A measurement and control algorithm for GMA welding with constant energy input

To ensure consistent mechanical and technological material properties, maintaining precise cooling times during welding is crucial. Tolerances in weld groove preparation may introduce localized variations in arc power due to varying contact-tip-to-work distance, resulting in changing energy input and thus variable cooling times. This phenomenon is shown at the example of tack welds using a sensor system comprising a moving thermal camera along with current and voltage measurements. Using tack welds as an example of discontinuities in the weld seam, it was shown that the heat input increases by approximately 8 to 10% depending on the magnitude of the contact-tip-to-workpiece-distance variation, resulting in an extension of the cooling time by 10–20%. To mitigate these variations in energy input, an algorithm was developed to identify these discontinuities within 0.1 s. The algorithm works as a semi-supervised discontinuities measurement and control algorithm based on the Mahalanobis distance and detects discontinuities in-situ with a height of 3 mm and a length of 7 mm. The application of this algorithm-supported adaptation of the energy input enables constant cooling times and weld seam geometries. Beyond such conventional welding applications, the presented measurement and control algorithm is also applicable in additive manufacturing scenarios.