Optimization of a centrifugal blower using a gradient-based approach
Please login to view abstract download link
Nowadays, developing more efficient turbomachines is required to minimize their energy consumption. Among the various optimization approaches, the adjoint method (gradient-based) is gaining popularity because of its low computational cost. It can compute the derivatives of an objective function to a large number of design variables [1]. This study proposes an automatic multi-objective optimization framework based on the adjoint method to optimize a centrifugal blower. The aim is to improve hydraulic power while reducing mechanical power, leading to an enhanced total efficiency. Multiple Python scripts link the automatic framework, which comprises Salome (CAD and mesh generators), OpenFOAM v2206 (CFD solver and optimizer), and Python scripts for post-processing. Reynolds-Averaged Navier-Stokes (RANS) equations closed by the k-ω SST model are used to simulate the turbulent flow inside the centrifugal blower. The whole numerical procedure has been first carefully validated against experimental data for two autonomous underwater vehicles [2]. Figure 1 displays a result sample for the centrifugal blower. It shows the initial and optimized blade profiles. After four optimization cycles, the mechanical power decreases by 0.71% and the hydraulic power rises by 0. 73%, resulting in a 1.5% gain in total efficiency.