3D Inverse Design Based Optimization of Multi-Blade Row Axial Fans Used for Distributed Propulsion

Development of Analytical and Computational Methods

In many fan applications such as electronic cooling or tunnel ventilation rotor/stator or contra-rotating fans are used. In many cases because of packaging limitations the blade rows have to operate with quite small axial spacing and hence there is strong interaction between the blade rows. It is hence important to be able to design and optimize these fans taking account of the strong interactions between blade rows.
In this paper, a new methodology is presented for design of multi-blade row fans where the blade rows are designed by using a viscous 3D inverse design method which takes account of the strong interaction between blade rows. In this inverse design method the blade geometry is computed for a specified distribution of blade loading subject to a specified spanwise work distribution. The unique feature of the method is to combine a 3D viscous inverse design method with a robust mixing plane method satisfying interface flux conservation, non-reflectivity and retaining interface flow variation. The method can be used in all Mach numbers by using a low Mach number preconditioning technique.
In this paper the methodology used is presented and then applied to a rotor/stator axial fan application. The method is used to design the fan rotor both in stage mode ( rotor/stator) and in single blade row mode by using the same blade loading for the rotor. Hence only the rotor geometry is changed and the stator geometry is not changed from the baseline design. The flow field in the resulting stages (New rotor designed in stage mode with baseline stator and the new rotor designed in single blade row mode with baseline stator) are then analysed by using commercial CFD code CFX and significant improvement in stage efficiency is observed for the stage design.