NUMERICAL ANALYSIS AND OPTIMIZATION OF HYDROTHERMAL PERFORMANCE OF MICROCHANNEL HEAT SINK

Abstract

The continuous increase in power density and miniaturization of modern electronic systems has intensified the demand for advanced thermal-management technologies capable of dissipating high heat fluxes while maintaining thermal reliability and energy efficiency. Although microchannel heat sinks have emerged as one of the most effective cooling technologies for electronic applications, their performance is often constrained by the competing requirements of heat-transfer enhancement, pressure-drop minimization, and thermodynamic efficiency. Consequently, the development of optimized cooling architectures capable of achieving superior thermo-hydraulic and exergetic performance remains a critical challenge in electronic cooling. To address this challenge, convergent-divergent microchannel heat sinks (CD-MCHS) were selected as the baseline cooling architecture because the periodic flow acceleration and deceleration induced by convergent-divergent passages can enhance fluid mixing, suppress thermal boundary-layer growth, and improve heat-transfer performance. The present research develops an integrated computational framework combining Computational Fluid Dynamics (CFD), Artificial Neural Networks (ANN), the Non Dominated Sorting Genetic Algorithm-II (NSGA-II), and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) for the multi-objective optimization of CD-MCHS. The optimization objective was to maximize thermal performance while minimizing entropy generation and thermodynamic irreversibility. The first phase of the investigation focused on the optimization of the baseline CD-MCHS geometry. The TOPSIS-selected configuration achieved a 20.1% enhancement in heat transfer performance and a 38.1% reduction in entropy generation compared with the conventional straight microchannel. Furthermore, the optimized geometry reduced the outlet temperature by approximately 8 K while exhibiting the lowest exergy destruction ratio (8.53%) and the highest exergy efficiency (91.47%). These findings demonstrate that geometric optimization alone can significantly enhance cooling performance while simultaneously improving thermodynamic behavior. Building upon the optimized CD-MCHS platform, three complementary passive enhancement strategies representing geometric modification (micro-pillars), coolant modification (tri-hybrid nanofluids), and porous-medium augmentation were systematically investigated to explore different mechanisms of thermal-performance Numerical Analysis and Optimization of Hydrothermal Performance of Microchannel Heat Sink P a g e | vii enhancement. For each enhancement strategy, ANN surrogate models were developed and coupled with NSGA-II and TOPSIS to identify the optimal configurations based on thermo-hydraulic and exergetic performance criteria. The optimized micro-pillar-enhanced CD-MCHS achieved a 71.45% increase in heat transfer performance with only a 15.64% increase in entropy generation, while reducing the maximum wall temperature by approximately 3.5 K. The optimized tri-hybrid nanofluid composition enhanced heat transfer by 16.36% while maintaining the lowest thermodynamic penalty among the enhancement techniques and reducing the maximum wall temperature by 1.89 K. In contrast, the optimized porous-medium configuration delivered the highest cooling capability, achieving a 265.65% increase in heat-transfer performance and approximately 7 K reduction in maximum wall temperature relative to the non-porous channel. However, this substantial thermal enhancement was accompanied by significantly higher entropy generation and exergy destruction due to increased flow resistance within the porous matrix. A comprehensive comparative assessment revealed that the three passive enhancement strategies represent distinct thermal-management philosophies. Porous media maximized cooling capability, whereas tri-hybrid nanofluids minimized thermodynamic penalties. The micro-pillar-enhanced CD-MCHS effectively bridged these two extremes by delivering substantial heat-transfer enhancement with moderate entropy-generation and exergetic penalties. Consequently, the micro-pillar configuration emerged as the most balanced passive enhancement strategy, offering the most favorable compromise among cooling effectiveness, thermodynamic efficiency, and pumping-power requirements. Overall, the present research demonstrates that the optimization of advanced microchannel heat sinks should not be guided solely by heat-transfer enhancement but must simultaneously consider hydrothermal and thermodynamic performance. The proposed CFD-ANN-NSGA-II-TOPSIS framework provides a robust methodology for the design and optimization of next-generation thermal-management systems. Furthermore, the findings establish convergent-divergent microchannel heat sinks as a promising platform for high-performance electronic cooling and provide practical design guidelines for advanced computing, power-electronic, and other energy-intensive applications.