The primary objective of system identification is to obtain an accurate analytical model of a vibrating structure which can effectively predict the dynamic behaviors and before and after the modification of the structure. The success of the dynamic analysis and reanalysis of complicated modern vibrating structures is heavily dependent upon the adequacy of the models used. Most of discretization methods commonly used in recent years still rely on the analysts`` intuition and experience to adequately model the essentially continuous structures. However, the results based on the models often do not cope with actual test results. Therefore, it is necessary to identify the structural parameters, the models, strictly based on the experimental data. This research is mainly concerned with developing techniques to model the vibrating structures by utilizing the measured frequency response functions or the estimated modal parameters and to enhance the dynamic reanalysis of the modified vibrating structures.
The proper treatment of truncated higher modes is an important area in system identification. The model which can compensate for the truncated modes effects should not only render the accurate modal parameters of structures but also predict the dynamic properties of the structure subject to modification to a reasonably good accuracy. In this research, an effective mode concept is introduced to compensate for the truncated higher modes in the measured frequency response functions of structures. Simulation results show that the proposed method is superior to other methods available so far in dynamic reanalysis of vibrating structures.
The second method developed is an efficient method for identifying the joint structural parameters directly from the measured frequency response functions directly by using the relationship between the frequency response functions of the coupled system and the uncoupled systems. The proper mathematical model of joints is chosen according t...