一个新的超塑变形内部损伤预测模型  

Damage Prediction During Superplastic Deformation Using Artificial Neural Networks

作　　者：刘马宝[1] 高宏 李淼泉[1] 吴诗[1] 

机构地区：[1]西北工业大学,西安710072

出　　处：《材料科学与工艺》1998年第2期79-82,共4页Materials Science and Technology

基　　金：国家自然科学基金!59471071

摘　　要：超塑变形往往具有空洞敏感性，对空洞的研究引起国内外学者的重视并取得较大进展，但现有描述起塑变形时空洞损伤行为的力学模型普遍存在精度问题。利用神经网络对超塑变形时的空洞损伤程度进行预测，不仅可提高精度，同时亦能充分反映超塑变形工艺参数对损伤的影响规律。因此，这就为研究超塑变形时的空洞损伤提供了一种新方法。Superplastic materials generally cavitate during superplastic deformation.Cavitation may lead to decreasing of forming limit and worsening of formed workpieces.We think description of such cavitation requires a suitable model.We now present such a model.Which is established by training the artificial neural network with the damage values obtained on half-hardened brass H62 under different processing conditions of superplastic deformation.It is found that the damage levels predicted by the trained artificial neural network agree closely with actual experimental results and that the model improves the predicting precision in comparison with the traditional model.

关 键 词：超塑变形 空洞 损伤 超塑性材料 预测模型 

分 类 号：TB301[一般工业技术—材料科学与工程] TG111.7[金属学及工艺—物理冶金]