Research Of Sheet Metal Forming Based On Defence Node Algorithm

Sheet metal forming is a very important technology in automotive manufacturing.Most automotive panels are obtained by sheet metal forming.In sheet metal forming process, the external forces which make deformation ofsheet metal are blank holder forces and contact forces. Blank holder forces are fixed,therefore, the calculation of contact forces directly affect the simulation results offorming process. Internationally accepted calculation methods of contact forces are:①penalty function method;②Lagrange multiplier method. Penalty function methodis an approximate method which allows contact borders penetrate each other. It makesrelationship between contact force and penetration by penalty parameter. Penaltyparameter multiply penetration is the size of contact force. This method is simple andless calculation, but is difficult to select penalty parameter. To make less borderpenetration, we should choose large value of penalty parameter which will increasethe system stiffness, lead to increase the system frequency and affect the critical timestep of the system. Lagrange multiplier method does not allow the borders topenetrate each other, and is an accurate calculation method of contact force. But it isincompatible with the explicit algorithm, and requires special numerical treatment. Inthis thsis, we choose the defence node algorithm which is proposed by ProfessorZhong Zhihua to calculate contact force. It can calculate contact force accurately, canuse Lagrange multipier method to satisfy constraints accurately, and also can avoid tosolve simultaneous equations.Around thesubject of the defence node algorithm inapplication of sheet metal forming, contents of this thesis are as follows:(1) Finite element program of sheet metal forming based on defence nodealgorithm was developed.(2) Tested and verified the advantages and disadvantages of the penalty functionmethod. These illustrated the necessity to use defence node algorithm to calculatecontact force.(3) Compared the calculation accuracy between developed program andcommercial softwares and experimental results. Verified the feasibility of theprogram.(4) Based on the developed program, an optimization design method based onmulti-island genetic algorithm is proposed to control springback. Calculation exampleshows that this method can better control springback.