Crack Detection of Thin Plate Structures Utilizing Measured Vibration

AbstractThis paper reports the development of a crack detection method to identify cracks on thin plate sstructures using measured vibration responses on only a few points of the target structure. One of the most difficult problems for model-based crack detection methods is the estimation of the number of cracks. In the proposed methhod, the number of cracks is identified by following the Bayesian moddel class selection method. Unlike methods following the deterministic approach, the objective of the proposed method is to calculate the posterior (updated) probability density function (PDF) of the crack parameters, such as crack locations, lengths and depths are identtified following the Bayesian statistical identification framework. This paper presents a series of numerical case studies for illustrating and verifying the proposed method as well as the theoretical development. The numerical case study results show that the proposed method can correctly identify the numbeer of cracks, the corresponding crack parameters and the associated uncertainties. Several factors affecting the accuracy of the crack detection results will also be discussed at the end of this paper