Classic Lightweight Design

Abstract

Along the process chain, Fig 1.3, new materials, better material properties, improved and more accurate design methods, new design concepts, new joining technologies, more efficient production technologies, and new automated quality assurance processes make diverse contributions to optimised and more cost-effective lightweight design. The basis is carbon-fibre reinforced polymers (CFRP), Sect. 1.3, whose lightweight design potential can be exploited much more extensively with the research results cited below as examples.

Along the process chain, Fig. 1.3, new materials, better material properties, improved and more accurate design methods, new design concepts, new joining technologies, more efficient production technologies, and new automated quality assurance processes make diverse contributions to optimised and more cost-effective lightweight design. The basis is carbon-fibre reinforced polymers (CFRP), Sect. 1.3, whose lightweight design potential can be exploited much more extensively with the research results cited below as examples.

2.1 New Material Hybrids and Semi-Finished Products

The development of new lightweight materials is characterised by strong dynamics. There are currently a large number of knowledge platforms and research programmes in materials development in Germany. The potentials are manifold, as the following examples show, but only sufficiently high advantages justify the extensive qualification efforts required in aircraft design before introduction in primary structures.

Material hybrids are a combination of different material classes. The material classes referred to here are metals, fibre composites and different matrix materials. Metal-fiber hybrids are more in the performance range of metals (compare Fig. 1.2), but by combining their properties, they can be advantageous in many applications where strength and stiffness are not dimensioning factors.

Fibre-metal laminates (FML, Fig. 2.1) utilise the specific properties of metals (isotropy, ductility, electrical conductivity) in combination with those of fibres (high strength and stiffness, no fatigue, no corrosion). The best-known fibre-metal laminate is GLARE (fibreglass plies with aluminium foils) used in the A380, developed to increase the fatigue and residual strength of aluminium. In contrast to GLARE, however, the use of aluminum should be consistently avoided when using carbon fibers. Nevertheless, FMLs have many advantages for lightweight design.

Fig. 2.1

Structure of a fiber-metal laminate

For CFRP tensile specimens with inserted titanium foils, up to 30% higher bearing strength can be demonstrated compared to a pure CFRP laminate [24], and with inserted steel foils a 15% higher strength against compression load after impact (CAI) [106]. When metallic foils are inserted in the outer layers of a laminate, a 50% increase in mass-specific energy absorption can be demonstrated [21].

Further applications of fibre-metal hybrids are discussed in Sect. 3.2.

Hybridisation of FC is also possible with elastomers (ethylene propylene diene monomer: EPDM) and is useful for various applications. An FML with absorber layers of EPDM exhibits an up to 35% smaller delamination area at impact compared to pure CFRP laminate [29, 30]. Furthermore, elastomer-CFRP hybrids are very effective in the area of shape variability (morphing), Sect. 4.2.

Matrix modifications improve the properties of epoxy resins, for example. To minimise matrix-based failure mechanisms in FC such as crack growth and delamination, nanoparticles are added to the resin. Basically, the smaller the particles, the larger the specific area and the better the fracture toughness and energy release rate. Other mechanical properties can also be positively influenced with nanoparticles.

For example, boehmite (aluminum hydroxide) can increase the fracture toughness of an epoxy resin by 39% and the energy release rate by 66%, as well as the stiffness of the resin by 17% [54]. Mineral nanoparticles with 30% filler content increase the thermal conductivity by up to 40% and reduce the fire spread rate to one third [67]. Taurine-modified boehmites with 15% filler content improve not only the E-modulus, fracture toughness and energy release rate but also the fatigue behaviour of the resin [75].

Thin ply laminates can be produced by special spreading technology with basis weights of 20 g/m² to 100 g/m² compared to standard laminates with 160 g/m² to 600 g/m² (Fig. 2.2). Coupon specimens exhibit better mechanical properties, better fatigue behaviour and smaller delamination areas [13] as well as increased resistance to cracking [98]. Omega stringers made from thin-film laminates also exhibit significantly smaller delaminations than those made from standard laminates [22]. A good overview of the potentials of thin-film laminates and the challenges of their processing is given in [88].

Fig. 2.2

Thin film laminates—advantages and disadvantages

High-temperature CFRP materials offer the advantage of load-bearing capacity even at elevated temperatures, which can save heavier protective materials.

Studies conducted as part of an EU project promise up to 15% weight and up to 17% cost reduction potential and give a selection of possible matrix systems [121]. Calculations show that a stiffened CFRP shell with conventional epoxy matrix still carries 53% of the load under axial pressure at 210 °C compared to the shell at room temperature [72].

Long-fibre-reinforced polymers have lower lightweight potential compared to continuous-fibre-reinforced polymers, but can be used for secondary structures or cabin applications (cf. Fig. 1.1), and can be well adjusted in terms of their stiffness by aligning the long fibres (up to 25 mm in length). The longer the fibres, the better the load transfer between fibres through the matrix material. Using long-fibre GFRP as an example, it can be shown that with the same orientation and fibre volume content (FVC), 50% of the stiffness can also be achieved [65] compared to a composite with continuous fibres.

Recycled materials and natural fibres are becoming increasingly important in secondary structures for aircraft design. Currently, recycled carbon fibres (rCF) or offcuts from manufacturing are used in various applications. These are long fibres, which are mostly processed in the form of non-wovens with a rather low FVC of about 30%. They also allow the addition of natural fibres. Composites of rCF non-wovens and epoxy matrix with 30% FVC have about 170% of the specific strength and 70% of the specific stiffness of lightweight aluminium (AlMgZn) [15]. Pure sisal fibres show up to 22 GPa stiffness and a composite with 30% FVC of 75% flax fibers and 25% rCF shows a stiffness of 12 GPa. These and other results were recently obtained in the EU-China project ECO-COMPASS [14].

2.2 Better Structural Properties

Structures are dimensioned against failure depending on structural properties. Structural properties are often not the mechanical properties of a material or a structure (strength, stiffness, etc.) measured on specimens or (better) components, but result from multiplying them by knock-down factors (KDF < 1). A KDF compensates for uncertainties in the description of a failure mechanism, load or manufacturing deviations. Often, a mechanical property is multiplied by several KDFs. More accurate and reliable failure mechanism calculation methods and new manufacturing quality assurance methods need to be developed for future weight savings.

Structural properties for the longitudinal compression load of a FC shell after an impact on the surface (Compression After Impact—CAI structural properties) often destroy lightweight construction potentials of high-performance CFRP semi-finished products. This is due to possible delaminations as a result of the impact, which cannot be detected visually due to lack of plasticity. Today, the results of simple CAI coupon tests are directly transferred to a real structure. However, shells react differently to impact: if a thin shell is hit in the middle between longitudinal and transverse stiffeners, part of the energy is compensated elastically and does not turn into damage (delamination); if, on the other hand, the impact hits a rear-supported shell, the damage is much greater. Today’s CAI coupon tests do not take this variance into account. The elastic energy component of an impact can be determined by new methods and differentiated with respect to the CAI characteristics on component level, thus allowing effective weight savings [20]. On the basis of simple calculation approaches, different designs against impact damage can already be performed at the preliminary design level. An overview of the known approaches for the consideration of impact damage on FC shells and recommendations for simple application are given in [19].

The fatigue strength of FC is significantly better than that of metals, but the proof for a whole aircraft lifetime is not easy to provide. In order to investigate the fatigue strength of a FC, resonant test methods are required that realise a number of load cycles >10⁷ at a high frequency in a reasonable time. The test specimens must be suitably selected to avoid unrealistic boundary effects and to ensure cooling of the specimens to a constant temperature. For a GFRP-epoxy composite, fatigue strength up to 2 ⋅ 10⁷ load cycles was demonstrated in a resonant test procedure for a strain level of 1700 micro strain [74].

The permissible strains of FC are currently restricted by KDFs due to the poor detectability of damage. SHM systems (Structural Health Monitoring) make invisible damage in the structure detectable, Fig. 2.3. Piezoceramics can emit and receive guided ultrasonic waves (confined by the surfaces of the structure) over a wide area. These interact with changes in stiffness or delaminations, for example. Damage can be detected and localised by comparing the TARGET-ACTUAL sensor signals. Thanks to such systems, the KDFs for allowable strains of FC can be increased and the structure can be built thinner. Weight savings of at least 5% could be demonstrated using the example of a vertical stabiliser, taking into account the SHM system weight [27].

Fig. 2.3

Effect of guided ultrasonic waves on damage detection

2.3 New Design Concepts

In order to build lighter with CFRP, fibre-composite-compatible design methods are required. These include a load-oriented fibre placement, the use of the anisotropy of a laminate structure and the application of bonding technologies also for primary structures. The advantages of integral design and relatively free shaping should be considered early in the design process. Practically, there are some challenges in the implementation and use of such design concepts. Here, with a view to the further development of calculation tools and manufacturing technologies (digitisation), it is worthwhile re-evaluating already known concepts.

A fibre-composite suitable design saves weight and manufacturing costs. An overview of requirements and different FC fuselage concepts was developed in a LuFo project (German aeronautical research programme). For example, a sandwich (SW) skin-shell concept with integrated longitudinal stringers and continuous frames can save up to 30% compared to the current A320 fuselage, Fig. 2.4. If ‒ in a different concept ‒ the cargo compartment is located outside the pressurized cabin[55]there is a potential of 25% weight savings compared to an A320 [63].

Fig. 2.4

Integral fuselage segment consisting of longitudinal stiffeners (longerons) and sandwich panels

The surrounding structure of a passenger door in an aircraft fuselage is heavy because of the stiffening of the large cutout against shear deformations and the support of locally concentrated loads from the door fittings, and also expensive because of the complex structure. In the 7th Framework Programme of the EU, the project MAAXIMUS. [47] new possibilities of a fibre-composite, high-integrity door-frame structure (Door Surround Structure: DSS) were investigated. Considerable weight savings can be achieved in terms of shear stiffening and skin thickening in the corners of the frame [118]. An integral design method has also been developed, which allows to effectively reduce the cost in the assembly of a door frame [62].

Weight, design complexity and costs can be saved if airframe shells are designed according to the double-double (DD) laminate principle. In particular, doubler runouts can be significantly reduced, which saves weight [57].

Ultra-lightweight design for special applications, for example high-flying unmanned communication platforms, reveals the limits of what is possible. For a high-flying platform with a maximum take-off weight (MTOW) of 135 kg and a wingspan of 30 m, a construction of a tube spar in winding technology with airfoil-forming sandwich ribs and covering with a total wing area weight of 0.9 kg/m² is possible, Fig. 2.5, [116].

Fig. 2.5

Ultra-light wing structure of a high-flying platform

2.4 New Joining Technologies

The joining of fibre composite components deserves special attention in lightweight design. Process-induced deformations pose a challenge: in the case of structural bonding, the joining partners must fit together precisely over the entire contact surface. If riveted, this results in additional weight and additional costs: The CFRP shells have to be designed with additional layers e.g. weight in the area of holes for fasteners due to reduced bearing strength and the required corrosion resistante titanium rivets increase the costs.

In the case of profile-like components such as frames with angular cross sections, process-induced deformation (PID) occurs, Fig. 2.6, and, in the case of planar laminates, warping. A calculation of these deformations resulting from the thermal shrinkage of the matrix allows for compensation in the mould. Remaining residual stresses (process-induced stresses ‒ PIS) can be minimised by analysing the curing behaviour and suitable temperature control.

Fig. 2.6

Frame profiles of an aircraft fuselage without and with PID compensation

Efficient simulation methods for precise target contour setting, based on tests of representative small samples (coupons), are available [58] and demonstrated on the example of a complex CFRP structure Fig. 2.6) for the compensation of the validated PID [56]. The effect of process parameter scatter on PID and PIS can also be determined [71].

Structural bonding is still only possible to a limited extent for primary structures in aircraft design because defects in the bond cannot be detected using non-destructive testing methods (NDT). The Acceptable Means of Compliance (AMC) regulations of the aviation authority EASA (AMC 20–29), [31] require certification of an adhesive bond to be substantiated by tests for all load types and a clear and individual test of each bond against weakening (weak bonds). A process-safe and certifiable structural bond, based on a special surface activation (Fused Bonding) (Fig. 2.7) and a testing technology for the bond (Bondline Control Technology: BCT) [43], has now been developed for the assembly of primary structures [42] and vividly described in a YouTube video [44]. This process also allows for the bonding of metallic structures.

Fig. 2.7

Fused bonding ‒ detailed view of the peeling process of the activation foil

Repair bonding is subject to increased requirements as it often has to be carried out under conditions with limited control. Several processes are now available to enable adhesive repairs.

Process-reliable adhesive repair is possible with BCT using metallic meshes to test adhesive pretreatment [43]. A hybrid adhesive bond of epoxy with thermoplastic phase stops the propagation of delamination at levels up to 5000 micro strains [73].

2.5 New Manufacturing Technologies

This chapter describes new manufacturing technologies that enable the accurate and cost-effective realisation of new lightweight structures.

Continuous fibres are deposited in flat components by automated tape laying (ATL) or automated fibre placement (AFP) processes with pre-impregnated fibre layers (prepregs) or in dry deposition with subsequent resin infusion (LCM). AFP with narrow fibre tapes allows for the deposition of radii. For complex linear components in profile form with variable cross-sections, dry deposition of textile semi-finished products with injection in closed moulds (resin transfer moulding: RTM) is usually used. Pultrusion and winding processes are also used.

For dry deposited fibres, impregnation is achieved by different variants of infusion or injection [49].

Different compacting methods (pure ambient pressure, mould, autoclave) are used to generate a defined fibre volume content (FVC). Up to approx. 65% FVC, complete impregnation and enclosure of the fibre filaments with resin is possible.

Manufacturing costs are influenced by the efficiency of fibre deposition, by handling, or better avoiding of manufacturing deviations, and by minimising the consumption of auxiliary materials and energy used.

A good overview of the current status of efficient and cost-saving CFRP manufacturing technologies is given in the project report EFFEKT (LuFo V-2) [60].

Near end-shape RTM technologies (Fig. 2.8) offer high automation potential and many advantages in the production of complex components. They can be taken directly from the fixed mould cavity without time-consuming and costly postprocessing such as edge sealing. Isothermal process control allows for minimisation of the required energy input and the use of cheaper mould metals, whose thermal expansion coefficient does not have to be minimised, as is the case with expensive INVAR steels.

Fig. 2.8

RTM technologies ‒ process chain with individual processes

An overview of new sub-technologies for the efficient application of RTM and advantages in interaction with Industry 4.0 is presented in [117]. Flexible manufacturing concepts and high levels of automation in the use of RTM technology even for low volumes have also been tested [110].

A process time reduction of 60% through isothermal process control and other advantages have been demonstrated using the production of fuselage frame segments [96].

To ensure fibre placement within aerospace tolerances (0.1 mm maximum gap between deposited fibre tapes), mostly rigid gantry systems are used today in which an AFP or an ATL depositing head places fibre material on horizontal forming tools. The deposition rate is currently around 10 kg/h (AFP) to 20 kg/h (ATL) due to material storage changes, quality controls and the elimination of laydown errors.

The required tolerances can also be achieved with robots. Such a system with eight mobile and coordinated robot units, which places fibres arranged on a rail system in vertically positioned forming tools, significantly increases the deposition rate and saves space. A demonstrator plant was set up in Stade in 2013, Fig. 2.9, [7]. In addition to vertical placement, special features are realized including parallel work of several robots on one component, uninterrupted continuation of work by transferring tasks when individual robots are out of phase (for example, for changing the material storage), online quality assurance, and collision-free and event-controlled system control.

Fig. 2.9

Plant for multi-robotic continuous fibre placement at DLR Stade

The concept of this multi-robot plant with a placement rate of 150 kg/h is described in [66].

With a second, parallel placement unit, it is already possible to achieve 30% higher placement rates [26].

Path planning and collision-free distribution of the work orders to the individual robots are essential [100]. The multi-robotic fibre placement system in action can be seen here: [122].

The fibre-metal hybrid GLARE has already been mentioned (Sect. 2.1). Automation of GLARE production has been developed to reduce manufacturing costs and times. Layer by layer, glass-fibre UD layers are positioned by robots in the correct angular position, thin aluminium foils of up to 1.3 m width and 2 m length as well as adhesive films are deposited without wrinkles in the correct position, and consolidated together with stitched stringers in the autoclave. The production rate could be increased by a factor of 5 [112].

For faultless impregnation (no air inclusions), the resin must be introduced into the FC component through suitable sprue channels. Dispersions in the permeability of the dry deposited fibres result in the flow front developing individually for each component. The infusion should therefore ideally be sensor-controlled (Sect. 2.6) according to the determined resin front development, i.e., the resin channels should be “switchable”. At the same time, it is desired to avoid costs and waste and to use such switchable channels repeatedly for the impregnation of many components. The vacuum differential pressure process (Fig. 2.10) allows for selective activation of reusable distribution channels without component imprints while reducing resin consumption [48] and is described in a tutorial [28].

Fig. 2.10

Operating principle of the vacuum differential pressure process

The autoclave is the most efficient tool for generating a high FVC for large-area components. Disadvantages are the vacuum bagging, the thermal inertia, which makes targeted control difficult, limited observability of the process material during curing, long process times (up to 9 h) and high power requirement (up to 6 MW). The latter can be largely overcome by predicting the thermal processes in the autoclave during the process time by means of a coupled thermodynamic flow simulation and by using a suitable sensor system on the component for early detection of TARGET-ACTUAL deviations and to determine the degree of cure. Up to 50% time and 30% energy savings are possible by using a virtual autoclave and control via dielectric sensors to determine the degree of cure [111].

To realise lightweight design with continuous filaments in 3D printing, the fibres must be embedded in a fusible matrix material. However, thermoplastics require special impregnation processes to evenly enclose the filaments due to their high viscosity.

With a new ultrasound-based impregnation technology (Fig. 2.11), the cost of continuous fibre-reinforced thermoplastic semi-finished products for 3D printing can be reduced by 80% [109]. For learning more on the development of continuous fibre 3D printing, cf. [34].

Fig. 2.11

Fibre impregnation in a melt bath a) without and b) with ultrasonic action

2.6 New Quality Assurance Procedures

Depending on the technology and component, quality inspection of FC components can account for up to 30% of manufacturing time. Today’s quality checks can be replaced by online quality control with increasing sensor quality and computing power. Online detection of deviations and quick identification of necessary corrections save considerable time and costs. Here are just a few examples of recent developments.

A typical manufacturing deviation is fibre waviness, which occurs when co-bonding pre-cured components with not yet cured FC shells. Understanding how these deviations interact with loads and potential damage during aircraft operation helps minimise rework or select KDFs for structural properties less conservatively. How fibre waviness interacts with impact damage under subsequent longitudinal compressive loading of an FC shell can now be described and has been validated by testing [12].

Fibre placement errors in ATL can be gaps or overlaps of the fibre tapes and, in addition, twists in the deposit with AFP. Since black fibres reflect little light, optical defect detection is difficult. Currently, laser line scan sensors (LLSS) are most commonly used for optical deposit monitoring (Fig. 2.12).

Fig. 2.12

Common fibre placement defects, schematic and exemplary as LLSS image [78]

Today, the camera image of a laser scan can be suitably modelled, the quality of the LLSS signal evaluated for defect detection [77] and a classification with up to 100% hit rate can be realised [76].

If a deviation from manufacturing specifications is detected, an in-situ evaluation is required to decide directly in the process whether tolerances can be applied or corrections must be made.

The influence of a missing fibre tape during the placement of a wing shell on the stress distribution can be seen in real time by comparing a local nominal and actual FEM [45].

A possible defect in infusion processes is the retention of dry inclusions in the FC component (dry spots) due to unevenly advancing resin fronts. The resin channels attached to the component are determined by an infusion strategy based on a precise knowledge of the flow behaviour of a resin in a cavity filled with fibre material under parameters such as pressure and temperature. How the parameters of matrix viscosity, fibre permeability and infusion pressure affect the formation of dry spots in an infusion process is now well understood [18].

For a high FVC, the FC component is completely encapsulated by vacuum bags in open moulds, a vacuum is applied, and the encapsulation is additionally pressed against the component by external pressure ‒ 6 to 15 bar in the autoclave. For faultless compaction and ‒ in the case of resin infusion ‒ complete component impregnation, the vacuum bag must not contain any leaks. However, due to the size of the component and its geometric complexity, for example in the case of shell structures, leaks often occur. Fast vacuum leak detection is therefore a crucial cost factor. Using thermography and piezoelectric pressure sensors, leaks can be found automatically with 20% time savings compared to conventional methods [40].

Monitoring the flow front development of an infusion requires suitable sensors when the mould is closed. Piezoceramic ultrasonic (US) sensors are suitable for pulse-echo applications (Fig. 2.13). The density change in a fibre structure before and after impregnation with resin leads to significant amplitude jumps in the signal response. The sensors can be built very small and integrated into the mould without coming into direct contact with the resin. They can be used to determine flow front curves very accurately, flow front velocities to within 5% and flow directions to within 19% [70].

Fig. 2.13

Principle of ultrasonic sensor integration for flow front monitoring [70]

Conclusion

New material combinations and semi-finished products are undergoing dynamic development. A major hurdle for their use in aircraft design is the extensive and thus cost-intensive qualification by specially certified testing laboratories (e.g., according to NADCAP [8], [50]). Cross-industry standards and qualification programs are needed to guarantee sufficient sales volumes of new semi-finished products.

In order to be able to build significantly lighter with FC, the detectability of damage in this class of materials must be improved. With higher allowable strains in CFRP structures, primary structure weight savings of as little as 5% to 10% could be achieved.

New designs show the greatest lightweight potential, but have far-reaching implications for manufacturing concepts, system integration and in-service maintenance. Therefore, the decisions here are clearly the responsibility of the aircraft manufacturers themselves.

Adhesive bonding as a joining technology saves weight and costs. Accuracy of fit of the components and process reliability as indispensable prerequisites of adhesive bonding are now feasible.

Highly automated and industrial-scale proven manufacturing technologies are available for the economical production of weight-optimised FC structures with less energy and lower quantities of auxiliary materials. New manufacturing technologies are opening up further application possibilities for fibre composites.

Modern sensor technologies and evaluation algorithms allow for online monitoring of production and fast, automated evaluation and correction of production deviations. Today’s manufacturing costs can be significantly reduced, and component scrap avoided.