Improved Formulation to Determine Minimum Sealing Flow – Cw, min – for Different Sealing Configurations

Abstract

Steadily increasing HP turbine inlet temperatures underline the need for effective sealing and cooling technologies. This paper presents a new approximation to determine the minimum sealing air flow Cw,min for sealing efficiencies of different rim seal geometries in the upstream cavity of an axial turbine stage. In contrast to the iteration method of Phadke and Owen the approximation does not require the knowledge of Cw,min. Only with information about the main stream and the geometry of the sealing configuration the minimum sealing air can be estimated. The approximation is based on experimental results performed in a 1.5 stage axial turbine at rotating speeds up to 9000 rpm with different Reynolds numbers 8·105 < Rec1 < 1.5·106 and 4.5·105 < Reu < 8·105 as well as different cooling gas mass flows. In order to get results, regarding the physical phenomena connected to hot gas ingestion, different measurement techniques were applied. Results of CO2 concentration measurements as well as steady pressure measurements are used for the approximation. Different sealing geometries were investigated: 1st a simple axial gap between a flat rotor disk and a flat stator disk, commonly used for industrial gas turbines, 2nd an axial lip of the rim seal on the stator combined with flat rotor disk, often found in aero engine applications, 3rd an axial lip of the rim seal on each stator and rotor and 4th a double lip on stator and single lip on rotor side. The results of Cw,min show very well the linear behavior as in the equation given by Phadke and Owen. Nevertheless the results of the approximation show a strong dependency of the gradient of the Cw,min due to the seal geometry. The gradient of Cw,min of the first geometrical configuration without any sealing lips is about 10 times higher than the gradient of configuration four with altogether 3 sealing lips. The other two configurations are in between these extremes. The result of the 2D–3D approximation with a polynomial equation η=exp∑i=02∑j=02∑k=01(Pi,j,k·(1/Cw)i·Cp1j·Gck) extend the rule of Phadke and Owen. This formulation gives the capability to optimize the sealing mass flow for different sealing configurations.