Study On Twin-wire CMT Welding System And Process For All-aluminum-alloy Semi-trailer Tank Body

In the global environment of green environmental protection,the tanker manufacturing industry has gradually embarked on a lightweight development path due to the social energy demand and the sustained rapid development of the oil transportation industry.The domestic market mainly uses 5083 aluminum alloy instead of traditional steel.However,Al-Mg aluminum alloy thick plates have problems such as high welding difficulty,low efficiency and poor weld consistency.There is no intelligent welding technology that can balance quality,efficiency and cost,and the mechanism and control method of high-efficiency arc welding of aluminum alloy are not very clear.According to the actual engineering requirements of the manufacture of all-aluminum semi-trailer tanks,the high-power high-efficiency series eutectic pool twin-wire CMT welding system and advanced function software are built by integrating and optimizing system hardware and software control resources based on the comprehensive consideration of the equipment composition,hardware interface,topology structure,communication protocol and software composition of twin-wire CMT mechanical execution,welding power,process control and signal acquisition subsystems in the paper.Based on the 32 sets of flat welding process and forming experimental data of full coverage equipment performance,keras was used to establish a 5083 twin wire CMT forming prediction neural network model.The results show that:The penetration depth,melting width and residual height of the twin-wire CMT weld are smooth and undistorted with the measured values.The model shows excellent multi-dimensional nonlinear fitting and generalization effects.After the measured data and the predicted data are serialized,the linear regression analysis can be used to evaluate the accuracy of the complex nonlinear model,and it is not subject to any limitation on the model dimension and data range,and has strong engineering application value.The core influence parameters of model prediction accuracy are the number of hidden layer neurons,network training times and deep network learning rate,and there is optimal parameter matching.A composite heat source model of the twin-wire CMT double ellipsoid heat source with Gaussian surface heat source was established by comparing and analyzing the high-speed imaging and synchronous electrical signal data and weld forming parameters of the 5083aluminum alloy twin-wire CMT welding process.And the grain coarsening index and t_5/3 of the welded joint were analyzed on the basis of the welding temperature field of MOOSE simulation and the welding thermal cycle data extracted by Paraview and Matlab.Then the improved phase field method is used to numerically calculate the evolution process of the joint microstructure under the welding thermal cycle.The results show that:The composite heat source model performs well in the simulation of the twin-wire CMT welding temperature field,and the calculation process is stable and reliable.The final microstructure state that determines the joint strength and toughness is determined by the welding thermal cycle through the process of grain coarsening and continuous cooling transformation affecting the original structure of the base metal,wherein the degree of grain coarsening is more significant than the cooling rate.The grain coarsening process simulation in the HAZ is highly matched with the EBSD test results.For the 5083 aluminum alloy twin-wire CMT welded joints,this paper firstly quantifies the matrix and mesophase according to the digital image features using microscopic quantitative analysis technology,and then analyzes the grain,grain boundary and texture by EBSD test technology.Finally,the mechanical properties test was carried out.The results show that:The intermetallic compound in the eutectic structure of the weld zone accounts for 12.48%,the average area is 3.74?m²,and the average width of the aluminum matrix sheet is about 5.58?m.The grain boundary of 45°⁵5°has the greatest influence on the strength and toughness of the joint.The heat-affected zone cubic texture{001}<100>has the largest single component content?20.73%?.The weld zone has the highest microhardness(about 93 HV_0.3),and the mechanical properties of the joints at different weld positions are similar,exceeding 80%of the base metal.An intelligent control method for aluminum alloy welding process parameters-weld formation-microstructure distribution-mechanical properties is proposed at the end of the paper.