Design Parameters of Brush Seals and Their Impact on Seal Performance

Abstract

Brush seals, which were originally designed for gas turbine applications, have been successfully applied to large-scale steam turbines within the past decade. From gas turbine applications, the fundamental behavior and designing levers are known. However, the application of brush seals to a steam turbine is still a challenge. This challenge is mainly due to the extreme load on the brush seal while operating under steam. Furthermore, it is difficult to test brush seals under realistic conditions, i.e. under live steam conditions with high pressure drops. Due to these insufficiencies, 2 test rigs were developed at the University of Technology Braunschweig, Germany. The first test rig is operated under pressurized air and allows testing specific brush seal characteristics concerning their general behavior. The knowledge gained from these tests can be validated in the second test rig, which is operated under steam at pressure drops of 45 bar and temperatures up to 450 °C. Using both the air test rig and the steam test rig helps keep the testing effort comparably small. Design variants can be pre-tested with air, and promising brush seal designs can consequently be tested in the steam seal test rig. The paper focuses on a clamped brush seal design which, amongst others, is used in steam turbine blade paths and shaft seals of current Siemens turbines. The consequences of the brush assembly on the brush appearance and brush performance are shown. The clamped brush seal design reveals several particularities compared to welded brushes. It could be shown that the clamped bristle pack tends to gape when clamping forces rise. Gapping results in an axially expanding bristle pack, where the bristle density per unit area and the leakage flow vary. Furthermore, the brush elements are usually assembled with an axial lay angle, i.e. the bristles are reclined against the backing plate. Hence, the axial lay angle is also part of the investigation.