Comparison on constitutive relationship of superalloy GH4742 based on Arrhenius equation and machine learning
Based on the isothermal constant strain rate compression tests of superalloy GH4742 at strain rates of 0.001-1 s-1,deformation temperatures of 950-1160℃,and high reduction rate of 60%,its rheological stress behavior was analyzed,and Arrhenius equation,sup-port vector machine and GWO-BP network constitutive models of the alloy were constructed respectively.The results show that the rheolog-ical stress curve of superalloy GH4742 presents a significant softening phenomenon at high strain rates and low deformation temperatures.As the strain rate decreases and the deformation temperature increases,the rheological stress curve gradually shows steady-state flow char-acteristics.The correlation coefficients of the peak stress and strain compensation Arrhenius models are 0.993 and 0.991,respectively,and the average absolute relative errors are 8.986%and 9.813%,respectively.The correlation coefficient of the test sample support vec-tor machine model is 0.997,and the average absolute relative error is 5.626%.While the correlation coefficient of the test sample GWO-BP model is 0.997,and the average absolute relative error is 5.471%.Thus,support vector machine and GWO-BP models have higher prediction accuracy and can better describe the high-temperature rheological behavior of GH4742 superalloy.
NETL
NSTL
万方数据
