Date of Award

2008

Degree Name

M.S. in Mechanical Engineering

Department

Department of Mechanical and Aerospace Engineering

Abstract

Coated bearings in high performance turbine engines have the potential to improve engine reliability and extend service life. One of the coating performance indicators is the level of vibration generated by rolling element and raceway impacts during bearing operation. Vibrations in engine bearings can be difficult to monitor because they are located in a cramped and high temperature environment, and there is often system noise caused by other components located on the same shaft. This thesis investigates the effectiveness of a non-intrusive cage-mounted sensor which detects cage vibration and transmits wirelessly to a nearby receiver. The sensitivity of the cage-mounted sensor was compared to a housing-mounted accelerometer in detecting seeded faults in coated and uncoated bearings. The test bearings were coated with varying thicknesses of TiCN with 3.2 mm diameter masks on the inner races. After coating, the masks were removed to reveal uncoated raceway, simulating a failed coating. In addition, a 3.2 mm wide flaw was machined into the inner race of an uncoated bearing. The depth of this flaw was increased to simulate spalling. For each coating thickness and spall depth, vibration data were recorded using the accelerometer and cage sensor and studied in the frequency domain for the presence of the inner race fault frequency. The smallest spall on an uncoated bearing that the sensor could detect was 28 pm deep with a reliability of -40%. In contrast, the accelerometer consistently detected all levels of flaws, the smallest being 0.4 pm deep on the coated bearing, with a reliability of 55%. This research demonstrates the sensitivity level of the RF cage sensor and support mounted accelerometer to inner race bearing flaws on this bearing tester.

Keywords

Bearings (Machinery) Testing, Turbines Vibration, Detectors Testing

Rights Statement

Copyright © 2008, author

Share

COinS