In many electronic cooling applications there is in increasing demand for high pressure rise compact axial fan design. This can only be achieved by increasing fan rpm, which can result in higher noise. In this paper, a methodology is presented based on coupling a 3D inverse design method together with Design of Experiments, Response Surface modeling and Multi-objective Genetic algorithms in order to improve the efficiency and tonal noise from high rpm cooling fans.
In total, 7 design parameters, related to blade loading, spanwise rV0 and stacking condition were used to parameterize the blade. The Design of Experiment method was used to create an experimental table consisting of 36 different geometries. All of these geometries were analysed in CFD at 3 different points and also their noise was predicted.