| Base on the engineering practice, this paper contains three major aspects. Firstly, it presents how to calculate the rail panel's springback accurately by 2D cross-section method and compensate the cross-section nodes by cross-section compensation algorithm. Secondly, panel surface after springback compensation can be achieved by SGD algorithm the surface is smooth and can be used to NC machining. Thirdly, the accurate trimming line is not easy to obtain especially when there is a large springback, so, the scale coefficient method is presented to optimize the trimming line and it's verified that the accurate result can be obtained.Main works of this paper are given as follows:1. Three springback modes which include angular change, sidewall curl, and twist are researched based on experiments of HSS. And methods will be given to deal with the three types of springback separately. An engineering example—front side rail outer is presented, twist springback of this rail will be investigated and some solutions including module process and correction will be proposed to minimize the twist. One of the most effective methods to solve the HSS springback is changing addendum shape and structure. It can improve material's stress distribution.2. Based on the characteristics of rail panel, it is presented to calculate the springback amount of rail panel accurately by 2D cross-section method in the condition of assumed plane-strain etc. The springback of all the cross-sections will be compensated by self-development section compensation algorithm. Some bugs of original section compensation algorithm are found in the practical applications of HSS. Considering the longitudinal springback and undercut compensation problems occur easily in HSS, the compensation algorithm is corrected based on the different reasons. After correction, the cross-section compensation algorithm is better fit for engineering application. From the contrast of springback results of different sections, we can deal with 3D rail panel's springback by 2D section method and compensation algorithm. It is proved to be universal significant and applicable in practice.3. The SGD (surface global deformation) algorithm which can modify the surface shape and ensure the quality of the deformed surface is presented. The characteristics of SGD algorithm are that it can keep the topology of panel surface invariable before and after deformation, keep the surface of connected areas smooth and ensure the quality of deformed surface as the original. The precise of the spingback compensation is also correspondingly increased in the condition of ensuring surface quality and deformation precise. The research of the panel's springback compensation and deformation by SGD algorithm contains three aspects. Firstly, it research how to use SGD algorithm to modify the original part surface and obtain the compensation surface through actual results of springback. Secondly, in the FEA simulation, matching section line of before and after springback will be gained by 2D section method and then compensation surface will be gained by SGD algorithm. Thirdly, the compensation surface can be gained by the SGD algorithm through all matching nodes of before and after springback.Because of using the SGD algorithm, the precise compensation surface can be gained easily and the quality and precise of panel surface have greatly increased.4. Scale coefficient method is presented in this paper. It doesn't give the same adjustment to each node but give the corresponding adjustment which are calculated by each node's scale coefficientωand shape error. This adjustment have better pertinency and sensitivity. And the spatial nodes can be searched effectively, rapidly and full ranged. So, more precise result can be obtained and less iterative time can be needed. The optimization algorithm makes iterations in accordance with the finite element analysis results. It considers stamping conditions such as part's material, thickness, forming force of flange and friction force etc., which is much closer to the practical process. Besides, the initial contour which is obtained by one-step algorithm or section method can reduce the iterative search region. So, the scale coefficient method not only has greatly shortened the optimization time, but also has greatly improved the accuracy.
|
PDF Full Text Request