiber prior to resin infusion, the veil acted to stabilize the structural fibers and to improve compressive properties.

Stitched joints

One example of reducing the number of fasteners is by using more unitized structures. The NASA Composite Wing Program [75, 99] demonstrated through-the-thickness stitching of dry preforms, combined with resin infusion for addressing cost and damage tolerance issues related to using carbon-fiber composites in primary structures of commercial aircraft. Stitching together fabric-based subcomponents was used to incorporate various elements of a wing torque box (i.e., wing skin, stiffeners, intercostal clips, and spar caps) into an integral structure that eliminated the requirement for thousands of mechanical fasteners by replacing mechanical fasteners with stitching threads. A photo of the first lower cover panel produced in this program is shown in Fig. 8.19. The technology used to fabricate this panel, in addition to demonstrating an approximate 75% reduction in fastener count, also showed much improved impact-damage resistance compared with other composite designs.

It is important to note that, because stitches provide a highly redistributive way to transfer load, they do not require any additional structure to carry the very small stress concentrations caused by the stitch threads. Furthermore, because the number of stitches is at least one to two orders of magnitude larger than any amount of fasteners that would be used, there is far larger redundancy in load transfer between the two structures being joined. Indeed, this large redistribution in load changes local failure modes. In impact, the joint is more robust against impact than the remaining structure. By contrast, in bonded joints the most susceptible area for impact damage is the bonded joint because, in the bonded case, only resin (co-cured) or an adhesive (co-bonded) is available to carry load, whereas in the stitched case multiple high-strength stitch threads can transfer the load through the joint without any damage. And because this means that the impact occurs, by definition, on a larger mass (bonded stiffening element + skin vs. skin alone), the impact energy is better dissipated when the impact is centered beneath stiffening elements instead of between them.

Z-pinned joints

Another method to reinforce joints is the use of Z-pins. These are pre-cured, stiff pins or rods that are inserted either into a preform or, more typically, an uncured prepreg to provide out-of-plane reinforcement between plies. Like stitching, it is highly load redistributive and minimizes stress concentrations typically caused by fasteners. Partridge and Cartié [100] give an overview of Z-pinning, pointing out that Z-pins should be inserted prior to bagging in the case of autoclave cure of prepreg-based parts or just prior to resin injection in the case of liquid molding of Z-pinned preforms. For prepreg, the specific part preparation required for pinning was stated to be minimal, usually limited to a longer de-bulk phase in order to compact the laminate as close as possible to its final cured thickness. However, their conclusion fails to note two principal problems with Z-pinning. First, the insertion of the Z-pins can be quite time-consuming, especially for large parts. Second, specialized tooling must be used to insert Z-pins in parts with complex geometries. For these reasons, Z-pins have been used mainly in small- to medium-sized parts that are not overly complex. Partridge and Cartié do note that Z-pinning is not particularly suitable for the reinforcement of dry-fiber preforms unless the preform is stabilized by some means, such as a binder to facilitate both compaction and pinning. The benefits of Z-pins to improving fracture toughness is clearly shown in Fig. 8.20 [100] where both Mode I and Mode II fracture toughness are significantly improved with a sufficient loading of Z-pins. It is well known that improvements in Mode I improve joint strength, particularly pull-off loads, and that improvements in Mode II correlate strongly with improved impact-damage resistance.

The improvement in fracture toughness from Z-pins (as for stitching) is largely attributed to increasing fiber pull-out, which increases the amount of energy required for a crack to grow per unit area. Dai et al. [101] did work to establish a bridging law for Z-pins; the functional relationship between the delamination crack-opening displacement and the closure force from a single pin being the bridging law. Because Z-pin pull-out is a complicated process that is affected by many variables such as material properties, geometry, and interfacial parameters between the pin and the laminate, these must all be accounted for in any bridging law. In order to build their bridging law, Dai et al. needed accurate experimental data, which they generated themselves. Their results concluded the following:

The first of these conclusions is the most important as it relates to improving damage toughness as it agrees with observations given above for stitching. The large load redistribution possible by having more, but smaller, out-of-plane reinforcements is clearly beneficial to providing a more damage-resistant structure. Additionally, as pin diameters grow, they become more and more like the fasteners they are meant to replace. Small pins at the same areal density, just like small stitches, will have far larger contact surface areas between the out-of-plane reinforcement than do larger pins, stitches, or fasteners. It is most likely that this very large increase in contact area accounts for a large portion of the improvement in fracture and pull-off loads from Z-pins and stitches. Extending this thinking further, it is also likely that, using the right type of out-of-plane reinforcement, further decreasing the size while maintaining the areal density will continue to provide benefits in improving damage tolerance.

Advanced preforms

The knowledge gained from understanding the roles of fiber, matrix, and out-of-plane reinforcements on improving damage tolerance is being used to create structures and structural concepts that have built-in damage tolerance features to leverage this knowledge. One recent structural concept is called PRSEUS [102] (for pultruded-rod-stiffened, efficient, unitized structure) shown in Fig. 8.21. This concept takes advantage of well-known mechanics principles like stiffening a structure by offsetting a stiff element (the pultruded rod) from the principal plane of bending (the skin) in a manner similar to an I-beam, but using a combination of composite materials, including a pultruded rod, closed-cell structural foam, a fibrous preform, and a matrix resin that is infused into the assembly and cured to make a part.

Yovanof et al. [103] examine the potential for the use of PRSEUS to solve some of the structural and weight problems posed by aircraft with non-circular fuselages, such as the blended-wing body, shown in Fig. 8.22. In addition to targeting non-circular pressure vessels like a fuselage, the PRSEUS design shows some of the multifunctional capabilities of composites as the stitched design is inherently damage-arresting such that damage in one structural bay will not easily propagate to another. Furthermore, the linking of the stiffened sandwich stiffening elements allows for efficient load redistribution that further enhances damage tolerance. Li and Velicki [104] show that the type of structural concept shown in Fig. 8.22 can be tailored and optimized to exploit fully the orthotropic nature of carbon fibers and damage arrest design philosophies compared with existing composite designs and practices.

Dang et al. [105] present a method for the optimal design of unitized structures under damage tolerance constraints by considering both unitized structures with straight and curvilinear stiffeners. The objective of their study was to minimize the weight of the structure with constraint on static fracture strength, whereas panel thickness and stiffener sizes (height and thickness) are design variables. The study included damage-tolerance-based optimization of unitized structures with periodic boundary conditions to demonstrate the effectiveness of unitized structures, especially curvilinear-stiffened structures in terms of the load-carrying capacity in the presence of damage. The authors claimed that the method developed in the study could be used to optimize the unitized structure with static fracture strength or fatigue-life constraints coupled with buckling constraints to better exploit the effects of shape variables on the reduction of mass of the structure.

Scaling and hybrids

Ilcewicz [106, 107] discusses some of the issues regarding damage tolerance and the introduction of composites into primary aircraft structure. Scaling from small-scale test coupons to full-scale structures as well as scaling from small- to large-notch behavior are among the key scaling issues identified in furthering the use of composites in airframes. Fig. 8.23 shows the interrelationships of scaling steps and damage tolerance. It was noted that a number of factors, including damage size, spacing of stiffening elements (frames or stringers), and the ratio of skin to stiffening-element load sharing all will affect progressive damage growth, load redistribution, and structural failure.

Structural response is dependent on the type of damage, such as outlined in Table 8.1, which differentiates between damage that the structure should be able to withstand without repair and those types of damage that will require either minor or major repair. In certain cases, replacement of the damaged component will be preferred to repair. In all these cases, consideration of the likelihood and severity of the different categories of damage and the regulatory requirements [1] will guide the ultimate design. For example, in stiffened-skin designs, as mainly considered here, damage growth is dependent on the fracture behavior of the structure. In this case, for damage between stiffening elements, the damage growth will slow as the increased load redistributes into the adjacent stiffening elements. At this point, damage can proceed in several ways: the stiffening element may fail; the stiffening element may become detached from the skin; damage may propagate beneath the stiffener into another bay; or damage may be redirected along the stiffening elements. All of these, with the exception of the latter, which is typical of stiffening elements attached to skins via stitching or Z-pinning, typically result in catastrophic failure.

Different material combinations, in addition to their structural arrangement, may also affect damage tolerance behavior. Most often these material combinations are in the form of so-called hybrid composites made up of two or more different fiber types. Strictly speaking, many of the composites listed previously, like stitched or Z-pinned composites, are hybrids, but in the present context this will only refer to those with multiple fiber types or mixtures of fabric forms (e.g. braids, weaves, etc.). This section will also neglect nanocomposites-based hybrids as there are insufficient data to date to assess their ability to modify damage tolerance in aerospace composites. Lastly, as the focus of the present chapter is on polymeric composites, fiber-metal laminates like GLARE (glass-aluminum-reinforced epoxy) and ARALL (aramid-aluminum laminate) will likewise not be considered here.

Hybrids may be made from commingled yarns, e.g. Ref. [108], alternating layers of different fibers, e.g. Refs [109]–[111], or woven, non-woven, or braided materials containing two or more fiber types. Many of these hybrids for aerospace use glass-carbon, glass-aramid, or carbon-aramid combinations. It has been found in multiple instances that hybrids of glass and carbon can provide improved impact resistance, e.g. Refs [110] and [111], depending on the specific material combinations and lamination scheme. The different properties of the glass and carbon will also have an effect on other mechanical properties such that the improved damage tolerance may not provide the most cost-effective or weight-efficient design [109].