An Analytical Approach to Support High Cycle Fatigue Validation for Turbocharger Turbine Stages

Abstract

As the pursuit of improved aerodynamic performance on turbochargers continues to push the boundaries of mechanical design, the risk of high cycle fatigue (HCF) failures of turbine wheels is elevated and drives the need for improved methods of analytical predication. Turbine wheel HCF is caused by the asymmetrical aerodynamic loading associated with the application of variable geometry turbines and compromised inlet/outlet geometries that accommodate packaging turbochargers in engine compartments. Historically, the turbocharger industry has focused on blade pass as the primary cause of HCF. BorgWarner presents an analytical technique to limit the risk of HCF by calculating all critical orders that intersect blade natural frequencies and quantify the relative energy of each forced excitation frequency. The geometry of a turbine volute mainly determines the inflow angle into the turbine wheel. Thus, for the thermodynamically optimized turbine a variable inflow angle dependent on the engine operating point is desirable. The Variable Turbine Geometry (VTG) concept applies adjustable turbine inlet guide vanes to approach the ideal velocity triangle. One of the inevitable disadvantages associated with either fixed or variable turbine nozzle vanes is the generation of wakes and pressure fluctuations upstream of the turbine wheel inducer. The resulting circumferentially non-uniform flow conditions apply a transient load on the rotating turbine wheel. Due to complexity of the VTG design including non-uniformly spaced vanes and struts, the excitation sources and resulting excitation orders are not readily apparent. The analytical method described in this paper applies a transient 3D Computational Fluid Dynamics (CFD) model of the rotor-stator interaction to calculate the time-dependent pressure fluctuations experienced by the turbine wheel blade. This data is used to extract the forced excitation function at the turbine wheel. Finite Element (FE) analysis is applied to determine the mean and dynamic stress. In case of dynamic stress, system vibration modes and the influence of local harmonic excitation on blade stress amplitudes is analyzed. The fixed-speed FE results are scaled for the effect of flow and speed by use of empirical data from strain-gauge measurement. Hence, computational methods combined with experience from experimental measurements are used to determine critical rotational speeds for a given turbocharger geometry. This method allows the analyst to predict the highest energy excitation orders and reduce the risk of turbine wheel fatigue damage. Since the durability of a turbine wheel cannot yet satisfactorily be quantified by the described computational method, the analysis results are used as rotational speed input in subsequent durability tests in order to reduce the necessary amount of testing resources.