论文部分内容阅读
本构关系是联系塑性加工过程中材料的动态响应与热力参数之间的媒介，因此，本构关系是用有限元对金属塑性加工过程进行数值模拟的前提条件。由于变形高温合金本构关系的高度非线性，在试验数据基础上，按Ａｒｒｈｅｎｉｕｓ型方程用数理统计方法建立的本构关系，结构十分复杂，不便工程应用。本文探讨用人工神经网络这种新一代信息处理工具表征变形高温合金本构关系的方法和效果。研究结果表明，应用人工神经网络表征的本构关系比用数理统计方法建立的能较宏观地反映材料在塑性加工过程中的动态行为，是一种有广泛应用前景的表征工程材料本构关系的方法。
The constitutive relation is the medium between the dynamic response and the thermal parameter of the material in the process of plastic working. Therefore, the constitutive relation is a prerequisite for the numerical simulation of the metal plastic working process by the finite element method. Due to the high degree of nonlinearity of the constitutive relation of the deformed superalloy, based on the experimental data, the constitutive relation established by mathematical Arrhenius equation is very complex and inconvenient for engineering application. This paper discusses the method and effect of using artificial neural network, a new generation of information processing tools, to characterize the constitutive relationship of deformed superalloy. The results show that the constitutive relation represented by artificial neural network can reflect the dynamic behavior of material in the plastic working process more macroscopically than that established by mathematical statistics. It is a promising method to characterize the constitutive relation of engineering materials method.