TY - JOUR
T1 - A multi-step topology optimization framework of fin structures for accelerated melting in a phase change material (PCM)-filled cavity for thermal management
AU - Fragnito, Andrea
AU - Andreasen, Casper Schousboe
AU - Iasiello, Marcello
AU - Mauro, Gerardo Maria
AU - Bianco, Nicola
N1 - Publisher Copyright:
© 2025 The Author(s)
PY - 2026
Y1 - 2026
N2 - The rising power density of modern electronics requires effective thermal management to maintain performance and reliability. Phase change materials (PCMs) provide passive cooling via latent heat but are limited by low thermal conductivity, which hinders their efficiency. This study investigates the application of topology optimization for designing thermally efficient fin structures in a PCM-filled domain. The PCM numerical model is validated against experimental data, ensuring code fidelity. The optimization process is conducted within a 2D finite element framework, where fins are designed on a fixed-temperature wall while maintaining the same material volume as a baseline configuration with three rectangular fins. Three objective functions – convective heat flux, diffusive heat flux, and thermal compliance – are considered using a steady-state approach. Their influence on the melting process is systematically evaluated. Then, the best performing one is used in a multi-step approach (MSTE) that resembles the unsteady PCM melting but allows to obtain performing design through an iterative process. Results demonstrate that the choice of a convective heat flux as objective accelerates the melting process by 32.1% compared to the baseline. Then, the MSTE shows that with two iterations it is possible to generate a design that fasten the melting process by 48.9% compared to the baseline. Further analysis includes comparisons of melt fraction, Nusselt number, energy storage and mean thermal power across all cases to provide valuable insights into the advantage of the MSTE approach for optimizing thermal management systems.
AB - The rising power density of modern electronics requires effective thermal management to maintain performance and reliability. Phase change materials (PCMs) provide passive cooling via latent heat but are limited by low thermal conductivity, which hinders their efficiency. This study investigates the application of topology optimization for designing thermally efficient fin structures in a PCM-filled domain. The PCM numerical model is validated against experimental data, ensuring code fidelity. The optimization process is conducted within a 2D finite element framework, where fins are designed on a fixed-temperature wall while maintaining the same material volume as a baseline configuration with three rectangular fins. Three objective functions – convective heat flux, diffusive heat flux, and thermal compliance – are considered using a steady-state approach. Their influence on the melting process is systematically evaluated. Then, the best performing one is used in a multi-step approach (MSTE) that resembles the unsteady PCM melting but allows to obtain performing design through an iterative process. Results demonstrate that the choice of a convective heat flux as objective accelerates the melting process by 32.1% compared to the baseline. Then, the MSTE shows that with two iterations it is possible to generate a design that fasten the melting process by 48.9% compared to the baseline. Further analysis includes comparisons of melt fraction, Nusselt number, energy storage and mean thermal power across all cases to provide valuable insights into the advantage of the MSTE approach for optimizing thermal management systems.
KW - Natural convection
KW - Nusselt number
KW - Phase change material
KW - Rectangular enclosure
KW - Thermal management
KW - Topology optimization
U2 - 10.1016/j.ijheatmasstransfer.2025.127694
DO - 10.1016/j.ijheatmasstransfer.2025.127694
M3 - Journal article
AN - SCOPUS:105013676834
SN - 0017-9310
VL - 254
JO - International Journal of Heat and Mass Transfer
JF - International Journal of Heat and Mass Transfer
M1 - 127694
ER -