Key points are not available for this paper at this time.
This research introduces a novel approach to thermal energy storage (TES) tank design by integrating physics-based modelling with genetic algorithms (GAs). Traditional TES designs face inherent limitations in balancing thermal performance, hydraulic efficiency, and material usage due to restricted exploration of complex parameter spaces. This research developed a computational framework that autonomously generates fractal-inspired TES tank configurations for phase change material (PCM) applications, evaluating designs based on thermal effectiveness, energy density, pressure drop, and geometric uniformity. Unlike previous approaches that merely optimise predefined structures, this methodology generated entirely new configurations through first principles. The optimised design features a three-level hierarchical branching structure with a main pipe length of 110 mm, radius of 13 mm, and branch angle of 49.5°. Performance evaluation across multiple operating conditions demonstrated thermal effectiveness of 0.313 at 1.5 L/min flow rate, energy density reaching 200 MJ/m 3 (15–45 °C temperature differential), and remarkably low pressure drops (maximum 10.4 Pa at 6.0 L/min). The design maintained consistent geometric uniformity (0.950) across all tested conditions while achieving 1693 Wh storage capacity in a compact tank. This genetic algorithm optimised design transcended conventional approaches by discovering non-intuitive configurations that would be difficult to identify through traditional engineering methods. The results demonstrated how computational intelligence integrated with physical principles yielded high-performance thermal energy storage systems while maintaining practical manufacturing constraints, potentially advancing renewable energy integration and thermal management applications.
Mehraj et al. (Thu,) studied this question.