Influence of nozzle clearance of VG turbocharger turbine on performance and internal flow

Abstract

Abstract To realize a sustainable society, it is necessary to reduce exhaust gas from automobiles and improve fuel efficiency. Turbochargers are useful for the purpose and especially variable geometry (VG) turbochargers are often installed because their nozzle blade angle can be optimally changed according to the rotational speed of the engine. To change the angle, it is needed to provide the nozzle clearance on both hub and shroud sides. In this study, CFD (Computational Flow Dynamics) calculations of a turbocharger turbine were conducted. The rotational speed of the turbine was set to 10,000 rpm. The diameter of the wheel blade is 60 mm and the height of the nozzle blade was 9.71 mm. The following were mainly investigated: the effect of the nozzle clearance position or blade angle on turbine performance, and the effect of leakage vortex from nozzle clearance on the downstream flow in the wheel domain. First, the steady-state CFD analysis on the nozzle and wheel domain was conducted. Four models of nozzle clearance were used: No Clearance (NC), Hub Clearance (HC), Shroud Clearance (SC), and Hub and Shroud Clearance (HSC). And four nozzle blade angles were used. With the Shroud Clearance model, the torque efficiency was worst with every nozzle blade angle. It was also found that the leakage flow rate from the nozzle clearance affects the loss increase in the wheel domain. Second, the unsteady-state CFD analysis on the full domain (casing + nozzle + wheel +diffuser) was conducted to investigate the leakage flow from nozzle clearance. The turbulence model was DES (Detached Eddy Simulation). Hub Clearance model and Shroud clearance model were used. The leakage vortex generated from nozzle clearance was found to flow into the wheel domain. With each nozzle model, the rotating directions of those vortexes were opposite. By the rothalpy loss investigation with the Hub Clearance model, it was confirmed that the portions of wheel domain where two leakage vortexes flow into showed larger losses than other portions. Those portions were located diagonally in two places, and they rotated in the opposite direction to the wheel rotation. On the other hand, with the Shroud Clearance model, the portions where the rothalpy loss was large were not shown regularly.