In order to extend the operating range of a centrifugal compressor, inclined discrete cavities located upstream of the impeller leading edge were optimized in this work. Aerodynamic performance analysis was performed using three-dimensional Reynolds-averaged Navier-Stokes equations with the shear stress transport turbulence model. A parametric study on aerodynamic performances of the centrifugal compressor with the inclined discrete cavities was conducted with six geometrical parameters. Through the parametric study, three geometric parameters were selected as design variables for optimization. Peak adiabatic efficiency and stall margin were selected as objective functions. The Latin hypercube sampling method was used to select the design points, and the radial basis neural network was used to construct surrogate models of the objective functions. A hybrid method combining the particle swarm optimization showed better overall performance in finding global optimum than the genetic algorithm. Pareto-optimal solutions provided the designs which enhance considerably both the performance parameters compared to the reference design.

This content is only available via PDF.
You do not currently have access to this content.