Estimating bead width and depth of penetration during welding by infrared thermal imaging

The key to the success of robotic welding systems lies in the implementation of adaptive/intelligent welding. Adaptive welding enables dynamic altering of the welding parameters to compensate for changing environment. In a manual welding process, the welder ensures the quality of the weld by monitoring and suitably manipulating the process parameters according to the changing environment and based on his knowledge-base and experience. In the case of adaptive welding, sensors play the role of welder. Welding being a thermal processing method, infrared thermal imaging (thermography) using IR sensors have been a natural choice for sensing and weld process monitoring. This paper highlights the estimation of bead width and depth of penetration during Tungsten Inert Gas (TIG) welding using infrared thermal imaging. The results confirm that IR technique is a sensitive and convenient tool ideally suited to monitor the welding process by mapping and analysing the surface temperature distributions. Comparison of various methods to determine the bead-width during the welding process indicates that using the first derivative, it is possible to obtain a better estimate of the bead-width compared to the full width half maximum or point of inflection methods.