Laser vibrometry based detection of delaminations in glass/epoxy composites

The significant progress in sensing and data processing technology has made monitoring and damage detection of engineering structures increasingly attractive. This paper presents a reliable in-situ damage detection technique, which is based upon dynamic analysis of a composite structure using bonded piezo-ceramic patches as actuators and a Scanning Laser Doppler Vibrometer as a sensor. In addition, Neural Networks have been considered to be a viable tool for handling the large number of data. A multilayer perceptron (MLP) neural networks, was trained and tested using the slope, the y-intercept of the linear fit of the root mean square of the Frequency Response Function (FRFrms) and the Deviation of the FRFrms of a candidate composite structure.