摘要
以0.05～0.33C(wt%,下同)、0.004～0.099V的3种微合金钢分别在1000和1050℃、0.01～10s-1应变速率下以Gleeble-1500热/力模拟实验应力-应变数据为样本,构建了C、V含量不同的微合金钢成分对动态再结晶峰值应变εp影响的误差反向传播(BP)人工神经网络模型,利用建立的BP模型研究了在不同应变速率下C、V含量对εp的影响规律。研究结果表明,C、V对含钒微合金钢动态再结晶峰值应变的影响与应变速率相关,高应变速率和低应变速率下元素的影响规律不同。
The models of errorback propagation（BP） neural network,which was the relationship between the content（C,V） and peak strain（εp）,were established on the basis of the data in Gleeble-1500 thermo-mechanical simulated experiment on the condition of isothermal compression at 1000℃,1050℃ and strain rate range of 0.01～10s-1 for3 micro-alloyed steel with 0.05wt%C～0.33wt%C,0.004wt%～0.099wt%V.The effects of C,V content on εp of vanadium micro-alloyed steel were researched using BP neural network models.The results show that the influences of C,V content on εp are related to strain rate.The influence law at high strain rate is different from that at low strain rate.
出处
《热加工工艺》
CSCD
北大核心
2010年第18期47-49,142,共4页
Hot Working Technology
关键词
变形奥氏体
峰值应变
神经网络
deformed austenite
peak strain
artificial neural network
