Welding seam inspection by non-destructive testing

Problems in Production

Avoidable machine downtime, a shortage of skilled workers and low automation rates are common problems in the industrial environment. The need for reliable, automated procedures for machine, process and product monitoring is cited by both users and machine manufacturers. Users generally focus on stabilizing processes, reducing part cycle times and optimizing tool life.
 

Acoustic AI for non-destructive testing (ndt) of welding types

Feedback from specialists shows that the sounds generated during welding can provide important information about the stability of the process, the quality of the welded joint and tool wear. The goal is therefore to integrate automated acoustic process to detect weld irregularities reliably and at an early stage in the future. This is the subject of current research at Fraunhofer IDMT.
 

Successful feasibility studies for acoustic welding seam inspection

The publicly funded project "AKoS - Acoustic Control of Welding Seams" has shown that airborne sound-based monitoring is a promising approach for detecting process irregularities and welding defects, such as in laser welding and other welding types. Based on laboratory tests, a measurement system was developed that synchronizes sensor and process data in a real environment. With these data, irregularities in the welding process in MAG welding could be detected with an accuracy of up to 95%. In friction stir welding (FSW), tool wear could be determined from the acoustic characteristics.
 

Looking for R&D partners

Based on the current results, joining processes have been identified as a promising field of development for the ndt acoustic detection of welding seams. Fraunhofer IDMT in Ilmenau is therefore looking for interested and reliable partners from the fields of mechanical and plant engineering, tool manufacturing and users from the field of joining technology for joint development and implementation.