This paper proposes a novel approach for nonlinearity detection in vibrating systems.
IntroductionIn engineering practice, many mechanical and structural systems require more than one set of coordinates to describe the systems' behaviours and, consequently, need a multi-degree of freedom (MDOF) model to represent the systems. MDOF systems can behave nonlinearly simply due to the nonlinear characteristics of one component within the systems. In many practical cases, such a nonlinear component may represent a fault. Typical examples of such a case are beams, as it is well known that beams can behave nonlinearly due to the presence of internal breathing cracks [1,2].Therefore, locating nonlinear components in MDOF systems has considerable significance in fault diagnosis for a wide range of MDOF engineering systems and structures, which can, like beams, behave nonlinearly due to the existence of faults.The detection of faults in MDOF structures has been studied by Zhu and Wu [3] where the structure with faults is still considered to be linear and the location and magnitude of the fault are estimated using measured changes in the natural frequencies. Based on a one-dimensional structure model, Sakellariou and Fassois [4] PDF created with pdfFactory Pro trial version www.pdffactory.com
2[5] have used a stochastic output error vibration-based methodology to detect faults in structures where the faulty elements are modelled as components of cubic stiffness.The present study is concerned with the development of a novel approach for detecting the nonlinearity in MDOF vibrating systems, which can represent a fault in the system. This work is the latest in a series of investigations conducted by the authors on this subject. For simplicity in demonstrating the main ideas, simple onedimensional MDOF systems with only one nonlinear component will be considered, as shown in Fig. 1. However, compared with existing methods, the present study addresses the issues of detecting nonlinearity for a much wider class of MDOF systems. The basis of this work is a totally new concept the author proposed in paper