Abstract
Low pressure axial fans have a very wide range of applications in the field of heating, ventilation and air conditioning as well as refrigeration (HVAC&R). Due to stricter legal requirements, increasingly higher minimum fan efficiencies are required. That is why the optimized design of the fan is a crucial point in the development process. Therefore, the simulation-driven preliminary design strategy and optimization is very important in the product development of low-pressure axial fans.
For a specified design point, the blade work distribution along a finite partial section of the impeller flow channel was predetermined. From that, the distribution of the meridian velocities was computed applying the simplified radial equilibrium and the Euler equation for turbomachinery. Then, the blade angles, i.e. the shape of the blade of the axial impeller, were determined. Using the continuity equation and the simplified radial equilibrium, the method has also been extended for non-constant hub shapes, for example a conical hub shape. The blade shape was parameterized by 9 angles, i.e. three angles at the impeller inlet (leading-edge section), three angles at the mid-streamwise section and three angles at the impeller outlet (trailing-edge section).
The angles for the hub plane, the mid-span plane and the shroud plane were determined accordingly. The geometry of the impeller was generated, modeled and solved for the performance characteristics with the numerical flow simulation software ANSYS CFX.
In a subsequent optimization process, based on these pre-designs the parameter space of the leading edge was populated by a DOE analysis using the Latin Hypercube sampling method. Then a multi-objective optimization procedure (MOGA-II) has been performed with the optimization constraints of maximum efficiency and maximum pressure and, at the same time, minimum flow blocked area in the blade channel. From this optimization approach the optimal design was determined, i.e. the optimal angles at the impeller inlet (leading-edge). The same optimization procedure was then also conducted for the mid-streamwise angles and finally for the angles at the impeller outlet at the blade trailing-edge.
Then, for the optimized axial fan, the complete performance characteristics, i.e. at the design and off-design points, have been calculated with CFD showing very good improvements over the full performance characteristics. The method and the results are discussed in detail.
