On the Failure of a Gas Foil Bearing: High Temperature Operation Without Cooling Flow

Abstract

Implementing gas foil bearings (GFBs) in micro gas turbine engines is a proven approach to improve system efficiency and reliability. Adequate thermal management for operation at high temperatures, such as in a gas turbine or a turbocharger, is important to control thermal growth of components and to remove efficiently mechanical energy from the rotor mainly. The paper presents a test rotor supported on GFBs operating with a heated shaft and reports components temperatures and shaft motions at an operating speed of 37 krpm. An electric cartridge heater loosely inserted in the hollow rotor warms the test system. Thermocouples and noncontact infrared thermometers record temperatures on the bearing sleeve and rotor outer diameter (OD), respectively. No forced cooling air flow streams were supplied to the bearings and rotor, in spite of the high temperature induced by the heater on the shaft outer surface. With the rotor spinning, the tests consisted of heating the rotor to a set temperature, recording the system component temperatures until reaching thermal equilibrium in similar to 60 min, and stepping the heater set temperature by 200 degrees C. The experiments proceeded without incident until the heater set temperature equaled 600 degrees C. Ten minutes into the test, noise became apparent and the rotor stopped abruptly. The unusual operating condition, without cooling flow and a too large increment in rotor temperature, reaching 250 degrees C, led to the incident which destroyed one of the foil bearings. Post-test inspection evidenced seizure of the hottest bearing (closest to the heater) with melting of the top foil at the locations where it rests on the underspring crests (bumps). Analysis reveals a notable reduction in bearing clearance as the rotor temperature increases until seizure occurs. Upon contact between the rotor and top foil, dry-friction quickly generated vast amounts of energy that melted the protective coating and metal top foil. Rather than a reliability issue with the foil bearings, the experimental results show poor operating procedure and ignorance on the system behavior (predictions). A cautionary tale and a lesson in humility follow.