Automated identification and classification of internal defects in composite structures using computed tomography and 3D wavelet analysis

The necessity of detection and identification of damages in structural elements intensifies the development of non-destructive testing (NDT) methods and techniques suitable for different composite materials and sensitive to particular types of damages. One of the most sensitive and accurate methods is the computed tomography (CT), which has already found a wide application in medical diagnostics and materials testing. However, there are some difficulties during testing of composite materials, whose internal structure is often quite complex. Other difficulties of application of CT are related to correct identification of diverse composite-specific defects. The following study presents the method of identification and classification of internal defects of polymeric composite structures after the water-jet cutting based on CT scanning and 3D wavelet analysis. The application of wavelet transform allows extracting meaningful information about the size of the defect, their locations and orientations and reducing the measurement noise. An applied classification procedure is based on the analysis of geometric properties of identified defects and the magnitudes of detail coefficients resulted from the wavelet transform. Proposed method could find an application both in quality control of composite components as well as their NDT during operation.