Friction stir welding line forming prediction optimization method based on numerical simulation and deep learning

Abstract

The invention provides a friction stir welding line forming prediction optimization method based on numerical simulation and deep learning, and belongs to the technical field of friction stir welding.The method comprises the following steps: step 1, setting three times of simulation tests as a data test set; step 2, calculating distribution conditions of a material flow field and a temperature field in a welding process; step 3, calculating the fracture failure condition of the friction stir welding tool under different parameters, and calculating the forming quality and defect distribution condition of the welding line under different parameters; and step 4, traversing all the process parameters and the welding line forming result of the welding tool structure by using the generative adversarial network deep learning model to obtain a welding line forming optimization result on the premise of ensuring the reliable work of the welding tool. The method aims at providing an effective universality prediction method for optimizing the friction stir welding process in production, and has the advantages that time consumption is reduced, the material cost is reduced, and the prediction accuracy is high.

Description

technical field [0001] The invention relates to a method for predicting and optimizing the formation of friction stir welding seams based on numerical simulation and deep learning, and belongs to the technical field of friction stir welding. Background technique [0002] As the most eye-catching welding method since the advent of laser welding, friction stir welding (Friction Stir Welding) has significant advantages in overcoming welding defects such as cracks and pores that are easy to occur in the process of traditional fusion welding of light alloys, making it difficult to weld Some of the welded materials have achieved reliable high-quality connections. The basic working principle is to use a welding tool with a shoulder (or with a stirring pin) for high-speed rotation, and a large amount of frictional heat and plastic deformation heat are generated between the tool and the workpiece. Make the workpiece material partially plasticized. When the welding tool moves forward ...

AI Technical Summary

Problems solved by technology

[0003] However, the immature friction stir welding has the following disadvantages: (1) There is no universal theory for welding tool design

In the friction stir welding process, the welded material is mainly driven by the welding tool to flow and form a dense weld structure. At the same time, the welding tool will be strongly resisted by the flowing welded material. Unreliable welding tool design will lead to poor weld formation , welding tool breakage and other problems; (2) At present, the research on the optimization of friction stir welding process parameters is mainly based on experiments. The number of times of experimental design is large, the workload is huge and the repeatability is weak, etc. The cost of experimental research and prediction of optimization prediction is extremely high, and the parameter optimization process in actual engineering applications is usually relatively simple.

Images