Experimental Study of Ingestion in the Rotor-Stator Disk Cavity of a Subscale Axial Turbine Stage

Abstract

While it is widely recognized that ingestion of hot gas from the main annulus of axial gas turbine stages into rotor-stator disk cavities depend strongly on the unsteadiness of the prevailing flow field, the large computational effort needed to simulate the flow field renders its use in the design of turbine internal air system and seals difficult. As an alternative, considerable effort has been devoted in recent years to develop simple orifice models of disk cavity rim seals based on time-averaged flow information; these models contain empirical discharge coefficients for ingestion into and egress from the cavities. The present experimental work in a subscale axial turbine stage reports a simple orifice model of an axially-overlapping radial-clearance seal at the disk cavity rim and values of the discharge coefficients over a range of purge flow rate supplied to the cavity. In the experiments, the ingestion process was dominated by the main gas annulus flow. Time-averaged static pressure distribution was measured in the main annulus and in the disk cavity; the driving force for ingestion and egress was taken to be the pressure differential between the main annulus and the rim cavity at prescribed locations. Time-averaged ingestion was measured using the tracer gas technique; the pressure and ingestion data were combined to obtain the ingestion and egress discharge coefficients at several purge flow rates. The location on the vane platform 1mm upstream of its lip represented the main gas annulus pressure in the calculation of discharge coefficients. In the rim cavity, two locations on the stator, one in the ‘seal region’ and the other slightly inward radially, were prescribed to represent the rim cavity pressure as well as the sealing effectiveness. Two corresponding sets of ingestion and egress discharge coefficients are reported for the various purge flow rates. The ingestion discharge coefficient obtained using the seal region location in the rim cavity decreased as the purge flow rate increased; the corresponding egress discharge coefficient increased with purge flow rate. For the rim cavity location slightly inward radially from the seal region, the egress discharge coefficient maintained the same trend; however, the ingestion discharge coefficient decreased only slightly as the purge flow rate increased. It is suggested that the seal region location in the rim cavity is the more appropriate location in calculating the rim seal discharge coefficients. The ratio of ingestion to egress discharge coefficients exhibited considerable variation with purge flow rate.

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

“…These results indicate that the flow field produced in the cavity by the experiments is different than the flow field generated by the CFD simulations. In order to gain insights into the fluid structure in the cavity, Balasubramanian et al [16] obtained radial and tangential velocities using PIV over a square laser sheet which encompasses approximately 90 deg inside the rotor-stator disk outer cavity. For the purge flow rate of C w ¼ 1540, ten instantaneous images were created at a capturing frequency of 3.3 Hz, or one image every 12.12 rotor revolutions.…”

“…On the other hand, the corresponding data for the larger clearance show a better match with the data across all radii. Since the difference falls within the uncertainty in pressure measurements [16], the CFD simulations were performed with 0.6 mm across the range of purge flows.…”

“…Craft et al [15] showed in his simulation of disk cavities using a two-equation turbulence model that the flow does not reach steady state even after 70 revolutions. The present work simulates the experimental rig geometry of Balasubramanian et al [16] in a 360 unsteady CFD model, runs the simulation under rig conditions, and then compares the CFD predictions of pressure, velocity, and concentration-based cavity effectiveness (g c ) with the experimental data of [16]. This data includes PIV measurements to monitor the fluid velocity inside the cavity between the lab seal and the rim seal.…”

“…Both the axial and radial overlap clearances are each 2.6 mm. Test data obtained in this rig (see Balasubramanian et al [16] for details) were compared with CFD simulations.…”

“…The turbine rim seal rig at ASU [16], shown in Fig. 1 was used to run experiments investigating main gas path ingestion in the forward disk rim cavity.…”