||Devon Goodspeed, Bradley Jared, William Hamel, Josh Penney, Shems Belhout, Ethan Rummel, John Codevilla, Dylan Lewis
With the average age of a weldor in the United States reaching fifty-five, the welding industry will face a shortage of roughly 400,000 welders by 2024. Because of this, the need for automated welding systems has never been more relevant, especially in the power-generation industry, where automated welding systems can be employed in locations that humans cannot due to spatial or environmental concerns. For these applications, TIG (tungsten inert gas) welding is preferred for its cleanliness, precise heat control, and consistency of weld quality. The University of Tennessee (UTK), partnered with the Electric Power Research Institute (EPRI), has developed a TIG welding robot capable of operating in these environments.
The robot we have developed is driven along an industry-standard track, and has five degrees of freedom, which include translation parallel to the weld groove, translation perpendicular to the weld groove, vertical translation, which also incorporates automatic voltage control (AVC) that controls tungsten height to maintain a certain arc voltage, as well as vertical translation of the filler material being deposited into the melt pool, and, finally, a rotational axis controlling where the filler is deposited across the melt pool. This system is also expandable to include pitch and roll axes that will allow for increased flexibility with weld geometries, as well as torch positioning.
This is accomplished by using an iterative design process, where the efficacy of various drive mechanisms and solutions are explored, validated, and iterated upon until the desired performance characteristic is obtained. These performance metrics include velocity, positional repeatability/maintainability, and overall weld quality on a v-groove geometry with a simulated root-weld in 316 Stainless steel.
The robot also carries an onboard suite of sensors, including two Cavitar cameras focused on the leading and trailing edges of the weld and that produce high-resolution imagery of the melt pool while removing arc glare, a Keyence laser profilometer that scans the groove and produces a 3d topology of the solidified weld, as well as the ability to record and log weld parameters from the Liburdi power supply via LabView for further data analysis.
As a result of our iterative design process, we have identified ideal solutions to the locomotive needs of our TIG welding robot, and future developments will be focused on refinement of the overall packaging and form factor of the robotic welding system, as well as expansion to seven degrees of freedom with weld defect detection capabilities and an adaptive welding capability.
With this robot, we have demonstrated the ability to develop and implement a five degree-of-freedom automated welding system incorporating datalogging capabilities and an onboard sensor suite to better understand and react to possible weld defects. The mechanical design, iterations, and implementation of the robot will be discussed and demonstrated, as well as a timeline for future development.