The precision forging process of spiral bevel gear has high requirements for pre forming and forging process parameter design. Reasonable process design and analysis of pre forming process and forging process of precision forging of spiral bevel gear can realize the comprehensive purpose of reducing forming load, improving forming quality and ensuring die life, and can realize the efficient, accurate and high-quality forming and manufacturing of driven spiral bevel gear of automobile rear axle, It has important theoretical and engineering application value.

Taking the precision forging of automobile rear axle driven spiral bevel gear as the research object, the forging process is optimized from the aspects of machining accuracy, quality and energy consumption, mainly including the design and selection of process scheme, the optimization design of preform process parameters, preform shape optimization The main research work and conclusions are as follows:

(1) The stress-strain curve of 22CrMoH gear steel is obtained through thermal compression test. It is found that when the temperature rises and the deformation rate decreases, it is more conducive to the occurrence of dynamic recrystallization, refine the grains inside the material and show good comprehensive properties. The constitutive equation of 22CrMoH gear steel is fitted and the hot working diagram is drawn. It is found that the power dissipation efficiency exceeds 35% in the temperature range of 930-1130 ° C and the strain rate range of 0.22-1s-1, which is the best processing area of the material.

(2) The precision forging process of driven spiral bevel gear, including blanking, upsetting and punching, ring rolling and final forging, is proposed. The thermal mechanical coupling finite element model of radial dairy ring is established on deform platform, and the distribution and variation laws of physical fields such as stress, strain and temperature during the rolling process of preformed ring are analyzed. Through theoretical analysis and experimental verification, it is concluded that feed speed, rotating speed and milk roller size are important factors affecting the return width and fishtail shape of preform. Through orthogonal experimental analysis, it is found that feed speed, rotating speed, drive roll radius and core roll radius have effects on the spread coefficient and fish tail shape.

(3) The neural network model of ring rolling process parameters, ring spread and fishtail phenomenon is established, and the prediction of ring spread and fishtail phenomenon in the rolling process of preform blank is realized. The average error of the prediction model is less than 6%. Taking the minimum spread coefficient and fishtail coefficient as the comprehensive optimization objective, a multi-objective optimization design model for the rolling process parameters of preform blank is established. The ant colony algorithm is used to optimize in the solution space of the established neural network model, and the optimal parameter combination is obtained: when the feed speed is 15mm / s When the rotating speed is 16.21rad/s, the radius of driving roller is 500mm and the radius of core roller is 36.75mm, the comprehensive goal of small ring spread coefficient and moderate fishtail coefficient can be achieved, and the good forming of preform can be realized.

(4) The final forging process of spiral bevel gear is numerically simulated and analyzed. It is found that the material gradually fills the cavity with the downward movement of the die. Firstly, the large end tooth root of the tooth shape is formed, and finally the small end tooth top of the tooth shape is formed. After the cavity is filled, the excess material flows to the flash. The equivalent stress and equivalent effect variable at the tooth root are large, and the maximum forming load is about 76900kn. The tooth shape of the formed forging is fully filled and the flash is uniform, which shows that the forming process is reasonable.

(5) The die stress and elastic deformation of die tooth profile in forging process are simulated and analyzed. The die stress value is distributed between 1500-2000mpa, and the equivalent stress distributed at the tooth root of the big end of die tooth is large. The displacement at different positions on the die tooth profile changes regularly along the helix and height direction. This change leads to the deformation that the height of the tooth profile decreases, the helix angle at the midpoint decreases, and the shape of the tooth profile narrows. The uneven elastic deformation on the die tooth profile will reduce the height of the forging tooth profile, the helix angle at the midpoint and the shape of the tooth profile section, which will affect the forming accuracy. It is proposed that in the process of die design, the die tooth profile is corrected by using the inverse compensation method to reduce the error of die elastic deformation.

(6) A shape and size scheme of preformed blank is proposed. Fillets R1 and R2 are designed on the upper end face of the blank, and a platform with length L and angle is designed on the lower end face α The conical surface of R1, R2, l α As design variables, the maximum value of equivalent stress and forming load is taken as the optimization objective, and the response surface model is established. The optimal solution after iterative optimization is: R1 = 10.40mm, R2 = 15.60mm, l = 19.50mm, α＝ At 16.80 °, the equivalent stress and forming load after optimization are reduced by 50.9% and 40.4% respectively.