Transmissibility Function-based diagnosis of a class of bolted Beam-like structures with nonlinear Fault-induced loads and nonlinear boundary conditions

Bolted beam-like engineering structures are frequently encountered with dynamic load-caused bolt faults like fatigue crack and loosening during operation. Bolt faults can seriously affect structural safety and integrity and thus health diagnosis as early as possible is very essential and meaningful. To this aim, various fault feature-based methods have been proposed, which, however, usually have some limitations, for example, neglect of nonlinear boundary condition, requirement of whole structure model, and need of benchmark structure data and so forth. To overcome these limitations, a systematic diagnosis method is proposed in this study. In the new method, the dynamic model of bolted beam-like structures is built as a discrete chain-type multiple degree-of-freedom (MDOF) model with nonlinear damper-spring connections simulating nonlinear boundary conditions and potential bolt faults. By stimulating the model three times with the same signal of different excitation amplitudes, local diagnosis features using nonlinear fault-induced loads and transmissibility functions of the local substructure are established. Then, a novel diagnosis approach with sensitive indicators which combines diagnosis features with relative change is developed. The results from a simulation example about a 20-DOF model with quadratic stiffness properties and extensive experimental studies on a lab bolted beam-like steel structure with nonlinear boundaries and bolt faults demonstrate and validate the effectiveness and feasibility.