A Perlin-Noise-Based Method for Multiscale Bio-Inspired Porous Materials

  • Yuan, Hand (Cardiff University)
  • Li, Zeyang (Cardiff University)
  • Lou, Chenxukun (Cardiff University)
  • Wu, Zhangming (Cardiff University)

Please login to view abstract download link

A Perlin-noise-based framework is proposed for the generation of porous microstructures by controlling the vector distributions at the vertices of minimum control units and extending them throughout a representative volume element (RVE). Within a unified formulation, the method enables the construction of several classical porous architectures with distinct mechanical characteristics, including isotropic, cubic, lamellar, and columnar configurations. Numerical homogenization is performed to systematically evaluate the effective elastic properties and anisotropic behavior of these structures. The resulting homogenized responses are further incorporated into a multiscale topology optimization framework for the design of a dual-scale MBB beam. In addition, the inherent local tunability of the Perlin noise approach allows spatially heterogeneous porous architectures to be generated within a single RVE without the need for additional stitching or transition procedures. Finite element simulations under impact loading demonstrate that localized structural modulation leads to distinct deformation mechanisms and macroscopic mechanical responses. These results indicate that the proposed framework provides a flexible and efficient tool for porous structure design in multiscale mechanical applications.