Defect Detection and Identification on an Embedded Beam Using Modal Deformation Parameters and the Dempster–Shafer Data Fusion Technique

Abstract

AbstractModal analysis was performed on an embedded beam with numerical simulation using Ansys finite element software. Strain modal parameters such as modal strain energy were calculated for undamaged and damaged models of the embedded beam. It is well known that the apparition of the defect in a structure will affect its strain modal parameters and their derivatives, and this effect is more significant in the defect region than in the other far regions. The most sensitive parameters to defects are the strain mode shapes, the frequency response function, and the modal strain energy. In this study, by using the changes in the modal strain energy before and after the fault insertion on the beam model, the position and severity of the fault can be identified. In addition, these parameters can be measured in practice on real structures with high fidelity using strain gauges. To ensure high efficiency of fault identification by the modal strain energy change method, a multi-step Dempster–Shafer data fusion technique was applied.KeywordsDefect detectionDempster–Shafer data fusion techniqueModal strain energy