Study On Welding Deformation Analysis And Optimization Of Car Sub-frame

With the development of China's automobile manufacturing industry,the improvement of people's living quality,the change of automobile manufacturing and mass consumption pattern,consumers demand that the precision of the product should be higher and higher.The sub-frame is an important part of the suspension assembly,therefore it is significant to improve the dimension accuracy of the sub-frame.At the same time,the improvement in performance of the suspension could bring better driving experience to the driver and passengers,and increase the competitiveness of product.The assembly process of the sub-frame mainly depends on the form of welding which is the most commonly used method in the modern automobile manufacturing process.Welding is a very complex procedure,and the welding deformation is a multidisciplinary problem involving electromagnetics,materials science,thermodynamics and mechanics.Generally,the welding deformation will affect the assembly accuracy of the sub-frame,some of which beyond the scope of qualified is useless,and the workload of rectification is large.Therefore,the study of the welding process of the sub frame in finding out the law of welding deformation and optimization control method of welding deformation,is valuable to improve the quality of the sub-frame.The following aspects are the main content of this paper:Firstly,this study introduce the current research situation of welding numerical simulation at home and abroad,and explain the general measures and means of welding deformation control.The theory of welding process is introduced,and the principle of welding deformation is studied.Finite element analysis model and mechanical simplification model were studied.Then,the paper introduces the main components of sub-frame structure and also described the welding process.Meanwhile,taking a vehicle sub-frame welding process as an example,we used the Weld Planner to make a infinite element model.By the finite element simulation of the welding process,we obtained the deformation after welding products.The results show that the traditional welding deformation is above the range of engineering application.By analyzing the rules of welding deformation,we take the form of adding the constraints and adjust the welding technology to control the welding deformation.This optimization,which makes the five welding process meets the product requirements.However,the upper and lower plates welding deformation can't access the above methods to reach the welding deformation standard.Finally,the buckled welding of the whole crossmember is studied.We analyzed the welding length during spot welding,the welding sequence of long welding seams,and the relationship between the welding sequence of long welding seams and welding deformation.The orthogonal experiment is for drawing samples.All the data obtained form the text are used to establish the response surface model of technological parameter and welding deformation in ISIGHT.By using the Multi-island Genetic Algorithm,the response surface model can calculate the minimum of the welding deformation.On that basis,an optimum design is carried out on the length of spot welding,the welding sequence of long welding seams and spot welding,the welding sequence in the long welding seams.By the adjustment and optimization of the welding technology,the welding deformation of the sub-frame of that vehicle is controlled in an acceptable range.The results show that the minimum welding distortion was obtained by the conditions as follows:the length of the welding joint is 20mm,the welding sequence of long welding seams is symmetric welding,the welding sequence of welding spot and the welding sequence in the long welding seams is also symmetric welding.The experimental results also demonstrate that the traditional method is not suitable to the modern welding process design.It is desirable to take a combination of numerical simulation and the experience to optimize the technical parameters by the means of building a agent model and numerical simulation experiment.