摘要
利用Gleeble1500热模拟试验机对TB8合金进行等温压缩试验,获得不同变形条件下的流变应力数据,在对数据进行摩擦修正的基础上建立了3×10×1的3层BP神经网络形式的本构关系模型。结果表明,在隐层神经元数为10、学习率为0.05、动量因子为0.4时,网络模型具有优良的性能,能精确反映热变形条件下温度、变形速率、变形程度与流变应力之间的关系,为TB8合金热加工工艺的合理制定和热变形过程的数值模拟提供依据。
Isothermal compression tests on TB8 alloy were conducted using Gleeble 1500 thermomechanical test system and the flow stress values at deferent deforming condition were obtained.Based on the experiment data corrected for friction,the constitutive relationship model was constructed in the shape of 3×10×1BP artificial neural network.The results show that the artificial neural network models have excellent capability under the conditions that the number of nodes in the hidden layer is 10,the learning rate is 0.05 and the momentum factor is 0.4.It can describe the influences of the temperature,strain rate and true strain on the flow stress accurately,and also provide basis for determining hot forming process and numerical simulation.
出处
《锻压技术》
CAS
CSCD
北大核心
2010年第1期130-133,共4页
Forging & Stamping Technology
关键词
TB8合金
本构关系
BP神经网络
TB8 alloy constitutive relationship BP artificial neural network
