Elsevier

Current Opinion in Solid State and Materials Science
固体与材料科学进展

Volume 19, Issue 5, October 2015, Pages 305-314
第 19 卷，第 5 期，2015 年 10 月，第 305-314 页
Current Opinion in Solid State and Materials Science

Mechanism of dislocation channel-induced irradiation assisted stress corrosion crack initiation in austenitic stainless steel
位错通道诱导奥氏体不锈钢辐照辅助应力腐蚀裂纹萌生机制

https://doi.org/10.1016/j.cossms.2015.04.001Get rights and content  获取权利和内容
Full text access  全文访问

Highlights  要点

  • •
    Three classes of dislocation channel–grain boundary intersections were observed.
    观察到三种类型的位错通道-晶界交点。
  • •
    Discontinuous channels undergo the least amount of displacement.
    不连续通道位移量最小。
  • •
    Discontinuous channels cause the greatest amount of stress at the grain boundary.
    不连续通道在晶界处产生最大应力。
  • •
    Discontinuous channels are the most likely to induce crack initiation.
    不连续通道最有可能引发裂纹萌生。
  • •
    IASCC is due to these high levels of stress at the channel–boundary intersection.
    辐照辅助应力腐蚀裂纹（IASCC）是由于通道-晶界交界面处的高应力引起的。

Abstract  摘要

The mechanism by which dislocation channeling induces irradiation assisted stress corrosion cracking was determined using Fe–13Cr15Ni austenitic stainless steel irradiated with protons to a dose of 5 dpa and strained at high temperature in both argon and simulated boiling water reactor normal water chemistry environments. Straining induced dislocation channels that were characterized by digital image correlation and confocal microscopy. Dislocation channels were found to be either continuous across the boundary, discontinuous, or discontinuous with slip in the boundary. Discontinuous channels were found to contain the least amount of strain but have the highest propensity for initiating cracks. Discontinuous dislocation channel–grain boundary intersections were shown to have the highest local stress. TEM in-situ straining of irradiated steels and atomistic simulation of dislocation–grain boundary interaction provided supporting evidence that channels that were unable to transfer strain underwent cracking. The inability of channels to relieve stress, by either slip in the adjacent grain or in the grain boundary, resulted in high local stresses and increased susceptibility to stress corrosion cracking initiation.
通过使用经过质子辐照至 5 dpa 剂量并分别在氩气和模拟沸水反应堆正常水化学环境中高温拉伸的 Fe–13Cr15Ni 奥氏体不锈钢，确定了位错通道诱导辐照辅助应力腐蚀裂纹的机制。拉伸诱发了通过数字图像相关和共聚焦显微镜表征的位错通道。研究发现，位错通道要么跨越晶界连续，要么不连续，要么在晶界处有滑移的不连续。不连续通道包含的应变最少，但最容易引发裂纹。不连续位错通道-晶界交界面显示出最高的局部应力。辐照钢的原位拉伸透射电子显微镜（TEM）观察和位错-晶界相互作用的原子模拟提供了支持证据，表明无法传递应变的通道发生了开裂。 通道无法通过相邻晶粒的滑移或晶界滑移来缓解应力，导致局部应力升高，并增加了应力腐蚀开裂的敏感性。

Keywords  关键词

Irradiation assisted stress corrosion cracking
Dislocation channels
Intergranular cracking
Discontinuous slip
Grain boundaries

辐射辅助应力腐蚀开裂位错通道晶间开裂不连续滑移晶界

1. Introduction  1. 引言

Austenitic stainless steel is used in light water reactors for various core components because of its high resistance to corrosion. However, under irradiation it is susceptible to intergranular stress corrosion cracking (IGSCC), termed irradiation assisted stress corrosion cracking (IASCC) [1], [2], [3]. The exact mechanism behind IASCC is not well understood. Irradiation causes a number of changes to the microstructure of the steel, including radiation induced segregation (RIS), radiation hardening, and a change in the deformation mechanism from relatively homogeneous slip to heterogeneous slip, with deformation confined to coarsely spaced bands referred to as dislocation channels. The complexity of the irradiated microstructure makes it difficult to separate the individual effects on IASCC.
奥氏体不锈钢因其高耐腐蚀性，被用于轻水反应堆的各种核心部件。然而，在辐照条件下，它易发生晶间应力腐蚀开裂（IGSCC），称为辐射辅助应力腐蚀开裂（IASCC）[1]，[2]，[3]。IASCC 的确切机制尚不清楚。辐照导致钢的微观结构发生多种变化，包括辐照诱导偏析（RIS）、辐照硬化，以及变形机制从相对均匀的滑移转变为非均匀滑移，变形被限制在粗大的带状区域，称为位错通道。辐照后微观结构的复杂性使得难以分离对 IASCC 的各个影响。
Early research focused on chromium depletion due to RIS at the grain boundary as the key factor responsible for increased cracking susceptibility, similar to the increased cracking susceptibility in sensitized steel. More recent data, however, has showed that, while chromium depletion plays a role in IASCC, it is not likely the dominant factor [4], [5]. In the work by Busby et al. [4], irradiation effects were found to recover at different rates during annealing, and in particular, it was observed that cracking susceptibility recovers much more rapidly than RIS. Cracking correlates with yield strength, but the correlation lacks a physical basis. More recent work has identified dislocation channeling as a potential factor controlling IASCC [1], [2], [6], [7], [8], [9] that is linked to the yield strength of the material. Dislocation channeling is a phenomenon that occurs as applied stresses reach the critical resolved shear stress and dislocations are pushed through the irradiation damaged microstructure. As the dislocations move along the slip plane, they clear some of the defect clusters formed by irradiation. Subsequent dislocations are more likely to follow the same path due to the lower density of defects, causing deformation to be highly localized and heterogeneous. The critical resolved shear stress is higher in irradiated metals [10], causing an increase in yield strength prior to relief through the creation and slip of dislocations.
早期研究主要关注辐照助应力腐蚀开裂（RIS）导致晶界处的铬贫化是增加开裂敏感性的关键因素，类似于敏化钢中开裂敏感性的增加。然而，最近的数据表明，虽然铬贫化在 IASCC 中发挥作用，但它不太可能是主要因素[4], [5]。在 Busby 等人[4]的研究中，发现辐照效应在退火过程中以不同速率恢复，特别是观察到开裂敏感性比 RIS 恢复得更快。开裂与屈服强度相关，但这种相关性缺乏物理基础。最近的研究已经确定位错通道是控制 IASCC 的潜在因素[1], [2], [6], [7], [8], [9]，它与材料的屈服强度相关。位错通道是一种现象，当施加的应力达到临界解理剪切应力时发生，位错被推过辐照损伤的微观结构。随着位错沿滑移面移动，它们清除部分由辐照形成的缺陷团簇。 后续位错更可能沿着相同路径移动，因为缺陷密度较低，导致变形高度局部化和异质化。辐照金属中的临界解理应力更高[10]，在位错产生和滑移通过之前，屈服强度会先增加，然后通过释放来缓解。
Dislocation channels are characterized by their width (∼0.1 μm), spacing (generally 1–3 μm), and the amount of shear strain in the channel [11], [12]. The amount of shear strain in the dislocation channels is difficult to analyze experimentally. Strain in the channel has been estimated by measuring the step height caused by the channel on the specimen surface [6], as well as the offset the channel caused when it intersected a grain boundary [13]. Using these methods of measurement, strain within dislocation channels was found to be around two orders of magnitude higher than the bulk applied strain.
位错通道的特征在于其宽度（~0.1 μm）、间距（通常 1-3 μm）以及通道中的剪切应变[11][12]。位错通道中的剪切应变难以通过实验分析。通道中的应变通过测量通道在样品表面造成的台阶高度[6]以及通道与晶界相交时造成的偏移[13]来估计。使用这些测量方法，发现位错通道内的应变比整体施加的应变高约两个数量级。
The high amount of strain in the dislocation channels is accompanied by elevated levels of stress as predicted by computer modeling. Evrard and Sauzay [14], [15] used a finite element model to simulate the intersection between channels and grain boundaries. Dislocation channels were modeled as soft regions (low critical resolved shear stress) in a hard grain (high critical resolved shear stress). Areas of high stress were found at the intersection of the channel with the grain boundary. Atomistic modeling has also shown high levels of stress due to dislocation impingement on grain boundaries [16], [17].
位错通道中的高应变伴随着计算机建模预测的高应力水平。Evrard 和 Sauzay[14][15]使用有限元模型模拟了通道与晶界的交点。位错通道被建模为硬晶粒（高临界解理应力）中的软区域（低临界解理应力）。在通道与晶界的交点处发现了高应力区域。原子尺度建模也表明，由于位错与晶界的碰撞导致高应力水平[16][17]。
The degree of localized deformation in the form of discontinuous channels correlates strongly with cracking [17], [18]. Was et al. [19] discussed possible mechanisms relating localized deformation and IASCC. Three forms of dislocation channel–grain boundary (DC–GB) interactions were considered: Continuous slip across the grain boundary, discontinuous slip with dislocation absorption into the grain boundary (disc. w/GB slip) resulting in grain boundary sliding, and discontinuous slip resulting in a dislocation pileup where slip is not accommodated in any form. Of these three, the latter two were considered the likely locations for crack initiation due high strain in the case of grain boundary sliding, and high stress in the case of the dislocation pileup.
局部变形的程度以不连续通道的形式存在，与开裂密切相关[17], [18]。Was 等人[19]讨论了局部变形与 IASCC 之间可能的联系机制。考虑了三种位错通道-晶界（DC-GB）相互作用形式：沿晶界连续滑移、位错被吸收到晶界的不连续滑移（与 GB 滑移）导致晶界滑动，以及导致位错堆积的不连续滑移（滑移未以任何形式得到适应）。在这三种形式中，后两种被认为是裂纹萌生的可能位置，因为在晶界滑动的情况下存在高应变，而在位错堆积的情况下存在高应力。
West and Was [20] identified a connection between IASCC susceptibility with stress at the grain boundary by using a Schmid-Modified Grain Boundary Stress (SMGBS) model to correlate grain boundary stress with IASCC. The SMGBS model used not only the orientation of the boundary plane with respect to the tensile axis, but also factored in the propensity of the grain for deformation, which was determined by the Schmid factor. Grains with low Schmid factors were determined to have a low propensity to deform, and this resulted in higher stress at the grain boundary, as well as a high propensity for cracking. DC–GB intersections where slip is not transmitted are also likely to cause localized areas of high stress, and are considered likely candidates to induce cracking, as described by Was et al. [19]. In this case, the dislocations are not able to be accommodated through slip transmission before the critical stress needed to induce cracking is reached.
West 和 Was[20]通过使用 Schmid 修正晶界应力（SMGBS）模型，将 IASCC 敏感性与晶界处的应力联系起来。该 SMGBS 模型不仅考虑了晶界平面相对于拉伸轴的方向，还考虑了晶粒变形的倾向，该倾向由 Schmid 因子决定。Schmid 因子较低的晶粒被认为变形倾向较低，这导致晶界处应力较高，并具有高开裂倾向。滑移未传递的 DC-GB 交点也可能导致局部高应力区域，正如 Was 等人[19]所述，这些交点被认为是引发开裂的潜在候选点。在这种情况下，在达到引发开裂的临界应力之前，位错无法通过滑移传递来适应。
Recently, digital image correlation (DIC) has been used to confirm that when a channel terminates at a grain boundary (discontinuous), two possible interactions may result [16]; the slip may be transmitted into the grain boundary, inducing grain boundary slip, or the slip may terminate at the grain boundary, resulting in unaccommodated slip. This work seeks to differentiate the roles of stress and grain boundary strain in the IASCC crack initiation mechanism.
近来，数字图像相关（DIC）技术已被用于证实当通道终止于晶界（不连续）时，可能产生两种相互作用[16]：滑移可能传递到晶界，诱导晶界滑移，或者滑移可能终止于晶界，导致未协调的滑移。本研究旨在区分应力与晶界应变在辐照辅助应力腐蚀裂纹萌生机制中的作用。

2. Experimental  2. 实验

A controlled purity austenitic Fe–13Cr–15Ni alloy (Table 1) was used in this study. Electric discharge machining (EDM) was used to cut a tensile bar with a 21 mm gage length and a 2 mm × 1.5 mm cross section. The EDM damage was removed with a mechanical polish, followed by an electropolish at 30 V in a solution of 10% perchloric acid and 90% methanol at −40 °C for 90 s in order to remove all residual mechanical damage from the mechanical polish.
本研究使用了控制纯度的奥氏体 Fe–13Cr–15Ni 合金（表 1）。采用电火花加工（EDM）切割一根长度为 21 mm、横截面为 2 mm × 1.5 mm 的拉伸试样。通过机械抛光去除 EDM 损伤，然后在-40 °C 的 10%高氯酸和 90%甲醇溶液中以 30 V 进行电解抛光 90 s，以去除机械抛光留下的所有残余机械损伤。

Table 1. Composition (wt%) of Fe–13Cr15Ni austenitic steel used in this study.
表 1. 本研究使用的 Fe–13Cr15Ni 奥氏体钢的成分（wt%）。

Material designation  材料标识FeCrNiMn  锰Si  硅PC
13Cr15Ni  13 铬 15 镍Bal.  平衡13.4115.041.030.1<0.010.016
After polishing, the sample was irradiated using 3 MeV protons to 5 dpa and at a temperature of 360 °C at the Michigan Ion Beam Laboratory. The temperature was monitored by a two-dimensional thermal imager (IRCON® Stinger thermal imaging system) that tracked surface temperatures of the sample at high spatial resolution throughout the irradiation. The temperature was controlled using a combination of an electric heater inserted into the stage, and air flow through the stage. The sample temperature was maintained to within 9 °C of the set temperature (360 °C).
样品经过抛光后，在密歇根离子束实验室使用 3 MeV 质子辐照至 5 dpa，温度为 360 °C。温度通过二维热像仪（IRCON® Stinger 热成像系统）监测，该系统在整个辐照过程中以高空间分辨率追踪样品表面温度。温度控制通过将电加热器插入载台并与载台气流结合实现。样品温度保持在设定温度（360 °C）的±9 °C 范围内。
The depth of the proton penetration into the sample was about 40 μm. The damage rate profile is shown in Fig. 1, as determined using the full cascade model in SRIM™. The damage rate, as well as the total damage, is typically determined in the flat region prior to the peak. In this work, the damage rate was calculated to be ∼9 × 10−6 dpa/s at a depth of ∼24 μm and represented an average damage rate value in the flat region. It should be noted that recent results have shown that the full cascade method overestimates the damage by about a factor of 2 [21].
质子穿透样品的深度约为 40 μm。损伤率分布如图 1 所示，该数据通过 SRIM™中的全级联模型确定。损伤率以及总损伤通常在峰值前的平坦区域确定。在本研究中，在深度约为 24 μm 处，损伤率计算为∼9 × 10 −6 dpa/s，代表了平坦区域的平均损伤率值。需要注意的是，最近的研究结果表明，全级联方法高估了损伤率约 2 倍[21]。
  1. Download: Download high-res image (62KB)
    下载：下载高分辨率图像（62KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 1. Damage rate profile for Fe–13Cr15Ni irradiated with 3 MeV protons. The damage rate at a depth of 24 μm was used to determine final dpa of the experimental sample.
图 1. 3 MeV 质子辐照的 Fe–13Cr15Ni 的损伤率分布。24 μm 深度的损伤率被用于确定实验样品的最终 dpa。

In-plane displacements of strained samples were determined using DIC. To perform DIC, a gold speckle pattern was deposited on the tensile bar surface using 40 nm gold nano-particles. The nano-particles were created by combining 1 mL of a solution consisting of 99 wt% water and 1 wt% Na3-citrate to 100 mL of 0.01 wt% HAuCl4 in distilled water which had been heated to a boil [22]. The tensile specimen was coated with (3-aminopropyl) trimethoxysilane and the gold nano-particles were deposited on the surface creating a speckle pattern, as shown in Fig. 2. This deposition technique was originally developed for surface enhanced Raman spectrometry [23], but later used by Kammers [24] for DIC.
应变样品的面内位移通过数字图像相关法（DIC）测定。为进行 DIC，使用 40 nm 金纳米粒子在拉伸试样表面沉积金点阵图案。这些纳米粒子是通过将 1 mL 由 99 wt% 水和 1 wt% 柠檬酸钠组成的溶液与 100 mL 加热至沸腾的 0.01 wt% 氯金酸（HAuCl₄）蒸馏水混合制备的 [22]。拉伸试样被涂覆 (3-氨基丙基) 三甲氧基硅烷，并在表面沉积金纳米粒子形成点阵图案，如图 2 所示。这种沉积技术最初是为表面增强拉曼光谱法开发的 [23]，后来被 Kammers [24] 用于 DIC。
  1. Download: Download high-res image (242KB)
    下载：下载高分辨率图像 (242KB)
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 2. Gold (white) speckle pattern on Fe–13Cr15Ni steel (black).
图 2. Fe–13Cr15Ni 钢（黑色）上的金（白色）点阵图案。

The tensile sample was imaged in a JEOL JSM-6480 SEM both before and after straining, so that the two images could be correlated and displacement measured. As there is some distortion in the SEM imaging process, an Ultrasharp TGX01 AFM calibration grid (3 μm grid spacing) was imaged alongside the tensile bar at the same working distance. Distortion in the grid was corrected using Matlab® and the same image correction was applied to the images of the tensile bar surface.
拉伸样品在应变前后均使用 JEOL JSM-6480 扫描电子显微镜进行了成像，以便将两张图像进行关联并测量位移。由于扫描电子显微镜成像过程中存在一定程度的畸变，因此在与拉伸试样在同一工作距离下，使用 Ultrasharp TGX01 原子力显微镜校准网格（3 μm 网格间距）进行了成像。使用 Matlab®对网格的畸变进行了校正，并将相同的图像校正应用于拉伸试样表面的图像。
The tensile bar was strained in two increments; to 1.5% plastic strain, followed by an additional 1% plastic strain increment for a total strain of 2.5% in high temperature (288 °C) argon at a strain rate of ∼3 × 10−7 s−1. Slow strain rate testing was used to provide good control over the level of strain induced into the samples. DIC was performed after each strain increment, with the primary focus on the 2.5% straining step where a larger number of measurements were taken. Gold particles were removed by lightly cleaning with a polishing pad wetted with distilled water after the second strain increment so that the gold particles would not affect the corrosion of steel when placed in the high temperature BWR water environment. DIC does not provide any topographical information (out-of-plane displacement), so an Olympus OLS4000 LEXT Confocal microscope was used to take the topographical measurements of the sample surface after the 2.5% strain in argon. Changes in displacement were measured across dislocation channels and grain boundaries, and the magnitude of all three vectors was calculated to determine the total displacement caused by the dislocation channel. Shear strain within the channel was calculated by assuming the dislocation channels were 100 nm wide [11], and dividing the total displacement by the width of the channel.
拉伸试样分两次应变；首先应变至 1.5%塑性应变，随后再增加 1%塑性应变增量，总应变为 2.5%，在高温（288°C）氩气中以约 3 × 10^0 s^-1 的应变率进行。采用慢应变率测试以对样品产生的应变水平进行良好控制。每次应变增量后进行数字图像相关（DIC）测量，主要关注 2.5%应变步骤，在该步骤中进行了大量测量。在第二次应变增量后，用浸有蒸馏水的抛光垫轻轻清洁以去除金颗粒，以防止金颗粒在高温 BWR 水环境中影响钢的腐蚀。DIC 不提供任何形貌信息（面外位移），因此使用奥林巴斯 OLS4000 LEXT 共聚焦显微镜在氩气中 2.5%应变后对样品表面进行形貌测量。测量了位错通道和晶界处的位移变化，并计算了三个矢量的幅度，以确定位错通道引起的总位移。 通道内的剪切应变通过假设位错通道宽度为 100 纳米[11]，并将总位移除以通道宽度来计算。
Residual stress measurements were made at various DC–GB intersections after straining to 2.5% using BLG Production’s CrossCourt 3 electron backscatter diffraction (EBSD) analysis software. This technique relies on the cross correlation of areas of interest in the EBSD patterns between the measurement pattern and a reference pattern, as explained by Wilkinson et al. [25]. Reference patterns were taken from areas well removed from the GB–DC intersections, where stress is likely to be close to zero. The small shifts in the EBSD pattern are correlated to distortions in the crystal lattice caused by elastic strain. From these shifts, the elastic strain tensor is determined, and the elastic stress tensor calculated using Hooke’s law. EBSD patterns were collected at points 100 nm apart, with the electron beam maintained on each measurement point for 3 s to collect high quality patterns. Twenty regions of interest were selected on each EBSD pattern, which were correlated with the reference pattern to measured elastic strain.
在将样品应变至 2.5%后，使用 BLG Production 的 CrossCourt 3 电子背散射衍射（EBSD）分析软件在多种直流沟道-晶界交点处进行了残余应力测量。该技术依赖于测量图案和参考图案中 EBSD 图案中感兴趣区域的交叉相关性，如 Wilkinson 等人[25]所述。参考图案取自远离晶界-直流沟道交点的区域，这些区域的应力可能接近于零。EBSD 图案中的微小偏移与弹性应变引起的晶体点阵畸变相关。根据这些偏移，确定了弹性应变张量，并使用胡克定律计算了弹性应力张量。EBSD 图案在相距 100 nm 的点处采集，电子束在每个测量点保持 3 秒以收集高质量的图案。在每个 EBSD 图案上选择了 20 个感兴趣区域，这些区域与参考图案相关联以测量弹性应变。
After analysis of the samples strained in argon, additional straining to 4.5% and then 7.2% was conducted in simulated boiling water reactor normal water chemistry (BWR NWC) conditions (288° water with a conductivity maintained at 0.2 μS/cm and dissolved oxygen at ∼2 ppm). These slow strain rate tests allow for cracking to be accelerated over what would occur normally in the light water reactor components under a high load. The slow strain rate tests also allow of good control so that straining could be stopped while cracks were small. Straining was performed incrementally so that crack initiation sites could be located before the cracks propagated to larger sizes and obscured the initiation site. After straining in water, cracks were located and characterized based on prior DC–GB information determined after the argon straining steps.
在分析了在氩气中受拉的样品后，在模拟沸水反应堆正常水化学（BWR NWC）条件下（288°水，电导率维持在 0.2 μS/cm，溶解氧为~2 ppm）进行了额外的 4.5%和 7.2%的拉伸。这些慢速应变率试验允许在高负载下加速裂纹的形成，这比轻水反应堆部件在正常条件下的裂纹形成要快。慢速应变率试验还允许进行良好控制，以便在裂纹较小时停止拉伸。拉伸是逐步进行的，以便在裂纹扩展到较大尺寸并掩盖起始位置之前找到裂纹起始点。在水中拉伸后，根据氩气拉伸步骤后确定的先前直流-晶界（DC-GB）信息定位和表征裂纹。
The statistical significance of the cracking results was determined using a binomial distribution model from which the standard deviation (σ) was used to determine the amount of error in the measurements. Standard deviations for binomial distributions are calculated as
使用二项分布模型确定了裂纹结果的统计显著性，该模型使用标准差（σ）来确定测量中的误差量。二项分布的标准差按以下方式计算(1)σx=x(1-p),where x is the number of cracks at the DC–GB type being studied and p is the probability of the crack occurring at the DC–GB type in question, defined as
其中 x 是所研究的 DC-GB 类型中的裂纹数量，p 是所讨论的 DC-GB 类型中裂纹发生的概率，定义为(2)p=xn.The value n is the total number of DC–GB intersections of the type being studied. In terms of the fractional uncertainty of the probability of cracking,
n 的值是所研究类型的 DC-GB 交叉的总数。在裂纹概率的分数不确定性的方面，(3)σpp=1n(1-p)p.
This statistical analysis was used to calculate error in the measurements taken of cracking at each of the three types of DC–GB intersections.
这种统计分析用于计算三种类型的 DC-GB 交叉中裂纹测量的误差。

3. Results  3. 结果

3.1. Characterization of DC–GB intersections following straining in an argon gas environment
3.1. 在氩气环境中受力后的 DC-GB 交叉表征

All measured DC–GB intersections were characterized as one of three possible types: Continuous, disc. w/GB slip, or discontinuous. Continuous intersections are those where slip is transferred from one grain to the next and are identified by channels in adjacent grains that meet at the grain boundary. They may occur by direct transmission of the dislocation across the grain boundary or by the formation of a new dislocation source in the adjacent grain at the point where the incoming channel intersects the grain boundary. Disc. w/GB slip are channels that terminate at the grain boundary but which give rise to measurable slip in the boundary. Discontinuous channels terminate at the grain boundary with no evidence of slip transmission into an adjacent grain or in the grain boundary.
所有测量的直流-晶界交点被归类为三种可能类型之一：连续型、带晶界滑移的圆盘型或不连续型。连续型交点是指滑移从一个晶粒传递到下一个晶粒的情况，其特征是相邻晶粒中的通道在晶界处汇合。它们可能通过位错直接穿过晶界形成，也可能在入射通道与晶界相交处，在相邻晶粒中形成新的位错源。带晶界滑移的圆盘型通道是指在晶界处终止，但能在晶界中产生可测量的滑移的通道。不连续型通道在晶界处终止，没有证据表明滑移传递到相邻晶粒或晶界中。
Total (sum of in-plane and out-of-plane) displacement in the dislocation channels at the DC–GB intersection, shown in Fig. 3 after straining to 2.5%, was largest at continuous intersections, and smallest at discontinuous, which is in agreement with earlier work on Fe–13Cr15Ni strained to 3.5% in an argon gas environment [16]. The average and maximum displacements of each DC–GB intersection type are given in Table 2, and clearly show that discontinuous channels exhibited far less displacement than channels that were classified as continuous or disc. w/GB slip, both on average displacement and in the maximum. Grain boundary displacement measurements (GB in Fig. 3) are also shown for cases of disc. w/GB slip. On rare occasion, a continuous DC–GB intersection will also experience grain boundary slip. Two cases were observed in this work, of the 126 cases of continuous DC–GB intersections identified after the 2.5% strain was applied to the specimen. These two cases (∼120 and 30 nm displacement) were not included in the grain boundary displacement measurements of Fig. 3.
在直流道与晶界的交点处，经过 2.5%应变后的位错通道总位移（平面内和面外位移之和），如图 3 所示，在连续交点处最大，在非连续交点处最小，这与先前关于在氩气环境中将 Fe–13Cr15Ni 应变至 3.5%的研究结果一致[16]。每种直流道与晶界交点类型的平均位移和最大位移如表 2 所示，清晰地表明非连续通道的位移远小于被归类为连续或圆盘形的通道，无论是在平均位移还是在最大位移上。图 3 中也显示了圆盘形 w/GB 滑移情况下的晶界位移测量值。在极少数情况下，连续的直流道与晶界交点也会发生晶界滑移。本研究中观察到了两个这样的案例，在将样品应变至 2.5%后识别出的 126 个连续直流道与晶界交点中。这两个案例（位移约为 120 和 30 纳米）未包含在图 3 的晶界位移测量值中。
  1. Download: Download high-res image (64KB)
    下载：下载高分辨率图像（64KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 3. Total displacement in the grain boundary after 2.5% strain in argon by DC–GB intersection type (continuous, disc. w/GB slip and discontinuous). Grain boundary displacement measurements from Disc w/GB slip are also included and account for only the displacement due to a single dislocation channel.
图 3. 在氩气中经 2.5% 应变后，不同位错通道与晶界交截类型（连续、带晶界滑移的圆盘形和不连续型）下的晶界总位移。圆盘形带晶界滑移的晶界位移测量结果也包含在内，仅考虑单个位错通道引起的位移。

Table 2. Average and maximum displacement in dislocation channels for each DC–GB intersection type (including both channel and grain boundary displacement measurements for disc. w/GB slip). Measurements taken after 2.5% strain in an argon environment at 288 °C.
表 2. 每种位错通道与晶界交截类型的平均位移和最大位移（圆盘形带晶界滑移类型包含通道位移和晶界位移测量结果）。测量在氩气环境中 288 °C 经 2.5% 应变后进行。

Empty CellContinuous  连续Disc. w/GB slip  带晶界滑移的圆盘形Discontinuous  不连续
Empty CellChannel  通道Boundary  边界
Average displacement (nm)
平均位移 (nm)		18517213266
Maximum displacement (nm)
最大位移 (nm)		660541320174
The grain boundary displacements shown in Fig. 3 and Table 2 are the displacement measurements caused by a single dislocation channel. Multiple channels may intersect the boundary and contribute to the total amount of slip. Fig. 4 shows an SEM image of multiple disc. w/GB slip dislocation channels, each inducing grain boundary slip, as seen in the XY shear strain map on the right-hand side, causing the amount of slip in the boundary to increase toward the bottom of the image. In order to report grain boundary slip due to a single channel, displacement was measured along the grain boundary on both sides of a DC–GB intersection, and the difference in the two measurements was attributed to the boundary slip caused by the dislocation channel. Fig. 5 shows the total displacement in the grain boundary, resulting from all dislocation channels intersecting the grain boundary, rather than the displacement in the boundary from a single channel, as in Fig. 3. In terms of the effects of grain boundary slip on cracking, the results in Fig. 5 are more relevant, as they represent the actual slip experienced by the boundary, rather than just the portion due to a single channel, as shown in Fig. 3.
图 3 和表 2 中所示的晶界位移是由单个位错通道引起的位移测量值。多个通道可能相交于晶界，并贡献于总滑移量。图 4 显示了多个圆盘形 w/GB 滑移位错通道的 SEM 图像，每个通道都诱导晶界滑移，如右侧的 XY 剪切应变图所示，导致晶界处的滑移量向图像底部增加。为了报告单个通道引起的晶界滑移，在 DC-GB 交点两侧沿晶界测量位移，并将两次测量的差值归因于位错通道引起的晶界滑移。图 5 显示了所有相交于晶界的位错通道引起的晶界总位移，而不是像图 3 那样显示单个通道引起的晶界位移。就晶界滑移对开裂的影响而言，图 5 的结果更具相关性，因为它们代表了晶界实际经历的滑移，而不是像图 3 那样仅显示单个通道引起的部分。
  1. Download: Download high-res image (176KB)
    下载：下载高分辨率图片（176KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 4. Example of Disc w/GB slip intersection, where each dislocation channel contributes to additional slip in the boundary. The right-side image shows the XY shear strain map.
图 4. 圆盘与 GB 滑移交叉的示例，其中每个位错通道对边界处的额外滑移做出贡献。右侧图片显示了 XY 剪切应变图。

  1. Download: Download high-res image (64KB)
    下载：下载高分辨率图片（64KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 5. Total displacement in grain resulting from all DCs intersecting the grain boundary.
图 5. 所有与晶界交叉的 DC 在晶粒中产生的总位移。

Straining in the argon environment was performed in two increments, 1.5% and 1% strain, to a total of 2.5%. Fig. 6 shows the in-plane slip distribution after the two strain increments. All channels were found to undergo additional slip when the overall sample strain was increased from 1.5% to 2.5%. Of the 44 DC–GB intersections characterized, 33 were continuous, and the strain in the channels increased, on average, 144%. Seven were D/GB and strain in the channels increased by 91% on average. Strain in the four discontinuous channels increased, on average, by 151%. With additional strain, slip transmission from the DC–GB intersection occurs on occasion. Fig. 7 shows one of two intersections that were found to change classifications, this one from discontinuous to disc. w/GB slip (the other went from discontinuous to continuous). These were not included in the 44 slip measurements where average changes in slip were determined for each DC–GB intersection type.
在氩气环境中进行两次应变增量，分别为 1.5%和 1%，总应变为 2.5%。图 6 显示了两次应变增量后的平面滑移分布。当整体样品应变从 1.5%增加到 2.5%时，所有通道均发生额外滑移。在表征的 44 个 DC-GB 交点中，33 个是连续的，通道中的应变平均增加了 144%。其中 7 个是 D/GB，通道中的应变平均增加了 91%。四个不连续通道中的应变平均增加了 151%。在额外应变下，有时会发生从 DC-GB 交点向通道的滑移传递。图 7 显示了两个发生分类变化的交点之一，这个交点从不连续变为碟形 w/GB 滑移（另一个交点从不连续变为连续）。这些未包含在 44 次滑移测量中，其中为每种 DC-GB 交点类型确定了平均滑移变化。
  1. Download: Download high-res image (80KB)
    下载：下载高分辨率图像（80KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 6. In-plane slip distribution of the same channels measured after 1.5% strain and 2.5% strain in argon.
图 6. 在氩气中经历 1.5%应变和 2.5%应变后相同通道的平面滑移分布

  1. Download: Download high-res image (155KB)
    下载：下载高分辨率图像（155KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 7. SEM image and horizontal (tensile axis) strain map of DC–GB intersection at 1.5% macroscopic strain (left) and 2.5% macroscopic strain (right). At 2.5%, the channel, which initially appeared to be discontinuous, initiated grain boundary slip, becoming disc. w/GB slip.
图 7. 1.5%宏观应变（左）和 2.5%宏观应变（右）下 DC-GB 交叉处的 SEM 图像和水平（拉伸轴）应变图。在 2.5%时，最初看似不连续的通道启动了晶界滑移，形成带晶界滑移的圆盘。

High resolution EBSD measurements were taken at each of the DC–GB intersection types to determine the stress (based on elastic strain measurements) at the intersection. In total, 48 intersections were measured, 28 taken after the 2.5% strain and 20 from a second sample strained to 3.5% in argon [16]. Examples of each DC–GB type are shown in Fig. 8. Here the SEM image is shown in the left column and an inverse pole figure EBSD map showing grain orientation is in the center. The column on the right shows the Von Mises stress maps created using CrossCourt 3, as described previously in the experimental section. White areas of the stress maps are locations where the EBSD pattern was not clear enough to perform a cross correlation. The data from the most complete maps (with the fewest missing pixels) were collected in graphs shown in Fig. 9, where the average stress is given as a function of distance (average of the measurements taken in a circle of a given radius from the DC–GB intersection). The different colors on the graph represent different DC–GB intersections. A general rise is noted as r approaches zero, which is the location of DC–GB intersection. This rise is more noticeable in the graph showing stress near the intersection with a discontinuous channel.
在每种直流沟道与晶界（DC–GB）交点处进行了高分辨率 EBSD 测量，以确定交点处的应力（基于弹性应变测量）。总共测量了 48 个交点，其中 28 个在 2.5%应变后测量，20 个来自在氩气中应变至 3.5%的第二块样品[16]。每种 DC–GB 类型的示例如图 8 所示。此处左侧列显示 SEM 图像，中间列显示显示晶粒取向的反极图 EBSD 图。右侧列显示了使用 CrossCourt 3 创建的 Von Mises 应力图，如实验部分先前所述。应力图中的白色区域是 EBSD 图案不够清晰无法进行交叉相关分析的位置。从最完整的图（缺失像素最少）中收集数据，如图 9 所示的图中，平均应力表示为距离的函数（平均测量值来自以 DC–GB 交点为中心、给定半径为半径的圆内）。图中的不同颜色代表不同的 DC–GB 交点。当 r 接近零时，即 DC–GB 交点的位置，观察到应力普遍上升。 这种上升在显示应力与不连续通道交点附近的图中更为明显。
  1. Download: Download high-res image (345KB)
    下载：下载高分辨率图像（345KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 8. Determination of elastic stress at DC–GB intersections. In the first column, an SEM image of the intersection is shown, with grain boundaries marked in blue. The second column shows EBSD measurements depicting orientation (each grain is a distinct color), with black pixels representing locations where orientation could not be determined. The third column shows the Von Mises stress results of the EBSD stress analysis. White pixels represent areas where data could not be collected. Shown in this figure are examples of (a) continuous, (b) disc. w/GB slip, and (c) discontinuous DC–GB intersections.
图 8. DC-GB 交点的弹性应力测定。第一列显示了交点的 SEM 图像，晶界用蓝色标记。第二列展示了 EBSD 测量结果，描绘了取向（每个晶粒为不同颜色），黑色像素代表无法确定取向的位置。第三列显示了 EBSD 应力分析中的 Von Mises 应力结果。白色像素代表无法收集数据区域。本图展示了（a）连续、（b）带晶界滑移的圆盘形以及（c）不连续 DC-GB 交点的示例。

  1. Download: Download high-res image (120KB)
    下载：下载高分辨率图像（120KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 9. Results from 10 continuous, 9 disc. w/GB slip, and 9 discontinuous DC–GB intersections. Average stress was determined at radii every 100 nm from the intersection points. Each color represents a different DC–GB intersection that was characterized with EBSD.
图 9. 10 次连续、9 次带 GB 滑移的圆盘和 9 次不连续 DC–GB 交点结果。在交点每个 100 nm 的半径处确定平均应力。每种颜色代表一个不同的 DC–GB 交点，该交点通过 EBSD 进行了表征。

3.2. Characterization of cracking in BWR-NWC
3.2. BWR-NWC 中的裂纹表征

The specimen strained in argon to 2.5% was then strained additional increments of 2% and 2.7% in simulated BWR NWC water to a total strain of 7.2%. Crack initiation sites were characterized according to the DC–GB intersection type (as determined after the 2.5% straining in argon). Fig. 10 shows example SEM images of (a) continuous, (b) disc. w/GB slip and (c) discontinuous DC–GB intersections taken after the 2.5% argon strain increment (left) and after cracks initiated during the 7.2% BWR NWC strain increment (right). This data, and a general characterization of DC–GB intersection densities performed after the 2.5% argon straining step, are shown in Table 3. All cracks characterized were initiation sites, and on the order of several microns.
将试样在氩气中应变至 2.5%，然后在模拟 BWR NWC 水中增加 2% 和 2.7% 的应变，直至总应变为 7.2%。根据 DC–GB 交点类型（在氩气中应变 2.5% 后确定）表征裂纹起始位置。图 10 显示了在氩气应变增量 2.5% 后（左）和在 BWR NWC 应变增量 7.2% 时裂纹起始时（右）的（a）连续、（b）带 GB 滑移的圆盘和（c）不连续 DC–GB 交点的典型 SEM 图像。这些数据以及在氩气应变步骤后进行的 DC–GB 交点密度的一般表征显示在表 3 中。所有表征的裂纹都是起始点，尺寸约为几微米。
  1. Download: Download high-res image (488KB)
    下载：下载高分辨率图像 (488KB)
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 10. Examples of (a) continuous, (b) disc. w/GB slip and (c) discontinuous GB intersections that resulted in crack initiation. Post 2.5% argon strain is shown on left and same area after the 7.2% total strain (2.5% argon and 4.7% BWR) is shown on the right.
图 10. 导致裂纹萌生的 (a) 连续型、(b) 具有晶界滑移的圆盘型和 (c) 不连续型晶界交界的示例。左侧显示 2.5% 氩气应变后的情况，右侧显示 7.2% 总应变（2.5% 氩气和 4.7% BWR）后的相同区域。

Table 3. Characterization of cracking at DC–GB intersections. The first column of data shows the bulk DC–GB densities, as measured after the 2.5 argon straining step. The next two columns show the number of cracks that occurred at each of the different location classifications.
表 3. DC–GB 交界面处裂纹的特征。数据的第一列显示在 2.5% 氩气应变步骤后测得的体相 DC–GB 密度。接下来的两列显示在每个不同位置分类中发生的裂纹数量。

Crack location  裂纹位置Number of DC–GB intersections/mm2 (% of total intersections)
DC–GB 交界面数/mm 2 （占总交界面百分比）		Number of cracks after 4.5% strain (% of total cracks)
4.5%应变后的裂纹数量（占总裂纹百分比）		Number of cracks after 7.2% strain (% of total cracks)
7.2%应变后的裂纹数量（占总裂纹百分比）
Continuous  连续385 (53)11 (12)27 (14)
Disc. w/GB slip  带晶界滑移的离散192 (27)8 (9)14 (7)
Discontinuous  不连续147 (20)14 (16)27 (14)

Unknown DC–GB intersection type
未知 DC-GB 交点类型		–7 (8)18 (9)
No visible DC  无可见 DC–9 (10)28 (15)
TJ–40 (45)79 (41)
Cracks also formed away from DC–GB intersections and at the intersection of three grains. Cracks that appeared to originate at these intersections where characterized as triple junction (TJ) cracks, regardless of whether the crack was also intersected by a dislocation channel. Of the 40 TJ cracks that occurred after 4.5% total strain, 50% of them were noted to have also been intersected by a channel (5 continuous, 4 disc. w/GB slip, and 11 discontinuous). However, since it is impossible to factor out the effects of the TJ from the channel, the TJ cracks were characterized separately from the cracks at DC–GB intersections. The importance of TJ in cracking will be examined later in the discussion section. The category “Unknown DC–DB intersection type” captures cracks that occurred at DC–GB intersections in which the dislocation channel formed after characterization (during straining in BWR-NWC). Cracks that occurred along grain boundaries at which there was no dislocation channel apparent in the SEM images were classified as “No visible DC”.
裂纹也形成在 DC-GB 交点和三晶粒交点处。这些交点处形成的裂纹被归类为三重结点（TJ）裂纹，无论裂纹是否也被位错通道相交。在 4.5%总应变后形成的 40 个 TJ 裂纹中，50%被观察到也被通道相交（5 个连续，4 个带晶界滑动的圆盘形，11 个不连续）。然而，由于无法分离 TJ 和通道的影响，TJ 裂纹与 DC-GB 交点处的裂纹被单独分类。讨论部分将进一步探讨 TJ 在裂纹形成中的重要性。"未知 DC-DB 交点类型"类别涵盖了在 DC-GB 交点处形成的裂纹，这些裂纹的位错通道在表征后形成（在 BWR-NWC 的应变过程中）。在 SEM 图像中未显示出位错通道的晶界处的裂纹被归类为"无可见 DC"。

4. Discussion  4. 讨论

The results are analyzed with respect to the slip oxidation model for crack initiation, focusing on the potential roles of both strain and stress in crack initiation. That is, the question to be answered is whether local strains or local stresses are the key features responsible for initiation of IASCC cracks in the context of the slip oxidation model. The conclusions reached are compared with prior work and are shown to be supported by the work of others examining IASCC.
研究结果针对裂纹起始的滑移氧化模型进行了分析，重点关注应变和应力在裂纹起始中的潜在作用。也就是说，需要回答的问题是，在滑移氧化模型的背景下，是局部应变还是局部应力是导致 IASCC 裂纹起始的关键特征。得出的结论与先前的工作进行了比较，并表明这些结论得到了其他研究 IASCC 的工作的支持。
It is important to note that in all experiments in this study, cracking only occurred during the strain increments performed in the BWR NWC environment. No cracks were observed after the argon straining. This is an expected result as IASCC is a stress corrosion cracking process that occurs only in the presence of a corrosive environment. Without the aggressive environment, cracked oxide layers would repassivate before the cracks developed. According to the slip oxidation model, cracking of the passive oxide layer is a critical step in promoting IGSCC [26], [27]. Cracking occurs either by slip in grain boundaries below the oxide layer causing it to rupture, or by high local stress that overcomes the cohesive strength of the oxide layer at the grain boundary.
需要注意的是，在本研究所有实验中，裂纹仅在 BWR NWC 环境中进行的应变增量期间出现。氩气应变后未观察到裂纹。这是预期结果，因为应力腐蚀裂纹（IASCC）是一种仅在腐蚀环境中发生的过程。没有腐蚀性环境，裂纹形成的氧化物层会重新钝化。根据滑移氧化模型，被动氧化物层的开裂是促进 IGSCC 的关键步骤[26]，[27]。裂纹的产生或是由氧化物层下晶界处的滑移导致其破裂，或是由高局部应力克服了晶界处氧化物层的内聚力。
In particular, the values of the stress and strain at the DC–GB intersection are important as this is the location at which cracking preferentially occurs. Of the cracks that appeared at DC–GB intersections, it is observed that the majority occurred at continuous and discontinuous DC–GB intersections, in similar numbers. However, Table 3 also shows that the majority of channels that intersect a grain boundary will result in a continuous DC. Fig. 11 shows the cracking fraction normalized to the total number of each type of DC–GB intersection that exists in the area characterized for cracking. The error bars were determined using Eq. (3). It is clear that discontinuous DC–GB intersections are the most susceptible to cracking, with continuous and disc. w/GB slip resulting in similar levels of cracking susceptibility. The high susceptibility to cracking exhibited by the discontinuous DC–GB intersections compared to disc. w/GB slip shows that it is the local stress, not slip in the boundary, which controls cracking.
特别是，DC-GB 交叉处的应力和应变值很重要，因为这是裂纹优先发生的位置。在 DC-GB 交叉处出现的裂纹中，观察到大多数发生在连续和断续的 DC-GB 交叉处，数量相似。然而，表 3 也显示，大多数与晶界相交的通道会导致连续的 DC。图 11 显示了裂纹分数被归一化到该区域中存在的每种类型的 DC-GB 交叉处的总数。误差线使用公式(3)确定。很明显，断续的 DC-GB 交叉处最容易发生裂纹，而连续和盘状/GB 滑移导致相似的裂纹易感性。与盘状/GB 滑移相比，断续的 DC-GB 交叉处表现出的高裂纹易感性表明，控制裂纹的是局部应力，而不是边界处的滑移。
  1. Download: Download high-res image (60KB)
    下载：下载高分辨率图像（60KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 11. Fraction of DC–GBs were crack initiation was found. Fractions are estimated based on bulk DC–GB density measurements, found in Table 3. The error bars were calculated using Eq. (3).
图 11. 发现裂纹起始的直流沟道分数。这些分数基于表 3 中测得的体直流沟道密度估计。误差线使用公式(3)计算。

Previous work has showed that exposure to 288 °C BWR NWC for five days (about the length of the SCC tests in this study) results in an oxide layer of ∼100 nm [28]. For most all boundaries that were characterized, the amount of grain boundary slip (Fig. 5) exceeded the oxide thickness. The out-of-plane displacement is of particular interest since fresh metal is exposed if the displacement exceeds the oxide layer thickness. Fig. 12 shows the out-of-plane portion of the displacement measurements vs. the total displacement (from Fig. 5). The average out-of-plane displacement in the grain boundary was 104 nm and the largest was 618 nm. This is sufficient to rupture the oxide layer and cause IASCC if slip in the grain boundary were controlling. However, cracks initiating at a DC–GB intersection are not directly correlated to the amount of slip within the channel. Recall from Table 2 and Fig. 3 that discontinuous channels had the smallest displacement of the three types by about a factor of 3. This means that IASCC crack initiation does not occur as a direct result of strain, either in the grain boundary or the dislocation channel. We next consider the role of stress caused by the dislocation channel intersection with the grain boundary.
以往研究表明，在 288 °C BWR NWC 中暴露五天（约本研究中应力腐蚀开裂试验的持续时间）会导致形成约 100 nm 厚的氧化层[28]。对于所表征的大多数边界，晶界滑移量（图 5）超过了氧化层厚度。面外位移特别值得关注，因为如果位移超过氧化层厚度，就会暴露新鲜金属。图 12 显示了位移测量值的面外部分与总位移（来自图 5）的对比。晶界中的平均面外位移为 104 nm，最大值为 618 nm。如果晶界滑移是控制因素，这足以破裂氧化层并导致 IASCC。然而，在 DC-GB 交点处萌生的裂纹与通道内的滑移量没有直接相关性。回顾表 2 和图 3 可知，不连续通道的位移是三种类型中最小的，大约小了 3 倍。这意味着 IASCC 裂纹的萌生并不是应变（无论是晶界还是位错通道）的直接结果。 我们接下来考虑位错通道与晶界相交引起的应力作用。
  1. Download: Download high-res image (54KB)
    下载：下载高分辨率图像（54KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 12. Out-of-plane portion of the grain boundary slip measurements vs. total displacement (in-plane and out-of-plane). Like Fig. 5, these measurements represent all of the displacement in the boundary, which in some cases was due to multiple dislocation channels contributing to the total slip. The trend line and the expected oxide thickness are marked on the graph.
图 12. 晶界滑移测量值（面外部分）与总位移（面内和面外）的关系。与图 5 类似，这些测量值表示晶界上的全部位移，在某些情况下，这是由于多个位错通道共同导致总滑移。图中标出了趋势线和预期的氧化物厚度。

Evidence of the importance of stress in IASCC has been noted in studies examining the character of the grain boundaries susceptible to cracking [8], [9], [17], [29]. Grain boundaries normal to the tensile axis and adjacent to low Schmid factor grains, where the resolved normal stress from the applied tensile stress is highest, were found to exhibit a high propensity for cracking. Had shear slip at the grain boundary been controlling the crack initiation, it is expected that the boundaries near 45° to the tensile axis would have been more susceptible to cracking, as seen in Alexandreanu’s studies on Ni–16Cr–9Fe in 360 °C primary water [30].
在研究易开裂晶界的特性时，已有研究注意到应力在 IASCC 中的重要性[8]、[9]、[17]、[29]。研究发现，垂直于拉伸轴且邻近低 Schmid 因子晶粒的晶界，由于施加的拉伸应力产生的 resolved 法向应力最高，表现出高开裂倾向。如果晶界剪切滑移控制着裂纹萌生，预计拉伸轴附近 45°的晶界将更容易开裂，正如 Alexandreanu 对 360°C 一次水中的 Ni–16Cr–9Fe 研究[30]所示。
Strain in continuous channels is accommodated by either by the transfer of slip to the neighboring grain or by activating a dislocation source in the neighboring grain. In a similar manner, disc. w/GB slip accommodate strain by inducing slip in the grain boundary. In the case of discontinuous channels, the dislocations pile-up at the grain boundary, causing an increase of stress locally. A discontinuous channel in a sample strained to 2.5% macroscopic strain may contain up to ∼200% strain within the channel, based on the displacement measurements shown in Fig. 3, and assuming a channel width of 100 nm. This is equivalent to almost 800 dislocations in a channel. In work by Evrard and Sauzay [15], similar channels (100 nm wide) were modeled using finite element analysis, as this was believed to describe the deformation over multiple parallel slip planes better than simple dislocation pile-up models which assume a single slip plane. It was found that the dislocation channels increased the stress by a factor of ∼4 over the applied stress determined on the boundary.
连续通道中的应变通过滑移转移至邻近晶粒或激活邻近晶粒中的位错源来适应。类似地，与晶界滑移（disc. w/GB slip）适应应变的方式是通过在晶界中诱导滑移。在非连续通道的情况下，位错在晶界处堆积，导致局部应力增加。根据图 3 所示的位移测量结果，并假设通道宽度为 100 nm，在宏观应变为 2.5%的样品中，一个非连续通道内的应变可能高达约 200%。这相当于通道中几乎有 800 个位错。在 Evrard 和 Sauzay[15]的研究中，使用有限元分析对类似的通道（宽度为 100 nm）进行了建模，因为人们认为这比假设单个滑移面的简单位错堆积模型能更好地描述多个平行滑移面的变形。研究发现，位错通道使边界上确定的应力增加了约 4 倍。
High resolution EBSD was used to make direct measurements of the local elastic strain field, from which the local stress field was determined according Hooke’s law. As shown in Fig. 13, the ledges created by out-of-plane dislocation channel and grain boundary slip interfered with the pattern collection by blocking the path from the backscattered beam to the detector. This created significant difficulty in collecting the data in the region close to the DC–GB intersection. As a result, no data was collected at less than 100 nm from the DC–GB intersection. It should also be noted that the values of the stress were higher than expected. This is believed to be due to surface damage caused during the irradiation and subsequent straining that distorted the crystal lattice near the surface and caused Kikuchi bands to be slightly blurred. As such, the value of these numbers is in the relative stress levels of the different types of DC–GB intersections. The results do show that stress tends to be elevated at the GB, and that there are cases where stress at discontinuous DC–GB intersections reach higher levels (in one case, nearly double) than those observed at continuous or disc. w/GB slip (Fig. 8) intersections.
使用高分辨率 EBSD 直接测量局部弹性应变场，根据胡克定律确定局部应力场。如图 13 所示，面外位错通道和晶界滑移形成的台阶干扰了图案收集，它们阻挡了背散射束到达探测器的路径。这给靠近直流-晶界交点区域的 数据收集带来了显著困难。因此，在距离直流-晶界交点小于 100 纳米处没有收集到数据。还应注意，应力值高于预期。这被认为是由于辐照期间造成的表面损伤以及随后的应变导致表面附近的晶格发生畸变，从而使 Kikuchi 带略微模糊。因此，这些数值的值处于不同类型直流-晶界交点的相对应力水平。结果表明，应力倾向于在晶界处升高，并且存在应力在非连续直流-晶界交点处达到更高水平（在一种情况下，几乎是连续或盘状交点处观测到水平的近两倍）的情况。 与 GB 滑移（图 8）的交点。
  1. Download: Download high-res image (231KB)
    下载：下载高分辨率图像（231KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 13. EBSD map of a discontinuous DC–GB intersection. The uncharacterized (white) area along the length of the channel was where the EBSD pattern was partially blocked by the emerging channel.
图 13. 不连续的 DC-GB 交点的 EBSD 图谱。通道长度上的未表征区域（白色）是 EBSD 图案被新出现的通道部分阻挡的地方。

The observations of high stress due to discontinuous DC–GB intersections are supported by atomistic modeling that has examined the interactions of dislocations with grain boundaries [16], as well as in-situ TEM straining experiments [31]. Farkas et al. [17] modeled austenitic fcc metal using the Mishin nickel potential [32]. Polycrystalline digital samples were constructed, with a diameter of up to 50 nm. The simulated grain structure was based on actual grain orientations from experimental samples, as determined using EBSD. Uniaxial strain-controlled virtual tensile tests at constant temperature were performed using on the large-scale atomic/molecular massively parallel simulator (LAMMPS) code [33] and the Virginia Tech supercomputer infrastructure, at a strain rate of 3 × 108 s−1.
由于不连续的 DC-GB 交点引起的高应力观察结果得到了原子尺度建模的支持，该建模研究了位错与晶界的相互作用[16]，以及原位 TEM 拉伸实验[31]。Farkas 等人[17]使用 Mishin 镍势[32]对奥氏体面心立方金属进行了建模。构建了多晶数字样品，直径高达 50 nm。模拟的晶粒结构基于实验样品的实际晶粒取向，该取向通过 EBSD 确定。使用大规模原子/分子大规模并行模拟器（LAMMPS）代码[33]和弗吉尼亚理工大学超级计算基础设施，在 3 × 10^0 s^-1 的应变速率下进行了恒温单轴应变控制虚拟拉伸试验。
Dislocations were observed [16] to impinge on a boundary, and the buildup of stress in the boundary region was followed quantitatively. The stress increased locally to values as high as 9 GPa, decreasing when stress was relieved by the dislocations moving into the boundary and eventually transmitting across the boundary into the adjacent grain. Areas where dislocations were transmitted across the boundary generally exhibited lower stress levels than areas where transmission did not occur. Furthermore, if the dislocations were not transmitted into the neighboring grains, the high stresses typically resulted in crack initiation. This was particularly true near a TJ, which acted as stress concentrators and where a number of cracks were found to initiate.
观察到位错[16]撞击边界，并定量地追踪了边界区域的应力累积。应力局部增加到高达 9 GPa，当位错移动到边界并释放应力时，应力会下降，最终通过边界传递到相邻晶粒。位错通过边界的区域通常比未发生传递的区域表现出较低的应力水平。此外，如果位错没有传递到相邻晶粒，高应力通常会导致裂纹萌生。这在 TJ 附近尤其明显，TJ 作为应力集中点，发现这里萌生了许多裂纹。
These results show that despite large differences in the grain size and strain rate, stress localization in this type of microstructure can lead to crack initiation when dislocations are not transmitted across a grain boundary. The atomistic simulations also point out to the importance of triple junctions in crack initiation. Quantitative comparisons have not been attempted due to the above mentioned differences in length and time scales, but it is certainly possible that stress buildup at the boundaries and triple junctions plays a critical role across length and time scales.
这些结果表明，尽管晶粒尺寸和应变率存在很大差异，但在这种微观结构中，当位错未跨过晶界传递时，应力集中会导致裂纹萌生。原子模拟也指出了三重结点在裂纹萌生中的重要性。由于上述长度和时间尺度的差异，尚未进行定量比较，但可以肯定的是，在边界和三重结点处的应力积累在长度和时间尺度上都起着关键作用。
Kamaya et al. [34], using finite element modeling, have also observed the high levels of stress at TJ, due to the deformation constraints caused by the adjacent grains. As the deformation constraints in an irradiated sample are significantly higher than that of unirradiated one, it is expected that the relative level of stress at the TJ will be even high in the irradiated steel. It is believed that the cracks that occurred at a TJ in the experiments (Table 3) were a result of this elevated stress, similar to cracks that form at discontinuous DC–GB intersections. The simulations clearly showed that un-accommodated and un-transmitted slip arriving at the boundary created very high local stresses. Even though the simulations were carried out at the extremely high strain rates necessitated by the molecular dynamics technique, the results of high stress buildup and subsequent crack initiation provide clear support for the idea that it is these areas of high stresses at discontinuous DC–GB intersections that result in crack initiation.
Kamaya 等人[34]通过有限元建模也观察到，由于相邻晶粒的变形约束，TJ 处存在高应力。由于辐照样品中的变形约束显著高于未辐照样品，预计辐照钢中 TJ 处的相对应力水平会更高。实验中在 TJ 处出现的裂纹（表 3）被认为是这种高应力的结果，类似于在非连续直流晶界交点处形成的裂纹。模拟清晰地表明，未协调和未传递的滑移到达边界时会产生非常高的局部应力。尽管模拟是在分子动力学技术所要求的极高水平应变率下进行的，但高应力累积和随后的裂纹萌生结果为这一观点提供了明确的支持，即正是这些非连续直流晶界交点处的高应力导致了裂纹的萌生。
Additional confirmation on importance of stress at DC–GB intersections is provided by in-situ TEM straining experiments performed on austenitic stainless steel samples irradiated in-situ at in the IVEM-Tandem microscope at Argonne National Laboratory [35], [36] at room temperature with 1 MeV Kr ions to a doses between 0.1 and 1 dpa [31], [37]. The in-situ straining revealed dislocation pile-ups in an irradiated 304 stainless steel [31]. When the pile-up could not be accommodated by slip transfer, stress was eventually relieved by crack nucleation. Similar to the atomistic model results, the TEM experiments performed by Cui et al. [37], revealed over 60 dislocations in a pile-up at the grain boundary before stress was relieved through slip transmission into the adjacent grain.
在直流-晶界交点处应力重要性的进一步确认是通过在阿贡国家实验室的 IVEM-Tandem 显微镜中，对室温下用 1 MeV 氪离子辐照的奥氏体不锈钢样品进行的原位 TEM 应变实验[35]，[36]，辐照剂量在 0.1 至 1 dpa 之间[31]，[37]。原位应变实验揭示了辐照 304 不锈钢中的位错堆积[31]。当堆积无法通过滑移转移来适应时，应力最终通过裂纹成核而释放。与原子模型结果相似，崔等人[37]进行的 TEM 实验也显示，在通过滑移传递到相邻晶粒释放应力之前，晶界处的堆积中有多于 60 个位错。
The in-situ TEM studies also provide a possible explanation for the cracking at a TJ. Cui et al. [37] reported that dislocations within a channel, once incorporated in the grain boundary, caused dislocation emission from the vicinity of a triple junction. If dislocation propagation away from the sources in the vicinity of the TJ was prohibited by the presence of the irradiation defects, the response may be for the triple junction to crack.
原位 TEM 研究也为 TJ 处的开裂提供了一种可能的解释。崔等人[37]报道，当通道内的位错一旦被纳入晶界时，就会导致位错从三叉结附近发出。如果由于辐照缺陷的存在，位错在 TJ 附近源处的传播被禁止，那么三叉结可能会开裂。
The final categories of cracks to address are those that formed where no channels intersections were previously observed. As noted earlier, straining in BWR NWC resulted in the creation of additional channels, an example of which is shown in Fig. 10a, where a dislocation channel is observed on the left side of the SEM image only after straining in BWR NWC to 7.2%. Thus, they can only be characterized as either continuous or discontinuous. Fig. 14 shows the characterization of the slip (continuous or discontinuous) for the 25 cracks in the “Unknown DC–GB intersection type” category. As expected, and consistent with results in Fig. 11, cracking propensity at discontinuous (discontinuous or disc. w/GB slip) boundaries was statistically higher than at continuous boundaries.
需要处理的最终裂缝类别是那些之前未观察到通道交叉的地方。如前所述，BWR NWC 中的应变导致了额外通道的形成，图 10a 展示了其中一个例子，其中仅在 BWR NWC 应变至 7.2%后，在 SEM 图像左侧观察到位错通道。因此，它们只能被归类为连续或非连续。图 14 显示了“未知 DC-GB 交叉类型”类别中 25 条裂缝的滑移（连续或非连续）特征。正如预期的那样，并且与图 11 的结果一致，非连续（非连续或带 GB 滑移的圆盘状）边界处的开裂倾向在统计上高于连续边界。
  1. Download: Download high-res image (88KB)
    下载：下载高分辨率图像（88KB）
  2. Download: Download full-size image
    下载：下载完整尺寸图像

Fig. 14. Number of crack DC–GB intersection in the classification “Unknown DC–GB intersection type” which are continuous, or discontinuous, based on SEM observations.
图 14。“未知 DC-GB 交叉类型”分类中，基于 SEM 观察，连续或非连续裂缝 DC-GB 交叉的数量。

While the data solidly supports localized stress at discontinuous dislocation channel-grain boundary intersections as the key factor in IASCC, it may be noted that only 18% of the most susceptible locations (discontinuous DC–GB intersections) crack. If a threshold in stress is required for cracking, then not all boundaries or DC–GB intersections will experience the same local stress for a given applied stress. Obtaining a quantitative measure of the local stress will be required to determine the value of this threshold stress. Further, variability in local grain boundary structure or chemistry may also exert an influence on the cracking propensity. Thus, full understanding of the conditions for IASCC requires a quantitative characterization of the local stress state and the influence of other factors. These results also examine only crack initiation, and additional work is needed to understand and characterize the crack propagation. At that point, the tools will be in place to develop predictive capabilities as well as mitigation strategies.
虽然数据有力地支持了局部应力在非连续位错通道-晶界交点处是 IASCC 的关键因素，但值得注意的是，只有 18%的最易发生断裂的位置（非连续 DC-GB 交点）发生了断裂。如果断裂需要应力达到某个阈值，那么在给定的外加应力下，并非所有晶界或 DC-GB 交点都会经历相同的局部应力。为了确定该阈值应力的值，需要获得局部应力的定量测量。此外，局部晶界结构或化学成分的变化也可能影响断裂倾向。因此，全面理解 IASCC 的条件需要对局部应力状态进行定量表征，并考虑其他因素的影响。这些结果仅考察了裂纹萌生，还需要进一步研究以理解和表征裂纹扩展。届时，将具备开发预测能力和缓解策略的工具。

5. Conclusions  5. 结论

Partitioning localized deformation into three fundamental modes, based on how the channels interact with the grain boundary, revealed that high local stress occurring as a result of dislocation pile-ups in discontinuous dislocation channels at the intersection with grain boundaries is the likely cause of IASCC in irradiated steel. It is the inability of these channels to relieve stress by either slip in the adjacent grain or in the grain boundary, which results in high local stresses and an increased susceptibility to IGSCC initiation. Where more slip has occurred in a discontinuous channel, a larger number of dislocations will pile up, exerting higher stress on the grain boundary. Channels that induce slip across or within a grain boundary relieve the local stresses and thus, show less susceptibility to cracking. The observation that discontinuous channel-grain boundary intersections are locations of high stress is supported by both in-situ straining of irradiated steels in the TEM and atomistic simulation of dislocation–grain boundary interaction in which channels that were unable to transfer strain underwent cracking.
基于通道与晶界相互作用的特性，将局部变形划分为三种基本模式，揭示出在晶界交点处不连续位错通道中位错堆积导致的高局部应力是辐照钢中应力腐蚀裂纹起始的潜在原因。这些通道无法通过邻近晶粒的滑移或晶界滑移来释放应力，导致高局部应力并增加了对辐照诱应力腐蚀裂纹起始的敏感性。在连续通道中滑移更多的地方，位错堆积数量更多，对晶界的应力作用更大。那些诱导晶界跨越或晶界内滑移的通道能够缓解局部应力，因此表现出较低的裂纹敏感性。不连续通道-晶界交点是高应力区域的观察结果，这一发现得到了透射电子显微镜中辐照钢原位变形实验和位错-晶界相互作用原子模拟的支持，后者显示无法转移应力的通道发生了开裂。

Acknowledgements  致谢

The authors acknowledge Alexander Flick for his assistance conducting the constant extension rate tensile tests in the Irradiated Materials Testing Laboratory at the University of Michigan, as well as the staff of the Michigan Ion Beam Laboratory, Ovidiu Toader and Fabian Naab for their assistance in performing the proton irradiations. The electron microscopy was accomplished at the Electron Microscopy Center at Argonne National Laboratory (a U.S. Department of Energy Office of Science Laboratory operated under Contract No. DE-AC02-06CH11357 by U Chicago Argonne, LLC). The NSF IRD program is also acknowledged for support. The authors acknowledge Advanced Research Computing at Virginia Tech for providing computational resources and technical support that have contributed to the results reported within this paper. URL: http://www.arc.vt.edu. Research (experimental work performed by McMurtrey, Was, Robertson and Cui and computational work performed by Farkas) was supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering under Award DE-FG02-08ER46525.
作者感谢亚历山大·弗里克在密歇根大学辐照材料测试实验室进行恒定伸长率拉伸试验时的协助，以及密歇根离子束实验室的员工、奥维迪乌·托德和法比安·纳布在执行质子辐照时的帮助。电子显微镜分析在阿贡国家实验室的电子显微镜中心完成（该中心是美国能源部科学办公室运营的实验室，由芝加哥阿贡公司，有限合伙公司根据合同号 DE-AC02-06CH11357 运营）。作者也感谢国家科学基金会创新研究计划的支持。作者感谢弗吉尼亚理工大学的先进研究计算中心提供了计算资源和技术支持，这些资源有助于本论文报告的结果。URL: http://www.arc.vt.edu。研究（实验工作由麦克马特里、瓦斯、罗伯特森和崔完成，计算工作由法尔卡斯完成）由美国能源部，基础能源科学办公室，材料科学与工程部门在 DE-FG02-08ER46525 奖项下支持。

References

Cited by (71)

  • Element-resolved electrochemical analysis of the passivity of additively manufactured stainless steel 316L
    对增材制造不锈钢 316L 的钝化过程的元素解析电化学分析

    2021, Corrosion Science

  • Characteristics of hydrogen embrittlement in high-pH stress corrosion cracking of X100 pipeline steel in carbonate/ bicarbonate solution
    碳酸盐/碳酸氢盐溶液中 X100 管线钢在高温 pH 应力腐蚀开裂中的氢脆特征

    2020, Construction and Building Materials

  • The effect of grain boundary structure on the intergranular degradation behavior of solution annealed alloy 690 in high temperature, hydrogenated water
    晶粒边界结构对溶液退火合金 690 在高温、氢化水中晶间降解行为的影响

    2020, Acta Materialia

  • Nanoindentation and in situ microcompression in different dose regimes of proton beam irradiated 304 SS
    不同剂量区间的质子束辐照 304 不锈钢的纳米压痕和原位微压缩

    2017, Journal of Nuclear Materials

  • Motivation for utilizing new high-performance advanced materials in nuclear energy systems
    利用新型高性能先进材料在核能系统中的驱动力

    2016, Current Opinion in Solid State and Materials Science
    2016，《固体与材料科学前沿》

  • Intergranular stress distributions in polycrystalline aggregates of irradiated stainless steel
    辐照不锈钢多晶聚集体中的晶间应力分布

    2016, Journal of Nuclear Materials
    2016，《核材料杂志》
    Citation Excerpt :  引用摘录：

    Numerous studies have been conducted in the recent past on IGSCC of (ion-irradiated) austenitic stainless steel, mainly with applied stress well above yield stress. It has been shown (see, e.g., [38]) that dislocation channelling (or clear bands) resulting in strain localization at the grain scale may be a factor controlling IGSCC of irradiated stainless steel. The constitutive equations used in this study do not display instabilities and subsequent strong strain localization at the grain scale, and therefore the results and methodology proposed in this study would probably show limitations when applied to highly strained materials.
    近年来，针对（离子辐照）奥氏体不锈钢的应力腐蚀开裂（IGSCC）进行了大量研究，主要是在远高于屈服应力的外加应力条件下。研究表明（例如参见[38]），导致晶粒尺度应变局部化的位错通道（或清晰带）可能是控制辐照不锈钢 IGSCC 的一个因素。本研究使用的本构方程不显示晶粒尺度应变局部化后的不稳定性，因此本研究提出的结果和方法在应用于高度应变材料时可能会存在局限性。

View all citing articles on Scopus
在 Scopus 上查看所有引用文章
1
Address: 1906 Cooley Building, 2355 Bonisteel Blvd, Ann Arbor, MI 48106, United States. Tel.: +1 734 936 0266; fax: +1 734 763 4540.
地址：美国密歇根州安娜堡市博尼斯蒂尔大道 2355 号库利大楼 1906 室。电话：+1 734 936 0266；传真：+1 734 763 4540。
View Abstract  查看摘要
Copyright © 2015 Elsevier Ltd. All rights reserved.
版权所有 © 2015 Elsevier Ltd. 保留所有权利。