Analysis of a Variable Geometry Radial Inflow Turbine Using 1D Mean-line Model

Abstract

Despite great progress in 3D computational fluid dynamics (CFD), mathematical modeling used for turbomachinery analysis, the prediction of flow parameters in the rotor (impeller) of radial inflow turbines using existing turbulence models often yields unsatisfactory results. This is because real physical processes related to turbulence are very complex. Other limitations of CFD tools include numerical errors due to finite difference approximations. Thus, for the analysis and optimization of radial inflow turbines, empirical loss correlations applied to a mean-line 1D mathematical model can provide faster and more accurate results compared to CFD simulations. In this study, a 1D approach was used for the mathematical modeling of gas flow in a variable geometry radial inflow turbine with a vaneless volute distributor (VVD). The variable geometry turbine (VGT) was part of a turbocharging system. Combustion engine operation mode with high-power and low-crankshaft rotational speed was chosen, as engine performance under this condition is sensitive to turbocharger adjustment in terms of break specific fuel consumption (BSFC) and other critical parameters. An experimentally-based approach was used where gas flow characteristics in the VGT were calculated under real engine-turbocharger operating conditions. Accordingly, exhaust gas parameters were first measured on an experimental setup. Mathematical modeling then allowed to assess how adjustment in VGT influenced turbine indices, including loss distribution across the stage. It was also validated in this study that during high-load and low-crankshaft operating conditions of the compression ignition (CI) engine, it is recommended to minimize the cross-sectional area at the outlet of the volute acceleration section in the VGT. This adjustment results in better fuel efficiency and lower thermal loads on turbine components, despite a slight reduction in the turbine’s overall efficiency.