Effect of electric parameters on weld pool dynamic behavior in GTAW

Assembling robot with human welder's intelligence is a good way to rapidly realize intelligent welding manufacturing. How to finely and accurately control dynamic weld pool, however, is a key issue for the intelligent robotic welding. To this end, an improved experimental system is built to simultaneously sense the weld pool dynamics and the electric parameters. The ROI signal of reflection laser images was obtained by a novel image processing algorithm, and the correlation between the weld penetration and the ROI signal was further established. The human welder's responses on the weld pool were detailly analyzed by the experiments with varying electric parameter. The effect of the electric parameters on the weld pool equilibrium was also studied by the established model which included the latent heat of phase change. The results indicate that the welding current, welding speed and arc length mainly changes the heat input to the weld pool, and easily loses its stability, resulting in a lack of penetration or excessive penetration. The fastest change of the weld pool occurs on the adjustment of the welding current, and the weakest change of the weld penetration happens in the adjustment of the arc length. Simulations ensure that the new equilibrium of the weld pool can be achieved within 1.8 s via adjusting the welding current, within 2.1 s via adjusting the welding speed, within 8 s via adjusting the arc length, respectively, which might be used to optimize the intelligentized controller and improve the accurate of the human welder's intelligence model.

Automated summary

An improved experimental system is used for sensing weld pool dynamics and electric parameters, establishing the correlation between weld penetration and ROI signals. The study finds that welding current, speed, and arc length influence weld pool equilibrium and require fine adjustments.