| Currently, the ship hull plate forming is mostly used in line heating technology in the major shipyards. This processing methods have many disadvantages and its production efficiency is relatively low, To some extent, it restricted the development of the shipbuilding industry. For this reason, Many scholars now begin to research machining technology for ship hull plate forming, changing the technology of ship hull plate forming from line heating processing into cold-formed processing. However, cold-formed process of ship hull plate forming is a very complex process and is influenced by many factors. So the springback is inevitable in this process. Springback will influence the forming quality of the ship hull plate, so it is important to do research and prediction of the springback.This paper used DYNAFORM to do numerical Simulation of the crash forming of ship hull plate. On the basis of the results of the simulation of the crash forming to do the simulation of springback and analysis of the result of the simulation of springback. It also studied the impact of various factors to springback, mainly studied the material properties of sheet metal, forming radius of curvature and the thickness of plate. By the study of the results of the springback simulation shows that:The bigger yield strength of the plate, the greater the amount of springback; with decreasing radius of curvature forming the springback of plate decreasing and decreasing trend gradually leveling off; with increasing of thickness of plate the springback of plate decreasing and decreasing trend gradually leveling off.In the prediction of springback, used BP neural network as the theoretical basis, in accordance with numerical simulation data to make springback prediction of different thickness of plate by MATLAB neural network toolbox. The results of prediction were compared with the numerical simulation analysis results, the deviation is very small. Leading to the conclusion:the prediction of springback by MATLAB neural network toolbox is accurate. |
