摘要
为研究表面超声滚压(SURP)处理对EA4T车轴钢疲劳性能的影响,首先,采用SURP技术对EA4T车轴钢试样进行表面处理,并对处理后的试样进行表面性能测试,分析表面三维形貌、粗糙度、硬度、残余应力、半高宽(FWHM)和晶粒尺寸的变化;然后,采用旋转弯曲疲劳试验机对EA4T车轴钢试样进行疲劳试验,获得应力-疲劳寿命(S-N)曲线,并分析裂纹扩展规律,研究SURP处理对EA4T车轴钢疲劳性能和裂纹扩展行为的影响;最后,采用BP(back propagation)神经网络建立了以加载应力幅值、表面粗糙度、表面半高宽、表面硬度、硬化层深度、表面残余应力和残余应力层深度为输入的超声滚压EA4T车轴钢疲劳寿命预测模型,并对超声滚压EA4T车轴钢试样进行寿命预测.研究结果表明:SURP处理可以使试样表面粗糙度降低为0.17μm,并去除表面梨沟形貌;试样表面硬度提升至420 HV,试样表面引入约−500 MPa的残余应力以及约550μm深的残余应力层;研磨试样和研磨抛光试样以及SURP处理试样均具有传统疲劳极限,研磨试样和研磨抛光试样的疲劳性能基本一致,且疲劳极限均为355 MPa,SURP处理试样疲劳性能显著提升,其疲劳极限为455 MPa,相比研磨试样提升了28%;疲劳断口观察表明,所有试样的疲劳裂纹均萌生自表面,SURP处理没有改变试样的疲劳破坏机制;SURP处理使试样的裂纹扩展门槛值从6.29 MPa·m^(1/2)增加到11.21 MPa·m^(1/2),同时减缓了裂纹萌生以及短裂纹扩展,从而显著提高了EA4T车轴钢疲劳性能;超声滚压EA4T车轴钢疲劳寿命预测模型预测精度为88.5%.
In order to study the influence of surface ultrasonic rolling processing(SURP)on the fatigue properties of EA4T axle steel,the EA4T axle steel specimens were firstly treated by SURP technology,and the surface properties of the treated specimens were analyzed.The 3D surface morphologies,roughness,hardness,residual stress,full width at half maximum(FWHM),and crystal size were investigated.Then,the fatigue tests were carried out on the EA4T axle steel specimens using a rotary bending fatigue testing machine.The stress–fatigue life(S-N)curves and crack propagation behaviors were obtained,and the effect of SURP on fatigue properties and crack propagation behaviors of EA4T axle steels was analyzed.Finally,a back propagation(BP)neural network was used to establish the fatigue life prediction model of SURP EA4T axle steel,which took load stress amplitude,surface roughness,surface FWHM,surface hardness,hardened layer depth,surface residual stress,and depth of residual stress layer as input.In addition,the life of SURP EA4T axle steel specimens was predicted.The results show that SURP can reduce the surface roughness of the specimens to 0.17μm and remove the groove morphology.Meanwhile,the surface hardness of the specimens is improved to 420 HV,and a surface residual stress of about−500 MPa and a residual stress layer with a depth of about 550μm are introduced.There are conventional fatigue limits for grinding specimens,grinding with polishing specimens,as well as SURP specimens.The fatigue properties of grinding specimens and grinding with polishing specimens are basically the same,with a fatigue limit of 355 MPa.The fatigue property of SURP specimens is improved significantly,with a fatigue limit of 455 MPa,an increase of 28%compared with the grinding specimens.Fatigue fracture observations show that the fatigue cracks of all specimens initiate from the surface,and SURP cannot change the fatigue damage mechanism of the specimen.SURP can increase the threshold value of crack propagation of the specimen from 6.29 MPa·m^(1/2) to 11.21 MPa·m^(1/2),and it can slow down the crack initiation and propagation of short cracks,thus significantly improving the fatigue property of EA4T axle steel.The fatigue life prediction model of SURP EA4T axle steel has a prediction accuracy of 88.5%.
作者
张继旺
张浩楠
杨冰
苏凯新
李行
ZHANG Jiwang;ZHANG Haonan;YANG Bing;SU Kaixin;LI Hang(State Key Laboratory of Rail Transit Vehicle System,Southwest Jiaotong University,Chengdu 610036,China;Aerospace Science and Industry Inertial Technology,Beijing 100070,China)
出处
《西南交通大学学报》
EI
CSCD
北大核心
2024年第6期1305-1313,共9页
Journal of Southwest Jiaotong University
基金
国家自然科学基金项目(51675445)。
关键词
EA4T车轴钢
超声滚压
疲劳性能
裂纹扩展门槛值
BP神经网络
EA4T axle steel
ultrasonic rolling
fatigue property
threshold value of crack propagation
BP neural network
作者简介
第一作者:张继旺(1983-),男,研究员,博士,研究方向为材料和结构的疲劳与断裂,E-mail:zhangjiwang@swjtu.edu.cn。
