This CFD benchmark activity for a box fan has been conducted through the Industrial Committee for Supercomputing Promotion (ICSCP) and the Fan Noise Prediction Working Group from the “Promotion of Practical Use of HPC in Turbomachinery” Subcommittee established in the Turbomachinery Society of Japan. In this benchmark activity, various CFD software has been used.
In this study, ANSYS CFX 18.2 is used and the difference in results are investigated by changing the turbulence model and the mesh model. The results between the k-epsilon model and the Shear Stress Transport (SST) model show a reverse trend. That is to say, the k-epsilon model overestimates the pressure difference at the high flow range and the SST model underestimates it at the low flow range compared with the experimental result. In order to improve these numerical results, CFD analysis is conducted by changing mesh models. The number of the blade surface mesh, and the initial height, height ratio, and number of prism layers are changed, but there is no improvement through these approaches. In this paper, CFD verification and validation are conducted for the box fan.