Numerical analysis of a turbocharger compressor

Abstract

The automotive industry is under obligation to meet regulations for emission control that has resulted in further use of turbochargers in passenger cars to enable downsizing and increase engine power density. In this study, a set of numerical simulations are conducted along two turbocharger compressor speed lines of 150,000 rpm and 80,000 rpm to analyse and validate the results against experimental data. The domain includes the full compressor stage comprising intake, impeller as a Multiple Reference Frame, diffuser and outlet. The k-omega SST turbulence model with three different mesh sizes is used tosolve the compressible flow using ANSYS Fluent software. Three points on each speed-line are selected:one point each in regions close to surge and choke and a point in the stable zone of the compressor map. The simulations predict compressor performance in terms of the total-to-total pressure ratioand total-to-total efficiency. Results reveal the predicted pressure ratio error is in the range of 1-6%. At 150,000 rpm the pressure ratio is underpredicted for the point close to the surgebut overpredicted for the point close to the choke. However, the pressure ratio results are within 1% difference for 80,000 rpm. In all cases, the predicted efficiency increased when a finer mesh is used.While results are close to the experimental data in both the surge and stable areas of the map, the efficiency wasoverpredicted up to 20% in the region close to the choke. In conclusion, the finer mesh leads to higher pressure ratio and efficiency values that overpredict the performance, especially for the pointclose to choke.