Thermo-Hydraulic Performance Optimization of a Disk-Shaped Microchannel Heat Sink Applying Computational Fluid Dynamics, Artificial Neural Network, and Response Surface Methodology

Abstract

Specifying the optimum cooling capacity of the microchannel heat sinks (MCHSs) to cope with the overheating problem in high-performance microelectronic equipment is essential in the modern technological world. The artificial neural network (ANN) and response surface methodology (RSM) have been newly developed as alternative approaches to model the thermal and hydrodynamic properties of mentioned equipment. In the present research, a bionic fractal heat sink with tree-like microchannels with trapezoidal cavities on the channels was selected to investigate the Nusselt number (as a criterion of heat transfer) of the device and pressure drop (∆P) of the water. The appropriate design of the ANN and RSM models was exerted separately to determine the optimum Nusselt number (Nu) and ∆P in the device. The aspect ratios of t/b (cavities’ upper side to bottom side ratio) and h/b (cavities’ height to bottom side ratio), as well as the Reynolds number, were set as the independent variables in both ANN and RSM models. All the codes were written using Python. The interactive effects of independent variables and their consequent impact on the responses were studied by RSM. While the ANN model was utilized to reach the precise equation which has the best fitting value. The obtained results indicated that both ANN and RSM models could exactly anticipate heat transfer and ∆P to a large extent. Considering the coefficient of determination value (99.9%), the Nu and ∆P variations were well predicted by the RSM and ANN models. To gain the optimum design of the MCHS with the approach of having the maximum Nu and the minimum ∆P, the efficiency index of the device was examined. It was found that the maximum efficiency index (1.070 by RSM and 1.067 by ANN methods) was obtained at the aspect ratios of t/b=0.2, h/b=0.2, and the Reynolds number of 1000.

Keywords: Bionic fractal heat sink, Thermal and hydrodynamic properties, Response Surface Methodology, Artificial Neural Network, Python