Sealing Performance of a Turbine Rim Chute Seal Under Rotationally-Induced Ingestion

Abstract

This study focuses on the sealing capability of a turbine rim seal subject to hot gas ingestion driven purely by the rotor disc pumping effect rather than that induced by mainstream features such as vane and rotor blade passing. The aim is to provide useful data for conditions in which rotation dominates, and to clarify the flow physics involved in rim sealing. Experimental measurements of sealing effectiveness for a chute seal are presented for the first time without and with an axial, axisymmetric mainstream flow external to the seal. The test matrix covers a range of rotational Reynolds number, Reø, from 1.5x106 to 3x106, and non-dimensional flow rate, Cw, from 0 to 4x104 with the mainstream flow (when present) scaled to match engine representative conditions of axial Reynolds number, Reax. Results from steady pressure and gas concentration measurements within the rotor-stator disc cavity and the rim seal gap are presented and compared to published data for other seal designs. Sealing performance of the chute seal is somewhat similar to that of axial clearance seals with the same minimum clearance.

Flow and Ingestion in a Turbine Disc Cavity under Rotationally-Dominated Conditions

“…The results of an investigation of the disc pumping effect in a turbine disc cavity with a chute rim seal arrangement are presented in this paper extending the preliminary study by Bru Revert et al [9]. The effect of annulus pressure asymmetries created by the presence of NGVs in the gas path is compared by Bru Revert et al [1] to the results of this study.…”

“…The results of an investigation of the disc pumping effect in a turbine disc cavity with a chute rim seal arrangement are presented in this paper extending the preliminary study by Bru Revert et al [9]. The effect of annulus pressure asymmetries created by the Initial studies of the disc pumping effect, such as that of Bayley and Owen [3], derived empirical correlations to quantify the minimum amount of purge flow necessary to fully seal the cavity.…”

“…This suggests a superior sealing capability of the open rim geometry as shown in Figure 7b (please note the wider range of purge flows tested at off-design). Here, the empirical constant in the model, k, is set to 1, as it was shown to best fit the experimental data for a chute seal by Bru Revert et al [9]. Therefore, the underprediction for the larger clearance suggests that a slightly lower value of k is needed to represent this seal.…”