58901 - Aeromechanical Optimization of Scalloping in Mixed Flow Turbines 

Scalloping of radial and mixed flow turbocharger turbine rotors has been commonplace for many years as a means of inertia reduction and stress relief. The interest in inertia reduction is driven by transient loading requirements of turbocharged internal combustion engines, where the time taken for the engine to reach the required torque output during a load step is of great interest in many applications. A key factor in the time taken to meet transient engine torque requirements is the turbocharger inertia. Due to the high density materials used in turbine rotors, any material removal from the turbine wheel has a significant impact on turbocharger inertia, and thus the transient response of the engine.

It is well known that scalloping not only reduces inertia, but also efficiency. Naturally this is an unwanted side effect; however, previous literature has shown that when considering transient performance alone this is generally a worthwhile trade off, as inertia is the dominant influence on performance. For applications where transient response is of secondary importance, such as base load power generation, or cases where turbine wheels are scalloped for mechanical considerations, the impact on efficiency is of significant importance. This study aimed to identify if it was possible to produce a new scallop design which reduced the scalloping efficiency penalty without increasing inertia, or compromising mechanical constraints. This was carried out with the view to providing design recommendations for scalloping where a complete minimization of inertia is not the design goal.

A multipoint, multi-physics numerical optimization, with constraints on inertia and back disc stress, was carried out to determine what efficiency benefit could be realised by aerodynamically designing mixed flow turbine scallops. An efficiency benefit was identified across the entire turbocharger operating line, with increased benefit towards part load, whilst not exceeding the design constraints. Scalloping losses for the baseline design were found to be greatest at part engine load, explaining why an aerodynamic redesign yields the greatest benefit under those operating conditions. These performance predictions were experimentally validated on the QUB cold flow test rig, with good agreement between simulated and measured data.

To conclude the study, a detailed loss audit was carried out to identify key loss generating flow structures, and also how changes in geometry affected the formation and development of these flow structures throughout the passage. A large vortex which enters the passage from the scalloped region and interacts with leading edge and tip leakage vortices along the suction surface of the blade was identified as the main source of scalloping loss. The optimized design was found to better control the entry of this vortex into the passage, thus reducing the associated loss, and facilitating a performance improvement. Geometric design guidelines are then outlined based on these findings.