IN LINE DEFECT DETECTION IN STEEL WELDING PROCESS BY MEANS OF THERMOGRAPHY

In welding processes, the in-line identification of defects during manufacturing is a key aspect of a safe, reliable and economical process. Among the various non-destructive techniques, thermography is a particularly suitable technique as it is a non-contact full-field technique. However, this technique still has aspects to be improved, in particular as regards the definition of the resolution and the quantification of the extent of welding non-conformities in industrial production contexts. In the present work, an approach based on real-time defect detection on welded joints supported by its numerical model is proposed and analyzed for two arc welding processes, SMAW and GMAW on a steel alloy by means of active thermography, in order to reduce the time and costs of inspections. Different defects are identified, outlined, and discussed. Furthermore, a metallographic analysis is carried out to verify the accuracy of the model.