Multi-objective optimization of a porous diverter plate for a liquid-cooled micro-jet heat sink via surrogate modeling

Abstract

This work reports on the optimization of a porous diverter plate for a micro-jet heat sink to obtain the optimum jet nozzle diameter and flow opening ratio, which play a significant role in the cooling characteristics of heat sinks for electronic chips. An artificial neural network (ANN) model was used to optimize and predict the sensitivity of different parameters (coolant flow rate, opening ratio, and jet nozzle diameter) on the operation of the heat sink design. It was found that smaller opening ratios and jet nozzle diameters in the range of 2.0–3.2 mm had a positive effect on improving the heat sink's thermal performance. Moreover, the diverter plate's opening ratio had a major sensitivity to pumping power, while the jet nozzle diameter had the lowest sensitivity. Finally, the overall cooling characteristics of all the designs were evaluated using the PPTR parameter, and the results demonstrate that the porous diverter plate with an opening ratio of 0.08 and a jet nozzle diameter of 1.81 mm yielded the optimum design for the porous diverter plate of the heat sink.

Keywords

• Multi-objective optimization
• Porous diverter plate
• Artificial neural network
• Heat sink