Per the CDC's recommendations on social distancing during the COVID-19 crisis, most LIFT staff is working remotely until Monday, July 6.
If you need assistance, please contact
Member Login

LIFT: Lightweight Innovations for Tomorrow

LIFT: Lightweight Innovations for Tomorrow

Robotic GTA Welding of Thin Complex Contours with Variable Fit-Up


Project Summary

The program will develop fully automated robotic welding of thin sheet metal parts with complex contours. Such parts have proven difficult to weld automatically because it has not been practical to achieve the precise fit-up needed to weld with a static set of weld parameters. The program will develop an adaptive weld approach wherein the joint is scanned for actual fit-up conditions and the weld schedule adjusts along the joint according to a database of predetermined weld schedules versus fitup conditions.

Technology Gap/Need

Automated welding of complex sheet metal shapes has proven extremely difficult. Normally an automated weld follows a constant weld schedule (set of parameter settings) established for a particular joint configuration. This weld schedule is established based on tests with ideal fit-up conditions – zero gap and zero mismatch. Industry experience shows that if the gap or mismatch exceed a small fraction of the material thickness, the pre-established weld schedule will not produce an acceptable weld. It will either fail to achieve full penetration or, more likely, it will burn through the material catastrophically.


This project will enhance the system to enable real- time adjustment of the weld parameters to adapt to variable fit-up and other part specific factors. The project will also produce a physics based model that will enable the approach to be efficiently adapted to multiple joint configurations and materials, streamlining the future advancement of the technology and enabling the robotic welding of complete systems.

Project Benefits

Today these complex structures must be welded manually. Even with a very skilled welder, defects are frequent. Weld speeds are slow so heat input and distortion are high. An automated system offers more than three times the throughput with no labor and much improved weld quality and repeatability.

Educational and Workforce Impact

Graduate students involved in the process will review the large volume of data from the experimental welds and develop a physics based model of the process. This will enhance the fundamental understanding of the weld process as they continue in the classroom and/or enter the workforce.

Project Duration

Start: April 2017
End: April 2018


Total Project Value: $1.45M


Industry Partners

  • GKN Aerospace
  • UTRC
  • Lockheed Martin
  • Comau

Research Partners

  • University of Notre Dame
  • EWI