热压缩Ti-4.5Al-3Mo-1V合金的流变应力行为  

Flow stress behavior of Ti-4.5Al-3Mo-1V alloy during hot compression deformation

作　　者：宗影影[1] 单德彬[1] 吕炎[1] 

机构地区：[1]哈尔滨工业大学材料科学与工程学院,黑龙江哈尔滨150001

出　　处：《锻压技术》2005年第3期50-52,55,共4页Forging & Stamping Technology

摘　　要：采用Gleeble-1500热模拟机对Ti-4.5Al-3Mo-1V合金在α+β相区进行了等温热压缩实验,根据摩擦修正后的流变应力曲线,研究了此合金在α+β相区恒温压缩时的动态软化规律,分析了热变形参数对该合金流变应力的影响,并采用BP人工神经网络的方法建立了该合金高温变形抗力与应变、应变速率和温度对应关系的预测模型。结果表明:合金的流变应力曲线在低应变速率下达到极值后逐渐软化,在高应变速率下,出现极值后连续振动,然后再逐渐软化的现象;软化的主要机制为动态再结晶;流变应力随温度的升高和应变速率的减小而急剧降低;神经网络方法能够较精确地预测材料的流变应力。Samples of Ti-4.5Al-3Mo-1V alloy are compressed in α+β phase field on a Gleeble-1500 Simulator. The dynamic softening model in the upsetting process is investigated based on the corrected stress-strain curves. The influences of hot working parameters on the flow stress are systematically studied. A predicting model has been developed by a BP neural network method for the relation between flow stress and deformation strain, strain rate and temperature of Ti-4.5Al-3Mo-1V alloy. The experimental results show that flow curves of the alloy are characterized by a linear increase until regular oscillations at high strain rates or single peak at low strain rates after which flow softening is observed. The softening mechanism is dynamic recrystallization. And the flow stress decreased with increasing temperature and decreasing strain rate. Moreover, the trained network is able to predict the flow stress very accurately.

关 键 词：合金 热压缩 BP人工神经网络 流变应力曲线 神经网络方法 热变形参数 低应变速率 高应变速率 动态再结晶 压缩实验 热模拟机 软化规律 变形抗力 预测模型 对应关系 连续振动 相区 温度 极值 

分 类 号：TG115[金属学及工艺—物理冶金]