Numerical and Experimental Investigation of the Impact of Mixed Flow Turbine Inlet Cone Angle and Inlet Blade Angle

Abstract

Mixed flow turbines (MFTs) offer potential benefits for turbocharged engines when considering off-design performance and engine transient behavior. Although the performance and use of MFTs are described in the literature, little is published on the combined impact of the cone angle and the inlet blade angle, which are the defining features of such turbines. Numerical simulations were completed using a computational fluid dynamics (CFD) model that was validated against experimental measurements for a baseline geometry. The mechanical impact of the design changes was also analyzed. Based on the results of the numerical study, two rotors of different blade angle and cone angle were selected and manufactured. These rotors were tested using the Queen's University Belfast (QUB) low-temperature turbine test rig, which allowed for accurate and wide-range mapping of the turbine performance to low values of the velocity ratio. The performance results from these additional rotors were used to further validate the numerical findings. The numerical model was used to understand the underlying physical reasons for the measured performance differences through detailed consideration of the flow field at the rotor inlet and to document how the loss mechanisms and secondary flow structures developed with varying rotor inlet geometry. It was observed that large inlet blade cone angles resulted in strong separation and flow blockage near the hub at off-design conditions, which greatly reduced efficiency. However, the significant rotor inertia benefits achieved with the large blade cone angles were shown to compensate for the efficiency penalties and could be expected to deliver improved transient performance in downsized automotive engine applications.

Evaluating the Use of Leaned Stator Vanes to Produce a Non-Uniform Flow Distribution Across the Inlet Span of a Mixed Flow Turbine Rotor

“…The maximum mesh y+ value achieved was below 11, and for the majority of the surfaces the y+ value was resolved to below 3. To maintain continuity with existing hardware on the QUB cold flow test rig, the geometry and meshes for the pre-swirl vanes and exhaust diffuser were carried over from a previous study at QUB which had already conducted a grid independence study and validation with experimental results [15]. The numerical model used in this study also closely matched that used by Leonard et al [15] as this modelling approach demonstrated good agreement with experimental results.…”

“…The numerical analysis employed the shear stress transport (SST) turbulence model, as it has provided good agreement with experimental test results in previous studies at QUB [15,16]. The meshes used in the numerical analysis consisted only of structured hexahedral elements.…”

“…This allowed for flow field measurements to be gathered across the span of the rotor outlet. Further details on the test facility are provided by Leonard et al [15].…”

Conjugate heat transfer study of backdisc cooling of a radial turbine

“…CFX ® 16.0 was applied to solve Reynolds-averaged Navier–Stokes equations and shear-stress transport (SST) two-equation turbulence model which has been extensively used in the studies of turbomachinery. 15,19,20 The computational model of the turbine included turbine housing, the full 360° channel of the rotor, and the backdisc cavity and solid regions of the turbine. Frozen-rotor method was employed at the rotor–stator (housing) interface.…”

The influence of a radial turbine volute on back-disk impingement cooling performance

“…ANSYS-CFX 16.0 was used to solve the full three-dimensional N-S equations. An SST-kω two-equation turbulence model, which has been widely tested and verified for use with turbines, was selected for the turbulence model [10,16,17]. A frozen-rotor method was used to process the data transfer between the turbine channel domain and the turbo wheel channel domain.…”