Back to Dr Markaki's page

Layered Metal Foam / Ceramic Composites

The project was concerned with the deformation and fracture behaviour of laminates made up of alternate layers of metallic foam and ceramic. However, before treating the deformation behaviour of these laminates, an understanding of the mechanisms by which the foams themselves fail was needed. This, in turn, required an assessment of the role of the cell morphology and cell wall microstructure in governing the foam properties and behaviour.

Commercial metallic foams invariably contain imperfections and inhomogeneities in their cell structure and, as a direct consequence of this, their properties are markedly inferior to those predicted for periodic, defect-free cellular materials. Also, the nature of the microstructure within the cell walls can have a significant effect on their performance. Cell walls often contain high volume fractions of coarse eutectic, fine oxide films and large brittle particles, which can affect both local and global plastic deformation characteristics.

It has become clear that optimisation of the mechanical performance of metallic foams requires an understanding of the interplay between processing conditions, cell structure, cell wall microstructure and mechanical response characteristics under various types of applied load.

Fig.1 Backscattered scanning electron micrographs showing the cell wall microstructure of Alporas foam. The cell walls consist of Al dendrites (indicated A) and a network of Al-Ca-Ti eutectic (B and C), with incorporated Al-Ca-Ti-Fe particles (D). The darker areas in the eutectic domain correspond to regions with lower concentrations of titanium (region B).

For more information see:

The Effect of Cell Wall Microstructure on the Deformation and Fracture of Aluminium-Based Foams. Markaki, A.E. and Clyne, T.W., Acta Mater. (2001), Vol. 49, Issue 9, pp. 1677-1686.

In-plane tensile and flexural loading has been carried out in order to explore the fracture response and measure the fracture energy of the layered metal foam / ceramic composites. Comparisons were made between measured fracture energies in bending and simple model predictions based on the deformation behaviour of bridging ligaments under small-scale yielding conditions. It was found that part of the energy is absorbed by plastic deformation of the metallic constituent but there is some evidence that another process, such as frictional sliding between the crack flanks, may also have made a significant contribution.

Inherent limitations in the mechanical performance of the foam laminates are particularly relating to the effects of low triaxial constraint and difficulties in generating multiple cracking. The effective yield strength of constrained foam layers has been measured to be very close to the unconstrained uniaxial yield stress. This is a consequence of the limited degree to which constraint can be imposed on metallic foams, which arises from relaxation of the normal stress at all the internal free surfaces (pores). Multiple cracking, which can be stimulated in laminates made from dense metallic layers, and can have significant contribution to the toughness, has not been observed with foam laminates. This is consistent with model predictions, which indicate that the strength of the foam would have to be raised substantially in order for multiple cracking to occur.

Fig. 2 Plot of the ratio of bridging metal stress to ceramic strength against the metal / ceramic layer thickness ratio. The solid curve of the LEFM model (Huang and Zhang, 1995) separates the single cracking regime from that of multiple cracking. Experimental data obtained from dense and foamed metal / ceramic laminates are also shown.

For more information see:

Energy Absorption During Failure of Layered Metal Foam / Ceramic Laminates. Markaki, A.E. and Clyne, T.W., Mat. Sci. & Eng. A, vol.323, pp.260-269, (2002).

The laminate response to localised deformation has been investigated in the low and intermediate velocity regimes. It was found that there is good correlation between low velocity impact and quasi-static responses. In both cases, penetration of the layered targets resulted characteristically in the formation of a distinctive plug. Increasing impact velocity (intermediate velocity range), plug formation was inhibited and the penetration was dominated by a fragmentation process. Transition in the penetration mode gave rise to a substantial increase in the absorbed energy.

Fig. 3 Energy absorption per unit volume versus impact velocity for impacts on foam laminates.

Results on impacts at velocities in the intermediate range suggest that foam laminates can absorb more energy per unit volume or mass of damaged material than (dense) metal laminates of equivalent areal densities.

Fig. 4 Energy absorption per unit mass for both foam and dense metal laminates impacted at about 200 m s-1, using a gas gun.

For more information see: Characterisation of Impact Response of Metallic Foam / Ceramic Laminates, A.E. Markaki and T.W. Clyne (2000), Mat. Sci. & Tech., 16 (7-8), p 785-791.
Back to Dr Markaki's page