Influence of Purge Flow Swirl and Passage Discrete Hole Injection Angle on Turbine Blade Endwall Cooling

Abstract

Gas turbine blade endwall cooling presents a significant challenge due to the complex secondary flow structures within blade passages. Purge flow coolant, which passes through the gap between the stator and rotor, and discrete cooling holes, which are positioned strategically within the coolant passage, are often utilized together to provide complete cooling coverage on the blade endwall surface. The relative motion between the stator and rotor surfaces gives the purge flow a degree of swirl. Discrete injection holes are usually angled in close alignment with the near-wall flow direction, but this is often not achievable due to manufacturing constraints. The effects of purge flow swirl and discrete hole injection angle are not well documented but are expected to influence the endwall cooling significantly. Therefore, it is the focus of this work to investigate the influence of purge flow swirl and passage discrete hole injection angle on endwall cooling. An experimental study is performed to investigate these endwall cooling phenomena using the naphthalene sublimation technique in a linear cascade composed of five high- pressure turbine blades. Detailed measurements for the endwall Sherwood number and film cooling effectiveness are made over a range of blowing rates, purge flow swirl angles, and discrete hole injection angles. The work is split into four complementary experimental cases. In the first case, a 45° straight injection slot is used to investigate the effect of purge flow blowing rate and swirl angle on endwall cooling. The purge flow swirl is simulated using turning vanes within the slot. In the second case, a more realistically shaped slot, featuring a gradual ramp leading to the endwall, is used to investigate the effect of purge flow blowing rate and swirl angle on endwall cooling. The purge flow swirl is again simulated with turning vanes, which are now located immediately upstream of the ramp. In the third case, 15 discrete endwall cooling holes are positioned along the endwall to investigate the effect of discrete hole injection blowing rate and discrete hole injection angle on endwall cooling. Two endwall plates are utilized: one that angles the discrete holes in alignment with the local near-wall flow direction and a second that angles the discrete holes 90° relative to the first, representing a near worse case misalignment effect. In the fourth and last case, the combined cooling sources from the realistic slot and the 15 discrete holes are studied together. This case investigates the interactions between the two coolant sources. The secondary flows in the blade passage limited the axial penetration of the purge flow coolant by sweeping the coolant toward the suction side of the blade passage. At high purge flow blowing rates, the effect of increasing the swirl angle led to significantly reduced axial penetration and thus overall reduced film cooling effectiveness levels. However, for low blowing rates, the effect of the swirl angle was weak. These results indicate that the blowing rate strongly influences to what degree the coolant follows its initial trajectory exiting the cooling slot, which is tied strongly to the swirl angle. Therefore, for low blowing rates, the coolant flow path is nearly unaffected by its swirl angle, whereas, at high blowing rates, the coolant flow path is closely aligned with its swirl angle. The purge flow results indicate that for realistic levels of purge flow blowing and swirl inadequate endwall cooling coverage will result, so supplementary coolant sources, like discrete cooling holes, must be utilized to provide complete endwall cooling coverage. The discrete hole injection angle misalignment effect led to multiple penalties: the Sherwood number was moderately enhanced, and the film cooling effectiveness was moderately reduced. This result indicates that the discrete hole injection angle misalignment effect promotes significant mixing and should be actively avoided where possible. The findings from this work demonstrate that the purge flow swirl and the discrete hole injection angle effects significantly influence endwall cooling and should not be neglected when designing endwall cooling schemes.