Analysis of hydrodynamic lift for gas turbine oil brush seals /

Official URL: http://risc01.sabanciuniv.edu/record=b1158359 (Table of Contents)

Brush seals are used in gas Turbine engine applications for their effectiveness in controlling leakage flows between stationary parts and rotating shaft With their superior leakage performance, they are becoming the successor to the labyrinth seals, which have always been widely used since the first gas turbine engines. After gaining popularity in secondary flow air sealing, applications extended to more challenging oil and oil mist sealing locations. Due to high rotor surface speed oil temperature rise and coking become main issues in addition to leakage performance. While each bristle forms very small bearing surface. high rotor surface speed and viscous sealing medium provide sufficient hydrodynamic lift force to overcome bristle reaction forces. The gap generated between the bristle tips and the rotor surface determines the oil temperature rise and leakage rate. Due to shear thinning in oil sealing applications lift force is inversely correlated to oil temperature rise leading to a self-balancing effect with surface speed. The aim of a proper design for oil brush seals would be to keep the leakage at the minimum rate while avoiding excess oil temperature rise. Inherent complexity of bristles make calculation of hydrodynamic lift forces under a brush seal rather complicated. To avoid costly experimental investigations, a closed form analytical solution would prove very useful to compare candidate designs and to study what-if scenarios. This work investigates analytical solutions to hydrodynamic- lift clearance and bristle lift forces based on bearing theory. After a detailed discussion on short bearing approach, possible closed form solutions based on long bearing approach have been explored. A more comprehensive yet numerical solution based on finite beating approach has also been provided to compare and understand pressure and lifting force results. Analyses have been performed for effective constant oil viscosity. Shear thinning effect has been later introduced through use of published oil temperature data for various rotor surface speeds. As bristle tip clearance data are not available, validity of analytical the results have been evaluated using beam bending forces and experimental leakage flow data for various surface speeds. The results indicate that long bearing solution underestimates the oil lift force. Although more comprehensive, finite bearing solution requires the knowledge of axial deflection/bloom of the bristle pack. Simple yet functional, short bearing solution appears as a viable tool for seal designers to compare various seal designs before doing detail experimental evaluation on a plausible candidate. Finally, possible further work areas have been discussed.