Analysis of an open foam generated from computerized tomography scans of physical foam samples. (23rd May 2022)

• Record Type: Journal Article
• Authors: Leblanc, Christophe; Kilingar, Nanda Gopala; Jung, Anne; Kamel, Karim Ehab Moustafa; Massart, Thierry Jacques; Noels, Ludovic; Béchet, Eric
• Is Part Of: International journal for numerical methods in engineering. Volume 123:Number 18(2022)
• Page Start: 4267
• Page End: 4295
• Publication Date: 2022-05-23
• Subjects: boundary representation -- finite element methods -- foam -- material science -- tomography; Numerical analysis -- Periodicals; Engineering mathematics -- Periodicals; 620.001518
• DOI: 10.1002/nme.7008
• Languages: English
• ISSNs: 0029-5981

Abstract

Abstract: The present article introduces an automated procedure to construct geometrical representative volume elements (RVE) of open‐foam cellular materials from computerized tomography (CT) images, with the final aim of generating meshable geometries usable in the finite element method (FEM) used in order to analyse their mechanical behavior. The methodology consists in growing and fitting a set of ellipsoids to each of the foam cells. These ellipsoids are seeded by local maxima of the distance to the struts obtained from computer tomography images. This methodology is thus fully voxel‐based and does not depend on any assumption about statistical distributions of the foam cells. Therefore, it is able to reproduce an accurate geometrical model of the foam's microstructure and its possible irregularities. Moreover, this procedure allows the processing of large 3D data sets that do not fit the random access memory (RAM) by slicing it into smaller independent chunks. The effectiveness of the proposed approach is illustrated by comparing it to FEM simulations for which meshes are obtained from a feature reconstruction approach. Both FEM simulations are then compared with experimental results of uniaxial compressions of an open foam.