Study on correlation methods for damage detection: Simulation and experimental validation

Abstract

In the current days there is an increment of interest in damage detection methods, aimed to assure the operating status of existing structures or for intensifier quality control on production line. These are only some of the applications whereby damage detection methods have been dev eloped. In the past several researches have been addressed towards damage detection using vibration analysis, especially through mode shape and natural frequencies changes. In the preset study correlation methods based on ODSs have been developed. The structure was taken under consideration is steel plate. The correlation methods presented are based on the comparison of the ODSs generated by two FEM models of the plate, one defined as pristine and the other as damaged. The latter has been modelled adding a single node mass element to the model surface. This mass element was chosen to simulate a magnet attached to the surface plate in the experimental case. Several simulations have been performed using combinations of mass and positions, for a total of 16 cases. Studying the correlations between a ODSs pair, given by the same excitation frequency and position, is possible to identify the presence of damage in the structure. The experimental model validation has been performed using the best excitation condition obtained by simulation, which can point out large differences between the damaged ODS and undamaged ODS.