Damage detection for composite plates using Lamb waves and projection genetic algorithm

A method of damage detection for composite plates using Lamb waves and a projection genetic algorithm is investigated in this study. This method first formulates the damage detection as an optimization problem of minimizing the error between the measured and calculated surface displacement response derived from Lamb waves. The calculated response is obtained by using the strip element method with the trial crack parameters (location, size). Then a projection genetic algorithm is used to solve the optimization problem and thus identify the actual crack parameters. The projection genetic algorithm is developed from the hybridization of the modified micro-genetic algorithm with a projection operator. It has an excellent convergence performance, taking only 9.9 similar to 14.2% of the number of function evaluations required by the conventional micro-genetic algorithm to obtain the global optima for six benchmark functions. Numerical examples are presented to verify the proposed method for detection of cracks inside composite plates. The maximal error of detected crack parameters is -4.3% for four simulated cases, which is achieved by running only 60 generations of projection genetic algorithm.