Surrogate-based multi-objective design optimization of tree-shaped fins with uniform branch end distribution for latent heat thermal energy storage.

This article focuses on optimizing tree-shaped fin geometries to enhance the performance of latent heat thermal energy storage (LHTES) systems, specifically targeting improvements in both energy and power densities. Using a surrogate model-based multi-objective optimization approach—evaluating models such as linear regression, artificial neural networks, and random forests—the study identifies an optimal fin design with a volume fraction of 33.9%, achieving a 61.6% increase in power density and a 38.18% reduction in melting time compared to conventional plate fins. The research introduces a novel constraint ensuring uniform distribution of the last branch tips, which improves thermal uniformity and reduces computational cost. The findings demonstrate that the optimized tree-shaped fin design significantly enhances thermal management efficiency in LHTES systems, offering a practical framework for future design innovations in renewable energy storage applications.