Studies on hydrothermal performance of minichannel heat sinks with MAL-MEPCM slurry: experimental analysis and machine learning modeling

In this study, heat transfer and fluid flow characteristics of myristyl alcohol microencapsulated with calcium carbonate (MAL-MEPCM) slurry in different minichannel heat sinks are investigated experimentally. The effect of mass flow rates, heat fluxes, and slurry concentrations on thermal performance in the straight minichannel (SMCHS), wavy minichannel (WMCHS) and semi-wavy minichannel (SWMCHS) heat sinks are examined and compared. The base fluid is water and the various concentrations of MAL-MEPCM slurry are 0.1, 0.25, 0.5, and 0.75%. The results reveal that average Nusselt number (Nu_avg) of the SMCHS with 0.5% MAL-MEPCM slurry is higher than the other concentrations. Also, the Nu_avg for WMCHS and SWMCHS using 0.1% slurry are enhanced by 24.1 and 26.5% when compared to water. Thermal resistance of the minichannels with slurries are lower than water. Machine learning models viz., Linear, Polynomial, Random Forest, Decision Tree, KNN and Gradient Boosting Regression are used to predict the Nusselt number and thermal resistance and are compared with experimental data. Among all the models, the prediction of Random Forest Regression model is closer to the actual data. The model gives RSME values of 0.212 & 0.0036 and R² values of 0.996 & 0.997 for Nusselt number and thermal resistance, respectively.