Journal Bearing “Morton Effect” Cause of Cyclic Vibration in Compressors

Abstract

A tilting pad bearing supporting an integrally geared compressor in an air separation service had gradually developed symptoms of very consistent cycling of high synchronous vibration levels. The vibration envelope increased and decreased over a roughly six-minute period. There was a difference in the phasing of vibration increase at one end of the rotor versus the other such that the two ends were not in phase but not 180 degrees out of phase either. Intensive site testing determined that no rotor critical speeds were close to the running speed of the pinion or of the bull gear shaft. However, it was observed with infrared thermography that the temperatures of the bearing pedestals were also cyclic with the same period as the vibration, as were the bearing pad and exit oil temperatures. The temperature cycle peaks were not in phase with each other or with the vibration. Various process parameters were changed on an experimental basis, indicating that the main parametric influence was bearing oil inlet temperature. If the oil temperature was optimized in a certain range, the cycling stopped. The conclusion was that the cyclic vibration behavior was a form of “Morton effect,” as discussed in recent research papers. The worn bearing clearances and shaft straightness were being changed by the bearing oil film temperature, which in turn was being affected by the vibration level, which was in turn affected by the instantaneous clearances and the degree of the instantaneous shaft bow. This chain of events, together with the significant time lag associated with transient thermal conduction, led to the cyclic but stable oscillation of the envelope of the amplitude of synchronous vibration. In a double overhung rotor such as the one involved here, this effect was found to happen with a peak cycle amplitude at each end occurring at different times, and with the apparent phase angle shift of the “hot spot” being limited in a deceiving, but ultimately explainable, manner. Field testing showed that maintaining an oil inlet temperature above a practical threshold temperature minimized viscous shearing forces in the bearing film enough to prevent the thermal cycling from occurring. Another option for eliminating the problem would be a bearing re-design to reduce the hot spot temperature severity and sensitivity. Review led by Gordon Kirk