【作者】 冷建成；

【导师】 张嘉钟； 徐敏强；

【作者基本信息】 哈尔滨工业大学 ， 一般力学与力学基础， 2012， 博士

【摘要】 无损检测技术对于保障设备可靠运行及人身安全具有重要意义。随着现代设备正日益向高载、高速、高温、高压的方向发展，为了防止突发性事故的发生，早期发现引起机械结构和设备失效的各种微观缺陷和局部应力集中就显得尤为重要，而这些在宏观裂纹或缺陷产生之前的隐性损伤对于传统的无损检测方法来说却无能为力。新兴的金属磁记忆技术，被认为在早期损伤检测方面极富开发潜力，但目前对铁磁构件早期损伤的磁记忆检测机理和方法还缺乏相应的系统研究。本文通过对金属磁记忆检测技术的研究，找到适用于铁磁性材料初期塑性变形及早期疲劳损伤诊断的方法，以达到尽早发现构件损伤、提前预防失效的目的。首先探讨了磁记忆技术用于应力状态和疲劳损伤检测的可行性，进而围绕铁磁性材料早期损伤检测这一核心，从静载和疲劳拉伸试验研究、塑性范围内的磁机械效应模型构建、面向早期疲劳损伤的磁场畸变建模等几个方面，较系统地研究了铁磁构件早期损伤的磁记忆检测问题。具体来说，本论文主要在以下几个方面展开了探索和研究：基于有效场理论和接近原理，详细推导了单向应力作用下铁磁性材料的磁机械效应模型，数值模拟了地磁场下磁化强度随应力的变化关系，表明应力致磁场的变化是一个由初始磁状态不断向非滞后磁化强度接近的过程。通过旋转弯曲疲劳试验，跟踪了磁信号在旋转一周不同位置的变化，发现与受检对象的实际应力-变形状态一致；同时研究了磁记忆信号随循环次数的变化特征，表明磁曲线与疲劳损伤之间具有相关性。在对磁记忆技术用于早期损伤可行性研究的基础上，通过对未退磁平板试件和退磁平板试件的静载拉伸试验，研究了加载过程中磁记忆信号的演变规律，分别提出了识别弹塑性不同变形阶段的磁信号特征；结合接近原理，分析了不同初始剩磁状态对应力致磁场变化的影响及原因，可为以后磁记忆检测的标准制定提供参考。通过拉-拉疲劳试验，研究了磁记忆信号随循环周次的变化规律，提出表征疲劳损伤的关键参量为应力集中区磁场梯度，基于建立的损伤变量模型计算结果与动态疲劳过程中四个阶段不同损伤程度的演化规律相一致。建立了适用于塑性变形阶段的改进的磁机械效应模型，得到磁化强度随应变之间的变化关系。针对目前磁机械效应模型仅在弹性范围内有效的局限性，从能量守恒的角度出发，推导出适用范围更广的磁化强度与有效场之间的关系模型。模型中考虑了参数钉扎系数和有效场度量系数在塑性范围内为位错密度的函数；塑性阶段加载时的有效场公式中明确区分了弹性变形和塑性变形所引起有效场分量的不同，而卸载后的有效场则要包括残余应力产生的附加磁场。通过对改进模型的数值模拟，结果表明在较小塑性变形时磁化强度会发生急剧变化；基于光滑平板退磁试件的静拉伸试验，发现在线卸载时的磁记忆信号在屈服时产生突变。理论计算结果很好地吻合实验现象，表明提出的模型可以解决铁磁性材料的初期塑性变形检测问题。采用XH-500现场金相显微镜加DIG300专用数字摄像头并同步配合磁记忆检测装置建立了显微疲劳试验系统，应用磁记忆显微疲劳观测试验方法，对45#钢试件进行了拉-拉疲劳试验，在细观尺度上动态观察了疲劳短裂纹萌生和扩展的演化过程，通过维氏硬度测试得到显微硬度在整个疲劳过程中的变化规律。同时分析了V型缺口处表面磁记忆信号随循环周次的变化，研究了不同裂纹尺寸下对应的磁信号曲线分布规律，进而确定了早期疲劳损伤的磁记忆信号特征。在此基础上，基于塑性区位错密度线性分布的假设，提出应用带磁偶极子模型等效应力集中区磁畴的定向排列，建立了裂纹尖端塑性区表面的漏磁场分布模型。分别对漏磁场切向分量和法向分量进行了数值模拟，提出了用于表征损伤程度和范围的两个关键参量分别为漏磁场切向分量的峰值Hxp和法向分量波峰-波谷之间的宽度△xHyp-p，实现了疲劳损伤的早期检测。更多还原

【Abstract】 Non-destructive testing techniques have very important influence onguaranteeing reliable equipment operation and life safety. With the increasingdevelopment of modern equipment towards high load, high speed, high temperatureand high pressure, it is vital significant to early discover various micro-defects andlocal stress concentration zones causing mechanical structure and equipment failureto avoid sudden accident, but these hidden damage occurring before macro-cracksor defects can not be detected by traditional non-destructive testing methods. Anewly arisen technique named metal magnetic memory is deemed to have potentialsin detecting early damage, whereas it is lack of systematic research on magneticmemory testing mechanism and method in early damage of ferromagneticcomponents.A method suitable to detect initial plastic deformation and early fatiguedamage of ferromagnetic materials is proposed via research on the metal magneticmemory testing technique, in order to discover component damage early and avoidfailure in advance. The feasibility of the magnetic memory technique applied todetect stress state and fatigue damage is first explored, and the early damage testingon ferromagnetic materials is kept as the central task. The main problems includingstatic and dynamic tensile tests, the magneto-mechanical effect modeling in theplastic region, and the magnetic field distortion modeling on the early fatiguedamage have been systematically investigated. To be specific, the main contents areexplored as follows.Based on the effective field theory and approaching principle, themagneto-mechanical effect model of ferromagnetic materials under the action ofunaxial stress state is derived in detail, and the variation of magnetization withstress in the earth’s magnetic field is numerically simulated, showing that the initialmagnetic state tends continuously towards the anhysteretic magnetization on theapplication of stress. The variation of the magnetic signal in one certain point atdifferent locations of the specimen rotated a period is recorded via the rotarybending fatigue experiment, which is consistent with its actual stress-strain state. Inthe meantime, the variation characteristics of the magnetic memory signal withcycle number are investigated, indicating that there is correlation between the magnetic curve and the fatigue damage.On the basis of the research on the feasibility of the magnetic memorytechnique applied to detect early damage, the evolution law of the magneticmemory signal on loading is studied via the static tensile experiments withundemagnetized and demagnetized plate specimens, and the magnetic signalcharacteristics to recognize elastic and plastic deformation stages are given.Combined with the approaching principle, the influence and reason of differentinitial residual magnetization on the magnetic field variation induced by stress areanalyzed, which helps to guide engineering application. The variation regularities ofthe magnetic memory signal with cycle number are investigated throughtensile-tensile fatigue experiments, and the magnetic field gradient in the stressconcentration zone as the key parameter to characterize fatigue damage is putforward. The result calculated by the established damage variable model coincideswith the evolution law of different damage degree in the four stages during thedynamic fatigue process.The modified magneto-mechanical effect model applicable to plasticdeformation stage is developed, and the variation relationship betweenmagnetization and strain is obtained. Considering the limit that the presentmagneto-mechanical effect models are only valid in the elastic range, the moreuniversal stress-magnetism coupling model correlating the magnetization and theeffective field is derived based on the law of conservation of energy. The piningcoefficient and the effective field scaling parameter are both regarded as thefunctions of dislocation density in the plastic range in the model, and the effectivefield components induced by elastic and plastic deformation are differentiateddefinitely in the effective field expression on loading, and the effective field afterunloading contains the additional magnetic field induced by residual stress. Thenumerical simulation results show that the magnetization by the rather small plasticdeformation decreased sharply, and the magnetic memory signal after unloadingonline also changed suddenly on yielding in the static tensile experiment withdemagnetized smooth plate specimens. The theoretical results have a goodagreement with the experimental phenomena, indicating that the proposed modelcan be used to detect the initial plastic deformation of ferromagnetic materials.The microscopic fatigue experimental system including XH-500metallograhpic microscope, DIG300special digital camera and magnetic memorytesting device, combined with the magnetic memory microscopic fatigue observation method is adopted, and the tensile-tensile fatigue experiment withspecimens made of steel45are conducted. The evolution process of fatigueshort-crack initiation and growth is observed dynamically from the microscopicscale, and the variation regularity of microscopic Vickers hardness during thefatigue procedure is also given. At the same time, the variation of the magneticmemory signal on the surface of the V-shaped notch with cycle number is analyzed,and the magnetic signal curve distribution regularity corresponding to differentcrack length is investigated. On this basis, assuming that the dislocation density inthe plastic zone is linear distribution, the leakage magnetic field distribution in thestress concentration zone is calculated in terms of the magnetic dipole model, andthe leakage magnetic field model on the surface of the plastic zone in the crack tipis set up. The tangential and normal components of the leakage magnetic field areboth simulated, and the peak Hxpof the tangential component and the width betweenabnormal wave crest and trough△xHyp-pof the normal component as two keyparameters to represent damage degree and range are proposed respectively,achieving early diagnosis on fatigue damage.更多还原