Reduction Method for the Secondary Flow Loss in Turbine Cascade

The aerodynamic loss accounted to the secondary flow, or secondary loss, is one of the most prominent cause of the internal losses in turbine cascades. Secondary losses are mostly due to the interaction of Horse-Shoe Vortex and endwall crossflow. Authors have developed a so called endwall fence to reduce the secondary loss in gas turbine cascade by mitigating this interaction. However, the improvements resulting by the application of this method to a steam turbine cascade proved to be not so remarkable as for the gas turbine cascade. By conducting total pressure measurements at the endwall, previous results showed that the stream wise evolution of the losses presents two peaks; Horse-Shoe Vortex peak and crossflow peak. The abovementioned endwall fence is blocks the Horse-Shoe Vortex, but its screening effect on the crossflow is not sufficient. So in this research the use of two kinds of fences is proposed to reduce the secondary loss: one is to block the pressure side leg of the Horse-Shoe Vortex, the second is to block and guide the crossflow, which starts from the middle of the blade passage, toward the outlet. Therefore, they are called respectively Horse-Shoe Vortex fence and crossflow fence. The tests are carried out using the wind tunnel cascade and the test fences are made by 3D printer. The total loss is estimated by means of an automatically moving pitot tube located downstream the cascade. Many kinds of fences are tested and the optimum fence shape and position to minimize the secondary loss is obtained.

Reduction in Secondary Losses in Turbine Cascade Using Contoured Boundary Layer Fence

Present work deals with reducing secondary losses in turbine cascade by using boundary layer fences in two ways. Firstly, to reduce the strength of vortex which is incident at the leading edge of airfoil and hence reduce the strength of horse shoe vortex, and secondly, to reduce the pressure gradient between the pressure side and the suction side in the flow passage region between airfoils. In previous works, the boundary layer fence followed the profile of airfoil. In this publication, boundary layer fence does not follow the profile of airfoil i.e stagger and camber of boundary fence is different when compared to airfoil. A profiled boundary layer fence is proposed in the present work which reduces the incident vorticity and also reduces pressure gradient from pressure side to suction side. Such boundary layer fence was checked on T106 test cascade which is available as open literature. Numerical work is carried out using commercial software Ansys CFX. Viscous RANS simulations are carried out using k-ω SST turbulence model with yplus value around unity on all walls. Coefficient of secondary kinetic energy (CSKE) and Secondary Kinetic energy helicity (SKEH) are used as target functions. Total pressure loss is also monitored. All the three functions show a reduction in secondary loss. The strength of horse shoe vortex is reduced by the fence protruding in front of leading edge. The converging flow passage created by the fence near the pressure side of airfoil reduces the pressure gradient from pressure side to suction side. The total pressure loss was reduced by 1.5 % and CSKE was improved by 36 % when the boundary layer fence was adopted.

The Effect of Leading Edge Diameter on the Horse Shoe Vortex and Endwall Heat Transfer

The endwall of the first stage vane / blade of modern high temperature gas turbine has been exposed to severe heat transfer environments. Due to the formation of a horse shoe vortex (HV), the flow field of a vane and blade leading edge juncture to endwall is especially complicated and it is difficult to estimate the heat transfer coefficients and the film cooling effectiveness levels in this area. This paper describes the results of experimental and numerical studies on the heat transfer and flow dynamics in the leading edge endwall region of a symmetric airfoil. The effects of inlet velocity, boundary layer thickness and leading edge diameter of a symmetric airfoil were investigated on the endwall heat transfer in a low speed wind tunnel facility. The time averaged local heat transfer coefficients were measured by naphthalene sublimation method and the instantaneous velocity field was obtained by Particle Image Velocimetry (PIV). As the leading edge diameter of symmetric airfoil decreases, the heat transfer coefficients on an endwall increases and is proportional to Re0.71 that is base on the leading edge diameter. However, the boundary layer thickness was found to have a marginal effect on the endwall heat transfer.

Endwall Boundary Layer Control in Compressor Cascades

Recent investigations have shown a reduction of secondary losses in compressor cascades using a bulb like modification of the profile at the endwall. This paper is focussed on experimental work in comparison of 5 different endwall modifications at a compressor cascade. The cascade is modified near the endwall with a bulb, a medium and a large fillet. The fillet configurations are modified by an axial blunt cut-off at the leading edge. The investigations have been carried out at a profile developed from a hub section of the Dresden Low Speed Research Compressor (LSRC) blade, a compressor profile with a nominal turning of 18 deg. A datum configuration and the 5 other configurations were tested at the Low Speed Cascade Windtunnel (LSCW). For the bulb configuration, an intensified horse shoe vortex was suspected and observed counterrotating to the passage vortex with an influence on its propagation. The interaction of the passage vortex and the suction side profile boundary layer is influenced. The superposition of both is minimized and the losses developing from this effect are significant lower. For the fillet and blunt-fillet configurations, a fillet vortex develops and was observed co-rotating to the passage vortex with an influence on the mentioned interaction as well. Blunt leading edges produce additional losses but the superposition of the growing vortices may reduce the overall losses. The cases show a reduction in losses of 1.9% for 3 deg incidence and a range of 1.2% rise to 1.9% reduction in dependence of the incidence. This equals a reduction of the isolated secondary losses up to 28% with respect to the reference profile. Detailed results of the experiments are presented for the reference and all modified cascades.