| In today's competitive manufacturing industries, there is a constant demand to improve productivity without sacrificing the overall quality of the product. For many welded products, an increase in productivity often requires use of higher welding speeds. As the welding speed is raised, it will result in welding bead defect of undercut and "hump". Constructing suitable vision-based sensing system to grabbing clear images of high speed GMAW weld pool, and researching the variation features of weld pool are helpful for knowing through the physical mechanisms responsible for humping phenomenon. Capturing clear images of weld pool and explaining the reason of weld pool defect formation are of great theoretical and practical significance in preventing or restraining the hump phenomenon, and finally increasing the welding speed.The liquid flow situation during high speed GMAW welding is unlike the situation in normal speeding weld. This can be deduced by analyzing the different formation feature of solid-liquid interface, both in the situation of humping phenomenon and non-humping. According to the weld pool feature of high speed GMAW, a vision-based sensing system is designed. Base on the analysis of the weld pool images, a series of image processing algorithms are developed to eliminate both of the noise and dentate edge. The edge dynamic searching algorithm is improved to extract the whole weld pool edge. This gives software support for the theoretical analysis of the mechanisms of humping formation.Through altering welding speed, several groups of pool images are captured from side-view. The height and solid-liquid interface of weld pool, are compared for the conditions with and without humping. The weld pool images arc processed to obtain the pool surface height and its variation with time. Good weld bead formation has been obtained through DE-GMAW experiment. Combing the captured pool images with the wall jet model, it explains why DE-GMAW does not form humping bead. |
PDF Full Text Request