DOI: 10.1002/adem.200800128 The Influence of Cell Shape Anisotropy on the Tensile Behavior of Open Cell Aluminum Foam** By Emiel Amsterdam, Hedde van Hoorn, Jeff Th. M. De Hosson* and Patrick R. Onck There are numerous processes for the fabrication of open- and closed cell metal foams and some of these methods result in a strong anisotropy in cell shape. The mechanical proper- ties of these specific foams, such as the stiffness, yield stress and ultimate tensile strength (UTS) (see Fig. 1), also show an- isotropy which is often linked to the cell shape. [1,2,3,4,5] Gibson and Ashby (G&A) constructed an analytical beam bending model to calculate the influence of the cell aspect ratio, i.e. the long cell axis divided by the short cell axis, on the stiffness and plastic strength. [6] Descriptions based on continuum ap- proaches have been proposed as well. [7,8,9] Obviously, for an experimental validation an accurate experimental determina- tion is important. In fact, X-ray tomography can be used to measure a large number of cells with high precision. [10,11,12] From the measured aspect ratio in the cell shape the anisotro- py in the stiffness in tension was calculated in agreement with experiments. [11] In addition, by accounting for the an- isotropy in plastic behavior the peak stress and strain have been predicted based on a critical strain criterion. [9] In these studies, however, the role of damage accumulation during straining has not been incorporated in the modeling, nor has it been quantitatively measured as a function of density. In a previous study we have shown that damage accumu- lation starts well before the peak stress has been reached and that the onset and rate of damage evolution depends critically on the base material properties and the relative density of the foam. [13] The ability of the foam to accommodate the applied strain by strut bending and reorientation was identified as an important mechanism responsible for the large peak strain at small densities. The objective of this paper is to study the effect of cell shape anisotropy looked upon in the light of these mechanisms. We start by relating the yield stress to the cell orientation at small strains, followed by a study of the damage evolution at large strain by measuring the electrical resistance in-situ. The material used is ERG Duocel open cell aluminum foam (20 PPI, alloy AA6101). The relative densities of the samples are grouped according to the following three ranges: 3–5 % (low), 6–7 % (medium) and 10–13 % (high). The relative density of each sample was calculated by measuring the mass and dividing it by the volume and the mass of pure Al (2.7 g/cm 3 ). The samples were received in the annealed (AN) condition (3 hrs at 412 °C, followed by slow cooling to 260 °C and maintaining that temperature for 0.5 hr followed by slow cooling to RT). Some of the samples were subjected to an ad- ditional precipitation hardening heat treatment (HT) to in- crease the yield stress. First the samples were exposed to a solid solution heat treatment (527 °C for 8 hrs) followed by quenching in RT water. After 16 hrs at RT the samples were artificially aged for 8 hrs at 177 °C, followed by cooling inside the furnace. Table 1 shows the mechanical properties in ten- sion of the Al alloy for the two heat treatments. [13] The samples were machined with electro-discharging into a dog bone shape. The area of the reduced section was 25 × 25 mm 2 and the length was 60 mm; the total length of the sample was 100 mm. The long axis of the cells was orient- ed longitudinal (LD) and transverse (TD) to the loading direc- tion. COMMUNICATIONS ADVANCED ENGINEERING MATERIALS 2008, 10, No. 9 © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 877 – [*] Dr. E. Amsterdam, H. van Hoorn, Prof. J. T De Hosson, Prof. P. Onck Department of Applied Physics Netherlands Institute for Metals Research and Zernike Institute for Advanced Materials University of Groningen Nijenborgh 4, 9747 AG Groningen, Netherlands E-mail: j.t.m.de.hosson@rug.nl [**] Financial support from the Foundation for Fundamental Re- search on Matter (FOM-Utrecht) and the Netherlands Insti- tute for Metals Research is gratefully acknowledged. Fig. 1. Stress-strain curve of two identical samples (AN, 6.1 %) loaded in the LD and the TD.