Pareto-Driven Multiobjective Design of Axial-Flow Automotive Fan with Response Surface Modeling

Abstract Automotive cooling fans play a vital role in thermal management, yet conventional designs often struggle to balance efficiency, pressure, and flow requirements. This work presents a multiobjective optimization of an axial-flow fan using response surface methodology and a genetic algorithm. Four critical parameters (the root and tip installation angles and sweep angles) were optimized with respect to volumetric flow rate (Q), static pressure (P), and efficiency (η). A surrogate model built from 25 Latin Hypercube Sampling points achieved high accuracy (R2 > 0.99). Sensitivity analysis showed that the tip angle predominantly affects Q and P, while the root angle strongly influences η. Optimization yielded Pareto solutions, where the efficiency improved from 18.31% to 21.19% without reducing the flow or pressure. The flow-field analysis demonstrated that the enhanced aerodynamic stability is addressed in the enhanced aerodynamic stability, characterized by smoother velocity profiles and reduced regions of separation and recirculation. The proposed framework not only improves fan aerodynamic efficiency but also establishes a generalizable strategy for systematic multiobjective optimization of rotating machinery.