Timothy C. Germann

MOLECULAR DYNAMICS COMES OF AGE: 320 BILLION ATOM SIMULATION ON BlueGene/L

As computational power is increasing, molecular dynamics simulations are becoming more important in materials science, chemistry, physics, and other fields of science. We demonstrate weak and strong scaling of our classical molecular dynamics code SPaSM on Livermores BlueGene/L architecture containing 131 072 IBM PowerPC440 processors. A maximum of 320 billion atoms have been simulated in double precision, corresponding to a cubic piece of solid copper with an edge length of 1.56 μm.

Shock-induced spall in solid and liquid Cu at extreme strain rates

We investigate spallation in solid and liquid Cu at high strain rates induced by planar shock loading with classical molecular dynamics. Shock simulations are performed at different initial temperatures and shock stresses but similar strain rates (e∼1010–1011s−1). The anisotropy in spall strength (σsp) is explored for five crystallographic orientations, ⟨100⟩, ⟨110⟩, ⟨111⟩, ⟨114⟩, and ⟨123⟩. For liquid, we examine shock- and release-induced melts as well as premelted Cu. The acoustic method for deducing σsp and e is a reasonable first-order approximation. The anisotropy in σsp is pronounced for weak shocks and decreases for stronger shocks. Voids are nucleated at defective sites in a solid. For weak solid shocks, spallation occurs without tensile melting; for stronger shocks or if the temperature right before spallation (Tsp) is sufficiently high, spallation may be accompanied or preceded by partial melting. Tsp appears to have a dominant effect on spallation for the narrow range of e studied here. σsp...

Atomistic simulations of shock-induced alloying reactions in Ni∕Al nanolaminates

We employ molecular dynamics simulations with a first principles-based many body potential to characterize the exothermic alloying reactions of nanostructured Ni/Al multilayers induced by shock loading. We introduce a novel technique that captures both the initial shock transit as well as the subsequent longer-time-scale Ni3Al alloy formation. Initially, the softer Al layers are shock heated to a higher temperature than the harder Ni layers as a result of a series of shock reflections from the impedance-mismatched interfaces. Once initiated, the highly exothermic alloying reactions can propagate in a self-sustained manner by mass and thermal diffusion. We also characterize the role of voids on the initiation of alloying. The interaction of the shock wave with the voids leads not only to significant local heating (hot spots) but also directly aids the intermixing between Al and Ni; both of these phenomena contribute to a significant acceleration of the alloying reactions.

Trillion-atom molecular dynamics becomes a reality

By utilizing the molecular dynamics code SPaSM on Livermores BlueGene/L architecture, consisting of 212 992 IBM PowerPC440 700 MHz processors, a molecular dynamics simulation was run with one trillion atoms. To demonstrate the practicality and future potential of such ultra large-scale simulations, the onset of the mechanical shear instability occurring in a system of Lennard-Jones particles arranged in a simple cubic lattice was simulated. The evolution of the instability was analyzed on-the-fly using the in-house developed massively parallel graphical object-rendering code MD_render.

The importance of fluctuations in fluid mixing

A ubiquitous example of fluid mixing is the Rayleigh–Taylor instability, in which a heavy fluid initially sits atop a light fluid in a gravitational field. The subsequent development of the unstable interface between the two fluids is marked by several stages. At first, each interface mode grows exponentially with time before transitioning to a nonlinear regime characterized by more complex hydrodynamic mixing. Unfortunately, traditional continuum modeling of this process has generally been in poor agreement with experiment. Here, we indicate that the natural, random fluctuations of the flow field present in any fluid, which are neglected in continuum models, can lead to qualitatively and quantitatively better agreement with experiment. We performed billion-particle atomistic simulations and magnetic levitation experiments with unprecedented control of initial interface conditions. A comparison between our simulations and experiments reveals good agreement in terms of the growth rate of the mixing front as well as the new observation of droplet breakup at later times. These results improve our understanding of many fluid processes, including interface phenomena that occur, for example, in supernovae, the detachment of droplets from a faucet, and ink jet printing. Such instabilities are also relevant to the possible energy source of inertial confinement fusion, in which a millimeter-sized capsule is imploded to initiate nuclear fusion reactions between deuterium and tritium. Our results suggest that the applicability of continuum models would be greatly enhanced by explicitly including the effects of random fluctuations.

LARGE-SCALE MOLECULAR-DYNAMICS SIMULATION OF 19 BILLION PARTICLES

We have performed parallel large-scale molecular-dynamics simulations on the QSC-machine at Los Alamos. The good scalability of the SPaSM code is demonstrated together with its capability of efficient data analysis for enormous system sizes up to 19 000 416 964 particles. Furthermore, we introduce a newly-developed graphics package that renders in a very efficient parallel way a huge number of spheres necessary for the visualization of atomistic simulations. These abilities pave the way for future atomistic large-scale simulations of physical problems with system sizes on the μ-scale.

Anisotropic shock response of columnar nanocrystalline Cu

We perform molecular dynamics simulations to investigate the shock response of idealized hexagonal columnar nanocrystalline Cu, including plasticity, local shear, and spall damage during dynamic compression, release, and tension. Shock loading (one-dimensional strain) is applied along three principal directions of the columnar Cu sample, one longitudinal (along the column axis) and two transverse directions, exhibiting a strong anisotropy in the response to shock loading and release. Grain boundaries (GBs) serve as the nucleation sites for crystal plasticity and voids, due to the GB weakening effect as well as stress and shear concentrations. Stress gradients induce GB sliding which is pronounced for the transverse loading. The flow stress and GB sliding are the lowest but the spall strength is the highest, for longitudinal loading. For the grain size and loading conditions explored, void nucleation occurs at the peak shear deformation sites (GBs, and particularly triple junctions); spall damage is entirely intergranular for the transverse loading, while it may extend into grain interiors for the longitudinal loading. Crystal plasticity assists the void growth at the early stage but the growth is mainly achieved via GB separation at later stages for the transverse loading. Our simulations reveal such deformation mechanisms as GB sliding, stress, and shear concentration, GB-initiated crystal plasticity, and GB separation in nanocrystalline solids under shock wave loading.

Shock wave loading and spallation of copper bicrystals with asymmetric Σ3〈110〉tilt grain boundaries

We investigate the effect of asymmetric grain boundaries (GBs) on the shock response of Cu bicrystals with molecular dynamics simulations. We choose a representative Σ3〈110〉tilt GB type, (110)_1/(114)_2, and a grain size of about 15 nm. The shock loading directions lie on the GB plane and are along [001] and [221] for the two constituent crystals. The bicrystal is characterized in terms of local structure, shear strain, displacement, stress and temperature during shock compression, and subsequent release and tension. The shock response of the bicrystal manifests pronounced deviation from planar loading as well as strong stress and strain concentrations, due to GBs and the strong anisotropy in elasticity and plasticity. We explore incipient to full spallation. Voids nucleate either at GBs or on GB-initiated shear planes, and the spall damage also depends on grain orientation.

Twinning in bcc metals under shock loading: a challenge to empirical potentials

Using density functional theory (DFT), we found that high pressures intrinsically favor twinning in niobium by reducing the thickness of a stable twin. Five empirical interatomic potentials for niobium were considered in molecular dynamics (MD) shock simulations. The results show that two potentials exhibit the experimentally observed twinning behavior. Comparing with DFT under high pressure, we found that these two potentials are capable of reproducing the generalized stacking fault (GSF) curve, but the others predict several artificial metastable states along the GSF curve resulting in an artificial structural transformation.

Spall damage of copper under supported and decaying shock loading

We investigate spall damage of single crystal Cu under supported (square) and decaying (Taylor wave) shock wave loading with molecular dynamics simulations. Varying the target-to-flyer plate thickness ratio R (with target thickness fixed) as well as the impact velocity induces square and Taylor waves with different pulse shapes, durations and strengths, which are well correlated with prespall damage, spall strength, and spall damage. Taylor wave loading results in higher spall strength than the supported shock loading at the same impact velocities, and the spall strength can be similar for both loadings with the same peak free surface velocities, while Taylor wave loading induces less spall damage than square wave loading. Void nucleation is preceded by plasticity and solid-state disordering. Multiple spall events appear to be independent of each other at the early stage of spallation. In applying the acoustic method for deducing the spall strength from the free surface velocity histories, one should cons...