Automatic defects recognition in composite aerospace structures from experimental and theoretical analysis as part of an intelligent infrared thermographic inspection system

An original concept for IR thermography nondestructive testing is validated. The principles of image and data processing investigated and developed as well as the utilization of AI should be transposable to other nondestructive techniques such as ultrasounds and X-rays. It is shown that modeling can be used in different ways to play a great part in the detection, the interpretation, and the sizing of the defects. The original concept lies in the comparison of experimental data with theoretical ones in order to identify regions of abnormal behavior related to defects. A Laplace transforms analytical method is successfully implemented in the case of composite materials such as graphite epoxy to identify a set of thermal parameters which contributes to the expertise. This approach is extended to a more complicated composite material such as Kevlar, which presents semitransparent characteristics. This modeling technique, which expresses experimental data in terms of thermal parameters, makes it possible to increase SNR and reduce the number of thermal images to be processed.