A Method for Estimating the Influence of Time-Dependent Vane and Blade Pressure Fields on Turbine Rim Seal Ingestion

Abstract

A method of estimating the turbine rim seal ingestion rates was developed using the time-dependent pressure distributions on the hub of turbines and a simple-orifice model. Previous methods use the time-averaged pressure distribution downstream of the vanes to estimate seal ingestion. The present model uses the pressure distribution near the turbine hub, obtained from 2D time-dependent stage calculations, and a simple-orifice model to estimate the pressure-driven ingress of gas-path fluid into the turbine disk cavity and the egress of cavity fluid to the gas path. The time-dependent pressure distribution provides the influence of both the vane wakes and the bow wave from the blade on the pressure difference between the hub pressure at an azimuthal location and the cavity pressure. Results from the simple-orifice model are used to determine the effective Cd that matches the cooling effectiveness at radii near the rim seal with the amount of gas-path-ingested flow required to mix with the coolant flow. Cavity ingestion data from rim seal ingestion experiments in a 1.5-stage turbine and numerical simulations for a 1 vane, 2-blade sector of the 16-vane, 32-blade turbine were used to evaluate the method. The experiments and simulations were performed for close-spaced and wide-spaced half stages between both the vane and blade and between the blade and a trailing teardrop-shaped strut. The comparison of the model with a single Cd for axial gap seals and the experiments showed a reasonable agreement for both close- and wide-spaced stages.

A Comparison of Single and Double Lip Rim Seal Geometries

“…The majority model ingress and egress through the rim seal as an orifice flow moderated by empirical discharge coefficients found by fitting to experimental data for a given seal. Amongst others Phadke and Owen [7], Chew et al [18], Reichert and Leiser [19], Bohn and Wolff [20], Scanlon et al [21], Johnson et al [22], Owen et al [23] all present models of this type. As these models can only describe the variation in seal performance with dimensionless flow rate for a seal which has been experimentally characterised, they do not allow prediction.…”

Pressure Distortion Effects on Rim Seal Performance in a Linear Cascade

“…Phadke and Owen [3] shows that either the main gas path pressure field or disk pumping beneath the platform may dominate the ingress process depending on the axial Reynolds number in the mainstream, the rotational Reynolds number, and the seal geometry. Similarly, the pressure model of Johnson et al [4] shows that when the vane and blade rows are closely spaced, their pressure fields interact and become a dominant influence on the ingress of hot gas beneath the platform.…”

“…Radial overlap seals are widely used as they require much less purge flow to prevent ingress than more simple geometries like axial gap seals [16]. The overlapping geometry of radial overlap seals reduce the impact of the main gas path pressure field on the flow beneath the platform [2,4]. The benefits of a radial overlap seal are maintained as long as any overlap exists.…”

Use of Pressure Measurements to Determine Effectiveness of Turbine Rim Seals

“…Johnson et al [13] used an orifice model to predict ingestion rates in the Aachen rig, using CFD to compute the unsteady pres sure distribution in the annulus. A modified version of this model [14] was also successfully used to estimate the discharge coeffi cients for ingress and egress from experimental data from a rig at Arizona State University [15J.…”