“五害元素”对P20塑料模具钢组织及力学性能的影响  

Influence of five kinds of impurity elements "H,O,N,P,S" on microstructure and mechanical properties of P20 plastic mould steel

作　　者：左秀荣[1,2] 陈蕴博[2] 王淼辉[2] 李勇[1] 翁永刚[1] 

机构地区：[1]郑州大学物理工程学院材料物理教育部重点实验室,河南郑州450052 [2]机械科学研究总院先进制造技术研究中心&先进成形与装备国家重点实验室,北京100083

出　　处：《金属热处理》2012年第11期32-36,共5页Heat Treatment of Metals

基　　金：国家高技术研究发展计划("863计划")(2007AA03Z511)

摘　　要：用人工神经网络及材料微观分析方法研究了"五害元素"H,O,N,P,S对P20钢组织及性能的影响。GRNN人工神经网络能根据化学成分精确预测P20钢的力学性能,同时能用于研究"五害元素"对力学性能的影响规律。预测结果表明:"五害元素"对断面收缩率和伸长率均有影响,而对抗拉强度及屈服强度的影响不大。减少"五害元素"含量,从而减少夹杂物的含量及减轻杂质元素在晶界的偏聚,增大了裂纹形核和扩展阻力,可使P20钢得到较高的断裂韧性。本研究提供了一种研究"五害元素"与力学性能关系的较好方法。Effects of five kinds of impurity elements ＂H,O,N,P,S＂ on microstructure and mechanical properties of P20 steels were studied using generalised regression neural network（GRNN） and microstructural analysis methods.It is found that the GRNN can predict the mechanical properties of P20 steels accurately according to chemical compositions.The trained network was then applied to research the relation between the mechanical properties and the contents of H,O,N,P and S.According to the results of model,H,O,N,P and S have influence on percentage reduction in area and elongation percentage,but have less influence on tensile strength and yield strength.Reducing the contents of H,O,N,P and S,which reduces the inclusions and impurity elements segregated in grain boundary,enhances the resistance of crack nucleation and propagation and improves the fracture toughness.The results have sufficiently mined the basic domain knowledge of relationship of mechanical properties vs contents of H,O,N,P and S of P20 steel.Therefore,it provides a better way to research the effects of five kinds of impurity elements ＂H,O,N,P,S＂ on microstructure and mechanical properties of P20 steels.

关 键 词：人工神经网络 P20钢 H O N P S 组织与性能 

分 类 号：TG156[金属学及工艺—热处理]