Automatic 3D image based finite element modelling for metallic foams and accuracy verification of digital volume correlation

Abstract

3D image-based finite element (FE) modelling provides the feasibility to explore how the subtle structural features inside metallic foams affect their complex deformation and mechanical responses. But the modelling currently includes frequent manual modifications, which makes it a time and labor consuming task. The quality of generated model depends heavily on experienced creator. This paper demonstrates an automatic procedure to extract structure information from 3D images and generate reliable geometric models for meshing. With self-adaptive algorithms that fit local features of grayscale distribution and internal structure, high-fidelity FE models can be established efficiently without interference from subjective judgments. The generated FE models well reproduce the stress-strain relation and inhomogeneous deformation of aluminium foams, compared with experimental results. Using the simulation with a high-fidelity FE model as “ground truth”, the accuracy of digital volume correlation (DVC) in the deformation measurement of complex cellular materials is for the first time evaluated quantitatively and precisely, based on the definition of an equivalent displacement vector bridging the results obtained by FE simulation and by DVC.