Application of Unified Smart Classification and Modified Weight Elimination algorithms to damage evaluation in composite structures

Unified Smart Classification Algorithm (USCA) for the purpose of data processing and classification of data obtained from different testing techniques is designed and tested. The developed algorithm conditions data taken from damaged composite structures such as modern car bodies and Plane frame structure. It is used in conjunction with a Modified Weight Elimination Neural Networks Algorithm (MWEA) to provide predictive models for impact damage in composite structures. The developed neural models correlated between various NDT testing techniques, such that in the absence of one technique, its results are predicted by the Neural Network through interrogation of available data obtained using other testing methods. The real and predicted data showed good agreements in terms of classification and prediction.