Topological and multi-objective optimization of single-phase heat transfer and energy efficiency using manifold micro-channels for high-power electrics cooling

Abstract

To achieve energy-efficient cooling of high-power electrics, present work demonstrates combined methods using topological design and multi-objective optimization to reconstruct manifold micro-channels (MMC). The micro-channel domain is first topologized by density-based method under pressure drop and energy dissipation constraints. The multi-objective optimization is subsequently applied to optimize structure and operating parameters of reconstructed MMC using kriging model and genetic algorithm. The simulation results show that favorable thermal performance is obtained in the topological MMC with branch-channels, pin fins and root-channels, increasing average heat transfer coefficient by 19.4 % compared to conventional MMC. For the optimal flow rate, the lowest total thermal resistance (Rtotal) of 0.17 KW−1 is obtained for the optimized-topological MMC at a pumping power (Pp) of 9.53 mW, meanwhile the maximum temperature-rise is decreased by 15.2K. Furthermore, an extensive benchmark of different single-phase micro-fluidic cooling technologies is performed. For an extremely low Rtotal of 0.2 KW−1, present optimized-topological MMC only requires Pp of 3 mW, which is 11.7 times, 26.7 times and 70 times lower than that of conventional MMC, pin-fin and strip-fin micro-fluidic systems. This superior hydro-thermal performance successfully manages the heat flux of 100 W/cm2 to 400 W/cm2 with cooling coefficient of 104 to 107, significantly outperforming other micro-fluidic cooling technologies.