基于极性交互模型的MoLa合金锻造工艺优化及热变形行为  

Hot Deformation Behavior and Forging Process Optimization of MoLa Alloy Based on Polar Reciprocity Model

作　　者：汪雨佳 王克鲁[1] 鲁世强[1] 许求喜 Wang Yujia;Wang Kelu;Lu Shiqiang;Xu Qiuxi(Nanchang Hangkong University,Nanchang 330063,Chin)

机构地区：[1]南昌航空大学,江西南昌330063

出　　处：《稀有金属材料与工程》2018年第7期2225-2231,共7页Rare Metal Materials and Engineering

基　　金：国家自然科学基金(51164030)

摘　　要：采用Geeble1500型热模拟试验,对MoLa合金进行等温恒应变速率压缩实验,研究其在变形温度800～1150℃,应变速率0.001～10s^(-1)范围内的热变形行为。通过对不同变形参数下的流变曲线分析发现,随应变的增加,流变曲线大多呈现为缓慢上升或保持稳定,但在应变速率为0.001 s^(-1)时,1000～1150℃变形温度下流变应力随应变的增加而下降;采用PSO-BP神经网络建立MoLa合金本构模型,经过误差计算得出,该模型的相关系数和平均相对误差分别为:0.995和1.48%,具有良好的精度;基于极性交互模型绘制MoLa合金本征热加工性能参数ξ图,并通过对失稳区和稳定区组织分析发现,失稳区主要以局部流动为失稳形式,稳定区主要以动态回复为变形机制;通过ξ图和组织观察可知,MoLa合金最佳的变形参数范围为:变形温度1100～1150℃、应变速率0.001～0.05 s^(-1)。To investigate the flow stress characteristics and deformation mechanism of MoLa alloy, the isothermal compression tests were performed at the deformation temperature of 800～1150 ℃ and the strain rate of 0.001～10 s ^（-1）. The results show that the flow stress increases slowly or remains stable with the increase of strain. But the flow stress decreases with the increase of strain at the deformation temperature of1000～1150 ℃ and the strain rate of 0.001 s^（-1）. A PSO-BP neural network of MoLa alloy was established to describe the changing rule of flow stress with the strain rate and deformation temperature. The correlation coefficient and average relative error of the network are 0.995 and 1.48%,respectively. So it is clear that the PSO-BP model has good accuracy. The intrinsic hot workability maps were constructed based on the polar reciprocity model at the strain of 0.3, 0.6, and 0.92. Combing microstructure observation, it is found that the buckling form in instability area mainly is local plastic deformation, and the main deformation mechanism in stability region is dynamic recovery. The optimal processing parameter ranges of MoLa alloy are the deformation temperature of 1100～1150 ℃ and the strain rate of 0.001～0.05 s^（-1）.

关 键 词：MoLa合金 本构模型 PSO-BP神经网络 极性交互模型 锻造工艺优化 

分 类 号：TG146.412[一般工业技术—材料科学与工程]