NASA NTRS1996
Brush seals are compliant, contact seals that have long-life, low-leakage characteristics desirable for use in rocket engine turbopumps. 50.8-mm (2.0 inch) diameter brush seals with a nominal initial radial interference of 0.127-mm (0.005 inch) were tested in liquid nitrogen at shaft speeds up to 35,000 rpm and differential pressure loads up to 1.21 MPa (175 psi) per brush. The measured leakage rate of a single brush was 2-3 times less than that measured for a 12-tooth, 0.127-mm (0.005 inch) radial clearance labyrinth seal used as a baseline. Stage effects were studied and it was found that two brush seals with a large separation distance leaked less than two brushes tightly packed together. The maximum measured groove depth on the Inconel 718 rotor was 25.4 (mu)m (0.001 inch) after 4.31 hours of shaft rotation. The Haynes-25 bristles wore approximately 25.4-76.2 (mu)m (0.001-0.003 inch) under the same conditions. Three seal runner coatings, chromium carbide, Teflon impregnated chromium, and zirconium oxide, were tested in liquid hydrogen at 35,000 and 65,000 rpm with separate 50.8 mm diameter brush seals made of Haynes-25 bristles and having a nominal initial radial interference of 129 rpm. Two bare Inconel-718 rotors were also tested as a baseline. The test results revealed significant differences between the wear characteristics of the uncoated and coated seal runners. At both speeds the brush seal with the bare Inconel-718 seal runner exhibited significant bristle wear with excessive material transferring to the runner surface. In contrast, the coated seal runners inhibited the transfer and deposit of bristle material. The chromium carbide coating showed only small quantities of bristle material transferring to its surface. The Teflon impregnated chromium coating also inhibited material transfer and provided some lubrication. This coating, however, is self-sacrificing. The Teflon remained present on the low speed runner, but it was completely removed from the high speed brush seal, which was tested considerably longer. The tests of the Teflon coating revealed the importance of using a lubricating and low friction coating for brush seals to reduce bristle and seal runner wear. The zirconium oxide coating exhibited the greatest amount of coating wear, while the brushes incurred only slight wear. Further testing of ceramics is recommended before making a final judgement on the viability of ceramic coatings for brush seals because of the contrast between the results reported by Carlile and the results presented herein. Strictly based on the results presented hereinabove, the chromium carbide and Teflon impregnated chromium coatings were considered preferable to the uncoated Inconel-718 and zirconium oxide coatings because of their good wear resistance and characteristics to inhibit bristle material wear and transfer to the seal runner.