Damage localization in plate-type structure using 2D-Bayesian technique

Failure in c ivil, mechanical, and aerospace structures are commonly due to heavy loads, environmental hazards and not to mention their lifespan. Structural health monitoring (SHM) has become vital to maintain the design and strength of these structures. For detection of damage, prior data of the structure is crucial but in the absence of the prior data, for example in case of antiquated structures, this becomes quite challenging. Gapped smoothing method (GSM) has been used successfully to detect damage with the response data only. However, noise in the measurement data is detrimental to its effectiveness. Moreover, single damage has been identified with most of the techniques particularly in plate-type structures, but those methods are not as successful in case of multiple damage. In this paper, a 2D-Bayesian technique is presented to suppress noise along with 2D-GSM to detect multiple damage in plate-type structures without using any prior data. The technique uses the fusion approach to combine the information from the available mode shape data by reducing the noise and highlighting the damage locations. Different damage scenarios in plate-type structures are simulated in ANSYS. To evaluate the effectiveness of the proposed method, the simulated data is contaminated with noise and results are compared with the published work.