The vehicle fan is the major element of a cooling system, and its operating noise has a significant impact on vehicle noise, which has a significant impact on ride comfort. Axial flow cooling fans are commonly employed in current vehicles to achieve a high flow rate nevertheless, increased demands are placed on the fan’s energy consumption in order to save energy and preserve the environment.
Noise, flow rate, and power are the primary indications of an axial fan for automotive applications. Fan diameter, rotational speed, number of blades, blade angle, chord length of the blades, and blade arrangement are often factors influencing fan performance.
Based on the aforementioned theory and boundary constraints, a theoretical simulation model of the fan was created. The blade is divided into 12 equal parts along the radius direction, 13 sections, each section corresponds to 1 blade angle and chord length, a total of 26 parameters for improving the accuracy of the theoretical simulation, and direct parameter optimization will consume a lot of computing resources, so the parameters with high sensitivity for optimization are filtered out by the Sobol’ sensitivity analysis method to ensure the accuracy of the case and shorten the calculation time.
Vehicle fan optimal design When compared to the old approach, the genetic optimization algorithm provides better convergence and resilience. Non-dominated sorting genetic algorithms II (NSGA II) is a common multi-objective genetic algorithm. The NSGA II by K.DED is an upgraded version of the original NSGA method introduced in 2000, which was used for the multi-objective optimization study in this research.
The multi-objective optimization approach was used to optimize the fan blade structure in order to reduce fan noise.
The conceptual approach was first established and verified, with the fan’s noise, flow rate, and power as the objectives then, the global sensitivity analysis method based on the Sobol method was used to obtain the contributions of each parameter to the fan‘s performance objectives, with the fan blade angle and chord length as the analysis parameters.
Finally, the evolutionary algorithm selects the sensitive settings to achieve the optimum noise-oriented complete performance of the fan. After improving the fan blade structure, the noise is decreased by 3.1dB, the flow rate is enhanced by 7.7%, the power is lowered by 7.9%, and the overall performance of the fan is considerably improved.