Automatic image-based stress analysis by the scaled boundary polytope elements

Digital imaging technology is increasingly being applied in material, biomedical and other disciplines of engineering and science. Image-based stress analysis provides an attracting way to perform virtual testing of materials and structural analysis. Existing computational methods for stress analysis are developed with computer-aided design (CAD) models as input and are faced with difficulties, e.g. in mesh generation, when handing digital images. The paper presents a numerical technique to perform stress analysis directly from a 2D digital image segmented to its constituent materials. The quadtree algorithm is employed for mesh generation. The quadtree cells are modelled by scaled boundary polytope elements, which eliminate the issue of hanging nodes faced by standard finite elements. The whole analysis is fully automated. A numerical example is presented to demonstrate the high efficiency of the proposed technique for the analysis of high-definition images