| Keyhole Tungsten Inert Gas(K-TIG)welding is a new welding method that uses high current to form a "keyhole" inside the molten pool on the basis of traditional TIG welding.It has the advantages of high welding efficiency,single-side welding and double-side forming,etc.It is often used in the welding of thick plates in large equipment.In the welding of large equipment,closing butt welding is an important process,which directly affects the production efficiency and quality of manufacturing equipment.Due to the reasons of processing technology and precision,it is difficult for the equipment components to be closed butt welded to maintain uniform butt gaps at each location,and there is the problem of butt gap variation.Therefore,K-TIG welding method was adopted in this paper to carry out the research on the visual inspection of the preset butt gap welding and the recognition of the penetration state of the 12 mm thick 304 stainless steel,which can lay the foundation for the realization of the welding seam formation control under different butt gaps of large equipment components.In this paper,a K-TIG welding molten pool and keyhole entrance feature detection system based on vision sensing is built,and a high-dynamic CCD camera is used to collect images of the welding molten pool and keyhole entrance.Combined with K-TIG welding mechanism and the image features of the molten pool and the keyhole entrance,an algorithm for extracting the characteristic parameters of the molten pool and the keyhole entrance is proposed.The image is divided into the solidification region of molten pool,the middle region of molten pool and the keyhole entrance region.The edges of molten pool and keyhole entrance are extracted by filtering,enhancement,edge detection,false edge removal,edge fitting and other algorithms.According to the edge information,five parameters are defined,including the width of keyhole entrance,the half length of keyhole entrance,the length of molten pool,the width of molten pool and the trailing length of molten pool.Finally,the accuracy of the molten pool width is checked,and the detection accuracy is above 95.60%.Under the conditions of constant welding parameters such as welding current(580A),welding speed(240mm/min),and the distance between the tip of the tungsten needle and the surface of the weldment(CTWD),the K-TIG welding test with different butt gaps was carried out,and it was found that The butt gap has a great influence on the characteristic parameters of the molten pool and the entrance of the keyhole.With the increase of the butt gap,the penetration of the weld is gradually enhanced.The penetration of the weld is the best when the butt gap is 1.1mm,and the welding collapse defect occurs when the butt gap is 1.4mm.The influences of welding current and welding speed of K-TIG welding on the characteristic parameters of molten pool and keyhole entrance,weld morphology and weld penetration state at different butt gaps were analyzed.It is found that the welding parameters of different butt gaps to ensure good weld formation are different.As the butt gap increases,the welding heat input to reach the full penetration state decreases gradually.According to the variation of the characteristic parameters of the K-TIG welding molten pool and keyhole entrance of different butt gaps,a BP neural network penetration prediction model was established,which took characteristic parameters of the molten pool and keyhole entrance,welding current and welding speed as input,and weld penetration state as output.The prediction accuracy of the model can reach 93.20%.Finally,the experiment verifies the generalization performance of the model. |
PDF Full Text Request