| Summary: | Structural Health Monitoring (SHM) is traditionally concerned with fitting sensors inside structural systems and analyzing the features of signals from the sensor measurements using appropriate signal processing techniques in order to reveal the systems’ health status. A significant change of signal features is often considered to be an indication of damage. However, generally speaking, these techniques often cannot distinguish normal structural changes due to variations in system environmental or operating conditions from the changes which are induced by damage. For example, transmissibility analysis is a widely used signal analysis method for SHM. But traditional transmissibility is determined by the ratio of the spectra of two different system outputs, which generally depends on the location of loadings on the system and is, consequently, affected by system environmental conditions. In order to solve this challenge, a series of studies are conducted in this PhD project. The objectives are to develop new SHM and damage localization methods, which can effectively address the effects of changing system environmental or operational conditions and have potential to be applied in practice to more effectively solve practical SHM and damage localization problems. First, a general baseline model based SHM method is developed in this thesis. This method can be used to address a wide class of SHM problems via a baseline modelling and baseline model based analysis. The method can systematically take the effects of system’s operating or environmental conditions such as, e.g., environmental temperature etc. on signal analysis into account, and can therefore solve relevant challenges. Both simulation studies and field data analyses have been conducted to demonstrate the performance of the proposed new technique. Moreover, new transmissibility analysis methods are proposed for the detection and location of damage with nonlinear features in Multi-Degree-Of-Freedom (MDOF) structural systems. These methods extend the traditional transmissibility analysis to the nonlinear case. More importantly, the methods are independent from the locations of loading inputs to the systems and, to a great extent, provide effective solutions to the above mentioned problems with traditional transmissibility analysis. Again both numerical simulation studies and experimental data analysis have been conducted to verify the effectiveness and demonstrate potential practical applications of the new methods. Based on the results of nonlinearity detection and localization, new guidelines are proposed for the application of transmissibility analysis based modal identification method to nonlinear structural systems, which have potential to be further developed into a new approach to transmissibility based nonlinear modal analysis. In summary, the present study has addressed a series of fundamental problems with SHM, especially, problems associated with how to deal with the effects of changing system environmental or operational conditions on SHM results. Experimental studies have demonstrated the potential and significance of these results in practical engineering applications.
|
|---|
