Rotor bearing systems frequently utilise hydrodynamic bearings whose dynamic properties are generally influenced by the bearing reaction forces (which determine the bearing stiffness and damping coefficients). These reaction forces are frequently unknown and are generally determined from the solution of the Reynolds equation using rotor motion measurements as input. Of interest is the attainable accuracy of such bearing force determinations, and for experimental evaluation, a test rig was fabricated, the design specification of which required that the rotor system run stably over its operating speed range. This paper describes the commissioning of this rig for stability purposes with the aid of natural frequency analyses, noting the required design modifications to ensure stable operation. Stability was found to be significantly influenced by the extent of the continuous fluid film in the hydrodynamic circumferentially grooved bearings. It was concluded that the assumption of a 180 degree film extent was totally inappropriate even though the bearing ends were open to the atmosphere, whereas the assumption of fluid film break up at the lubricant saturation vapour pressure proved appropriate for stability predictions provided one ensured that the bearings were flooded. Preliminary bearing force evaluations proved inconclusive, primarily because the self aligning bearings nevertheless experienced angular misalignment; and because there was uncertainty as to how much air was entrained in the bearings, in spite of attempts to prevent air ingress.

This content is only available via PDF.
You do not currently have access to this content.