Large Scale 3D Topology Optimization of Conjugate Heat Transfer

This paper presents an approach for large scale density-based topology optimization (TO) of a three-dimensional thermal-fluid cooling system. The forced cooling system is simulated with weakly coupled stationary incompressible Navier-Stokes equation and the energy equation. Under a heat source, we study the distribution of the conductive solid materials in order to minimize the temperature of the solid components as well as the flow energy dissipation. The governing state equations and their adjoint equation for obtaining sensitivity are solved through a stabilized finite element formulation. The design domain is discretized with millions of tetrahedral elements. Numerical examples are presented to show the accuracy of our solver. Topologically optimized designs in several test cases are presented.