Data fusion in non destructive testing using fuzzy logic to evaluate friction stir welding

摘要

The Friction Stir Welding process (FSW) was inventednin December 1991 by The Welding Institute, and todaynrepresents one of the most important developments innwelding technology in recent decades. However, itsntransfer into regular production use also requires reliablenNDT techniques. The current state of development ofnsuch techniques does not allow the detection of defectsn(the more correct term, in NDT, is discontinuities, butnthe term defect will be used, on the assumption thatnany discontinuity present constitutes a harmful defect)nthat are important in structural performance, undernstatic pressure and for fatigue, as well as the corrosionnresistance of the welded components. [1].nGiven the location and the specific morphologynof characteristic FSW defects, their detection andndimensioning are very difficult. As a result, no NDTntechnique exists that can be used individually to ensurena complete and reliable inspection1,2. To resolve thisnproblem, some works have taken place, among whichnthe European QUALISTIR project3 stands out, whichnstudied the application of advanced ultrasound techniquesnvia phased array. One of the conclusions of this studynrefers to the need to integrate different NDT techniquesnwith a view to obtaining the best results.