Damage Detection and Quantification in Cantilevered Beam Using Modal Strain Energy and Genetic Algorithm

Structural health monitoring is known to be a demanding area of research which has its applications in large-scale load carrying structures like bridges, aircraft, automobiles, offshore platforms, and submarines. Numerous optimization-based algorithms are suggested for damage detection in literature. Evolutionary optimization algorithms require minimum number of modal parameters as inputs and these methods are also capable to deal with incomplete data and noise. Therefore, genetic algorithm is preferred over deterministic optimization algorithms in the current study. Using genetic algorithm-based optimization techniques has the drawback of being computationally expensive because it requires estimating the fitness for the entire set of solution space at the end of each iteration of GA. In the current study, a multistage methodology is used to reduce the GA solution parameters. Modal analysis was performed on both undamaged and damaged structures. Damage was localized using the ratio of change in modal strain energy (MSECR) in order to reduce the solution space for the GA. An evolutionary optimization algorithm based on changes in frequency and mode shape is used in order to calculate the true damage percentage. Numerical studies of cantilever beam have been carried out to validate methodology. Numerical studies indicate that the multistage approach is more computationally feasible, robust, and rapidly convergent.