Endwall profiling for the reduction of secondary flow in turbines
This thesis describes investigations into the use of a technique for improving the efficiency of axial flow turbines. The flow in the turbine component of axial flow machines is complex, with a number of three-dimensional features. In order to extract power from a stream of high pressure and high temperature flow this flow must be turned through a large angle, this high turning introduces a phenomenon know as "secondary flow". This secondary flow introduces additional loss, unsteadiness and regions of high heat transfer into the machine - all of which are undesirable features. Endwall profiling aims to reduce these undesirable features by shaping the end-wall between the turbine blades. The shaping either accelerates the flow which reduces the local static pressure or retards the flow which increases the static pressure. These effects are confined to a region near the endwall so the overall performance of the blade row is not affected. However due to the complexity of the flow it is easy to make things worse rather than better! - careful design is needed. This thesis aims to understand how and why the reductions in may be achieved so that they can be better exploited as well as providing information of the performance of a major engine manufacturers design system. The thesis describes pressure probe measurements inside and outside of the blade passage of a low speed linear cascade with a number of profiled endwall geometries. The aerofoils used in the cascade are already relatively efficient and the overall loss changes are small, accurate measurement is therefore very difficult. The current best profiled endwall reduces secondary loss by 30%±5% compared to the planar case. Hot film measurements have been conducted on the endwalls and suction surface of the blade to determine if these benefits are substantially due to changing the boundary layer state. The results from this thesis indicate that this is not the case. This thesis describes measurements on three generations of profiled endwalls, two of which successfully reduce loss, one does not. The success of the first two endwalls indicates the power of current CFD based design practices, the failure of the third design to reduce loss illustrates some of the shortcomings of current CFD based design practices. The information from this thesis is being used in the design of the next generation of aircraft engines to which non-axisymmetric profiled endwalls are being fitted.