Optimization of piezoelectric sensor location for delamination detection in composite laminates

The optimal placement of sensors is a critical issue in detecting damage in laminated composite structures. The aim is to use a minimum number of sensors, placed at the correct locations, so that the voltage signals received from the sensor set can be used to detect both the presence and the extent of damage. In this study, an optimization procedure is developed to detect arbitrarily located discrete delamination in composite plates using distributed piezoelectric sensors. The probability of damage distribution in the plate is determined using a statistical model. A genetic algorithm (GA) is used to detect the number and location of the sensors. The analysis uses a Monte Carlo method to generate the initial population. The simulation and signal processing is performed using a finite element procedure based on the refined layer-wise theory, which is capable of modelling ply-level stresses, and seeded delaminations are modelled with Heaviside step functions. A two-way electromechanical coupled field formulation is used to describe the induced strain. The objective function is a damage index which compares the voltage signals from a healthy (no delamination) and a statistically determined delaminated model. The voltage signals are affected by the local changes in the strain induced by the presence of delamination. The optimization solutions are verified by numerical simulation as well as with experiments conducted using customized piezoelectric sensors and a laser scanning vibrometer. The results presented show that the optimum sensor pattern is capable of detecting discrete seeded delaminations in moderately thick composite plates.