Pre-research on Enhanced Heat Transfer Method for Special Vehicles at High Altitude Based on Machine Learning

Abstract

Abstract The performance of an integrated thermal management system significantly influences the stability of special-purpose vehicles; thus, enhancing the heat transfer of the radiator is of great significance. Common research methods for radiators include fluid mechanics numerical simulations and experimental measurements, both of which are time-consuming and expensive. Applying the surrogate model to the analysis of the flow and heat transfer in louvered fins can effectively reduce the computational cost and obtain more data. A simplified louvered-fin heat transfer unit was established, and computational fluid dynamics (CFD) simulations were conducted to obtain the flow and heat transfer characteristics of the geometric structure. A three-factor and six-level orthogonal design was established with three structural parameters: angle θ, length a, and pitch L p of the louvered fins. The results of the orthogonal design were subjected to a range analysis, and the effects of the three parameters θ, a, and L p on the j, f, and JF factors were obtained. Accordingly, a proxy model of the heat transfer performance for louvered fins was established based on the artificial neural network algorithm, and the model was trained with the data obtained by the orthogonal design. Finally, the fin structure with the largest JF factor was realized. Compared with the original model, the optimized model improved the heat transfer factor j by 2.87%, decreased the friction factor f by 30.4%, and increased the comprehensive factor JF by 15.7%.

Keywords

• Louvered fins
• Numerical simulation
• Machine learning
• Comprehensive performance