NSF/DMR-BSF: Bridging the gap between atomistic simulations and fracture mechanics

美国国家科学基金会和美国--以色列双国科学基金会(BSF)共同支持一名美国研究人员和一名以色列研究人员之间的合作。美国国家科学基金会材料研究部资助了这一奖项，该奖项支持关于材料断裂的基本性质的研究和教育，这是由原子尺度上发生的过程控制的。该项目试图回答以下问题：为什么原子模拟预测材料中的裂缝与理论预测和实验观察到的不同？PI假设这是因为，到目前为止，由于计算限制，大多数原子模拟都是人为限制的，研究的系统太小。为了超越目前的能力，PI将扩展和应用一种新的计算方法，使断裂实验能够进行真实的模拟。模拟将集中在硅上，并将根据以色列合作者进行的独特的高分辨率动态断裂实验进行验证。选择硅是因为它在许多技术中都很重要，包括微电子设备、微纳机电系统、太阳能电池和仿生设备。这项研究通过阐明如何防止设备发生灾难性故障，有可能对工业产生重大的积极影响。在该项目中开发的所有计算机代码将通过专门的门户网站(qcmethod.org和Openkim.org)免费提供给研究界。该合作项目还将加强美国和以色列之间的研究联系，并让以色列研究生(包括犹太人和阿拉伯人)与他们的美国同行合作，在一个动荡不安的地区创造一个积极的合作范例。TECHNICAL SUMMARY国家科学基金会和美国-以色列双国科学基金会(BSF)共同支持美国研究人员和以色列研究人员之间的这种合作。美国国家科学基金会材料研究部资助了这一奖项，该奖项支持关于材料断裂的基本性质的研究和教育，材料断裂的基本性质受原子尺度上的过程控制。该项目的最终目标是开发一个可预测的多尺度框架，用于模拟断裂现象，明确地考虑各种因素的影响，包括晶体取向、加载速度、温度和先前存在的缺陷。这需要一种多尺度的方法，既包括对裂纹产生的远程应力场的正确处理，也包括涉及裂纹尖端键断裂的原子尺度断裂过程。主要集中在裂纹尖端区域的全原子模拟显示了一种称为“晶格陷阱”的效应，即加载装置必须克服与跨越解理平面的原子键断裂相关的能量障碍。这种过载会导致裂纹开始以较高的初始速度移动。相反，连续介质理论没有解释晶格陷阱，并预测裂纹可以以任何速度开始扩展。以色列合作者最近对硅晶体进行的高分辨率断裂实验与连续介质理论一致，并表明晶格陷阱效应很小，可能会出现较慢的初始裂纹速度。PI假设晶格陷阱是施加在原子模拟上的尺寸和2D约束的伪影。所提出的框架将使在实际加载速率下模拟具有足够大的原子化区域以发生曲率效应的试件的断裂实验成为可能。这项研究通过阐明如何防止设备中的灾难性故障，有可能对工业产生重大的积极影响。在该项目中开发的所有计算机代码将通过专门的门户网站(qcmethod.org和Openkim.org)免费提供给研究界。该合作项目还将加强美国和以色列之间的研究联系，并让以色列研究生(包括犹太人和阿拉伯人)与美国同行接触，在一个动荡不安的地区创造一个积极的合作范例。

NONTECHNICAL SUMMARYThe National Science Foundation and the United States -- Israel Binational Science Foundation (BSF) jointly support this collaboration between a US-based researcher and an Israel-based researcher. The NSF Division of Materials Research funds this award, which supports research and education on the fundamental nature of fracture in materials, which is governed by processes that occur at the atomic scale. The project seeks to answer the following question: why do atomistic simulations predict cracks in materials with different characteristics than those predicted by theory and observed in experiments? The PIs postulate that this is because, up until now and due to computational limitations, most atomistic simulations are artificially constrained and study systems that are too small. To go beyond current capabilities, the PIs will extend and apply a novel computational approach that will enable realistic simulations of fracture experiments. The simulations will focus on silicon, and will be validated against unique high-resolution dynamic fracture experiments performed by the Israeli collaborator. Silicon was chosen due its importance in many technologies including microelectronics devices, micro- and nano-electro-mechanical systems, solar cells, and bio-inspired devices. This research has potential for a significant positive impact on industry by elucidating how to prevent catastrophic failure in devices. All computer codes developed in this project will be made freely available to the research community via dedicated web portals (qcmethod.org and openkim.org). The collaborative project will also strengthen research ties between the US and Israel, and engage Israeli graduate students (both Jewish and Arab) with their US counterparts to create a positive example of collaboration in a troubled region.TECHNICAL SUMMARYThe National Science Foundation and the United States -- Israel Binational Science Foundation (BSF) jointly support this collaboration between a US-based researcher and an Israel-based researcher. The NSF Division of Materials Research funds this award, which supports research and education on the fundamental nature of fracture in materials, which is governed by processes that occur at the atomic scale.The ultimate aim of this project is to develop a predictive multiscale framework for simulating fracture phenomena that explicitly account for the effect of a variety of factors including crystallographic orientation, loading rate, temperature, and preexisting defects. This requires a multiscale approach that includes both the correct treatment of the long-range stress field generated by a crack, and the atomic-scale fracture processes that involve bond breaking at the crack tip. Fully-atomistic simulations that focus primarily on the crack tip region exhibit an effect called "lattice trapping" whereby the loading device has to overcome an energy barrier associated with breaking atomic bonds that span the cleavage plane. This overloading causes cracks to begin to move at high initial speed. In contrast, continuum theory does not account for lattice trapping and predicts that a crack can begin to propagate at any speed. Recent high-resolution fracture experiments on silicon crystals performed by the Israeli collaborator agree with continuum theory, and show that the lattice trapping effect is small and that slow initial crack speeds are possible. The PIs posit that lattice trapping is an artifact of the size and 2D constraints imposed on the atomistic simulations. The proposed framework will make it possible to simulate fracture experiments at realistic loading rates for specimens with a large enough atomistic region for curvature effects to occur.This research has potential for a significant positive impact on industry by elucidating how to prevent catastrophic failure in devices. All computer codes developed in this project will be made freely available to the research community via dedicated web portals (qcmethod.org and openkim.org). The collaborative project will also strengthen research ties between the US and Israel, and engage Israeli graduate students (both Jewish and Arab) with their US counterparts to create a positive example of collaboration in a troubled region.