A 3D inverse design based multidisciplinary optimization on the radial and mixed-inflow turbines for turbochargers

A methodology for designing radial and mixed-inflow turbines to meet multiple aerodynamic and mechanical requirements is presented in this paper. The method couples a 3D inverse design code and Design of Experiment Method (DoE) and the response surface method (RSM) to design turbines which meet various design criteria. Initially there are 17 design parameters and 15 performance parameters. The performance parameters include aerodynamic objective at multi operating points such as efficiency and mass flow factor as well as stress and vibration parameters and turbine inertia. A DOE table is created consisting of 25 different configurations. The resulting response surface is then used to investigate the sensitivity of different design parameters on the performance parameters. By using the sensitivity information two different impeller geometries are designed. In design1, the design parameters are set up in such a way as to improve the efficiency while in design2 the aim is to reduce the turbine inertia. CFD and FEA is run on both designs and the results confirm that the design1 has higher efficiency than the baseline impeller but at expense of higher stress and lower impeller stiffness.