A study of chaos in a rotor system supported by ball bearings

Abstract

Chaos is a new term coined to describe the ordered randomness that appears in many deterministic nonlinear dynamic systems. Although chaos has some desirable features, it is usually advantageous to control and eliminate the chaotic behavior. Numerous researchers have published a myriad of papers on the application of chaos theory to mechanical, electrical and biological systems. However, control theories of chaos have been in existence for only a decade and have not been applied to rotor systems. The original objective of this research was to use these control theories to eliminate the chaotic behavior of a rotor system supported by ball bearings. After a semester of challenging research, the aim of this research was modified to determine if chaos existed in a real-life rotor system supported by ball bearings. This research will later be used to continue the original objective, control and elimination of chaos in the system. A Bently Nevada rotor system was assembled and connected to proximity probes and a data-acquisition system to study the vibration of the rotor's shaft at varying speeds. Poincar ̌plots were initially used in an attempt to find chaos. When this approach was inconclusive, a wavelet analysis technique and Fast Fourier Transform yielded more concrete results. As the rotor speed increases, the behavior changes from quasi-periodic to weakly chaotic. The work done in this research project has been successful and will be applied to the control of chaos in this rotor system.