the struts composing their truss-like structures are typically not very straight or do not penetrate each other at the intersections as they do in truss PCMs that have ideal configurations. However, the struts are curved and pass each other within close contact, hence the use of the term ‘truss-like’ to describe their structures. Consequently, the mechanical properties of a wire-woven metal are likely to be somewhat inferior to the corresponding truss PCM. On the other hand, wire-woven metals have the potential for large size fabrication with multi-layered and fine cells through a continuous process, as mentioned above. Hence, wire-woven metals are worth considering. This paper reviews various wire-woven metals developed over the last decade, and the topologies and fabrication processes for various wire-woven metals are introduced. In addition, mechanical and thermal properties with analytic solutions, their variations, and applications are presented. Finally, the limits and prospects of wire-woven metals are discussed.

Topologies of wire-woven metals

Single-layered wire-woven metals are mainly used as low density cores of sandwich panels. With the exception of ordinary plain-woven wire meshes, the first single-layered wire-woven metal was proposed by Wallach and Gibson [27]. Fig. 3(a) depicts its configuration. It is composed of wires, bent into the shape of a triangular wave, that are arranged parallel in two orthogonal directions with a constant distance and have a pyramidal truss structure. Later, the same shape was fabricated by crimping an ordinary plain-woven wire mesh in the diagonal direction along the cross points of intersecting wires [25], as shown in Fig. 3(b).

As mentioned above, multi-layered structures with fine cells provide important advantages such as material homogeneity, high damping or absorption of vibration or sound, and resistance against face sheet buckling if used as a sandwich core. Also, their large surface area allows them to be used for heat transfer media, filters, and catalyst supports. The first multi-layered wire-woven metal was Textilecore proposed by Sypeck and Wadley [26]. In the Textile core, multiple plain woven wire meshes are stacked node-to-node and brazed to create a lattice core of cellular material, as shown in Fig. 11(a). Fig. 11(b) shows their projections in several directions, which reveals that the Textilecore can have a quite regular truss-like structure. The fabrication process is simple, and the metal has sufficient mechanical strength to permit its use inload-bearing applications [25,36]. However, Textilecore is generally too heavy (that is, the relative density is generally high) for a sandwich core, as is discussed below. Also, Textilecore has highly anisotropic mechanical properties and its strength depends heavily on the brazed joints. Fig. 12(a) depicts the Wire-woven Bulk Kagome (WBK) proposed by Lee et al. [37]. WBK is composed of helically formed wires with a constant pitch and helical radius in six different directions evenly distributed in space. Fig. 12(b) shows their projections in several directions.

In the ‘preforming’ step, wires are formed into a certain shape before being assembled or woven. Generally, wires need to be supplied in a straight shape for the weaving process. In traditional fabric weaving, a proper tension applied to yarns or threads is enough for this purpose. However, because metallic wires are often too stiff and are supplied coiled around a reel, they need to be straightened. Fig. 18(a) depicts a simple roller system used for this purpose. At least three couple rollers are employed with an application of a backward and forward tension. Generally, one of the couples acts as a wire feeder.

such as rotation and translation. In fact, the Dual Wire Octet can be fabricated in an alternative way [30]. Fig. 19 shows the process, whereby a triangular mesh is first prepared using a triaxial weaving technique such as Dow weaving [55]. Then, every second triangular hole in the mesh is deformed into hexagonal hole using an expanding device, as shown in Fig. 19(a) and (b). The mesh with hexagonal holes is then stretched in one of the three directions of the six sides of the hexagonal holes, as shown in Fig. 19(c). Finally, the mesh is crimped along the intersections, as shown in Fig. 19(d). The Dual Wired Kagome-1 can be fabricated similarly. The Zigzag core does not need any assembly or weaving process.

The assembling process for WBK, Strucwire, WBD, and WBC can be divided into two stages: inplane assembling and out-of-plane assembling. The in-plane assembling is very similar to that of the single-layered wire-woven metals, whereby in the in-plane assembling for WBK, three groups of parallel helical wires are assembled in three axes angled to each other in 60 or 120˚ at a plane in an overlapping arrangement and spin-inserting, similarly to assembly of the Circular Spring Kagome and Hexagonal Spring Kagome. Subsequently, the additional three groups of the parallel helical wires are assembled in the three out-of axes by spin-inserting. All six axes are evenly distributed in space and angled to each other at 60˚ or 120˚.

A plane weave mesh is crimped into be a pyramidal truss-like structure, as mentioned above. It is then joined with face sheets on the top and bottom to form a sandwich panel with a single-layered core. Alternatively, multiple layers are stacked peak to peak to form Textilecore. In this case, pre-crimping provides a convenient means for precise control of the cell size, especially at low relative density. Corrugation provided by crimping increases the separation of the layers and reduces the relative density. The resulting structure has a degree of anisotropy controlled by corrugation of the nodes prior to bonding [25]. In a plane weave mesh, the angle at which wires intersect each other in a layer is normally 90˚ and is readily modified after weaving by shearing the mesh. The shearing of a woven or assembled structure can be applied to any other statically unstable structures, i.e., Textilecore and WBC, mentioned above. If Textilecore or WBC is used as a core of a sandwich panel, which normally needs higher shear strength than compressive strength, it is applied in the so-called ‘diamond orientation’ or ‘X-orientation’.

Fig. 24(a) and (b) depicts these two configurations. By shearing meshes before stacking to form Textilecore or by shearing an assembled WBC, the included angle, denoted as x in the two figures, is adjusted to obtain an optimal ratio of the compressive strength to the shear strength of the core.

After weaving or assembling, the structure is fixed at the intersections between the wires by one of various methods to obtain its maximum strength. The most common fixing method is soldering or brazing, in which the filler metal normally has the least metallurgical reaction with the mother material, that is, the metallic wires. Transient liquid phase (TLP) bonding [58,59] is a relatively new bonding process used to join metals using an interlayer (a filler metal). On heating, the interlayer melts and the interlayer element (or a constituent of an alloy interlayer) diffuses into the substrate materials, causing isothermal solidification. The result of this process is a bond that has a higher melting point than the bonding temperature. In general, TLP bonding with a Ni ally filler metal is used to fix stainless steel wire-woven or assembled structures for high temperature applications. For steel wires, copper or brass is widely used as a filler metal due to its easy process and high quality. Commercially available filler metals include tin, lead, silver, lead-free, cadmium-free, sil-phos, copper, aluminum, nickel, and jewelers gold [60]. The choice of filler metal for a wire-woven metal is based not only on the metallurgy but also on the mass-productivity of the brazing process.

For mass productivity, brazing or soldering should be carried out simultaneously at all the intersections between the wires. For this purpose, two types of process can be used for applying a filler metal over a wire-woven metal. One process uses paste diluted in a liquid. Normally, a paste consists of the powder of a filler metal, flux, and polymeric binder. For example, a copper paste is mixed with water, and sprayed over a wire-woven or assembled structure, or the structure is dipped into the water– paste mixture and taken out, then dried in air. The other process involves plating a filler metal on wires. Another way to apply a filler metal before brazing was reported in which galvanic coating of Cu or chemical coating of NiP of 10 lm thickness is used over carbon steel or stainless steel wires [61]. Hence, plating a filler metal on wires before or after weaving or assembling would be a good option for a continuous process.

Brazing or soldering is completed by heating the filler metal-applied wire-woven or assembled structure in a controlled environment such as a vacuum furnace or a furnace with deoxidization gas or inert gas. Furnaces with closed chambers are preferred and continuous furnaces are often used for better mass-productivity. In cases of brazing with brass filler metal, employing an open hearth furnace with multiple gas torches after a weaving or assembling line would be desirable for large scale objects such as ship structures [62,63]. An infra-red light could be used because a wire-woven metal has a regular open and very light structure (i.e., low thermal capacity) which allows light to reach deep inside and allows rapid heating [64].

Brazing at a high temperature often induces annealing of the wires, which results in a substantial deterioration in the strength of wire-woven metals. In these cases, a post treatment is essential after brazing. Quenching and aging can be carried out on carbon steel and aluminum alloy, respectively [41]. Because a wire-woven metal has an open cell structure composed of thin wires, the post heat treatment is simply performed using a heating chamber followed by an air, water or oil sprayer on a continuous line after the brazing process. However, if, during their forming process (i.e., drawing), the strength of the metallic wire is achieved through a high level strain hardening, a substantial deterioration in strength of the wires is inevitable even with a post treatment for strength restoration such as quenching or aging. For example, although the yield strength of Cr–Si spring steel wires of d = 1 mm diameter was ro = 2110 MPa as the received state, it decreased to ro = 630 MPa after annealing at 1120 C in the de-oxidation atmosphere of the H2–N2 mixture during brazing. After quenching in oil at 50 C from 900 C and tempered at 450 C for 5 min, the yield strength recovered to only ro = 1010 MPa [41]. This is why wires of which the strength does not heavily depend on strain hardening, should be developed exclusively for wire-woven metals.

Another option of post treatment is anti-corrosion treatment by painting and plating. Because wire-woven metals have complicated 3D structures, methods driven by electrochemical reaction such as electrophoretic deposition (EPD) or electric plating are recommended for uniform coat and strong adhesion over the surface.

The four single-layered wire-woven metals before the Dual-wired octet have similar relative densities, i.e., the constants vary in a narrow range from 2.12 to 2.25. The second column in Table 4 lists equations for the relative density of multi-layered wire-woven metals. 3WEAVE is composed of closed packed wires intersecting each other in the three orthogonal directions. Its relative density is maximum, qrel = 0.589 for the case that the combination of numbers of wires intersecting each other in the three orthogonal directions is 1:1:1, as shown in the unit cell shown in Table 2. The relative density decreases somewhat for other combinations of numbers of wires; for example, qrel = 0.491 for a combination of 3:1:1. The relative densities of the other multi-layered wire-woven metals are given as functions of the ratio of d/D similarly to the single-layered wire-woven metals, where D denotes the cell size defined in the configurations of unit cells in Table 2. As for Table 3, all of the equations in Table 4 were derived assuming that the structures have ideal trusses composed of straight struts and joints with struts penetrating each other. As a result, the equations becomemore erroneous with the higher d/D, particularly for Strucwire for the same reason as that for Circular Spring Kagome mentioned above. Note that the unit cells of Textilecore and WBC are depicted for the orientations, in which they can resist best against shear loading, i.e., Diamond and X-orientation [48,67]. In contrast to the other four metals, the relative density of Textilecore is not proportional to (d/D)2, but is proportional to d/D, which is attributed to the fact that Textilecore is basically a 2D structure.

Here, c is the length of struts. The relative densities were calculated using CAD with consideration of wire curviness and brazed filler metals. The data of WBK, WBD, and WBC were sited from Lee et al. [48], and the data of the other structures were calculated for this particular article. The size of the filler metals brazed on the intersections between the wires varies depending on the geometry of the wire intersections and material properties such as viscosity of the filler metal or the wettability between the wires and filler metal at the brazing temperature. Based on observation of the real specimens, the sizes of the brazed filler metals were set to be as simple as possible (no filler metal for 3WEAVE), as depicted in Fig. 27(b). As expected by the equations in Table 4, the density of Textilecore is roughly proportional to d/c, whereas the densities of the other four structures are proportional to (d/c)2. However, the relative densities calculated using CAD for realistic geometries becomes deviated from the linearity as d/c increase, because the interference among wires becomes more prominent. Note that an upper limit is reached in the relative density of a wire-woven metal because of the interference between wires. Namely, the upper limits of relative densities of WBK, Strucwire, WBD, Textilecore, and WBC are qrel = 6.3%, 12.4%, 12.8%, 39.9%, and 42.5%, respectively.

The forth columns of Tables 3 and 4 show analytic solutions of the equivalent Young’s moduli, E, derived for unit cell models of single- and multi-layered wire-woven metals, respectively, under the same assumptions as those on which the solutions of the equivalent compressive strengths were based. In the equations, E denotes the Young’s modulus of the mother materials, i.e., the wires. The equivalent Young’s moduli of Circular Spring Kagome and Strucwire were derived by applying Carpinteri’s energy method [74] for a ring. Their equivalent Young’s moduli are proportional to the forth power of the stubbiness ratio, (d/c)4 (i.e., (d/D)4). In contrast, the moduli of Textilecore and of allthe remaining wire-woven metals are proportional to d/c (i.e., d/D) and its second power, (d/c)2 (i.e., (d/D)2), respectively.

Fig. 33 shows data of the equivalent Young’s moduli of various cellular metals according to the relative densities. In the figure, the equivalent Young’s moduli are normalized by the Young’s moduli of the mother materials and the relative densities. Similarly to Section 4.2, the experimental data for the Textilecores [67] were recalculated, and some data for the Textilecores and WBC with low densities were estimated by FEA with purely elastic material properties. Also, as conventional cellular metals, data for aluminum honeycombs [72] and foams [75] are included. In the figure, the specific equivalent Young’s moduli, E=Eqrel, are distributed over a wide range from 0.03 to 0.3. The range is much lower than the range for the specific strengths, ro=roqrel, from 0.2 to 1.5. Most specific moduli are located between the upper and lower boundaries of the aluminum honeycombs and aluminum foams, respectively. Strucwire at low densities may have lower specific moduli than some aluminum alloy foams. For Strucwire, WBD and WBK composed of significantly curved wires, the specific moduli are relatively low, but they increase with the density more rapidly than those of WBC or Textilecore. The rate of increase of specific Young’s modulus is the highest for Strucwire. Similarly to the strength mentioned above, the short and more curved struts are likely to provide more effective reinforcement by filling of the brazing metal at the joints. The specific moduli of WBC are intermediate. The Textilecore has the highest specific modulus. The specific moduli of WBC and the Textilecore are almost constant regardless of density. Note that the modulus of the partially-filled WBK is not as high as its strength is, as shown in Fig. 12, but similar to those of the Textilecores. The reason why the Textilecores have higher moduli per density than WBC (even though both have a similar structure) can be rationalized in the same way as that for the strength mentioned above.

The open topologies of truss PCMs accompanied with high surface area densities have thermal attributes that enable applications, which require simultaneous heat dissipation as well as mechanical stiffness/strength mentioned in the preceding sections. The thermal characteristics of sandwich panels with truss PCM cores, which are also called a lattice-frame material (LFM), have been examined experimentally and theoretically [15–18]. As a multifunctional structure (carrying both mechanical and thermal loads), the truss PCM thermally and hydraulically behaves in a manner similar to an array of cylinders, which has been widely used as a heat dissipation medium, due to structural resemblance between the truss and the cylinder arrays. However, the structural superiority, particularly under shear loading, of the truss PCMs over other heat dissipation media, such as the cylinder arrays, makes it unique and practically significant.

Generally, a truss PCM has a lower surface area than a stochastic foam for a given relative density. However, under forced convection, a pressure drop across the truss PCM with a typical relative density is likely to be much lower due to regularity in the structure, whilst high flow mixing occurs to promote local heat transfer from the solid ligaments to the convective flow [17]. As a result, the heat transfer performance of truss PCMs may be as high as that of foams. It has been reported that high flow mixing and low pressure drop are closely related to topological parameters such as porosity (or relative density), open flow area ratio, and surface area density.

Because wire-woven metals resemble truss PCMs in their geometry, they are similar in terms of the thermal and hydraulic behaviors. Some difference may be attributed to the fact that the ligaments of wire-woven metals have a tortuous shape and the ligaments (i.e., the wires) are joined by brazing, whereby the joints are much thicker than the corresponding truss PCMs with an ideal configuration. To date, measured data of the thermal properties of wire-woven metals are only available for Textilecore and WBK. Fig. 34(a) depicts the aluminum WBK specimen used to measure the thermal and hydraulic behavior under forced convection. Fig. 34(b) schematically shows the test setup for measuring steady state pressure drop and heat transfer in the WBK specimen. Air in ambient conditions was drawn into a rectangular channel. To impose a constant heat flux through the bottom face sheet, i.e., asymmetric heating, a heating element with AC input was used, while the other face sheet and the side-walls were thermally insulated. The air temperatures, pressure drop, and velocity were separately measured at the inlet and the outlet of the test section. The experiments were performed for two different orientations, O-A and O-B, as defined in Fig. 34(a). The technical details are given in the author’s previous papers [65,77].

As mentioned above, the curved struts in ordinary wire-woven metals inevitably result in some deterioration in mechanical properties, particularly, equivalent elastic moduli. The collinear core was developed by Wadley and his colleagues [14,67,85] to challenge the problem and even to enhanceproductivity. The collinear core, as will be discussed below, was used more often to fabricate even a lighter and stronger structure, the ‘hollow collinear core’, using hollow tubes rather than solid wires. Recently, the author’s group developed two variations of WBC [86]. Fig. 37(a)–(c) depicts the configurations of ordinary WBC, semi-WBC, and straight WBC, respectively, in X-orientation with solid sheets attached to their upper and lower faces, and show enlarged images of their respective unit cells.

maximum of the forces acting in the out-of-plane struts [88]. Therefore, in semi-WBK, straight wires are used to build out-of plane struts for higher load support, and helical wires are used to build in-plane struts for lower load support. Fig. 42 shows a sample of semi-WBK made of stainless steel wires, which reveals that the assembly is not self-supported but needs the external brace (i.e., the drilled acrylic plates in this specific case) to maintain the structure until the intersections are fixed by brazing or bonding.

As shown in Sections 4.2 and 4.3, the straight struts in out-of-plane directions substantially increase the equivalent strength and modulus of a wire-woven metal. This effect, however, would be even more significant in composite truss structures made from unidirectional fiber reinforced composite rods in the out-of-plane directions. In fact, in the author’s previous study, uni-directional pultruded carbon composite rods were used in out-of-plane directions to fabricate semi-WBK [88], which resulted in much higher strength than the ordinary composite WBK which was fabricated solely of helically formed uni-directional carbon poly-phenylene sulfide (PPS), a thermoplastic resin [89]. Microscopic observation revealed bent and twisted fibers and voids existing in the helically formed wires that significantly degraded the strength of the wires subjected to axial compression.

The term ‘‘Partially filled’’ is used when some of the cells of a wire-woven metal are regularly filled with another material. In fact, the only practical partially filled structure, known to date, is the partially- filled WBK, the configuration of which is shown in Fig. 46(a). If a WBK is idealized as an ideal 3-D Kagome truss composed of very thin and straight struts, the filler material occupies only 1/24 of the total volume, although the filler material contacts all the struts and restricts their deformations. In the realistic geometry of WBK, the volume taken by the filler metal can be even smaller because of the concave shape of tetrahedron-like cells and the volume of the wires.

A method for filling the tetrahedron cells in a practical fabrication process now needs to be determined. The best method is to use capillary action; during processes for bonding the wires after WBK is assembled from helical wires, the bonding material, such as liquid adhesive or molten metal, can be attracted into the tetrahedron-like cells by capillary action. Because each octahedral cell takes up a larger volume than the tetrahedron cells, if the viscosity, surface tension and density of the liquid adhesive or molten metal are properly controlled via a mixing ratio of adhesive chemical agents or brazing temperature, only the tetrahedral cells can be filled. The filling process could be as simple as dipping the WBK into a liquid adhesive or molten metal and then pulling it out. In fact, the partially- filled WBK was accidently discovered when too much filler metal was applied to braze the intersections among the wires in a WBK. Because a WBK with higher stubbiness ratio of struts, d/c, has tetrahedron-like cells of smaller inner volume, as shown in Fig. 46(b) of the unit cells, the liquid filler material is more likely to ‘fill’ the entire volume of only the tetrahedron-like cells, even if the capillary action of the liquid filler material is not very high.

Here, the term ‘hybrid’ means that wire-woven metal is combined with other solids in the cells. Because a truss PCM has regular open cells, they lend themselves to being filled with solids for a specific purpose. Wadley et al. reported extruded 6061T6 aluminum alloy sandwich panels with 2D triangular corrugated cores filled with alumina ceramic blocks [94,95]. In some cases, wire-woven metals are used as reinforcements of solid media, which are not regarded as hybrid wire-woven metals in this paper, but are mentioned in the following section of applications of wire-woven metals. The only case ever reported to date is the WBK-MHS hybrid [94].

Metal hollow spheres (MHSs) are hollow spheres with very thin walls that are produced using a powder metallurgical process. A feature of the MHS technology is its suitability for a wide range of materials. A broad cell size spectrum from 0.5 to 10 mm is possible whereas the cell size distribution is very narrow. The cell wall thickness can vary from about 20 lm to 1000 lm. The walls can be sintered to be fully dense as well as porous [95,96]. MHSs are known to have a good energy absorbing capability [97].

As mentioned in Section 4.2,WBKs often fail due to buckling of the struts composing the WBK with undulation of the strength after an initial peak. To enhance buckling resistance and to maximize deformation energy absorption, the ‘WBK-MHS hybrid’ was developed, in which the WBK contains MHSs in the internal space. This new cellular metal combines the periodic open cell architecture of WBK and the high energy absorption of MHS. Fig. 50(a) and (b) shows two-dimensional (2-D) schematics of Type-I and Type-II WBK-MHS hybrids, respectively. In three-dimensional (3-D) real shapes, Type-I holds the MHSs in octahedron-like spaces, and Type-II holds the MHSs in tetrahedron-like spaces.

The main application of wire-woven metals would be the low density core of sandwich panel. In contrast to honeycombs or closed cell foams, wire-woven metal cores provide an opportunity for multiple functions in addition to load bearing using the interior space. A sandwich panel is a well optimized structure with respect to mechanics, i.e., thin, dense and high strength face sheets support stress due to bending moment, while a thick low density core supports stress due to shear force, which is usually much lower than the bending stress, and provides high areal moment of inertia, which makes it possible to reduce the level of the stresses in the structure. The scheme of optimal design has been well established by Ashby et al. [2] and Wicks and Hutchinson [12] for foam-cored and truss PCM-cored sandwiches, respectively. A sandwich panel subjected to lateral loading may fail by one of the various modes shown in Fig. 53. The optimization searches the design variables such as face sheet thickness, core height and stubbiness ratio of core which provide the best performance at given constraints. For example, Fig. 54 shows an optimization scheme of the sandwich panel for the maximum load capacity at a given total weight. For optimal designs of sandwich panels with Textilecore, hollow collinear core and WBK core, refer to Zok et al. [36], Rathbun et al. [100] and Lee and Kang [101], respectively.

Fig. 55(a) and (b) shows samples of sandwich panels with Textilecores and hollow collinear cores, respectively. In each figure, the sandwich panels have cores in two different orientations. Fig. 56(a) shows 3DWT cored sandwich panels with and without face sheets. Fig. 56(b) shows photos of sandwich panels with Strucwire cores. Fig. 57(a) and (b) shows photos of sandwich panels with Dual Wired Octet and Dual Wired Kagome-2 cores, respectively. Each photo has the close view of the structure. Fig. 58 shows the photos of sandwich panels with ordinary WBK and partially-filled WBK cores. As shown in Fig. 25, WBK can be used to fabricate curved sandwich panels due to its flexibility before being brazed. The same technique can apply to most of wire-woven metals.

To the author’s knowledge, the biggest sample of sandwich panel with a truss PCM core made ever to date is a prototype of lashing bridge platform with WBK cores developed for a container carrier ship. Fig. 59(a)–(d) shows a drawing, a photo of the WBK core placed in a furnace, a stack of sandwich blocks, and the finished sample, respectively. The length and width were 5 m and 0.6 m, respectively. The sample was made of five sandwich blocks positioned in a row. For a long sandwich panel such as this sample, several separate pieces of cores need to be serially placed with a gap between the blocks. While the lashing bridge platform was being developed, the effects of the gap on the mechanical behavior of the sandwich panel under bending load were investigated [102], revealing that the strength and stiffness were often substantially deteriorated depending on the gap width between the cores and the detailed geometry.

As mentioned in Section 4.4, for a truss PCM such as wire-woven metals, under forced convection, a pressure drop across the truss PCM with a typical relative density is likely to be much lower due to the regularity in the structure than that with foams, whilst the heat transfer performance of truss PCMsmay be as good as that of foams. This is a rather desirable characteristics for heat-dissipation media. In order to quantify the performance of a heat dissipation medium under forced convection, Tian et al. [82] suggested ‘thermal efficiency index’ as defined as follows;

Vehicle braking systems transform kinetic energy into heat using the sliding friction between the brake discs and pads (Fig. 64). Previous studies have shown that the heat generation can cause brake fade [103,104], increased wear of the rubbing surfaces [105], and other problems such as boiling of thebrake fluid [106] and disc cracking [107]. Therefore, sufficient and uniform cooling of the brake disc is desirable to improve braking performance and safety. Brake discs are, therefore, often designed with two rubbing discs sandwiching a ventilating core. These ‘vented’ or ‘ventilated brake discs’ draw cooling flow into the ventilated channel when the brake disc rotates, which is a type of forced convection through the ventilated channel. In addition, these disks should resist the high stress induced by braking forces applied on the surfaces of the disks. Hence, the core is regarded as a typical example of cellular media having bi-functions, i.e., heat dissipation and mechanical load support.

Recently, a new class of ventilated brake disc with a core of wire-woven metal, WBD has been introduced [108,109]. A single-layered WBD was first assembled using helically formed steel wires (SAE1006B) with the diameter of 1.5 mm and the pitch of 19 mm. Afterwards, the assembly was sprayed with copper paste (Cubond™ grade 17LR, from SCM Metal Products, Inc.) and brazed at 1120 C in a de-oxidation atmosphere of H2–N2 mixture. The single-layered WBD core was subsequently cut into an annular shape, and ground on the upper and lower surfaces. The WBD was then sandwiched by and brazed on to two mild steel rubbing discs. Fig. 65(a) shows two samples, in which two rubbing plates are attached on both the upper and lower faces of the core, and only one rubbing plate is attached on the lower face of the core.

Oxidation catalysts such as those used in exhaust gas purifiers for automobiles are placed across gas flow to induce oxidation reaction. Catalyst material is coated on the surface of a shaped substrate (i.e., support), inducing mass transfer between the gas phase and the catalytic active sites for adsorption and desorption. A catalyst-coated cellular support with high surface area per unit volume is the most general form of oxidation catalyst system for gas. The surface area per unit volume (i.e., the specific surface area) is known as the main factor used to rate the performance of a catalytic support. However, the higher specific surface area is likely to cause higher resistance against gas flow.

The higher flow resistance results in a higher consumption of power needed to operate the catalyst system.

Therefore, flow resistance is another main factor of a catalytic support, especially, for gas at high flow rate. Typical materials for catalyst supports are ceramic and metals. In particular, cordierite monolith with numerous honeycomb channels is widely used due to its relatively low price, hydrothermal stability, and resistance to oxidation and corrosion [110,111]. However, cordierite supports are fragile and degradable under thermal cycling. Metallic monolithic supports have been proposed for their higher mechanical strength, thermal conductivity, and the possibility of thinner walls [112–114]. Especially, Fe–Cr–Al alloys form a dense and hard coating of alumina on the surface at high temperature and have good durability, mechanical resistance, and chemical resistance.

Metallic wires have been regarded as effective reinforcements for soft or brittle solids. Typical examples are metal matrix composite (MMC) and ceramic matrix composite (CMC). Sennewald et al. [117] investigated the potential of 3DWT as reinforcement for a magnesium alloy to enhancethe brittleness and low stiffness. Specifically, they used 3DWT fabricated of 0.5% carbon steel wires and magnesium die cast alloy, AZ91. They found that the effect of the reinforcement on tensile strength and modulus was insufficient because the weight increased by the addition of the steel wire structure. However, the improvement of the specific impact strength and specific energy absorption capability was noticeable. In another of their investigations, the 3DWT was combined with carbon fibers which were woven directly into the structure during the weaving process, and embedded in magnesium alloy using a gas-pressure infiltration method. It was shown that the strength and stiffness of the metal matrix composite significantly increases when combined with the hybrid woven structure. Fig. 69 shows the mold, 3DWT installed in it, and the process of the infiltration. Notice that the 3DWT was combined with carbon fibers. An additional advantage of the hybrid woven structure was the supporting effect of the steel wire structure to keep the carbon fibers in position. Fig. 70 shows an example of another MMC developed by the same research group [118]. The fractured surface exposes the Strucwire used as the reinforcement of AZ91 matrix.

Lee et al. [119] pointed out that an important benefit of wire-woven metals as reinforcements in MMC or CMC is their good dispersion. In general, the difference in densities between reinforcements and a matrix, and the technical difficulty of uniform mixing of short or long fibers in a matrix in liquid state are likely to result in non-uniform distribution, which, in turn, causes lower strength than expectation. The regular and self-supporting structure of wire-woven metals guarantees good dispersion of the reinforcements. In their study, to fabricate the Al-WBK composites, molten aluminum was pouredinto a mold in the open air with stainless steel WBK preform placed.

Seismic load is a major factor in designing structures such as high-rise buildings and public facilities. The calculation procedures for shear forces and dynamic responses due to an earthquake have been theoretically developed and widely adopted in design codes and standards around the world over the last few decades [125]. Most seismic design concepts basically allow some degrees of damage in the building structures depending on the level of the ground acceleration and the importance of a building. An alternative design concept is to employ a vibration control system to keep a building structure undamaged under an earthquake. Over the past couple of decades, many types of vibration control systems have been developed and now are regarded as one of the most promising earthquake protection systems. In general, a passive energy dissipation system, particularly a damper with a compact or planar shape to allow insertion between walls or floors, is popular, because it is relatively simple, and does not require external power supply. A damper can be replaced easily at minimum cost after the earthquake. Various dampers are available for this purpose, i.e., fluid viscous dampers, visco-elastic dampers, friction dampers, and metallic yielding dampers. These dampers have been well reviewed for their performance, characteristics, and applications by Miyamoto and Hanson [126].

Ko et al. expected that WBK withstands repeated seismic loading because WBK is composed of thin wires, and carried out a preliminary investigation on the feasibility of WBK as a metallic yielding damper [127]. They studied the hysteresis characteristics of stainless steel WBKs under cyclic shear loading through experiments and numerical simulation based on the 3-D FEA. The results showed that the WBK had a stable hysteresis response after some initial softening, and that the total energy dissipation capacity was sufficiently high for use as a vibration control device. However, the brittle intermetallics between the stainless-steel wires and Ni alloy filler metal composing WBK caused premature failure under repeated shear loading. Hwang et al. [128] investigated the responses of low carbon steel WBK dampers brazed with copper filler metal under cyclic shear loading.

Fig. 76(a) shows hysteresis loops measured from one of the WBK dampers under alternating shear displacement of triangular wave with gradually increasing amplitudes. They found that the steel WBK damper yielded a constant response up to 25 cycles of shear displacement loading which produced a constant amplitude of the equivalent shear strain, c = ±0.15, and withstood 40 cycles until the shear strength decreased to half of the initial value. Based on the data of hysteresis responses, the effect of variables such as the stubbiness ratio of the struts, material properties of wires and the external dimensions of the damper were analyzed and the optimal values to maximize the energy dissipation capacity were derived. Also, they simulated the vibrational behavior of a five story building under a series of seismic loadings. Fig. 76(b) depicts a schematic of a building with WBK dampers installed on the side walls. Fig. 76(c) is a more specific example showing how the dampers can be installed on a building.

Wire-woven metals can be applied in many other areas. The following describes concept designs and prototypes for their potential applications. Fig. 79 depicts a possible configuration of high energy heat storage module based on Strucwire and phase changing material [118,130]. Heat storages are used to compensate the time differences between the supply of heat and its demand. Because the transformation on heat into other types of energy is mostly inefficient, its storage – (assuming a later heat demand) – is more effective. Thermal storages with high volumetric energy densities are only practicable using phase changing materials such as water, paraffin, and salt hydrates. Thermal storage is based on so-called latent heat represented by the melting enthalpy of the used material at a nearly constant temperature. In the specific case of Fig. 79, the wire-woven metal, Strucwire, was used as a medium to spread heat quickly into a phase changing material, because most phase changing materials have low thermal conductivity.

Fig. 80 shows Strucwire crash elements integrated in a car buffer bar [118]. The crash elements were designed to protect passengers and drivers from crash impact by using high energy absorption of the wire-woven metal during compressive deformation. In general, an impact buffer should satisfy three requirements: sufficient strength before and after an impact, high energy absorption with a displacement as long as possible, and soft touch. Foams are likely to lack sufficient strength particularly after impact (i.e., poor residual strength). Metallic tubes are too stiff to provide soft touch. Hence, a wire-woven metal composed of ductile wires could be an idea material for this application.