| description abstract | Abstract. Gas turbine secondary air systems enable elevated turbine entry temperatures for increased cycle efficiency and work output. To prevent the ingress of hot mainstream gas into the turbine cavity, purge flow is supplied to the cavity from the upstream compressor. It subsequently exits the cavity through a rim seal into the mainstream gas-path (egress). The interaction between egress and the mainstream alters the endwall secondary flow structures that form within the rotor blade passage. Purge has a significantly lower temperature than the mainstream flow and so a non-unity purge-mainstream density ratio (DR) exists, with unknown implications on the endwall secondary flow. Phase-locked, ensemble-averaged volumetric velocimetry measurements of the flow field within the rotor blade passage were conducted using a one-stage, optically accessible, rotating turbine test facility. The effect of DR was simulated by varying the concentration of purge carbon dioxide to achieve three DR conditions: 1, 1.26, and 1.54. Pitch-wise and radial positions of the endwall secondary flow vortices were tracked using a non-local vortex detection method. A significant pitch-wise shift in the egress vortex occurred when the cavity sealing effectiveness was increased. An independent increase in either the non-dimensional sealing flow parameter (Φ0) or DR resulted in increased radial migration (h), annulus blockage ratio (ξ), and circulation (Γ) of the passage vortex. A new cavity-derived blowing ratio, Φe*, was developed. This is proportional to the classical blowing ratio when in the purge-dominated interval, and has a strong positive correlation with Δh, Δξ, and ΔΓ. Therefore, measurements in the cavity can only be related directly to the mainstream gas-path if the non-dimensional purge level is normalized with respect to DR. | |
