Integrated optimization of turbomachinery via discrete adjoint and hierarchical free form deformation parameterization method

Aerodynamic optimization is an effective technology to improve the performance of turbomachinery. Traditional aerodynamic optimizations often focus on blade profiles and neglect the interactions between the blade and end wall, which may lead to suboptimal optimization result. To address this challenge, this study introduces an integrated optimization framework to simultaneously optimize both the blade and hub end wall aerodynamic shape. A hierarchical free form deformation (FFD) technique is employed to achieve integrated parameterization of the blade and hub, with a total of 580 design variables. To ensure efficient optimization with such a high-dimensional problem, the discrete adjoint method is used, which makes the optimization cycle less than one day. Under the framework, the blade optimization and integrated optimization of a typical compressor fan, rotor37, are conducted. With strict constraints on flow rate and total pressure ratio, the integrated optimization achieves a 2.85% improvement in isentropic efficiency—about 0.6% higher than blade optimization. The flow control mechanisms of two optimization approaches are compared, revealing that both blade optimization and integrated optimization effectively modify the blade profile to weaken the shock wave. Furthermore, by modifying the hub aerodynamic shape to improves the flow structure near the end wall, integrated optimization it is more efficient in turbomachinery aerodynamic optimization.