Thermal performance of microchannel heat sink integrated with porous medium, slip coefficient and phase change material and machine learning approach

Abstract

In this study, the thermal performance (TP) of the microchannel heat sink (MCHS) under the influence of the slip coefficient on the microchannel wall, the use of porous medium (PM) and phase change material (PCM), and the changes in the geometrical parameters and the shape of the microchannel have been discussed. The flow and energy equations were solved using the finite volume method. The validation of the melting process, slip coefficient, and porous medium with the previous literature demonstrate the high accuracy of the numerical solution, with the difference between the results being <8 %. The results indicate that using a wavy microchannel can improve TP by approximately10.6 % and 5 % compared to smooth and converging microchannels. Additionally, using non-Newtonian fluids with a power characteristic of 0.53 leads to improved TP about 2.8 % by reducing shear stress compared to Newtonian fluids. Furthermore, the presence of slip on the microchannel wall influences the flow pattern, resulting in reduced convection resistance and improved TP about 3.2–3.8 %. The highest TP was observed at a slip coefficient of 0.00001. Incorporating PM in the microchannel (either in the center or around it) can enhance TP by approximately 8–15 % compared to the base state, with greater improvements observed when PM is placed around the microchannel. The type of PM material and the porosity coefficient significantly impact TP. Varying the cross-sectional shape of the microchannel (square, rhombus, and triangle) has a substantial effect on improving TP compared to a circular cross-section, with the highest TP (with a 71 % increase) observed in microchannels with a square cross-section. Lastly, the use of PCM and PM on the active surface leads to a significant TP improvement of over 80 % compared to the base state. Using the Group Method of Data Handling algorithm, thermal resistance was estimated with an R-square value of 0.94 to validate the available data.