摘要
利用Zwick/Roell Z100材料试验机,对TC4钛合金进行等温恒应变速率下的单向拉伸试验。基于获得的试验数据,采用BP神经网络技术建立了该合金的高温本构关系模型,并对其预测性能进行分析。基于ABAQUS/Explcit平台进行材料子程序二次开发,将神经网络本构模型嵌入到有限元计算中,实现了TC4钛合金高温变形的数值模拟。结果表明,神经网络本构模型预测精度很高,可以准确地描述TC4钛合金在热态下的动态力学性能。神经网络本构模型应用于有限元模拟可行且有效。
Isothermal constant-strain-rate uniaxial tensile tests were conducted for Ti-6Al-4V alloy on Zwick/Roell Z100 dynamic materials testing system.According to the obtained experimental data,a high temperature constitutive model of the alloy was proposed with BP neural network.Based on ABAQUS/Explcit,the artificial neural network constitutive model was joined into finite element calculation through secondary development of user-defined material subroutine.It realized the numerical simulation of the deforming process for the material.The results indicated that the artificial neural network constitutive model had high prediction precision,and could accurately describe the dynamic mechanical property of Ti-6Al-4V alloy at elevated temperature.The FEM case proved that the application of the constitutive model established by artificial neural network to FEM simulation was effective.
出处
《塑性工程学报》
CAS
CSCD
北大核心
2013年第1期89-94,共6页
Journal of Plasticity Engineering
基金
中国商用飞机有限责任公司"大型客机制造类关键技术攻关"资助项目
