The duct is one of the key components of an aeroplane.At present,argon arc welding is the main method for the manufacture of aircraft welding tubes in China,which is characterized by high risk factor,low production efficiency and unstable welding quality.In view of the backward manufacturing status of aviation duct,this thesis aims to establish a set of automatic and intelligent aviation duct welding test system,and researches the key technologies in this system,such as off-line programming path planning of duct and flange,off-line programming sheet metal deviation and clamping deviation compensation,welding hot deformation compensation and so on.First,in view of the air duct,such as the complex shape,small diameter,wall thickness,established a set of suitable for different shapes and space orientation of general welding test system,the hardware part includes: an six axis machine and welding gun,a set of linear structured light laser vision sensor,a two-axis displacement machine,a dedicated digital inverter welding power source,etc.;The main functions include: a)off-line programming path planning for aerial conduits;b)Sheet metal deviation and clamping deviation compensation of conduit and flange;c)Welding thermal deformation deviation compensation in advance;d)Real-time compensation for weld tracking;e)Post-welding quality inspection.The virtual working environment is built to simulate the welding site of robot,the movement trajectory of welding robot is planned offline,and the program that can be executed by robot is generated initially.Multidirectional analysis of error sources of off-line programming path planning;In this thesis,the influence of the preweld deviation of the conduit and flange on the post-arrival process is expounded,and the inversion model of the sheet metal deviation and clamping deviation of the conduit and flange welding is put forward,and the error compensation idea of compensating multiple deviations is realized by adjusting the deviation size through the Angle of the positioner.A method of automatic compensation was obtained by introducing the deviation caused by thermal deformation into off-line programming path planning through finite element simulation of welding pipe and flange,which improved the alignment neutrality between the real position of weld after thermal deformation and welding torch.Deflection is deduced from the machine collaborative robot welding line structured light vision sensor laser and welding torch pose relationship,determine the structure creation time,make the welding torch is consistent with the line structured light precise spatial location,the traditional first generation "teaching--reappearance" type development for robot capable of real-time tracking of intelligent robot posture adjustment.Make welding test plate and test tube according to actual manufacturing process in factory;The welding parameters suitable for butt welding of 6063 aluminum alloy flange to medium thin wall aerial conduit were determined.When the welding current is 130 A,the welding voltage is 20.5V,and the average welding speed is 720mm/min,continuous and uniform welds with good forming state can be obtained.Finally,through the modified line structure light and welding torch reach the same space position,the weld tracking online compensation,after compensation,the maximum transverse and longitudinal deviation of the weld is ±0.5mm and ±0.6mm,respectively.The weld formed after welding is scanned by sensors for simple Three-Dimensional reconstruction to measure the weld width and residual height and observe whether there are cracks and other defects on the weld surface.Compared with other nondestructive testing,it is more cost saving and can effectively screen out the weld that does not meet the welding standards.The results show that the average residual height and weld width are2.244 mm and 7.789 mm respectively,which are in line with the requirements of actual production. |