A numerical study with surrogate-assisted multi-objective optimization of cross-flow liquid cooling for cylindrical Li-ion cell array

Key Points

• The study aims to optimize cross-flow liquid cooling for cylindrical Li-ion cell arrays using a surrogate-assisted framework.
• Developed a CFD-ML framework for simulating a cross-flow liquid-cooled battery thermal management system (BTMS)
• Conducted transient conjugate simulations to evaluate safety thresholds and performance metrics
• Utilized a log-scaled surrogate model to predict temperature differentials and pressure drops effectively.
• Achieved a maximum temperature (T max) of less than 308 K and a temperature difference (ΔT) of less than 5 K.
• Selected Pareto knee-point led to a compact design that maintained performance while optimizing parameters.
• The optimized design reduced temperature non-uniformity by 69.7% compared to the baseline without cooling.