Research On Combined Die For Cold Forging Of Bevel Gear Based On Deform

The recent years'rapid development in automobile industry contributes to the application of cold forging technology. Parts produced by cold forging has good mechanical properties which are are also cost-effective, energy-saving and eco-friendly. Research on this technology is necessary so as to serve the manufacturing industry better.This thesis aims to do research on forging technology of the bevel gear, study how to improve the cold forging die of the bevel gear in order to obtain better forming quality, higher die strength and longer life-span.Main contents of this thesis are:(1) Used Pro/E to design 3D model, imported it into Deform to do numerical simulation so as to study the basic rule of cold forging, indicated the limitation of the single die by simulation results, thus introduced combined die to optimize the stress distribution of the die.(2) Conducted mechanical analysis, determined the parameters and goals for optimization.(3) Adopted orthogonal method to figure out test groups that contain structural characteristics of the die, obtained training data for neural networks via FEA, established neural networks, obtained the interpolated function by using linear regression analysis on neural networks, adopted genetic algorithm to optimize the global optimal solution of the mathematical function. The optimal solution was used for FEA, and then the feasibility of collaborative optimization of the combined die could be verified.(4) Finally, in consideration to the restriction of experiment conditions, a 1/3 model of the combined die with the optimal solution was manufactured and was used for cold forging testing. An experiment for test strain was designed to analyze influence of the interference fit within the combined die to stress distribution. Results show that die strength can be improved significantly with the optimized combined die design.