Damage integrity assessment for beams using structural health monitoring technique

Structural health assessment of critical structure is one of the essential safety requirements. Three levels of structural health assessment are performed in this paper, i.e., damage identification, localisation and quantification. Among all the computational tools available in civil engineering for damage detection, artificial neural network (ANN) and support vector machine (SVM) are aimed to detect damage in beams. Cantilever I and hollow beams are modelled using finite element software and modal parameters are extracted. An attempt is made to demonstrate that parameters such as mode shapes and frequency are adequate for detection of structural damage using three classical techniques namely frequency-based damage detection method (FBDD), mode shape-based damage detection method (MBDD), mode shape curvature square (MSCS) damage detection techniques. Damage scenarios are created in the beam with various severities and locations along the beam. The modal parameters thus extracted from ANSYS for damaged beams are used as input for ANN and SVM algorithms for damage assessment. Thus, the results show that both algorithms are accurate for detecting damage in cantilever beams but ANN performed better than SVM.