【作者】 李建伟；

【导师】 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】 金属磁记忆方法是针对铁磁材料的一种新兴无损检测技术，该方法利用构件的表面磁场分布来评估材料的应力（应变）状态。应力集中将会引起表面磁场的畸变。同其他磁性检测技术相比，该方法不但可以用来准确确定应力集中的部位，而且不需要昂贵的磁化设备，不需要表面处理，检测速度快，设备轻便等优点，为无损评估技术在构件应力状态的评估方面提供了新的方向，因此具有广阔的应用潜力和研究意义。然而，磁记忆基础理论研究的缺乏已严重制约了磁记忆技术的发展。实质上，金属磁记忆研究的是地磁场下，机械应力或应变引起的磁性质的变化。该方法的本质是弱磁场下的磁机械效应。而对磁机械效应的研究已有100多年的历史，形成了比较成熟的理论体系。磁机械效应是机械参量引起的磁化强度的变化。本论文针对磁记忆现象机理的模棱两可，结合磁记忆检测的实验结果，基于磁机械效应的部分成果，建立了适合于材料不同的应力状态下的磁化模型。具体来说，本文在以下几个方面进行了探索和研究：（1）为了将拉压磁非对称性并入到J-A模型，弥补基本J-A模型在刻画拉压磁性非对称性上的不足，分别考虑了应力退磁项、应力相关的钉扎系数、磁畴耦合系数以及应力依赖的饱和磁致伸缩系数等因素引起的拉压磁性非对称性。并入这些非对称因素后，改进的J-A模型不仅能描述磁性检测中的反转现象和接近现象，更是增加了拉压应力下的磁性非对称性，改善了模型模拟结果的精度，特别是压应力下的磁感模拟结果；（2）塑性变形引起磁化变化的机制主要有三点，一就是塑性变形引起了位错的大量增殖，二就是塑性变形引起的磁塑性能，三就是塑性变形引起了残余应力。为了描述磁化随塑性变形的变化，将上述三种机制转化成塑性变形对有效场和模型参数的影响，其中在有效场方面引入了残余应力场和磁塑性场，在模型参数方面主要考虑了塑性变形对钉扎参数、磁畴耦合系数以及规划系数的作用。模拟的结果显示在塑性变形初期，磁化强度出现了“跳水”式的下降，随后随着塑性变形缓慢的减少，这与实验结果非常的吻合；（3）循环应力的磁化不仅要考虑磁化随应力历程的变化，还应该考虑疲劳微观损伤对磁化的作用。疲劳过程的各阶段引起磁化变化的机制是不同的：疲劳的第I阶段，是一个循环软化或硬化的阶段，在模拟这一阶段的磁化调整的过程中，考虑了软化或硬化对钉扎系数和磁畴耦合系数两个模型参量的影响；疲劳的第II阶段，是疲劳的稳定阶段，这一阶段的磁记忆参量也处于稳定的水平，从每循环周次的损伤-热和缺陷-铁原子比的角度讨论了这一阶段的不可测度性；疲劳的第III阶段，材料出现了宏观的裂纹，磁记忆参量也出现了较大的变化，这一阶段主要考虑了宏观裂纹位置处形成的强退磁场对磁化的作用；（4）断裂引起磁化出现了正负峰的突变，而且突变峰值的量级是其他变形阶段无法比拟的。借助磁偶极子模型模拟了这种突变。基于硬度-位错-磁偶极子分布一致的研究结果上，分别建立了拉伸断裂和拉拉疲劳断裂的位错-磁偶极子模型：拉伸断裂的磁偶极子分布是从远处至断口的线性增加，而拉拉疲劳断裂的磁偶极子分布则只是在断口有密集的存在。建立的模型不仅能很好的模拟拉伸断裂与拉拉疲劳断裂的正负峰值突变，而且很好的解释了拉拉疲劳断裂突变峰值一般大于拉伸断裂峰值的实验现象。更多还原

【Abstract】 Metal magnetic memory (MMM) method is a novel nondestructive testing(NDT) technology for ferromagnetic material. This method evaluates the stress (orstrain) state of component by testing its distribution of surface magnetic field. Thestress concentration induces the abnormality of the surface magnetic field.Compared to routine magnetic testing methods, the MMM method can not onlylocate the stress concentration accurately, but also this method has followingadvantages: unwanted expensive magnetization devices, unnecessary to clean testedsurface, easy in operation, portable equipment and etc. This method provides a newdirection for the non-destructive evaluation techniques in estimating the stress stateof the component, and so it has promising future. The shortage in the theoreticalresearch of MMM method has become the bottleneck hindering its application.In essence, the MMM can be defined as the change in magnetic parameters of amagnetic material resulting from a change in applied stress (or strain) undergeomagnetic field. The nature of MMM is magnetomechanical effect under weakmagnetic field. The research of magnetomechanical effect can be traced back to over100years ago, and the study on it has formed a relatively mature theory system. Themagnetomechanical effect is the change of magnetization of a magnetic materialresulting from a change in applied stress. In this dissertation, aiming at ambiguousmechanism of the MMM phenomenon, combined with the experimental results ofthe MMM test and based on partial results of the magnetomechanical effect, themagnetization models for different stress states are established, respectively. To bemore specific, the main works are as follows:(1)To incorporate the asymmetry in magnetic property behavior under tensileand compressive stress in the J-A model and make up defect of basic J-A model indescribing asymmetry, the asymmetry properties caused by stress demagnetizationterm, a variable pinning coefficient, stress-dependence domain coupling coefficientand stress-dependence saturation magnetostriction are incorporated in the basic J-Amodel. It is found that the modified model can not only present the phenomena offield reverse and magnetization approach, but also provides a much betterdescription of asymmetrical magnetic property under tension and compression.Moreover, the accuracy of calculations is improved considerably, particularly in thecompressive stress situation;(2) The change of magnetization caused by plastic deformation can beconcluded as following three mechanisms: first, the plastic deformation induces mass multiplication of dislocation; second, the plastic deformation causesmagnetoplastic energy; third, the plastic deformation gives rise to the residual stress.In order to describe the change of magnetization with the plastic deformation, theabove three mechanisms are equivalent to the responses of the effective field andmodel parameters to the plastic deformation, where the effective field incorporatesthe contributions caused by residual stress and magnetoplastic energy, and the modelparameters prime consider the magnetoplastic effect on the pinning coefficient, theinterdomain coupling coefficient and the scaling constant. The computedmagnetization exhibits sharp change in the preliminary stage of plastic deformation,and then decreases slowly with the increase of plastic strain, in agreement withexperimental results;(3) Magnetization simulation caused by cyclic stress not only should considercourse of magnetization change with stress, but also take into account the influenceof fatigue damage on magnetization. The magnetization mechanisms are differentfor different fatigue stages: In the stage I of fatigue, cyclic softening or hardeningoccur in material. In the process of simulating the magnetization accommodation,the effects of softening or hardening on the pinning coefficient and the interdomaincoupling coefficient are considered; in the stage II of fatigue, it is a stable stage offatigue and the parameter of MMM testing is also stable. From the views ofdamage-to-heat and defects-to-iron atoms per cycle, the immeasurability isdiscussed; In the stage III of fatigue, macroscopic cracks appear in material and theparameter of MMM testing changes greatly. The leakage magnetic field caused bymacroscopic cracks is considered;(4) Peak-to-peak saltation in magnetization occurs at the instant of fracture, andthe magnitude of saltation is without parallel. This saltation is simulated with thehelp of theory of magnetic dipole. Assume that the distribution ofhardness-dislocation-magnetic charge is consistent, based on the test result ofhardness, the models of dislocation-magnetic dipole for tensile fracture andtensile-tensile fatigue fracture are established, respectively: The distribution ofmagnetic charge for tensile fracture is linear increase from distant to the fracturezone, while the distribution of magnetic charge for tensile-tensile fatigue fracture isconverging only at the fracture zone. It is found that the established model can notonly present the saltation of peak-to-peak in magnetization for tensile fracture andtensile-tensile fatigue fracture, but also provides an explanation to the experimentalphenomena that the peak value of saltation for tensile-tensile fatigue fracture ishigher that that for tensile fracture.更多还原