Study On Forming Springback Of Clutch Cushion Segment And Its Simulation Prediction

Springback is a common phenomenon in the forming process of sheet metal, and it affects the precision and assembly of workpiece directly. Generally, springback is complicated because it involves material nonlinearity, contact condition nonlinearity and geometrical nonlinearity. Springback is inevitable in the forming process of sheet metal which has a relationship with material properties, the geometry of dies, the process conditions and other factors. Therefore, the problem of springback has attracted wide attention. Accurate prediction of springback is the prerequisite for control of springback. Traditionally, the "trial and error" method would be applied to adjust the process parameters and the die-face. With the development of numerical simulation, in order to shorten the cycle of die design and reduce the times of die repairing, numerical simulation is used to guide the die design widely. Therefore, it reduces the cost of production and provides strong competitiveness for enterprises.Dynaform is used to simulate the forming process and springback of CK75 for cushion segment, studying the relationship of forming height for cushion segment to process parameters, designing the forming mold of cushion segment to verify the forming height of cushion segment. Additionally, BP neural network is emploied to predict the forming height of cushion segment. The main research may be summarized as follows:(1) The basic theory of bend-forming is described in this paper, The flow of material, stress of springback-before and springback-after are investigated, the release of residual stress is an important reason for springback.(2) The influence of clearance, radius of bending, velocity, friction coefficient and thickness of material on the forming height of cushion segment are studied by numerical simulation, designing the forming mold of cushion segment to verify the relationship between clearance, radius of bending, thickness of material and forming height of cushion segment by the experiments.(3) Orthogonal experiment is applied to simulate springback of cushion segment at different process parameters. A conclusion is made that thickness and clearance have an significant influence on the form ing height of cushion segment, while the order of factors is clearance, thickness of material, friction coefficient, radius of bending and velocity respectively. The optimized combinations are acquired for the bend-forming of cushion segment by orthogonal experiment.(4) BP neural network is emploied to predict the forming height of cushion segment, which contains structural design, training and testing of BP neural network. The results show that BP neural network predicts the forming height of cushion segment effectively and it provides a certain basis for die design.The working achievement of this paper will enrich and develop the bend-forming technology of cushion segment, the obtained conclusions will offer instructional significance to application of this technique in enterprises.