Multi-objective optimization and performance assessment of microchannel heat sinks with micro pin-fins

In this study, laminar, steady-state microchannel flow and conjugate heat transfer in a microchannel heat sink (MCHS) with micro pin-fins were investigated numerically. Computational Fluid Dynamics (CFD) simulations were performed with ANSYS Fluent for heat and fluid flow in MCHSs with circular-, square- and diamond-shaped micro pin-fins of the same hydraulic diameter. Arrangements of micro pin-fins in the MCHS were optimized for each shape using a multi-objective genetic algorithm (MOGA) method called Non-dominated Sorting Genetic Algorithm (NSGA-II). The fluid entered the microchannel at channel Reynolds numbers ranging from 100 to 350, corresponding to circular- and square-shaped pin-fin Reynolds numbers ranging from 20 to 110 and diamond-shaped pin-fin Reynolds numbers ranging from 20 to 190. In all simulations, uniform heat flux of 60 W/cm2 was applied from the bottom surface of substrate. As the cooling fluid, water with temperature-dependent variable viscosity was assigned in the fluid domain and copper with constant thermophysical properties was considered for the entire solid domain. Design variables were selected in order to represent both the pin-fin arrangement and flow-characteristics: porosity number and pin-fin Reynolds number, respectively. As objective functions, pressure drop ratio was defined as the hydrodynamic performance indicator and the Nusselt number was chosen to represent the thermal performance. To represent the trade-off between two objective functions, optimization was conducted after a systematic parametric study and pareto-fronts were obtained. Optimized solutions of each pin-fin shape showed that square-shaped pin-fin was unfavorable compared with the performance of other shapes of pin-fins. Among all configurations, diamond-shaped pin-fins significantly enhanced heat transfer at an increased pressure drop ratio. Depending on the pin-fin shape, optimal configurations emerged over a wide range of porosity values mostly at the upper limit of the pin-fin Reynolds number, with average Nusselt numbers varying between 3 and 12 and corresponding pressure drop ratios varying between 1 and 12. •Thermo-hydrodynamic behaviour of pin-finned MCHS was numerically investigated.•Multi-objective optimization (NSGA-II) was utilized to obtain optimized configurations.•Objective functions for optimization were Nusselt number and pressure drop ratio.•Optimal configurations were obtained in a broad span of application.•Optimal configurations wer