Titanium alloy constitutive relation prediction method based on machine learning

Abstract

The invention relates to a titanium alloy constitutive relation prediction method based on machine learning, and belongs to the technical field of constitutive behavior prediction of metal materials.The prediction method comprises the following steps: obtaining and preprocessing stress-strain curves of various titanium alloys under different temperature and strain rate conditions; making a curvedata set which is independently used for training the a VAE-GAN model; constructing a prediction model part I based on the VAE-GAN model, and performing training; building a prediction model part II based on a polynomial regression model, and predicating coding of a stress-strain curve according to experimental conditions; and inputting the prediction codes into a VAE-GAN decoder, and outputting afinal prediction stress-strain curve. According to the prediction method, the change process and the failure strain of the stress of the titanium alloy material along with the strain are predicted atthe same time, the defect that a traditional constitutive model cannot predict the failure strain of the alloy material is overcome, and a new method is provided for alloy material constitutive relation prediction.

Description

technical field [0001] The invention relates to the application of machine learning technology in the field of material science, in particular to a method for predicting the constitutive relationship of various titanium alloys by using a neural network encoder model and a polynomial regression model, belonging to the technical field of constitutive behavior prediction of metal materials . Background technique [0002] Traditional empirical metal constitutive models, such as the Johnson-Cook model, can describe the constitutive behavior of metal materials at different strain rates and temperatures. However, due to the complex nonlinear relationship between the deformation of metal materials and the temperature and strain rate, the empirical model can only accurately fit the stress-strain curve within a limited range. Since the end of the last century, some scholars have tried to use the neural network (ANN) model to predict the constitutive behavior of materials. Since the n...

AI Technical Summary

Problems solved by technology

However, the data generated by the VAE model is often fuzzy and cannot meet the accuracy requirements of this study for data fitting.

However, there is no report on the use of VAE-GAN model to predict the constitutive of metal materials

Images