The entropy generation analysis of the influence of using fins with tip clearance on the thermal management of the batteries with phase change material: Application a new gradient-based ensemble machine learning approach. (July 2022)

• Record Type: Journal Article
• Authors: Shahsavar, Amin; Goodarzi, Abbas; Baniasad Askari, Ighball; Jamei, Mehdi; Karbasi, Masoud; Afrand, Masoud
• Is Part Of: Engineering analysis with boundary elements. Volume 140(2022)
• Page Start: 432
• Page End: 446
• Publication Date: 2022-07
• Subjects: Battery thermal management system -- Phase change material -- Fin, Entropy analysis, Computational fluid dynamics -- Gradient boosting decision tree; EV Electronic vehicle -- CART Classification and regression trees -- GBDT Gradient boosting decision tree -- ML Machine learning -- PCM Phase change material -- PVT Photovoltaic thermal -- TC Tip clearance -- TMS Thermal management system; Boundary element methods -- Periodicals; Engineering mathematics -- Periodicals; Équations intégrales de frontière, Méthodes des -- Périodiques; Mathématiques de l'ingénieur -- Périodiques; Boundary element methods; Engineering mathematics; Periodicals; 620.00151
• DOI: 10.1016/j.enganabound.2022.04.024
• Languages: English
• ISSNs: 0955-7997

Abstract

Abstract: The present paper deals with 3D numerical analysis of a battery thermal management system (TMS) including the Phase Change Material (PCM). The TSM comprises three annular fins located around the battery considering the tip clearance (TC) space between the fin tips and the alumina enclosure. The calculations were performed for four cases with different TCs (1.5 mm, 1 mm, 0.5 mm, and 0 mm). The entropy generation analysis was performed to determine the locations of the geometry with highest frictional and thermal irreversibilities. The results showed that the application of TC leads to improve the PCM free convection and thereby enhance the heat transfer rate. So that there is an optimum TC (0.5 mm) in which the highest heat transfer rate and lowest PCM melting time is obtained. Moreover, the magnitude of frictional entropy generation rate is much lower than that of the thermal term. For accurate estimation of the liquid fraction, fractional and thermal entropy generation rates, a new ensemble machine learning (ML), namely Gradient Boosting Decision Tree (GBDT), was developed based on the fin tip and flow time parameters as input features. The outcomes of ML-based simulation exhibited promising performance for the precision prediction of three understudy targets.