Investigation of a novel secondary flow feature in a turbine cascade with end wall profiling.

A novel secondary flow feature, previously unreported for turbine blading as far as the authors are aware, has been discovered. It has been found that it is possible to separate part of the inlet boundary layer on the blade row end wall as it is being over-turned and rolled up into the passage vortex. This flow feature has been discovered during a continuing investigation into the aerodynamic effects of non-axisymmetric end wall profiling. Previous work, using the low speed linear cascade at Durham University, has shown the potential of end wall profiling for reducing secondary losses. The latest study, the results of which are described here, was undertaken to determine the limits of what end wall profiling can achieve. The flow has been investigated in detail with pressure probe traversing and surface flow visualization. This has found that the inlet boundary locally separates, on the early suction side of the passage, generating significant extra loss which feeds directly into the core of the passage vortex. The presence of this new feature gives rise to the unexpected result that the secondary flow, as determined by the exit flow angle deviations and levels of secondary kinetic energy, can be reduced while at the same time the loss is increased. CFD was found to calculate the secondary flows moderately well compared with measurements. However, CFD did not predict this new feature, nor the increase in loss it caused. It is concluded that the application of non-axisymmetric end wall profiling, although it has been shown to be highly beneficial, can give rise to adverse features that current CFD tools are unable to predict. Improvements to CFD capability are required in order to be able to avoid such features, and obtain the full potential of end wall profiling