TP26-19 The Effect of Two-Stage Optimization Studies Using RSM and Adjoint on the Efficiency of Cooling Tower Fans

This study investigates the effect of two-stage optimization, based on Response Surface 
Methodology (RSM) and the adjoint method, on the efficiency of cooling tower fans. In the first stage, RSM was used to methodically explore the design space and find the most significant geometric parameters affecting fan efficiency and static pressure rise. Through optimization based on RSM, static pressure rise and efficiency improved by 20.6% and 18.2%, respectively, and the required power was reduced by 2.3% against the base design. The second phase involved adjoint optimizations for optimizing the blade geometry with high-fidelity gradient information obtained directly from CFD solutions. These adjoint optimized refinements yielded another 6.08% rise in static pressure and a further 9.23% improvement in efficiency compared with the RSM-optimized configuration, with another 4.43% lower power consumption. Based on flow field analyses, it was found that the adjoint-optimized blade redistributes aerodynamic loading well and changes the radial velocity structure, leading to reduced wall-related losses and better use of flow momentum. These findings show that the integrated RSM-adjoint design provides a robust and efficient approach for achieving significant aerodynamic performance gains and energy savings in cooling tower fans.