James N. Glosli

Extending stability beyond CPU millennium: a micron-scale atomistic simulation of Kelvin-Helmholtz instability

We report the computational advances that have enabled the first micron-scale simulation of a Kelvin-Helmholtz (KH) instability using molecular dynamics (MD). The advances are in three key areas for massively parallel computation such as on BlueGene/L (BG/L): fault tolerance, application kernel optimization, and highly efficient parallel I/O. In particular, we have developed novel capabilities for handling hardware parity errors and improving the speed of interatomic force calculations, while achieving near optimal I/O speeds on BG/L, allowing us to achieve excellent scalability and improve overall application performance. As a result we have successfully conducted a 2-billion atom KH simulation amounting to 2.8 CPU-millennia of run time, including a single, continuous simulation run in excess of 1.5 CPU-millennia. We have also conducted 9-billion and 62.5-billion atom KH simulations. The current optimized ddcMD code is benchmarked at 115.1 TFlop/s in our scaling study and 103.9 TFlop/s in a sustained science run, with additional improvements ongoing. These improvements enabled us to run the first MD simulations of micron-scale systems developing the KH instability.

Phase transformations of nanometer size carbon particles in shocked hydrocarbons and explosives

Estimates for the displacement of the phase equilibrium lines for small carbon particles containing from several hundred to several tens of thousands of atoms are made, and an error analysis of the uncertainties in these estimates is derived and evaluated using available experimental data. Hugoniot calculations for methane, benzene, polyethylene, and polybutene, based on a carbon particle surface energy adjusted equation of state, are in better agreement with shock pressure-volume and temperature data than those obtained with a bulk carbon equation of state. The results suggest that carbon particles, of order 103–104 atoms, can exist in the liquid state at lower temperatures than bulk carbon.

Kinetics and thermodynamic behavior of carbon clusters under high pressure and high temperature

Physical processes that govern the growth kinetics of carbon clusters at high pressure and high temperature are: (a) thermodynamics and structural sp ?-to- sp ? bonding) changes and (b) cluster diffusion. Our study on item (a) deals with ab initio and semi-empirical quantum mechanical calculations to examine effects of cluster size on the relative stability of graphite and diamond clusters and the energy barrier between the two. We have also made molecular dynamics simulations using the Brenner bond order potential. Kesults show that the melting line of diamond based on the Brenner potential is reasonable and that the liquid structure changes from mostly sp -bonded carbon chains to mostly sp ?-bonding over a relatively narrow density interval. Our study on item (b) uses the time-dependent clustor size distribution function obtained from the relevant Smoluchowski equations. The resulting surface contribution to the Gibbs free energy of carbon clusters was implemented in a thermochemical code.

Shear-induced anisotropic plastic flow from body-centred-cubic tantalum before melting

There are many structural and optical similarities between a liquid and a plastic flow. Thus, it is non-trivial to distinguish between them at high pressures and temperatures, and a detailed description of the transformation between these phenomena is crucial to our understanding of the melting of metals at high pressures. Here we report a shear-induced, partially disordered viscous plastic flow from body-centred-cubic tantalum under heating before it melts into a liquid. This thermally activated structural transformation produces a unique, one-dimensional structure analogous to a liquid crystal with the rheological characteristics of Bingham plastics. This mechanism is not specific to Ta and is expected to hold more generally for other metals. Remarkably, this transition is fully consistent with the previously reported anomalously low-temperature melting curve and thus offers a plausible resolution to a long-standing controversy about melting of metals under high pressures.

Molecular dynamics study of interfacial electric fields

Abstract Electric fields and potentials of an equilibrated assembly of ions and water molecules adjacent to a charged metal surface are calculated as a function of perpendicular distance z from the surface from data derived from molecular dynamics trajectories. The spatial distributions of atoms or molecules along direction z are found by ensemble averaging of trajectories followed by averaging with a localized function with a well defined length scale. Two methods were used calculate z dependent charge density distributions. In the first, to be called the atom method, the trajectories of charged atoms are averaged. In the second, called the molecule method, a Taylor expansion of charged atom positions relative to molecular centres is performed and the charge density separated into monopole, dipole, quadrupole, octopole, … components. These distributions are used to calculate the electric potential and in one example to study the progressive loss of structure due to water as the length parameter is scanned through the dimension of a water molecule. This latter result provides a link between simulations with detailed atomic modelling of intermolecular interactions and electric potentials derived from Gouy-Chapman theory. Illustrative examples are chosen from simulations of aqueous solutions of simple alkali halide electrolytes next to charged and uncharged flat metal surfaces. The smallest system has one ion and 157 water molecules, the largest 60 ions and 1576 water molecules.

Simulating solidification in metals at high pressure: The drive to petascale computing

We investigate solidification in metal systems ranging in size from 64,000 to 524,288,000 atoms on the IBM BlueGene/L computer at LLNL. Using the newly developed ddcMD code, we achieve performance rates as high as 103 TFlops, with a performance of 101.7 TFlop sustained over a 7 hour run on 131,072 cpus. We demonstrate superb strong and weak scaling. Our calculations are significant as they represent the first atomic-scale model of metal solidification to proceed, without finite size effects, from spontaneous nucleation and growth of solid out of the liquid, through the coalescence phase, and into the onset of coarsening. Thus, our simulations represent the first step towards an atomistic model of nucleation and growth that can directly link atomistic to mesoscopic length scales.

Molecular dynamics simulation of adsorption in electric double layers

Classical molecular dynamics is used to model the structure and dynamics of electric double layers that form when a metal electrode is in contact with an aqueous electrolyte solution containing simple ions and neutral organics. First attention is focused on the distribution of ions and neutrals next to an uncharged electrode. The electric field and potential across the system are calculated. The iodide ion adsorbs from neutral solution and causes a negative shift in the potential of zero charge (PZC) relative to fluoride. Adsorbed benzene causes a small positive shift of the PZC. Benzene desorbs when the metal electrode is charged and the surface electric field causes a layer of localized and oriented water to form next to the electrode. Non contact adsorbed hydrated sodium ions are more effective than contact adsorbed chloride in desorbing benzene.

Towards real-time simulation of cardiac electrophysiology in a human heart at high resolution

We have developed the capability to rapidly simulate cardiac electrophysiological phenomena in a human heart discretised at a resolution comparable with the length of a cardiac myocyte. Previous scientific investigation has generally invoked simplified geometries or coarse-resolution hearts, with simulation duration limited to 10s of heartbeats. Using state-of-the-art high-performance computing techniques coupled with one of the most powerful computers available (the 20 PFlop/s IBM BlueGene/Q at Lawrence Livermore National Laboratory), high-resolution simulation of the human heart can now be carried out over 1200 times faster compared with published results in the field. We demonstrate the utility of this capability by simulating, for the first time, the formation of transmural re-entrant waves in a 3D human heart. Such wave patterns are thought to underlie Torsades de Pointes, an arrhythmia that indicates a high risk of sudden cardiac death. Our new simulation capability has the potential to impact a multitude of applications in medicine, pharmaceuticals and implantable devices.

The melting line of diamond determined via atomistic computer simulations

The diamond melting line was determined for a model system based on Brenner’s bond order potential for hydrocarbon systems. The location of this first-order phase boundary was found by a free energy calculation of the diamond and liquid phases, using atomistic simulation methods. The melting line was found to have a positive slope consistent with the present understanding. The location at lower pressure was also consistent with experiment and a number of other theoretical approaches. The slope is found to increase with pressure. The structure of the liquid is examined as a function of density and is suggestive of a liquid–liquid phase boundary.

Beyond homogeneous decomposition: scaling long-range forces on Massively Parallel Systems

With supercomputers anticipated to expand from thousands to millions of cores, one of the challenges facing scientists is how to effectively utilize this ever-increasing number. We report here an approach that creates a heterogeneous decomposition by partitioning effort according to the scaling properties of the component algorithms. We demonstrate our strategy by developing a capability to model hot dense plasma. We have performed benchmark calculations ranging from millions to billions of charged particles, including a 2.8 billion particle simulation that achieved 259.9 TFlop/s (26% of peak performance) on the 294,912 cpu JUGENE computer at the Jülich Supercomputing Centre in Germany. With this unprecedented simulation capability we have begun an investigation of plasma fusion physics under conditions where both theory and experiment are lacking-in the strongly-coupled regime as the plasma begins to burn. Our strategy is applicable to other problems involving long-range forces (i.e., biological or astrophysical simulations). We believe that the flexible heterogeneous decomposition approach demonstrated here will allow many problems to scale across current and next-generation machines.