Monitoring of weld joint penetration based on weld pool geometrical appearance: The geometrical appearance of the weld pool contains sufficient information to determine full joint penetration in GTAW

Weld joint penetration monitoring and control are fundamental issues in automated welding. A skilled human operator can determine the weld penetration from the geometrical appearance of the weld pool. To emulate this using machine vision, a high-shutter-speed camera assisted with pulsed laser illumination is used to capture the clear image of the weld pool. The pool boundary is extracted by the developed real-time image processing algorithm. In order to emphasize the emulation of the human operator, general terms, i.e., size, shape and geometrical appearance, are used for the conceptual discussion, whereas more specific terms such as length, width, and rear angles are used in the detailed analysis. In particular, the size will be specified by the pool width and length, and the shape will be defined using the proposed rear angle of the weld pool. The geometrical appearance is described by a combination of the size and shape parameters. To investigate the relationships, which could be complicated, between the weld penetration and different parameters, neural networks are used because of their capability for modeling complicated nonlinear functions. Extensive experiments have been developed to measure the weld penetration from the captured image in 200 ms using the neural network and real-time image processing.