An Evolutionary Algorithm-Based Damage Detection in Structural Elements

Abstract

Damage in structural and rotating machine elements causes the local changes in the dynamical parameters of the system. To find the damage present in a system is becoming one of the important research topics in today’s civil and mechanical engineering field. This topic of research can mainly be used in bridges, offshore platforms, plates, shells, beams, aerospace and composite structures and other large civil structures to detect structural damages by analyzing the dynamic features of the system. It has been observed due to any slight physical change, the stiffness of the system changes which changes the modal responses of the system. The aim of this research work is to derive a simple method for estimating the failure parameters (crack depth and crack location) in structures based on a data-driven subspace identification technique. These changes in the modal parameters can be used as the input variables to find out the damage severity. The responses (natural frequencies) were obtained using finite element analysis, and then, the differential evolution algorithm (a type of evolutionary algorithm) is used to detect and characterize these defects. This work proposes a robust computational application of the differential evolution algorithm that more accurately takes into account the natural evolution with initial point and produces a good converging result.