Adaptive anisotropic porous structure design and modeling for 2.5D mechanical parts

As displaying many advantages including high specific strength, light weight, energy absorption, etc., porous structures have been widely used in aerospace, medical science, engineering and other fields. In this paper, an adaptive anisotropic porous modeling method is proposed for improving the mechanical performance of 2.5D mechanical parts whilst minimizing weight and maximizing product benefits. The method relies on the Anisotropic Centroidal Voronoi Tessellations (ACVTs) by offsetting closed B-spline curves. To enhance adaptability to a given parts, the finite element analysis (FEA) results of stress field is combined with stress-based weighted random sampling strategy to generate ACVTs according to Riemannian metric. A skin frame is established to improve geometrical quality of porous structure and ACVTs is tailored to be adaptive for concave and Non-zero genus. Besides, a parametric model between the stress tensor distribution and the relative density field is formulated, allowing the size, the distribution and the shape of pores to be controlled by stress mapping. Both the FEA results and the experimental data show that the adaptive anisotropic structures have significant advantages in mechanical properties, topological consistency and connectivity. Additionally, an example of wrench is provided to prove the great potential of the adaptive anisotropic structures in engineering applications.