B10铜合金高温流变行为及BP神经网络本构模型  

High temperature flow stress behavior of B10 copper alloy and BP neural network constitutive model

作　　者：韩茂盛[1] 陶欢[1] 

机构地区：[1]中国船舶重工集团公司第七二五研究所,河南洛阳471000

出　　处：《锻压装备与制造技术》2016年第6期112-115,70,共5页China Metalforming Equipment & Manufacturing Technology

摘　　要：利用Gleeble-1500D热模拟机对B10铜合金进行热压缩实验，研究了该合金在高温塑性变形过程中的流变应力行为。实验温度为800～950℃，应变速率为0．1～15s-1。研究结果表明，B10铜合金的流变应力随着变形温度的增加而减小，随着应变速率的增大而增大。基于BP神经网络建立了该合金的本构关系模型，预测值与实验值对比表明BP神经网络具有很高的预测精度，所建立的本构模型平均相对误差在1％以内。该模型能够客观真实地描述B10铜合金的高温塑性变形行为，为该合金热变形分析提供基础。The hot compression test has been conducted to B10 copper alloy by use of Gleeble-1500D simula- tor under temperature range from 800℃ to 950℃ andstrain rate range from 0.1s-1 to 15s-l. The flow stress behavior of this alloyduring plastic deformation at elevated temperature has been studied. The experimental re- sults show that the flow stress of B10 copper alloy decreases with the increase of deformation temperature but enhances with the increase of the stain rate. The constitutive relationship model has been built to this alloy on the basis of BP neural network. By comparison between the predicted value with the experiment value, it is shown that BP neural network has high prediction accuracy. The average relative error of the constitutive model is less than 1%. This model can objectively and truly describe the plastic deformation behavior of the B10 copper alloy at high temperature. It provides basis for hot deformation analysisof this alloy.

关 键 词：B10铜合金 流变行为 本构关系 BP神经网络 

分 类 号：TG146.1[一般工业技术—材料科学与工程]