Flow-field in a linear cascade of filleted vanes with endwall film-cooling

Abstract

The optimum performance of the turbine, while adhering to certain design limitations, is to improve the aerothermal performance of axial flow turbines by reducing the total pressure losses and increasing the film-cooling coverage on the endwall of the blade and vane passages. The secondary flows near the passage endwall and film-cooling hole configurations in the endwall primarily increase the passage aerodynamic losses. According to previous studies, a leading-edge fillet, which fills the intersection of the nozzle guide vane (NGV) and the endwall, has the potential to mitigate the secondary flows. The film flow coverage on the endwall from the endwall coolant holes is also negatively affected by the secondary flows. Past investigations indicate that the interactions between the endwall secondary flows and cooling flows drag the cooling flow towards the passage suction-side and away from the endwall region leaving the pressure-side endwall unprotected from the hot combustion gas. The interactions of secondary flows with film-cooling flow often add to the total pressure losses. This thesis experimentally investigated the detailed flow field in a filleted vane cascade, employing endwall film-cooling flow. The objectives were: (i) to mitigate the endwall secondary flows in the vane passage through the benefit of a leading-edge fillet, (ii) to combine the effects of fillet with the effects of some new configurations of the leading-edge coolant holes in the endwall in order to cover the pressure-side of the passage endwall with the coolant flow, and (iii) to reduce the total pressure losses in the passage when employing film-cooling flow. The experiments were performed in a low-speed linear vane cascade, employing the GE-E3 first-stage nozzle guide vane profile. A new leading-edge fillet configuration was used as a method of mitigating the negative effects of the endwall secondary flows and film-cooling flow on the total pressure losses. The film cooling flow was delivered through the holes located in the endwall, upstream of the cascade as well as inside the cascade passage. The upstream coolant hole configurations included continuous and discontinuous slots, discrete cylindrical holes, and cylindrical holes with diffused exits. The passage endwall employed discrete cylindrical holes at the non-zero compound angles along an isobar line. The injection angle of all the coolant hole configurations was 30° relative to the main flow. The inlet Reynolds number of the main flow was 2.0 E+05 based on the vane chord. The average blowing ratio of the film-cooling flow ranged from 1.2 to 2.8 and the temperature ratio and the density ratio of the coolant to main flow were 0.94 and 1.05 respectively. Measurements of the flow field were obtained in four pitch-wise normal planes along the cascade passage and presented as the velocity vectors, vorticity vectors, flow angles, total pressure losses and flow temperatures. The results included baseline data with and without fillet and film-cooling flow to evaluate the effects of fillet without film-cooling flow, the effects of film-cooling configurations without fillet, and the effects of fillet combined with film-cooling flow. The main observations from the measurements were as follows: The fillet reduced the total pressure losses in the vane passage because it weakened the passage vortex in the endwall region. The distribution of flow temperature showed that the presence of film cooling in the endwall region is better with LE discrete cylindrical holes and LE slots compared to other cases. When the coolant was added to the passage holes, the LE discrete cylindrical hole/slot film-cooling configuration still performed better than the other cooling configurations because the coolant slipped under the boundary layer separation region, cooling the entire endwall. The worst cooling behaviour was noticed for the LE discrete cylindrical hole configurations with the fillet and passage coolant holes. The coolant penetrated into the mainstream higher above the endwall before the endwall boundary layer separation region than when the LE diffused hole configuration was employed at all blowing ratios. The maximum increase in the (Cpt,loss)Mass-av of the LE discrete cylindrical holes and LE slots cooling configuration is by about 40% at M=1.4. The maximum decrease in the (Cpt,loss)Mass-av of the LE diffused cylindrical holes cooling configuration is by about 11% at M=1.2 The LE discrete diffused hole configuration provided the best performance without the filleted vanes. For the baseline configuration, the diffused cylindrical holes, with the passage coolant holes, provided slightly better performances than for the configuration provided by the LE cylindrical holes in terms of the spread of coolant coverage along the endwall, less turning of streamline coolant flows towards the suction-side, and reducing the pressure losses along the passage. The LE discrete cylindrical holes/slots with the fillet and passage coolant hole configuration showed a higher reduction in total pressure losses and exhibited the best cooling ability for pitch-wise (PS-SS) coolant spread. Keywords: LE fillet, film-cooling, secondary flows, nozzle guide vane, passage loss.