Topology optimization, additive manufacturing and thermohydraulic testing of heat sinks

This paper presents a set of high-performing heat sinks that exhibit twice the thermohydraulic performance in terms of conductance compared to conventional rectangular fin heat sinks. The heat sinks presented here are designed through three-dimensional topology optimization (TO), manufactured using additive manufacturing (AM), and their performance is validated through experimental testing. The heat sink design is governed by steady-state Navier-Stokes equations and the energy equation. The objective is to minimize the average temperature of the heat source surface with a constant heat flux. Our design process incorporates two constraints: the pressure drop constraint and the project undercut perimeter (PUP) based overhang angle constraint. The incorporation of the overhang angle constraint ensures that the optimized heat sink design is self-supported and amenable to additive manufacturing without the need for additional support structures. Post-optimization CFD investigations revealed that the optimized heat sink offers improved thermal performance, attributed to 2 kinds of three-dimensional convection effects, thermal boundary layer re-initialization, and efficient mixing. The optimized heat sink designs are manufactured using laser-powder bed fusion process, an additive manufacturing technique, and their superior performance relative to a conventional rectangular heat sink is validated through experimental measurements. The experimental tests are in good agreement with CFD simulations, confirming a remarkable 100% increase in conductance for the TO designs compared to a conventional heat sink. •Geometrically intricate heat sink design: achieved through topology optimization and additive manufacturing.•Improved thermohydraulic performance: twice the heat conductance over rectangular fin design.•Cooling mechanism: optimized designs disrupt boundary layer, enables full-field mixing of cold and hot flow.•Smaller surface area: optimized designs possess smaller surface area than rectangular fins.