Skip to Main Content
Skip Nav Destination

The classical dichotomy between structure and materials to meet design requirements has led to separate routes, microstructure optimization and shape optimization. The increasing demand on a combination of properties in materials, and the required multi‐functionality in present‐day design can often no longer be met with ‘classical materials’. Architectured materials are a combination of material(s), shape and topology, at length scales comparable to the scale of the components. These materials offer an alternative strategy for ‘materials by design’. The occurrence of several length scales and the central role of geometry in this new class of materials, as well as the hierarchical structure often seen in biomaterials, suggest that bio‐inspiration could be an interesting approach to the development of this innovative strategy.

From the engineering viewpoint, materials are matter with a function. In order to fulfil this function, the properties of materials are only one of the variables; the shape and the scale can also contribute to give to a component the required response to an external stimulus, which is the so‐called ‘function’. If a buoyancy device is required, a material such as cork, whose density is lower than that of water, might be sought, but a steel hollow sphere might also suffice. If a conductor that can carry electricity with little joule loss and yet will remain flexible is required, the natural solution will be to look for materials with low resistivity (such as copper) and to play on the scale (fragmenting the rod into wires) to obtain the required flexibility. In these very simple examples, it can be clearly seen that the function is related not to the property, but to a combination of materials, shape and scale. As a consequence, processes play a central role in proceeding from the ‘matter’ status to the ‘materials status’: processes allow us to obtain the shape, to control the scale and, in some materials such as metallic alloys, to obtain some properties.

Traditionally, the field of knowledge was clearly partitioned between different scientists. Optimizing properties via the chemical building block was devoted to solid state chemistry or organic chemistry, microstructure was the goal of metallurgists, macroscopic structures would be optimized by mechanical engineers and electrical engineers, and process engineers would tell how to shape a given material into a given shape. This division was based on a sort of ‘scale’ separation: the features in the microstructure controlling, for example, the yield stress were on length scales much smaller than the scales of the component, e.g. the sheet for an airplane wing. If a change to the stiffness of the wing was required, the composition (developing Al–Li alloys) or the shape (developing a whole anthology of stiffeners), or both could equally well be changed.

This ‘gentlemen's agreement’ between ‘composition/microstructure/structure’ was first challenged in engineering materials when the scale of the microstructure became commensurate with the scale of the components. This occurred at both ends of the length scale. At the macroscopic level, the use of fibre‐reinforced polymer composites, with the associations of different plies with different orientations could lead to shape‐dependent properties. At the microscopic level, the scaling down in microelectronics led to dimensions when electron scattering by the interface of the copper ‘vias’ with the dielectric and the diffusion barriers become very important: resistivity is no longer a property of the materials, but of materials at a given scale.

Now a second challenge appears with the understanding of ‘natural materials’: since they are, hierarchical, as many examples show it in this book, the decoupling between ‘structure, microstructure and composition’ becomes less and less relevant.

From an engineering perspective, the need to develop materials with conflicting properties (such as strength and toughness, or conductivity and flexibility) has led, first in a empirical way, and now in an emerging systematic manner, to the concept of ‘architectured materials’, i.e. associations of materials/shape/scales in order to fulfil multi‐objectives/multi‐constraints design requirements.1–3  The key concept is that the variability of properties (i.e. of composition and/or microstructures) occurs on length scales comparable to the dimensions of the component. If we go back to the traditional microstructure/structure dichotomy, that means developing gradients of microstructures (graded materials) and distribution of matter (hybrid materials) and all the possible variations and combinations of these two strategies. The positioning of this strategy is shown in Figure 1.1.

Figure 1.1

The different length scales and the position of architectured materials.

Figure 1.1

The different length scales and the position of architectured materials.

Close modal

To illustrate the overall strategy by using a simple example, let's consider the necessity to develop electric cables that should be flexible. Obviously, there is no single material which can be conductive enough in the longitudinal direction, insulating enough in the radial direction, and sufficiently bendable. Electrical conductivity comes with free electrons which are associated with a strong interatomic bonding which implies a high elastic modulus. Low electrical resistance imposes high conductivity and precludes the small cross section that would be required by a good flexibility. The reasoning to meet these conflicting requirements is shown schematically in Figure 1.2.

Figure 1.2

Schematic of the design reasoning leading to an ‘engineering architecture material’ and its application to the design of an electric cable.

Figure 1.2

Schematic of the design reasoning leading to an ‘engineering architecture material’ and its application to the design of an electric cable.

Close modal

The trick is then to decouple the function (longitudinal conduction and radial insulation), to select the conducting materials (metals) and the insulating materials (polymer or ceramic) and to ensure flexural elastic bendability for the conductor by fragmenting the rod into filaments, for the coating by selecting a polymer rather than a ceramic. If, in addition, high strength is required (for instance in the design of wirings for high magnetic fields for which a high magnetic pressure has to be contained), development can be carried out on a different scale – the microstructure of the conductor – to bypass the intrinsic contradiction between high conductivity (which requires high purity and few defects in the metal) and the high strength (which requires a large density of obstacles to dislocation motion). Systems such as poly‐nano‐twinned copper or hyper‐deformed copper–niobium composites provide such a compromise.4,5  An alternative solution for less critical requirements on strength (such as an electric cable which would have to sustain its own weight plus some external load (snow for instance) would lead to a mixed cable of steel/aluminium. Table 1.1 gives a few examples among many of classical conflicting requirements and their solution, and of specific engineering challenges and a possible architecture solution.

Table 1.1

Examples of multi‐objective engineering problems that can be solved by using architectured materials.

Conflicting requirementsArchitectured materials
Combining tensile strength and flexural bendability Cables 
Plate combining lightness, flexural stiffness in one direction, and bendability in the other Corrugated plates 
Combining strength and damage tolerance with respect to surface defects Soft layer on a hard substrate, such as a surface decarburised martensitic steel 
Combining strength and corrosion resistance in a light alloy Aluminium plates cladded with pure aluminium 
  • Engineering objective

    • Combining thermal conductivity and heat capacity for energy management

    • Combining high thermal conductivity and low thermal expansion coefficient

    • Combining strength, thermal insulation and small gas permeability

    • Combining strength and toughness

    • Combining deformability and magnetic strength, and possibly magnetostriction with large displacement

    • Thermally driven actuator

    • Optical transparency, strength and safety in fracture

    • Heat exchangers at minimum weight and resistant to oxidation

    • High strength cutting tool for fast machining

    • Tiles for ablation and thermal protection for aerospace reservoirs

    • Diverter for fusion reactors extracting heat and resisting to plasma ablation

    • Shock protecting helmet with comfort

 
  • Architectured solutions

    • Metallic wools or honeycomb filled with phase transformation materials

    • Mixture of metals and ceramics, copper/diamond, etc.

    • Multilayer polymer metal coatings

    • Phase transformation toughening ceramics, ceramic–polymer multilayers

    • Polymers reinforced with magnetic particles

    • Bimetallic strip

    • Glass plates with polymer films

    • Vapour deposited copper of a polymeric foam followed by cracking and surface treatment

    • Combination in a saw of a high conductivity copper core with a cermet blade

    • Graded foams in a sandwich structure

    • Copper–tungsten multi‐layers

    • Sandwich structure with a graded polyurethane foam

 
Conflicting requirementsArchitectured materials
Combining tensile strength and flexural bendability Cables 
Plate combining lightness, flexural stiffness in one direction, and bendability in the other Corrugated plates 
Combining strength and damage tolerance with respect to surface defects Soft layer on a hard substrate, such as a surface decarburised martensitic steel 
Combining strength and corrosion resistance in a light alloy Aluminium plates cladded with pure aluminium 
  • Engineering objective

    • Combining thermal conductivity and heat capacity for energy management

    • Combining high thermal conductivity and low thermal expansion coefficient

    • Combining strength, thermal insulation and small gas permeability

    • Combining strength and toughness

    • Combining deformability and magnetic strength, and possibly magnetostriction with large displacement

    • Thermally driven actuator

    • Optical transparency, strength and safety in fracture

    • Heat exchangers at minimum weight and resistant to oxidation

    • High strength cutting tool for fast machining

    • Tiles for ablation and thermal protection for aerospace reservoirs

    • Diverter for fusion reactors extracting heat and resisting to plasma ablation

    • Shock protecting helmet with comfort

 
  • Architectured solutions

    • Metallic wools or honeycomb filled with phase transformation materials

    • Mixture of metals and ceramics, copper/diamond, etc.

    • Multilayer polymer metal coatings

    • Phase transformation toughening ceramics, ceramic–polymer multilayers

    • Polymers reinforced with magnetic particles

    • Bimetallic strip

    • Glass plates with polymer films

    • Vapour deposited copper of a polymeric foam followed by cracking and surface treatment

    • Combination in a saw of a high conductivity copper core with a cermet blade

    • Graded foams in a sandwich structure

    • Copper–tungsten multi‐layers

    • Sandwich structure with a graded polyurethane foam

 

The application of the ‘architectured materials strategy’ is not always as trivial as the model example of the electric cable, and the search for a ‘materials‐by‐design route’, going beyond optimizing the elementary cell (as do solid‐state chemists) and/or the microstructure (as do metallurgists), often requires modelling materials with their architecture in order to identify a promising solution before going to the trouble of processing it. The two examples which follow illustrate this approach for an acoustic absorber and for a radiant burner. In both cases a fluid has to pass through the material. For the acoustic absorption, fluid shear will dissipate the pressure waves. For the radiant burner, the gas should be such that the flame front resulting from its combustion stays within the materials to generate radiative power. In both cases the set of requirements immediately drive the engineer toward cellular materials, where the degrees of freedom are the constitutive materials and the characteristics (scale and porosity) of the architecture. In both cases, modelling is crucial to evaluate a priori the potentially promising architectures, and materials selection methods using the Cambridge Engineering Selector (CES)6  database are used to select the appropriate constitutive materials.

The sound absorber for an airplane engine should dampen a spectrum of noise (both background noise and peaks) characteristic of the engine.7  It has to operate at high temperatures, which are different depending on the position with respect to the combustion chamber. The main objective is maximum acoustic absorption for a given spectrum of noise. Although mechanical properties are not the major concern, it still requires some mechanical integrity in order to be a ‘stand‐alone’ component; otherwise the additional device to hold it would increase the total weight. It requires therefore both sufficient stiffness and creep resistance. Since the materials must be porous to present an acoustic impedance compatible with air (otherwise the sound waves would be reflected instead of being absorbed) an open porosity is needed. But for a given quantity of matter in a porous structure, close porosity with shell‐like elements is preferable. This simple reasoning leads to a class of porous materials, the stacking of hollow spheres as shown in Figure 1.3. Once the geometry is chosen, the free variables of the problem are the constitutive materials, and the sphere shell inner and outer radii. The operating temperature requires a high‐temperature resistant alloy as a constitutive material. Computing by finite elements the elastic properties and the plastic yielding of the architectured material and searching for the minimum weight will lead to the specific properties (properties normalized by the density) as ‘performance indices’. Two families of materials emerge from this analysis: nickel‐based superalloys and austenitic stainless steels. The choice of these families also prescribes the ratio of thickness over sphere radius. The detailed choice between the different members of these two broad families depends on the detail of the creep strength, the oxidation resistance, the possible surface treatments, and the brazing methods to stick the spheres together.

Figure 1.3

Acoustic absorber made from (left) a regular, and (right) an irregular stacking of metallic hollow spheres.

Figure 1.3

Acoustic absorber made from (left) a regular, and (right) an irregular stacking of metallic hollow spheres.

Close modal

The acoustic properties are a dynamic fluid mechanics problem.8  The dissipation in the shell is negligible as can be shown experimentally by comparing a stacking of hollow spheres and a stacking of dense spheres. The acoustic properties are therefore controlled by the external geometry of the channels between spheres. In addition, the acoustic absorber is the locus of standing acoustic waves: from that appear a second variable, the thickness of the acoustic layer. The two variables are the sphere external radius, R, and the absorbing layer thickness, L. A simulation based on homogenization techniques allows the real geometric properties to be related to the phenomenological description of the porous medium (porosity, tortuosity, etc.) and therefore to provide a simple relation with the acoustic properties for a given frequency (Biot–Alard model).7  The next step is to optimize the absorption coefficient normalized by the absorbant thickness, with respect with the two variables R and L. The results of this optimization for two ‘simple spectra’ are given in Figure 1.4.

Figure 1.4

Computation of the absorption factor for two different noise spectra. The horizontal axis is the external radius of the spheres, R; the vertical axis is the thickness, L, of the absorbing layer. From blue to red the absorption increases.

Figure 1.4

Computation of the absorption factor for two different noise spectra. The horizontal axis is the external radius of the spheres, R; the vertical axis is the thickness, L, of the absorbing layer. From blue to red the absorption increases.

Close modal

This qualitative description of the whole strategy (for more details see Gasser et al.7 ) shows the ingredients which are typical of an ‘architectured materials approach’: geometry identification, materials selection, and modelling or simulations of the properties as function of both the constitutive material(s) and the architecture dimensions. This allows the identification, in a continuous and wide class of materials, of the promising ones, and then further studies have to be performed on these with regard to the best methods for processing them, protecting them against shocks and oxidation, and determining their fatigue behaviour, etc. These properties are indeed crucial in bringing the new ‘architectured material’ from concept to engineering reality, but the approach outlined has narrowed down the range of possibilities to those that are potentially interesting and for which these difficult issues are worth exploring. In that sense, we can say that we have applied a ‘materials by design’ approach.

In the previous example, the competence required was on solid and fluid mechanics. The modelling tools were both numerical simulations and analytical methods. Developing architectured materials is intrinsically a cross‐disciplinary exercise, as illustrated in the following example. Radiant burners are a classical way of producing heat via the combustion of a gas, making sure that the flame front is positioned within the porous material in order to maximize heat radiation9  (Figure 1.5). Again, an architectured porous material is the obvious solution.

Figure 1.5

Principle and example of a radiant burner.

Figure 1.5

Principle and example of a radiant burner.

Close modal

The objective is, of course, to maximize the efficiency for a given influx of fluid. But ‘green design’ also requires minimizing the production of CO and NOx due to incomplete combustion. The free variables are the constitutive materials (which limit the operating temperature, but will also enter through its optical emittance and its thermal conductivity), and also the architectural characteristic of the porous material (porosity, size of the pores etc.). The modelling tool will be fluid mechanics (to characterize the fluid velocity), combustion theory (to make sure that the flame front stands inside the burner), thermal physics (to compute the temperature distribution for a given position and intensity of the flame front), optical physics (to compute the radiative power of a porous solid as function of the constitutive materials and the geometry, and of course chemistry to calculate the heat generated by combustion and the rate of completion which controls emission of CO and NOx. The detailed calculations can be found in Randrianalisoa et al.9 

The results are shown in Figure 1.6. The impact of materials choice on both ‘pollution’ and ‘efficiency’ is clear. For a given porosity and cell size, the best choice would be FeCrAlY alloys to minimize pollution, and mullite to maximize efficiency. In a ‘green design’ approach, the choice is not obvious: should we insist on pollution control, or on minimizing the consumption of non‐renewable materials?

Figure 1.6

(Left) NOx and CO emission for different constitutive materials, with a porosity of 90% and a cell size of 1.2 mm; (right) radiated power for similar conditions. The radiant burner is dimensioned for a prescribed energy influx.

Figure 1.6

(Left) NOx and CO emission for different constitutive materials, with a porosity of 90% and a cell size of 1.2 mm; (right) radiated power for similar conditions. The radiant burner is dimensioned for a prescribed energy influx.

Close modal

The previous examples show that ‘engineering architectured materials’ can provide efficient solutions for complex set of requirements for which ‘classical materials’ are not well suited. As seen in many chapters in this book, natural materials are almost, by necessity, architectured.10  They combine a limited number of building blocks, since the operating conditions for life require water and mild temperatures: only polymers and ionic solid solutions can provide processing routes under such conditions.11  The variability of properties required by the variability of functions imposes the use of materials associations and architectures to provide this range of properties with such a limited variety of building blocks. In addition, natural materials are ‘growing’, and this growth is provided via living cells. This explains the ubiquity of layered and fibrous structures in natural materials.12  Transport of matter is mainly dominated by fluid transport, which requires a sort of porosity: that is a possible reason for the frequently observed cellular solids.13  The net result is that natural materials are an association of materials, at different scales, with different shapes, and often present in addition to a distribution of matter at scales comparable to that of the organ.

The parallel with the ‘engineering architectured materials’ is then obvious. Clear also is the engineering value of a ‘biomimetic’, or rather a ‘bio‐inspired’ approach. Association of the large variety of engineering materials (with metals, ceramics and polymers, and their combinations in composite materials, there are about 100 000 of them) with the variety of architectures observed in nature, one can expect to have a very efficient strategy to develop new ‘architectured materials’ enabling the engineer to ‘fill some holes’ in materials space.3  The most efficient way to implement this strategy is to start from a specific set of requirements in terms of ‘function/constraint/objectives’. This set of requirements is then translated into a ‘property/shape’ strategy, via the now classical ‘performance index method’.6  Of course, in any realistic industrial component, the set of requirements is multi‐constraint, multi‐objective and more and more often multi‐functional. This results in searching for combinations of properties which are physically incompatible in ‘classical materials’, such as ‘strength/toughness’, ‘electrical conductivity/flexibility’, ‘stiffness/vibration damping’. Many engineering requirements have no equivalent in nature, so it would be a poor understanding of bio‐inspiration to limit its approach to realize by engineering processes combinations of functions that nature does well. We would probably conclude that wood is by far the best material to make a tree! More interesting is the approach in terms of ‘generic contradictions’: what are the architectures that increase strength, at the same time as increasing flexibility, or increasing vibration damping capacity? In addition, the strategy of ‘architectured materials’ involves, almost by definition, interfaces between very different materials. These interfaces are regions of mechanical incompatibilities and are known to be sites for damage nucleation and accumulation. Even in engineering materials, weak interfaces are already used for crack deflection (for instance in ceramic/ceramics composites). But the optimal design of interfacial strength, and of the width and property gradients around the interface is a key issue in developing architectured materials. Bio‐inspiration, guided by the clear identification of the contradictions to be met, can be a precious source of innovation, both for the design of the appropriate architecture, and for the design of interfaces (Figure 1.7).

Figure 1.7

The reasoning behind architectured materials and the possible role of bio‐inspiration: architecture selection and interface engineering.

Figure 1.7

The reasoning behind architectured materials and the possible role of bio‐inspiration: architecture selection and interface engineering.

Close modal

In spite of various attempts (such as the TRIZ methodology14 ), inspiration is by definition non‐systematic, and analogy provides germs of ideas more often than solutions. Let's proceed in the following way:

  1. Start from some engineering challenges leading to conflicting requirements.

  2. Imagine possible situations in nature where similar conflicts might appear, and analyse the bio‐architecture developed.

  3. Propose an engineering solution stimulated by this excursion into nature.

In the last sections of this chapter we will address the question of the process necessary to implement these innovative solutions.

This is one of the oldest nightmares of engineers: high‐strength materials show little plasticity, thus little dissipation in the vicinity of cracks and almost no blunting at the crack tip. They have therefore a low toughness and poor damage tolerance. That is often a problem, for instance in landing devices which require high‐strength materials, but are ready to compromise (meaning to increase the necessary weight) in order to keep some ductility. In such materials, fracture is governed by the propagation of a single crack: it is a toughness‐controlled ductility. By contrast, in ductile materials cavity nucleation, growth and coalescence govern ductility. Of course the first engineering solution is to eliminate hard inclusions which may cause an incipient crack which can become rapidly fatal. But when, due to alloy cleanliness, all dangerous inclusions have been removed, there are still the surface defects, surface scratches which can be hard to eliminate on large components (electro‐polishing a whole landing device may be quite challenging!). How can we reduce the harm of these surface defects without losing the strength necessary for the function? In nature there are examples of such situations where an intrinsically brittle material is toughened by an architectural solution. The enamel of teeth is very hard (in order to be wear resistant), but a crack in it rarely propagates due to the soft layer of dentine existing below.15  Both the gradient in elastic properties and the gradient in dissipation act as ‘crack stoppers’. Inspired by this observation, we can think of stopping the surface cracks via a surface layer of soft materials. The bulk material is a high‐strength martensitic steel, the surface layer can be a thin low‐carbon ferritic steel … and it is quite easy to obtain (see below).

There is a simple and elegant solution to part of this problem: a sandwich structure with steel skins and polymeric core. Of course it is not easy to implement it ‘for true’; specific polymers have been developed to survive the heat treatments during further painting, special surface treatments are necessary to provide polymer/metal adhesion, but these solutions are affordable and are used in domestic appliances where vibrations might be annoying (such as a washing machine). Car manufacturing is also in demand of such devices, but with an additional requirement: weldability. And welding a polymer/steel sandwich is an insuperable challenge, and the car industry is not ready to change for gluing structural components. Exit then the metal/polymer sandwich. Weldability requires a whole metal solution, which is flexible (in order to be stamped), which is rigid after stamping, but still retains some dissipation processes enabling vibrational damping. Birds’ nests are natural examples that need to be both stiff and comfortable at the same time (besides providing thermal insulation) and use entangled materials made up of different constitutive materials. In parallel to this natural example one can think of a metallic core for a steel‐faced sandwich made from steel wool.16  The weldability is no longer a problem. The stiffness of the wool comes from the entanglement of the fibres. Stamping the sandwich is easy since a large elastic deformation at a moderate stress is possible and the final shape will be imposed by the deformed skin. Further stiffening of the wool can be obtained by sintering, whose effect is to cross‐link the fibres.17  A fully cross‐linked steel wool will have the properties of a steel foam, and will be stiff but with little vibration damping. Partial sintering will allow for an attractive combination between stiffness and vibration damping.

In applications such as the protective coating of a blast furnace, ceramics are the only materials which can sustain such elevated temperatures. Bulk ceramics are far too brittle, and damage tolerance is basically nil. That forces the engineer to consider tiles and bricks instead: as the volume of each is smaller, the likelihood of having a crack in one of them is lower, and if one element is cracked, the crack will not propagate into the neighbouring ones. What remains is the question of how to assemble the tiles together. Metallic nails are excluded, adhesion can be obtained by special cements, which deteriorate faster than the tiles themselves and limit the duration of the coating. Preferably, the tiles need to be assembled as in a jigsaw puzzle. In nature, that's the solution found in the suture of the red‐slider turtle.18  The carapace is formed from ‘interlocked’ elements whose geometry prevents extensive sliding in all of the three directions in space. In addition, the structure has flexibility. Similarly, the geometry of the tiles for the blast furnace can be designed such that their assembly, conveniently confined by a frame, makes a so‐called ‘interlocked material’.19,20  Such a structure has, in addition, the peculiarity of a ‘tunable stiffness’ depending on the stress imposed on the frame. To conclude this example, it can be seen that the variety of such ‘interlocked materials’ is almost infinite: materials, shape, surface friction, and clamping conditions are the variables to be explored.21–23  It is out of question to explore them without any guidance; that's where modelling comes into play. A Finite Element Method (FEM) calculation can be used directly on a complete interlocked structure, but having to deal with multiple sliding interfaces with solid friction is extremely demanding on computer time. A very convenient method to bypass this difficulty is to develop a ‘hybrid model’ (see Figure 1.8) where the interaction between two blocks is computed by an FEM calculation, whereas the collective behaviour of a block assembly is computed via a discrete element method (DEM, the macroscopic equivalent of molecular dynamics) with the interaction forces calculated above.24,25  This DEM+FEM approach may also be useful in understanding biomaterials where building blocks can be clearly identified.

Figure 1.8

Principle of a ‘hybrid simulation’ of an interlocked material (a): the interaction between blocks computed by FEM (b) is used as the input for a DEM simulation (c).

Figure 1.8

Principle of a ‘hybrid simulation’ of an interlocked material (a): the interaction between blocks computed by FEM (b) is used as the input for a DEM simulation (c).

Close modal

The production of architectured materials requires the development of new processing routes or important changes in conventional routes (Figure 1.9). The classical way is to propose a ‘bottom‐up’ process, and to mimic by ‘soft chemistry’ the route followed by nature. The studies on biomineralization follow this line of thought. We want here to suggest an alternative route worth exploring when the aim is to produce large quantities of materials cheaply. The purpose here is not to be exhaustive, but to stimulate by a few examples the potential of ‘classical’ industrial processes to create architecture materials, bio‐inspired or not.

Figure 1.9

The engineering challenge of architectured materials: the role of processing.

Figure 1.9

The engineering challenge of architectured materials: the role of processing.

Close modal

In order to do so, it is necessary to start from ‘macroscopic components’, and to create the architecture. Multi‐filament extrusion to large strain can provide a hierarchically architectured Cu–Nb cables which are currently used for high‐strength conducting wires needed for high‐field pulse electromagnets.5  Co‐casting using the solidification of an alloy serving as mould for another alloy is a simple and elegant way to develop graded sheets, and further rolling can lead to co‐laminated multi‐layers with good interfacial properties. An efficient way to develop heat exchangers is to create a polymer foam with open cells, to vapour deposit a metal and to burn the polymer in order to obtain a set of interconnected hollow struts with a high exchange surface. The metallic wools or felts discussed above are obtained by scratching rods of stainless steels followed by compaction of the chips. A variable stiffness can be obtained by a variable sintering treatment.26  The high‐strength carbon steel with an acceptable ductility can be obtained by decarburization in the austenitic regime and creation of a thin soft ferritic layer above the martensitic structure following the quench.27 

However, these ‘macroscale processes’, as most processes in structural applications, are based on deformation and heat treatments; they require a combination of materials that have seminal deformability, or are both compatible with similar temperature ranges. One can think, in certain conditions, of co‐deformations of metals and polymers (multi‐layers of aluminium polymers providing vapour proofness to vacuum super‐insulators are often processed that way). It is much more difficult with metal–ceramic couples. Heat treatments on metals (including chemical modifications by surface alloying) are hardly compatible with coexisting metals and polymers. Processing routes coming from surface engineering are an alternative, using vapour deposition which allows for layers of materials on any type of substrate. The processing rate is not unreasonable and industrial materials can be produced that way.

As can be seen, there is good hope that engineering architecture materials can be produced in an affordable way using standard processing techniques. The central issue on which very little has been done, and which will certainly be a key to future developments, is the question of interface engineering, proving ad libitum, strong or weak interfaces, high or low transfer coefficients.

Of course, with respect to architectured materials, it is possible to think of microstructure as an additional way of modifying the ‘local constitutive materials’. In principle, nothing prevents the heat treatment of a metallic foam made from a heat treatable alloy but, obviously, due to the thermal exchange imposed by the geometry, the optimization of heat treatment parameters does not have to be similar to those for the bulk constitutive materials. Beside the ‘interface engineering challenge’, designing process routes that allow simultaneously benefit from architecture optimization and microstructure control will be one on the main challenges to be overcome.

This chapter has outlined a possible route to ‘fill the holes’ in materials selection maps. Starting from a complex set of requirements, which cannot be efficiently reached by ‘on the shelf materials’, the functions can be decoupled, relevant architectures imagined, appropriate materials selected by using performance indices, and modelling used to predict the relation between, on the one hand, constitutive materials properties and characteristics of the architecture, and on the other hand the macroscopic properties of interest. These ‘architecture materials’ are ideal candidates for a ‘materials by design approach’. New architectures can be inspired by observing biological hierarchical structures, provided the ‘conflicting properties’ corresponding to the function is clearly identified. Finally, one of the main challenges is to design affordable processes allowing mass production of these objects. This is a necessary step to take from the demonstrator to the realistic engineering devices. Multi‐functional cables or deep ocean flexible pipes are real engineering devices, they have shown the way, they are just the beginning of a new family of materials, of a new way to think beyond the ‘materials/structure’ dichotomy, and bio‐inspiration may provide many others in the coming years.

This chapter was written during a research stay at the Max Planck Institute of Colloids and Interfaces supported by the Gay‐Lussac Humboldt Prize.

Close Modal

or Create an Account

Close Modal
Close Modal