Influence of heat source distribution optimization on gallium heat transfer characteristics in a annular tube based on MLA and CFD

ABSTRACT

In this study, innovative combined Machine Learning Algorithm (MLA), Genetic Algorithm (GA) and Computational Fluid Dynamics (CFD) was used to investigate the influence of nonuniform heat source distribution on gallium heat transfer performance in the annular tube, and an optimization scheme of heat source distribution was proposed. First, the heat source distribution scheme was quantified using four intervals A, B, C, and D from the inlet to the outlet, and the effect of different schemes on the heat transfer performance was investigated by CFD. Second, the prediction model of the average surface heat transfer coefficient under different heat source distributions was established based on various MLAs, and the effect of heat source interval on heat transfer performance was obtained through the model analysis. The results show that the intervals A and B have more influence on heat transfer performance, the heat transfer coefficient increases with the increase in A and B intervals. Finally, the optimal heat source distribution scheme is solved by GA. Compared with the traditional distribution scheme, the optimal scheme makes the inner wall temperature more uniform, reduces the high-temperature areas, reduces the maximum temperature by 2.6%, and increases the average surface heat transfer coefficient of the inner wall by 6.42%.

KEYWORDS:

• Numerical simulation
• heat source distribution
• machine learning
• genetic algorithm

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Natural Science Foundation of China under Grant [No. 62162050].