Adjoint lattice Boltzmann method-based multiscale topology optimization of thermo-fluid systems using anisotropic porous media

The escalating thermal management demands in advanced electronic systems necessitate innovative cooling solutions that optimize heat transfer efficiency. Compared to continuous materials, employing porous media as the optimization matrix is a promising method as it enables simultaneous weight reduction and heat transfer enhancement. This study proposes an innovative framework for multiscale topology optimization using anisotropic porous media by integrating the adjoint lattice Boltzmann method with graphics processing unit acceleration. The results demonstrate that at the macroscopic scale, the topological configurations of optimized designs exhibit similarity with solid materials distributed in high-velocity flow regions to maximize heat transfer performance, regardless of the characteristics of the porous media. At the microscale, the structural characteristics of porous media exert a critical influence on the flow and heat transfer performance of the optimized configurations. Compared to conventional continuous materials, optimization utilizing porous media under identical weight constraints demonstrates a 24.9% enhancement in heat transfer performance. The anisotropy of the porous media regulates heat transfer performance by modifying the fluid flow patterns within the media to optimize the synergy between velocity and temperature fields. In optimized design, priority should be given to using anisotropic porous media with principal permeability aligned with the primary flow direction because this configuration could achieve optimal flow resistance reduction and heat transfer enhancement. An increase in the permeability of porous media leads to greater concentration of the optimized solid accompanied by enhanced heat transfer performance.