An Atomistic-to-Microscale Characterization of the Kink-controlled Dislocation Dynamics in BCC Metals through Finite-Temperature Coarse-grained Atomistic Simulations
An Atomistic-to-Microscale Characterization of the Kink-controlled Dislocation Dynamics in BCC Metals through Finite-Temperature Coarse-grained Atomistic Simulations
复制标题
DOI：
10.1016/j.actamat.2023.119440
复制
发表时间：
2023-10
期刊：
影响因子：
9.4
通讯作者：
Rigelesaiyin Ji;Thanh Phan;Youping Chen;David McDowell;Liming Xiong
中科院分区：
文献类型：
--
作者：
Rigelesaiyin Ji;Thanh Phan;Youping Chen;David McDowell;Liming Xiong
以bcc钨(W)为模型材料，利用有限温度粗晶（FT-CG）原子模拟表征了60 nm ~ 1 μm位错线上的扭结动力学与温度和应力的关系。这项工作的主要新颖之处在于容纳了主要的突出方面，即μm长的位错线的运动，原子尺度的扭结动力学和声子动力学的全谱，所有这些都在一个FT-CG模型中。在分子动力学（MD）计算成本的一小部分，FT-CG模拟预测：(a) W的位错诱导声子态密度（PDoS）的退化；(b)扭结引起的位错核心结构由“软”（非平面、致密）向“硬”（平面、分裂）转变；(c)流动应力的温度依赖性从线张力（LT）状态过渡到弹性相互作用（EI）状态。结果表明：(1)扭结激活应力σf不仅与温度T有关，而且对位错线长度L敏感。对于μm位错，σf接近实验结果，但对于nm位错，σf不接近实验结果；(2)随着ofT的增大，μm位错试样的σ消减量显著大于nm位错试样；(3)基于“温度跳跃试验”的FT-CG模拟数据，表征了扭结激活焓ΔH的相关性。对于长度为几十nm的位错，它可以高达~ 3 eV，但当长度为0.3 μm或更长的位错时，它降低到实验上可比较的~ 1.5 eV水平。这表明更长的位错更容易激活扭结。这种位错线长度依赖ofΔHcan在更低的外加应力下进一步放大；(4)熵扭结激活势垒ΔHT与t成线性比例。然而，theΔHT-Trelation的斜率在纳米尺度的MD模拟中会被大大低估，但可以与lis ~ 0.3 μm或更长时的实验结果相媲美。这些发现突出了纳米尺度MD模型在模拟扭结控制的位错动力学方面的局限性。在这里获得的知识可以支持迁移率定律的发展，这些迁移率定律将应力、温度和线长依赖性都纳入一个公式，以理解bcc金属和其他高佩尔应力合金的塑性。
Adopting bcc tungsten (W) as a model material, we characterize the temperature and stress dependence of kink dynamics on a dislocation line with lengthLranging from 60 nm to 1 μm using finite-temperature coarse-grained (FT-CG) atomistic simulations. The main novelty of this work is to accommodate major salient aspects, namely the motion of μm-long dislocation lines, the atomic-scale kink dynamics, and the full spectrum of phonon dynamics, all in one single FT-CG model. At a fraction of the cost of molecular dynamics (MD) calculations, the FT-CG simulation predicts:(a)a dislocation-induced degeneration of the phonon density of states (PDoS) of W;(b)the kink-induced dislocation core structure transition from a “soft” (non-planar, compact) to a “hard” configuration (planar, split); and(c)the crossover from the line tension (LT) regime to the elastic interaction (EI) regime in the temperature dependence of the flow stress. Several findings arise from the simulations:(1)the kink activation stress,σf, not only depends on the temperature,T, but also exhibits a sensitivity to the dislocation line length,L. For μm-long dislocations, it approaches experimental results butσffor kink activation on nm-long dislocations does not;(2)upon an increase ofT, theσfreduction for the sample containing μm-long dislocations is significantly larger than that for the one with nm-long dislocations;(3)based on data extracted from FT-CG simulations of “temperature jump tests”, thel- dependence of the kink activation enthalpy,ΔH, is characterized. It can be as high as ∼3 eV for a dislocation with a length of tens of nm but reduces to an experimentally comparable level of ∼1.5 eV whenLis 0.3 μm or longer. This suggests an easier kink activation on a longer dislocation. Such a dislocation line length dependence ofΔHcan be further amplified at an even lower applied stress;(4)the entropic kink activation barrier, ΔHT,is linearly proportional toT. The slope of theΔHT–Trelation, however, will be largely underestimated in nanoscale MD simulations, but can be comparable with that from experiments whenLis ∼ 0.3 μm or longer. These findings highlight the limitations of nanoscale MD models in simulating kink-controlled dislocation dynamics. The knowledge gained here can support the development of mobility laws that incorporate the stress-, temperature-, and line length-dependence all into one formulation for understanding plasticity in bcc metals and other high-Peierls-stress alloys.