Damage Identification of a model frame-type structure and effects of offline tuning technique on the identification results

Abstract

In shake table tests, the main problem is the incompatibility between the command and feedback signals. This means that the shaker cannot track the command signal completely (a.k.a. trajectory tracking problem). The complexity of the components of the shaker and the shaker–specimen interaction are the main reasons for the aforementioned problem. Solving the problem is critical for the health of the test data. Offline tuning technique (OTT) is a method which can be used for this purpose. In this study, OTT and its effects on a shake table are examined without and with a three-story, one-bay aluminum specimen on it. System and damage identification studies of this specimen are performed by various operational modal analysis methods (Fourier amplitude spectra combined with half-power bandwidth, SSI, and EFDD) and a damage identification technique (damage index method) based on the changes in modal parameter estimations, respectively. This study aims (i) to investigate the performance of different system identification methods, (ii) to identify the predefined damage of a test specimen, and different from the existing state of the art, (iii) to investigate the effects of the implementation of OTT on the system and damage identification results. The main outcomes of this paper are: (i) Almost the same results are obtained by different system identification methods, (ii) OTT enables to obtain realistic system identification results, (iii) the predefined damage of the specimen is estimated by the damage index method, and (iv) OTT affects the damage parameters in value, but does not influence the determination of the damage location.