Topology optimization of thermal-fluid systems with non-uniform thermal loads using a novel objective function

Abstract

Topology optimization transcends the limitations of traditional engineering design and enables innovative structural concepts. This study integrates the adjoint lattices Boltzmann method with the Level-set method to perform a topology optimization on the convective heat transfer under non-uniform thermal loads. Firstly, a novel objective function is proposed, which achieves a tight coupling of forward parameters in the adjoint problem and the optimization performance can be improved accordingly. With the GPU acceleration, a tenfold increase in optimization efficiency is achieved. Secondly, by decoupling the evolution and boundary equations of the forward LBE, the adjoint LBE could be derived efficiently and it enhances the simplicity and generality of the adjoint LBM. Finally, the topology optimization under various thermal loads is investigated based on our proposed method. The results indicate that the optimized structures are significantly influenced by the distributions of thermal loads. The solid tends to concentrate in the high heat-flux regions to ensure a high efficiency of heat transfer in the area, since the low porosity leads to high-velocity flow to enhance the convective heat transfer. According to the optimized structures, there are two reasons for the enhanced heat transfer capability, one is the increased temperature gradient in the thermal boundary layer and the other is the increased tortuosity in high heat flux regions.