A brush-labyrinth sealing configuration consisting of two labyrinth fins upstream and one brush seal downstream is studied experimentally and theoretically. Two slightly different brush seal designs with zero cold radial clearance are considered. The sealing configurations are tested on the no-whirl and dynamic test rigs to obtain leakage performance and rotordynamic stiffness and damping coefficients. The no-whirl tests allow identification of the local rotordynamic direct and cross-coupled stiffness coefficients for a wide range of operating conditions, while the dynamic test rig is used to obtain both global stiffness and damping coefficients, but for a narrower operating range limited by the capabilities of a magnetic actuator. Modeling of the brush-labyrinth seals is performed using computational fluid dynamics. The experimental global rotordynamic coefficients consist of an aerodynamic component due to the gas flow and a mechanical component due to the contact between the bristle tips and rotor surface. The CFD-based calculations of rotordynamic coefficients provide however only the aerodynamic component. A simple mechanical model is used to estimate the theoretical value of the mechanical stiffness of the bristle pack during the contact. The results obtained for the sealing configurations with zero cold radial clearance brush seals are compared with available data on three-tooth-on-stator labyrinth seals and a brush seal with positive cold radial clearance. Results show that the sealing arrangement with a line-on-line welded brush seal has the best performance overall with the lowest leakage and cross-coupled stiffness. The predictions are generally in agreement with the measurements for leakage and stiffness coefficients. The seal damping capability is noticeably underpredicted.

This content is only available via PDF.
You do not currently have access to this content.