TY - JOUR
T1 - Design parameter optimization of a CPU heat sink using numerical simulation for steady-state thermal analysis and CFD-modeling
AU - Iglin, Pavel
AU - Iglina, Tatyana
AU - Pashchenko, Dmitry
PY - 2023
Y1 - 2023
N2 - This paper deals with the design of a CPU cooling system using a novel numerical modelling approach based on automatic calculation in a commercial software. A research object is an aluminium CPU heat sink with a thermal design power of 50 W with a new fin design. A numerical model of the cooling process has been developed, and the heat sink efficiency has been investigated. The main goal of optimization was to get the minimum temperature of the CPU processor at the minimum mass of the heat sink. The comparative analysis of the results that obtained via three methods (screening, adaptive multiple-objective, multi-objective genetic algorithm) was performed. This analysis showed that screening was the least time-consuming method, but it did not provide the required solution. Adaptive multiple-objective and multi-objective genetic algorithm solutions show similar results but significantly differ in time. It was established that the adaptive multiple-objective method is the best method for the heat sink optimization task. At the determined optimal design parameter, the CPU temperature is in the range 304-307K, while the mass was 81-87g. In comparison, the heat sink mass before optimization of the design parameters was 93g at the CPU temperature of 309-311K.
AB - This paper deals with the design of a CPU cooling system using a novel numerical modelling approach based on automatic calculation in a commercial software. A research object is an aluminium CPU heat sink with a thermal design power of 50 W with a new fin design. A numerical model of the cooling process has been developed, and the heat sink efficiency has been investigated. The main goal of optimization was to get the minimum temperature of the CPU processor at the minimum mass of the heat sink. The comparative analysis of the results that obtained via three methods (screening, adaptive multiple-objective, multi-objective genetic algorithm) was performed. This analysis showed that screening was the least time-consuming method, but it did not provide the required solution. Adaptive multiple-objective and multi-objective genetic algorithm solutions show similar results but significantly differ in time. It was established that the adaptive multiple-objective method is the best method for the heat sink optimization task. At the determined optimal design parameter, the CPU temperature is in the range 304-307K, while the mass was 81-87g. In comparison, the heat sink mass before optimization of the design parameters was 93g at the CPU temperature of 309-311K.
U2 - 10.1142/s0129183123500900
DO - 10.1142/s0129183123500900
M3 - 文章
SN - 0129-1831
JO - International Journal of Modern Physics C
JF - International Journal of Modern Physics C
ER -