An image-based multi-level hp FCM for predicting elastoplastic behavior of imperfect lattice structure by SLM

In metallic additive manufacturing (AM), the process-induced geometrical defects challenge the boundary-conforming mesh-based numerical methods for predicting the as-fabricated mechanical properties of the parts with geometrical deviation from their as-designed counterparts. In this study, a computed tomography (CT) image-based multi-level hp finite cell method (FCM) is proposed to predict elastoplastic behaviors of the imperfect 3D lattice structure by the selective laser melting AM. The voxel data of the CT scans of the build are used to generate the locally refined structured mesh with the composed quadrature points utilizing the refine-by-superposition concept to achieve h and p refinements. Therefore, the struts’ external and internal geometrical defects can be efficiently resolved in the FCM discretization model for an accurate prediction of the mechanical behaviors of the builds. The numerical results obtained by the proposed method have shown a good agreement with the experimental data. It is identified that the effect of external geometrical defects on elastoplastic responses of the AM products is significant, while the effect of the internal voids is relatively minor, partly because of their low volume fraction.