Cooling Injection Effect on a Transonic Squealer Tip—Part II: Analysis of Aerothermal Interaction Physics

Abstract

A basic attribute for turbine blade film cooling is that coolant injected should be largely passively convected by the local base flow. However, the effective working of the conventional wisdom may be compromised when the cooling injection strongly interacts with the base flow. Rotor blade tip of a transonic high-pressure (HP) turbine is one of such challenging regions for which basic understanding of the relevant aerothermal behavior as a basis for effective heat transfer/cooling design is lacking. The need to increase our understanding and predictability for high-speed transonic blade tip has been underlined by some recent findings that tip heat transfer characteristics in a transonic flow are qualitatively different from those at a low speed. Although there have been extensive studies previously on squealer blade tip cooling, there have been no published experimental studies under a transonic flow condition. The present study investigates the effect of cooling injection on a transonic squealer tip through a closely combined experimental and computational fluid dynamics (CFD) effort. The experimental and computational results as presented in Part I have consistently revealed some distinctive aerothermal signatures of the strong coolant-base flow interactions. In this paper, as Part II, detailed analyses using the validated CFD solutions are conducted to identify, analyze, and understand the causal links between the aerothermal signatures and the driving flow structures and physical mechanisms. It is shown that the interactions between the coolant injection and the base over-tip leakage (OTL) flow in the squealer tip region are much stronger in the frontal subsonic region than the rear transonic region. The dominant vortical flow structure is a counter-rotating vortex pair (CRVP) associated with each discrete cooling injection. High HTC stripes on the cavity floor are directly linked to the impingement heat transfer augmentation associated with one leg of the CRVP, which is considerably enhanced by the near-floor fluid movement driven by the overall pressure gradient along the camber line (CAM). The strength of the coolant-base flow interaction as signified by the augmented values of the HTC stripes is seen to correlate to the interplay and balance between the OTL flow and the CRVP structure. As such, for the frontal subsonic part of the cavity, there is a prevailing spanwise inward flow initiated by the CRVP, which has profoundly changed the local base flow, leading to high HTC stripes on the cavity floor. On the other hand, for the rear high speed part, the high inertia of the OTL flow dominates; thus, the vortical flow disturbances associated with the CRVP are largely passively convected, leaving clear signatures on the top surface of the suction surface rim. A further interesting side effect of the strong interaction in the frontal subsonic region is that there is considerable net heat flux reduction (NHFR) in an area seemingly unreachable by the injected coolant. The present results have confirmed that this is due to the large reduction in the local HTC as a consequence of the upstream propagated impact of the strong coolant-base flow interactions.