Abstract
Unsteady flow fields are generated in gas turbines due to interactions between rotating and stationary components. Increased understanding of unsteadiness and its effect on blade heat transfer is crucial to thermal durability predictions for turbine components. This study examines unsteady pressure, heat transfer, and film cooling on a rotating high-pressure turbine stage using experimental and unsteady computational tools. Comparisons between experimental and computational results identify several key elements necessary for successful computational models. Models benchmarked against experimental data provide valuable insight into the nature of flow field unsteadiness and its impact on blade heat transfer.
Comparisons presented in Part I of this paper identified a region near the leading edge of the suction surface of the airfoil where the computational and experimental results did not agree. Part II of the paper will examine the mechanisms driving unsteadiness in this region and how they can be better predicted. The source of this disagreement is that vane blade interaction causes large swings in incidence angle at the leading edge, which causes periodic separation on the suction surface. In addition, pressure fluctuations couple with unsteadiness in cooling flow at the leading edge to further amplify unsteadiness.
Examining the links between pressure and heat transfer fluctuations shows that the pressure surface and suction surface each had strong but unique correlations, and the leading edge is complicated by coupling between pressure and shower head cooling unsteadiness. This indicates a different turbulent Prandtl number may be required in different regions of the blade to accurately track heat transfer. Further complicating heat transfer predictions, small changes in the velocity field cause large levels of unsteadiness in film cooling. The relative movement of large-scale vortex structures over a vane pass increase time-averaged dissipation of cooling jets beyond steady prediction levels.
The results of this study shed light on the nature of unsteadiness, highlighting specific components of unsteadiness observed to augment heat transfer. Better characterizations of unsteadiness can help designers utilize computationally efficient design tools. With increased understanding of how unsteadiness impacts time-averaged performance, a more accurate representation of time-averaged performance can be leveraged from low-cost RANS computational tools.
