Measurements and Modeling of Ingress in a New 1.5-Stage Turbine Research Facility

In gas turbines, hot mainstream flow can be ingested into the wheel-space formed between stator and rotor disks as a result of the circumferential pressure asymmetry in the annulus; this ingress can significantly affect the operating life, performance, and integrity of highly stressed, vulnerable engine components. Rim seals, fitted at the periphery of the disks, are used to minimize ingress and therefore reduce the amount of purge flow required to seal the wheel-space and cool the disks. This paper presents experimental results from a new 1.5-stage test facility designed to investigate ingress into the wheel-spaces upstream and downstream of a rotor disk. The fluid-dynamically scaled rig operates at incompressible flow conditions, far removed from the harsh environment of the engine which is not conducive to experimental measurements. The test facility features interchangeable rim-seal components, offering significant flexibility and expediency in terms of data collection over a wide range of sealing flow rates. The rig was specifically designed to enable an efficient method of ranking and quantifying the performance of generic and engine-specific seal geometries. The radial variation of CO2 gas concentration, pressure, and swirl is measured to explore, for the first time, the flow structure in both the upstream and downstream wheel-spaces. The measurements show that the concentration in the core is equal to that on the stator walls and that both distributions are virtually invariant with radius. These measurements confirm that mixing between ingress and egress is essentially complete immediately after the ingested fluid enters the wheel-space and that the fluid from the boundary layer on the stator is the source of that in the core. The swirl in the core is shown to determine the radial distribution of pressure in the wheel-space. The performance of a double radial-clearance seal is evaluated in terms of the variation of effectiveness with sealing flow rate for both the upstream and the downstream wheel-spaces and is found to be independent of rotational Reynolds number. A simple theoretical orifice model was fitted to the experimental data showing good agreement between theory and experiment for all cases. This observation is of great significance as it demonstrates that the theoretical model can accurately predict ingress even when it is driven by the complex unsteady pressure field in the annulus upstream and downstream of the rotor. The combination of the theoretical model and the new test rig with its flexibility and capability for detailed measurements provides a powerful tool for the engine rim-seal designer.