Aerodynamic Optimization of an Axial Compressor Using Directly Manipulated Free-Form Deformation and Data-Driven Optimizer

Yi Liu,Jiang Chen,Hang Xiang
DOI: https://doi.org/10.1007/978-981-97-4010-9_111
2024-01-01
Abstract:Due to the massive design parameters and complicated flow conditions, the aerodynamic optimization of axial compressors is facing with two main difficulties: numerous variables and expensive evaluations. To address this issues, two main methods are developed in this paper: the parameterization method and a data-driven optimization algorithm. As a dimension reduction method, the directly manipulated free-form deformation (DFFD) method is introduced to parameterize the geometry of axial compressor. DFFD directly put the control points on the geometry of the blades, flow path and corners of the axial compressor. To save the total expensive evaluation times, this paper develops a data-driven optimization algorithm, named pre-screen surrogate model assisted particle swarm optimization algorithm, which can get competitive optimization result in limited true evaluate steps. Within these two methods, a fast aerodynamic optimization platform is established. A transonic axial compressor is optimized. The efficiency and the surge margin are increased by 3.83% and 6.04%, which verifies the effectiveness of the platform.
What problem does this paper attempt to address?