Evaluation of the Rotor Temperature Distribution of an Automotive Turbocharger Under Hot Gas Conditions Including Indirect Experimental Validation

Turbocharging is a key technology to improve performance and efficiency of combustion engines. Operating behavior of the turbocharger highly depends on the rotor temperature distribution as it directly modifies viscosity and clearances of the fluid film bearings supporting the rotor. Therefore, rotordynamic and bearing friction characteristics both change depending on the concrete thermal state. However, a direct experimental identification of the rotor temperature of an automotive turbocharger with acceptable expense is not feasible. Hence, a combination of numerical analysis and experimental identification of the accessible region is applied to investigate its temperature characteristic and level.

In a first step, a numerical model of the entire investigated automotive turbocharger including the most important solid components and the fluid domain of turbine and compressor is developed using a commercial CFD-tool. To obtain physically consistent rotor temperature distributions under consideration of the impact of the fluid film bearings, a bidirectional, thermal coupling of the CFD-model to thermo-hydrodynamic thrust and journal bearing codes is implemented. Secondly, experimental investigations of the numerically modelled turbocharger are conducted on a hot gas turbocharger test rig for selected operating points. Here, rotor speeds range from 64 to 168 krpm. Turbine inlet temperature is set to 600°C and the lubricant is supplied at a pressure of 300 kPa with 90°C to ensure practically relevant boundary conditions. The turbocharger housing features about 100 thermocouples to gain an insight on dominant effects of heat transfer in the solid domain. Moreover, pressure sensors and proximity probes are applied to monitor global parameters. While pressure measurements provide validation of the numerically determined thrust forces, the proximity probes enable to identify floating ring speed in the journal bearings.

Comparisons of measured and numerically predicted local temperatures of the turbocharger housing indicate good agreement of both analyses. Furthermore, local temperatures in the thrust bearing correspond very well. For validation of the predicted rotor temperature the calorimetrically determined frictional power loss of the bearings as well as the floating ring speed are used as additional parameters. To quantify the thermal impact of the bearings on local rotor temperature, a comparison between a diabatic and an adiabatic approach at the rotor-bearing interfaces is conducted and shows fundamental differences. Moreover, evaluation of heat flow of diabatic simulations indicates a high sensitivity of local temperatures on rotor speed and load. A significant cooling effect of the journal bearings is present, involving high temperature gradients between the two bearing ends, especially on the turbine side. Consequently, results confirm the necessity of the diabatic approach in heat flow analysis of turbocharger rotors.