Optimization Design Of Processing Parameters For Electric Upsetting Process Of Heat-resistant Alloy To Attain The Homogenized And Fine Grain

This paper mainly studied the optimization design of processing parameters for electric upsetting process of heat-resistant alloy to attain the homogenized and fine grain, providing a theoretical basis for practical electric upsetting production to reduce micro-defects. Firstly, a BP neural network model of constitutive relations was developed based on experimental true strain-stress data from the isothermal compression tests of 3Cr20Ni10W2 austenitic heat-resistant alloy. A macro and micro multi-scale finite element analysis platform for electric upsetting of 3Cr20Ni10W2 austenitic heat-resistant alloy was constructed on the basis of a combination of electrical-thermal-mechanical coupling finite element analysis method and mathematical model of dynamic recrystallization and grain growth. According to the finite element analysis results, the response surface methodology was used to establish the dynamic response relationship between processing parameters of electric upsetting and quantitative indicators of grain size and grain distribution, and the genetic algorithm was used for the optimization design of processing parameters for electric upsetting process to attain the homogenized and fine grain. The research results have important theoretical significance and practical application value for the setting and optimization of processing parameters for electric upsetting. The main contents and conclusions of this paper are as follows:(1) The isothermal compression tests of 3Cr20Ni10W2 austenitic heat-resistant alloy were conducted at temperature range of 930 ?~1130 ?, strain rate range of 0.01 s-1~10 s-1, and a BP neural network model of constitutive relations was developed based on the true strain-stress data extracted from the isothermal compression tests.(2) Based on the electrical-thermal-mechanical coupling analysis platform in FEM software MSC.Marc and the mathematical model of dynamic recrystallization and grain growth for 3Cr20Ni10W2 austenitic heat-resistant alloy, a macro and micro multi-scale finite element analysis platform for quantitative analysis of electric upsetting was constructed. The macroscopic temperature distribution and deformation state, and microscopic grain size distribution in electric upsetting process was analyzed qualitatively and quantitatively on the constructed platform.(3) The finite element simulation scheme was designed by using the four factor Box-Behnken experiment design method, and the dynamic response relationships between processing parameters of electric upsetting(heating current, clamping length, upsetting pressure and velocity of the anvil cylinder) and quantitative indicators of grain size and grain distribution was constructed by means of response surface methodology on the basis of the finite element analysis results. Taking the two constructed response surface models as objective functions, the optimization design of processing parameters for electric upsetting process was carried out by multi-objective genetic algorithm. A set of processing parameters for electric upsetting to attain the homogenized and fine grain was carried out according to the optimization results.(4) A trial-manufacture experiment of electric upsetting was conducted based on the optimized processing parameters. According to the metallographic analysis results, the simulation results match well with the experiment results, demonstrating that the macro and micro multi-scale finite element analysis method for quantitative analysis of electric upsetting and the optimized processing parameters are reliable.