In order to hasten the thermal design for forced convection electronic devices, cooling fans should be modeled to reduce a computational load. A fan-curve-model, which generates volumetric flow rate versus the characteristics pressure difference of a fan, is very simple and usually incorporated into commercial CFD codes. However, this model often results in an erroneous flow rate. In this work, both the experiments and the CFD simulation were performed around small axial-flow-fans of 30 and 40 mm in diameter. The measured PQ curve was applied to the fan model, and compared the result of the simulation to the experimental data. It was clarified that the major reason behind the disagreement was the difference in the pressure definition of the fan model from the PQ curve measured using a chamber. Based on this, a simple method was proposed to correct this definition. Also, the system effect, which is the impact of obstacles on the fan delivery curve, was investigated by setting a cylindrical obstacle at upstream or downstream proximity of the fan.

This content is only available via PDF.
You do not currently have access to this content.