A gradient-based optimization framework that designs cementless bone implants made of triply-periodic minimal surface (TPMS) lattices with spatially varying wall thickness to maximize predicted bone in-growth. Bone growth is estimated by a finite-element mechanobiological model driven by in vivo load magnitude and frequency and surrounding bone density; the TPMS lattice is homogenized and its properties supplied by a surrogate model of local wall thickness and in-grown bone density. Wall thicknesses are parameterized over a control-point grid, and design sensitivities are computed by the direct sensitivity method. Demonstrated on a cementless hip stem subject to manufacturability wall-thickness constraints to reduce stress shielding and promote osseointegration.
Proposes a gradient-based optimization method to design bone implants built from triply-periodic minimal surfaces (TPMS) of spatially varying wall thickness in order to maximize bone in-growth. Predi…