Discussion And Analysis On Several Issues About The Numerical Simulation Of The Axial Flow Fan

In the last decade, the research about the optimization and noise reduction of the axial flow fan, which based on CFD numerical simulation technology, has served as a hotspot. However, there has been no generally accepted standard for the selection of the meshing technology, turbulence model, and difference scheme, or for the setup of turbulence parameters belonging to the inlet boundary condition. In this paper, the impingement of the above factors for simulating the axial flow fan are discussed on the basis of low pressure axial flow fan with 810mm diameter. Suitable selection of above factors has been concluded for reference.First, two grid technology, structured mesh and unstructured mesh, are adopted for the geometry model of the fan. Then gird independence is tested based on the orthogonal design experiment. User-defined function is employed to analyze causes of the difference between two kinds of gird model in accordance with the point of dissipation vortex. Unstructured meshing model can reduce cell distortions which surrounding the largely distorted blade with its flexibility, so as to improve the accuracy of the fan performance simulation.The axial flow fan performance is simulated based on S-A turbulence model, Standard k-ε turbulence model, RNG k-ε turbulence model and Standard k-ω, SST k-ω turbulence model. Under the guidance of rapid-distortion theory, reasons for different simulation results in these turbulent models are explored based on vortex size and Sk/ε. It has been drawn that the RNG k-ε turbulence model is most suitable for the axial flow fan simulation. Meanwhile, the flow field of the axial flow fan with different inlet turbulence parameters is analyzed, through which, reasonable inlet turbulence parameters settings of different turbulence models are obtained.Finally, the influence of different difference schemes and algorithms on the performance of axial fan is discussed on the basis of two grid models. It results that difference schemes have a big impingement on the axial fan performance simulation. For the unstructured grid model of the axial flow fan, the simulated results of the Standard pressure difference scheme is more accurate; while for structured grid model, the PRESTO! difference scheme registers the more accurate one.