This paper investigates the high-temperature deformation behavior of thermoformed steel 38MnB5 with good comprehensive properties. The high-temperature deformation behavior of the steel is modeled by the JC model, ZA model and artificial neural network (ANN) model. The performations are evaluated by comparing the numerical simulation results to the experimental ones.
Isothermal unidirectional tensile tests were carried out in the temperature range of 25–600 °C. The plastic deformation of the material under different stress states was obtained by different specimens. Considering the effect of temperature on tensile properties, the JC model, ZA model and ANN model were used to calibrate the true stress–plastic strain curves of the material. Determination coefficient and average absolute relative error (AARE) were used to evaluate the accuracy of the models.
The determination coefficient of the ANN model is 0.99997, close to 1, and the AARE is 0.000007, close to 0, indicating that the prediction accuracy of the ANN model is far superior to the other two traditional models. To deal with the complex engineering stress states such as shear and plane strain, finite element simulation using a neural network model is carried out to predict the strength of hot-formed steel under uniaxial tensile, plane strain and shear stress. The experimental results are in good agreement with the simulated load–displacement curves, which further verify the accuracy of the ANN model.
The high-temperature behavior of thermoformed steel 38MnB5 is characterized and modeled by the JC model, ZA model and ANN model. The performations are evaluated by comparing the numerical simulation results to the experimental ones.
