Unsteady 360 Computational Fluid Dynamics Validation of a Turbine Stage Mainstream/Disk Cavity Interaction

The amount of cooling air assigned to seal high pressure turbine (HPT) rim cavities is critical for performance as well as component life. Insufficient air leads to excessive hot annulus gas ingestion and its penetration deep into the cavity compromising disk or cover plate life. Excessive purge air, on the other hand, adversely affects performance. Experiments on a rotating turbine stage rig which included a rotor–stator forward disk cavity were performed at Arizona State University (ASU). The turbine rig has 22 vanes and 28 blades, while the cavity is composed of a single-tooth lab seal and a rim platform overlap seal. Time-averaged static pressures were measured in the gas path and the cavity, while mainstream gas ingestion into the cavity was determined by measuring the concentration distribution of tracer gas (carbon dioxide) under a range of purge flows from 0.435% (C_w = 1540) to 1.74% (C_w = 6161). Additionally, particle image velocimetry (PIV) was used to measure fluid velocity inside the cavity between the lab seal and the rim seal. The data from the experiments were compared to time-dependent computational fluid dynamics (CFD) simulations using fluent CFD software. The CFD simulations brought to light the unsteadiness present in the flow during the experiment which the slower response data did not fully capture. An unsteady Reynolds averaged Navier–Stokes (RANS), 360-deg CFD model of the complete turbine stage was employed in order to increase the understanding of the swirl physics which dominate cavity flows and better predict rim seal ingestion. Although the rotor–stator cavity is geometrically axisymmetric, it was found that the interaction between swirling flows in the cavity and swirling flows in the gas path create nonperiodic/time-dependent unstable flow patterns which at the present are not accurately modeled by a 360 deg full stage unsteady analysis. At low purge flow conditions, the vortices that form inside the cavities are greatly influenced by mainstream ingestion. Conversely at high purge flow conditions the vortices are influenced by the purge flow, therefore ingestion is minimized. The paper also discusses details of meshing, convergence of time-dependent CFD simulations, and recommendations for future simulations in a rotor–stator disk cavity such as assessing the observed unsteadiness in the frequency domain in order to identify any critical frequencies driving the system.