Experimental and numerical analysis of scatter and anisotropy in closed-cell all-aluminum sandwich core materialThursday (08.10.2020) 16:30 - 16:50
In lightweight design sandwich structures are widely used components. For recycling reasons all-aluminum sandwich panels are advantageous because core and sheets consist of the same material. For the structural integrity analysis, the heterogeneous cellular microstructure is a challenge. Uncertainties in the cell size, shape and orientation implicate a pronounced scatter and uncertainty in the macroscopic material properties. In metallic foams this effect is even more important due to the limited number of cells in the thickness direction.
A sample batch of foam core material is characterized experimentally in a probabilistic manner, using a large number of small size specimens. In order to interrelate position, microstructure and properties, the original position and orientation is documented together with a photographic documentation of the microstructure in each specimen. Material scatter and anisotropy effects due to the manufacturing process are investigated. The material characterization is performed in three directions to analyse the influence of the loading direction in relation to the orientation of the microstructure on the effective material behaviour.
For numerical investigations an algorithm for generation of cellular structures consisting of stretched respectively truncated cells with completely random or preferred spatial orientation is developed. The process is an extension of the well-known Voronoï process in Laguerre geometry in which the packing of spheres is substituted with ellipsoids. The computation of the cellular structure is mathematically much more complex and time expensive since the curvature of ellipsoidal surfaces is a function of the position. The benefit is that structures consisting of cells which deviate from an optimum surface to volume ratio as observable in reality can be generated. In a subsequent homogenization analysis, the influence of the cell size, cell shape and orientation on the effective material behaviour is analysed numerically analysed. Numerical predictions and experimental data are found in a good agreement.