Ti-6.5Al-3.5Mo-1.5Zr-0.3Si合金本构关系的BP神经网络模型  

A Constitutive Relation Model for the Ti-6.5Al-3.5Mo-1.5Zr-0.3Si Alloy Based on BP Neural Network

作　　者：王克鲁[1] 鲁世强[1] 李鑫[1] 董显娟[1] 杨化娟[1] 

机构地区：[1]南昌航空大学材料科学与工程学院

出　　处：《特种铸造及有色合金》2008年第8期575-578,共4页Special Casting & Nonferrous Alloys

基　　金：江西省自然科学基金资助项目(2007GQC1575);江西省材料科学与工程中心基金资助项目(ZX200601002);江西省教育厅科技项目(GJJ08203)

摘　　要：利用THERMEcMASTOR-Z型热力模拟试验机，在变形温度为7801080℃，应变速率为0．001～70．0S。条件下对Ti6．5A1-3．5Mo-1．5Zr-0．3Si合金进行等温恒应变速率压缩试验，获得不同变形温度、不同应变速率和不同真应变下的流动应力数据。结合试验数据和神经网络知识，构建了采用BP算法的人工神经网络，训练结束后的神经网络即成为Ti-6．5A3．5Mo-1．5Zr-0．3Si合金的一个知识基的本构关系模型。利用所建立的BP网络模型对材料的流动应力进行了预测，发现预测值与试验数据吻合良好，说明该BP网络本构关系模型具有较高的精度．可用于指导Ti-6．5Al-3．5Mo-1．5Zr-0．3Si合金热加工工艺的制定。The hot deformation behavior of Ti-6.5Al-3.5Mo-1.5Zr-0.3Si alloy was investigated with THERMECMASTOR-Z tester at strain rates of 0.001 s-1~70 s-1 at temperature of 780~1 080 ℃, obtaining the flow stress data at varied temperatures and varied strain rates as well as different true strain. Based on experimental data and neural network knowledge, a BP neural network with error-back propagation algorithm was established, which forms a knowledge-based constitutive relation model after training. The flow stress in the alloys was predicted by the BP neural network model. The results show that the predicted results are well in agreement with the experimental ones,confirming the availability of the model.

关 键 词：Ti-6.5Al-3.5Mo-1.5Zr-0.3Si合金 BP神经网络 本构关系 流动应力 

分 类 号：TG146.23[一般工业技术—材料科学与工程]