Atomistic simulation is a useful method for studying material science phenomena. Examination of the state of a simulated material and the determination of its mechanical properties is accomplished by inspecting the stress field within the material. However, stress is inherently a continuum concept and has been proven difficult to define in a physically reasonable manner at the atomic scale. In this paper, an expression for continuum mechanical stress in atomistic systems derived by Hardy is compared with the expression for atomic stress taken from the virial theorem. Hardy's stress expression is evaluated at a fixed spatial point and uses a localization function to dictate how nearby atoms contribute to the stress at that point; thereby performing a local spatial averaging. For systems subjected to deformation, finite temperature, or both, the Hardy description of stress as a function of increasing characteristic volume displays a quicker convergence to values expected from continuum theory than volume averages of the local virial stress. Results are presented on extending Hardy's spatial averaging technique to include temporal averaging for finite temperature systems. Finally, the behaviour of Hardy's expression near a free surface is examined, and is found to be consistent with the mechanical definition for stress.

Structural reorientations in metallic fcc nanowires are controlled by a combination of size, thermal energy, and the type of defects formed during inelastic deformation. By utilizing atomistic simulations, we show that certain fcc nanowires can exhibit both shape memory and pseudoelastic behavior. We also show that the formation of defect-free twins, a process related to the material stacking fault energy, nanometer size scale, and surface stresses is the mechanism that controls the ability of fcc nanowires of different materials to show a reversible transition between two crystal orientations during loading and thus shape memory and pseudoelasticity.

Atomistic simulation is used to examine nanoindentation of a Au(111) crystal both near and far from a surface step. While the load needed to nucleate dislocations decreases significantly when indenting close to the step, the extent of the step's influence is not as great as seen experimentally. This behavior is explained by measuring the contact area from the simulation data. A new metric, the slip vector, shows material slip coinciding with the <112> directions of a lowest unstable stacking fault barrier. The slip vector is used to calculate an atomic critical resolved shear stress, which is shown to be a good dislocation nucleation criterion.

No abstract

No abstract

We present numerical simulations of gold nanowires under tensile loading at various strain rates and wire sizes at room temperature. The simulations were performed using molecular dynamics modeling the gold nanowires using various forms of the embedded-atom method, and concentrated on investigating the yield and fracture properties of the nanowires. It is clearly demonstrated that the accurate modeling of stacking fault and surface energies is critical in capturing the fundamental deformation behavior of gold nanowires. By doing so, phenomena which have been observed both experimentally and numerically in first-principles calculations, such as the formation of atom-thick chains ͑ATCs͒ prior to fracture, zigzag, helical rotational motion of atoms within the ATCs, structural reorientation of the ATCs to a hexagonal crystal structure, and ͑111͒ faceting of the nanowire in the yielded neck region by the ATCs, are accurately captured.

Atomistic calculations for the 112 -generalized stacking fault (GSF) energy curve are performed for various embedded atom models of FCC metals. Models include those by Voter and Chen; Angelo, Moody and Baskes; Oh and Johnson; Mishin and Farkas; and Ercolessi and Adams. The resulting curves show similar characteristics but vary in their agreement with the experimental estimates of the intrinsic stacking fault energy, sf , and with density functional theory (DFT) calculations of the GSF curve. These curves are used to obtain estimates of the unstable stacking fault energy, us , a quantity used in a criterion for dislocation nucleation. Curves for nickel and copper models show the theoretically expected skewed sinusoidal shape; however, several of the aluminium models produce an irregularly shaped GSF curve. Copper and aluminium values for us are underestimates of calculations from DFT, although some of the nickel models produce a value matching one of the available DFT results. Values for sf are either fitted to, or underestimate, the measured results. For use in simulations, the authors recommend using the Voter and Chen potential for copper, and either the Angelo, Moody and Baskes potential or the Voter and Chen potential for nickel. None of the potentials model aluminium well, indicating the need for a more-advanced empirical potential.

No abstract