Multi-objective optimization of plate-fin latent thermal energy storage for airborne high-power laser thermal buffering under mass and volume constraints

Airborne high-power laser systems subjected to periodic thermal loads require compact thermal buffers with high storage capacity and rapid transient response under stringent mass and volume constraints. This study develops a surrogate-assisted multi-objective optimization framework for a plate-fin latent thermal energy storage (LTES) device using condensing NH 3 as the heat-transfer fluid, with emphasis on cold-discharge performance. A two-dimensional transient distributed-parameter model was established to resolve coupled HTF condensation, plate-fin conduction, and PCM melting, and was integrated with Morris sensitivity analysis, Kriging surrogate modeling, NSGA-Ⅱ, and TOPSIS. Sensitivity analysis showed that the energy metrics were governed mainly by PCM-metal volume allocation, whereas the power metrics were more sensitive to operating conditions and fin height. The Kriging models achieved validation-set R 2 values of 0.991–0.999, and the relative errors of representative optimized designs re-evaluated using the original transient model remained below 5 %. Under the nominal operating condition, the mass-constrained design increased specific energy and specific power by 39.4 % and 67.0 %, reaching 198.17 kJ kg −1 and 0.496 kW kg −1 , respectively. The space-constrained design increased energy density and power density by 13.8 % and 44.7 %, reaching 173,847 kJ m −3 and 461.31 kW m −3 , respectively. Mechanistic analysis showed that the optimized geometries accelerated melting-front propagation, whereas late-stage melting remained limited by increasing PCM-side resistance and declining active heat-transfer area. The framework efficiently quantifies and improves the Ragone-type energy-power trade-off and provides constraint-specific design guidance for compact LTES thermal buffers.