Multi-objective optimization of oscillating cooling heat transfer performance in a horizontally OP2S engine piston based on the kriging model and NSGA-II

Horizontally opposed two-stroke (OP2S) engine pistons replace conventional cylinder heads, requiring internal oscillating oil-cooling cavities for effective forced cooling. Existing studies prioritize overall heat transfer enhancement but overlook the critical issue of deteriorated heat exchange uniformity caused by increased global heat transfer coefficients. This research gap is addressed by proposing a systematic methodology that combines numerical simulation with multi-objective optimization, specifically tailored for the cooling systems of small-bore OP2S pistons. First, we developed a dynamic mesh simulation model to describe the thermal transfer performance of the oil cooling cavity. Subsequently, a multi-objective optimization framework was developed, targeting the overall heat transfer coefficient and circumferential uniformity. The framework employs Maximin Latin hypercube sampling for initial sample generation, a Kriging surrogate model for response surface construction, and the NSGA-II algorithm to obtain a Pareto front quantifying the trade-off between the two objectives. The TOPSIS-EWM method was further utilized to select the optimal design from the Pareto front. Results demonstrate that the optimized design achieves an average heat transfer coefficient of 2278.58 W/(m²·K) (8.3% higher than the original model) and a circumferential heat transfer coefficient standard deviation of 175.04 W/(m²·K) (30.0% lower than the initial value). This work provides a robust computational optimization methodology for balancing heat transfer efficiency and uniformity, offering valuable engineering guidance for the structural design and performance improvement of oscillating cooling systems in OP2S pistons and similar engine components.