Bulletin of the American Physical Society
APS March Meeting 2014
Volume 59, Number 1
Monday–Friday, March 3–7, 2014; Denver, Colorado
Session M27: Focus Session: Petascale Science and Beyond: Applications and Opportunities for Materials Science III |
Hide Abstracts |
Sponsoring Units: DCOMP Chair: Jack Wells, Oak Ridge National Laboratory Room: 501 |
Wednesday, March 5, 2014 11:15AM - 11:51AM |
M27.00001: Exciting Quantized Vortex Rings in a Superfluid Unitary Fermi Gas Invited Speaker: Aurel Bulgac In a recent article, Yefsah et al., Nature \textbf{499}, 426 (2013) report the observation of an unusual quantum excitation mode in an elongated harmonically trapped unitary Fermi gas. After phase imprinting a domain wall, they observe collective oscillations of the superfluid atomic cloud with a period almost an order of magnitude larger than that predicted by any theory of domain walls, which they interpret as a possible new quantum phenomenon dubbed ``a heavy soliton'' with an inertial mass some 50 times larger than one expected for a domain wall. We present compelling evidence that this ``heavy soliton'' is instead a quantized vortex ring by showing that the main aspects of the experiment can be naturally explained within an extension of the time-dependent density functional theory (TDDFT) to superfluid systems. The numerical simulations required the solution of some 260,000 nonlinear coupled time-dependent 3-dimensional partial differential equations and was implemented on 2048 GPUs on the Cray XK7 supercomputer Titan of the Oak Ridge Leadership Computing Facility. [Preview Abstract] |
Wednesday, March 5, 2014 11:51AM - 12:03PM |
M27.00002: Basis Function Sampling for Material Property Computations Jonathan K. Whitmer, Chi-cheng Chiu, Abhijeet A. Joshi, Juan J. de Pablo Wang--Landau sampling, and the associated class of flat histogram simulation methods, have been particularly successful for free energy calculations in a wide array of physical systems. Practically, the convergence of these calculations to a target free energy surface is hampered by reliance on parameters which are unknown {\it a priori}. We derive and implement a method based on orthogonal (basis) functions which is fast, parameter-free, and geometrically robust. An important feature of this method is its ability to achieve arbitrary levels of description for the free energy. It is thus ideally suited to {\it in silico} measurement of elastic moduli and other quantities related to free energy perturbations. We demonstrate the utility of such applications by applying our method to calculation of the Frank elastic constants of the Lebwohl--Lasher model. [Preview Abstract] |
Wednesday, March 5, 2014 12:03PM - 12:15PM |
M27.00003: Performance of replica-exchange Wang-Landau sampling for the study of spin systems Ying Wai Li, Markus Eisenbach, Thomas Vogel, Thomas W\"ust, David P. Landau The recently proposed replica-exchange Wang-Landau sampling (REWL)\footnote{Phys. Rev. Lett. \textbf{110}, 210603 (2013)} is a novel, massively parallel Monte Carlo method which allows for the parallelization of Wang-Landau sampling based on a replica-exchange framework. The robustness of the scheme is demonstrated by its broad applicability on a variety of spin systems: from the simplest models with discrete or continuous energy domains, to complex systems captured by large-scale first principles density functional theory calculations. The accuracy of REWL is studied by comparing the thermodynamic properties with exact solutions and results obtained by the original, serial Wang-Landau sampling. The principles for the speed-up, the strong and weak scaling behavior of REWL are also investigated when different parameter settings are employed. We will show, with the aid of selected spin systems, that the method accelerates the simulations significantly with a possible improved accuracy.\footnote{This research was partly sponsored by the Office of Advanced Computing Research of the US Department of Energy. It used resources of the Oak Ridge Leadership Computing Facility at ORNL supported by the Office of Science of the DOE under contract DE-AC05-00OR22725.} [Preview Abstract] |
Wednesday, March 5, 2014 12:15PM - 12:27PM |
M27.00004: Magnetic entropy change calculated from first principles based statistical sampling technique: Ni2MnGa Khorgolkhuu Odbadrakh, Don Nicholson, Markus Eisenbach, Gregory Brown, Aurelian Rusanu Magnetic entropy change in Magneto-caloric Effect materials is one of the key parameters in choosing materials appropriate for magnetic cooling and offers insight into the coupling between the materials' thermodynamic and magnetic degrees of freedoms. We present computational workflow to calculate the change of magnetic entropy due to a magnetic field using the DFT based statistical sampling of the energy landscape of Ni2MnGa. The statistical density of magnetic states is calculated with Wang-Landau sampling, and energies are calculated with the Locally Self-consistent Multiple Scattering technique. The high computational cost of calculating energies of each state from first principles is tempered by exploiting a model Hamiltonian fitted to the DFT based sampling. The workflow is described and justified. The magnetic adiabatic temperature change calculated from the statistical density of states agrees with the experimentally obtained value in the absence of structural transformation. The study also reveals that the magnetic subsystem alone cannot explain the large MCE observed in Ni2MnGa alloys. This work was performed at the ORNL, which is managed by UT-Batelle for the U.S. DOE. It was sponsored by the Division of Material Sciences and Engineering, OBES. This research used resources of the OLCF at ORNL, which is supported by the Office of Science of the U.S. DOE under Contract DE-AC05-00OR22725. [Preview Abstract] |
Wednesday, March 5, 2014 12:27PM - 12:39PM |
M27.00005: Exact enumeration of an Ising model for Ni$_2$MnGa Markus Eisenbach, Gregory Brown, Don M. Nicholson Exact evaluations of partition functions are generally prohibitively expensive due to exponential growth of phase space with the degrees of freedom. An Ising model with $N$ sites has $2^N$ possible states, requiring the use of better scaling methods, such as importance sampling Monte-Carlo for all but the smallest systems. Yet the ability to obtain exact solutions for large systems can provide important benchmark results and opportunities for unobscured insight into the underlying physics of the system. Here we present an Ising model for the magnetic sublattices of the important magneto-caloric material Ni$_2$MnGa and use an exact enumeration algorithm to calculate the number of states $g(E,M_1,M_2)$ for each energy $E$ and sublattice magnetization $M_1$ and $M_2$. This allows the efficient calculation of the partition function and derived thermodynamic quantities such as specific heat and susceptibility. Utilizing resources at the Oak Ridge Leadership Facility we are able to calculate $g(E,M_1,M_2)$ for systems of up to 48 sites, which provides important insight into the mechanism for the large magnet-caloric effect in Mn$_2$NiGa as well as an important benchmark for Monte-Carlo based calculations (esp. Wang-Landau) of $g(E,M_1,M_2)$. [Preview Abstract] |
Wednesday, March 5, 2014 12:39PM - 12:51PM |
M27.00006: First-principles calculation of electronic stopping contributions from core electrons and off-channeling Alfredo Correa, Andre Schleife, Yosuke Kanai In order to understand the interaction of projectile atoms with targets under particle radiation in materials, e.g.\ in space applications or nuclear reactors, it is critical to investigate electronic and ionic contributions to stopping power. The goal of such efforts is detailed understanding of radiation damages as well as fundamental effects such as ion-electron interaction. While ionic stopping has been successfully modeled by molecular dynamics in the past, only recently a computational framework came within reach that is capable of accurately describing \emph{electronic} stopping from first principles. Using our large-scale implementation of real-time time-dependent density functional theory in non-adiabatic Ehrenfest molecular dynamics, we are able to gain deep insight into electronic stopping for systems with hundreds of atoms and thousands of electrons, taking into account their quantum-mechanical electron-electron interaction. We discuss distinct contributions of valence and core electrons of aluminum target atoms to electronic stopping, and we study their importance for different projectile (hydrogen and helium atoms) velocities. There is striking influence of the stopping geometry especially for fast projectiles, and we find excellent agreement with experiment. [Preview Abstract] |
Wednesday, March 5, 2014 12:51PM - 1:03PM |
M27.00007: Highly parallel implementation of non-adiabatic Ehrenfest molecular dynamics Yosuke Kanai, Andre Schleife, Erik Draeger, Victor Anisimov, Alfredo Correa While the adiabatic Born-Oppenheimer approximation tremendously lowers computational effort, many questions in modern physics, chemistry, and materials science require an explicit description of coupled \emph{non-adiabatic} electron-ion dynamics. Electronic stopping, i.e.\ the energy transfer of a fast projectile atom to the electronic system of the target material, is a notorious example. We recently implemented real-time time-dependent density functional theory based on the plane-wave pseudopotential formalism in the Qbox/qb@ll codes. We demonstrate that explicit integration using a fourth-order Runge-Kutta scheme is very suitable for modern highly parallelized supercomputers. Applying the new implementation to systems with hundreds of atoms and thousands of electrons, we achieved excellent performance and scalability on a large number of nodes both on the BlueGene based ``Sequoia'' system at LLNL as well as the Cray architecture of ``Blue Waters'' at NCSA. As an example, we discuss our work on computing the electronic stopping power of aluminum and gold for hydrogen projectiles, showing an excellent agreement with experiment. These first-principles calculations allow us to gain important insight into the the fundamental physics of electronic stopping. [Preview Abstract] |
Wednesday, March 5, 2014 1:03PM - 1:15PM |
M27.00008: Large-scale massively parallel atomistic simulations of short pulse laser interaction with metals Chengping Wu, Leonid Zhigilei Taking advantage of petascale supercomputing architectures, large-scale massively parallel atomistic simulations (10$^{8}$-10$^{9}$ atoms) are performed to study the microscopic mechanisms of short pulse laser interaction with metals. The results of the simulations reveal a complex picture of highly non-equilibrium processes responsible for material modification and/or ejection. At low laser fluences below the ablation threshold, fast melting and resolidification occur under conditions of extreme heating and cooling rates resulting in surface microstructure modification. At higher laser fluences in the spallation regime, the material is ejected by the relaxation of laser-induced stresses and proceeds through the nucleation, growth and percolation of multiple voids in the sub-surface region of the irradiated target. At a fluence of $\sim$ 2.5 times the spallation threshold, the top part of the target reaches the conditions for an explosive decomposition into vapor and small droplets, marking the transition to the phase explosion regime of laser ablation. The dynamics of plume formation and the characteristics of the ablation plume are obtained from the simulations and compared with the results of time-resolved plume imaging experiments. [Preview Abstract] |
Wednesday, March 5, 2014 1:15PM - 1:27PM |
M27.00009: Challenges in Modeling of the Plasma-Material Interface Predrag Krstic Recent work with lithium coatings deposited on a variety of metallic and graphitic surfaces, in a number of tokamak fusion machines around the world, has provided evidence of the sensitive dependence plasma behavior has on these ultra-thin deposited films. Our computer simulations, validated by recent experiments, have elucidated roles of lithium in carbon walls to the recycling of the plasma hydrogen [1]. We performed quantum-classical atomistic calculations on many thousands of random trajectories to clarify the interplay of lithium and oxygen in amorphous carbon. We show that the presence of oxygen in the surface plays the key role in the increased uptake chemistry and suppression of erosion, while lithium has a decisive role in achieving high concentrations of oxygen in the upper layers of the surface upon bombardment by deuterium. D atoms preferentially bind with O and C-O. The plasma-facing walls of the next-generation fusion reactors will be exposed to high fluxes of neutrons and plasma-particles and will operate at high temperatures for thermodynamic efficiency. To this end we have been studying the evolution dynamics of vacancies and interstitials to high doses of tungsten surfaces bombarded by self-atoms, using classical molecular dynamics. Results show surprising saturation of the defects upon cumulative irradiation of only 1 DPA, as well as the defects clustering at the tungsten surface. These findings are obtaining validation in recent experiments. [Preview Abstract] |
Wednesday, March 5, 2014 1:27PM - 1:39PM |
M27.00010: Beyond Petascale with the HipGISAXS Software Suite Alexander Hexemer, Sherry Li, Slim Chourou, Abhinav Sarje We have developed HipGISAXS, a software suite to analyze GISAXS and SAXS data for structural characterization of materials at the nano scale using X-rays. The software has been developed as a massively-parallel system capable of harnessing the raw computational power offered by clusters and supercomputers built using graphics processors (GPUs), Intel Phi co-processors, or commodity multi-core CPUs. Currently the forward GISAXS simulation is a major component of HipGISAXS, which simulates the X-ray scattering process based on the Distorted Wave Born Approximation (DWBS) theory, for any given nano structures and morphologies with a set of experimental configurations. These simulations are compute-intensive, and have a high degree of parallelism available, making them well-suited for fine-grained parallel computations on highly parallel many core processors like GPUs. Furthermore, a large number of such simulations can be carried out simultaneously for various experimental input parameters. HipGISAXS also includes a Reverse Monte Carlo based modeling tool for SAXS data. With HipGISAXS we have demonstrated a sustained compute performance of over 1 Petaflop on 8000 GPU nodes of the Titan supercomputer at ORNL, and have shown it to be highly scalable. [Preview Abstract] |
Wednesday, March 5, 2014 1:39PM - 1:51PM |
M27.00011: Hyperdynamics boost factor achievable with an ideal bias potential Chen Huang, Danny Perez, Arthur Voter Hyperdynamics has been proven to be very promising for bridging the time scale gap between simulations and experiments. Much effort has been devoted to developing valid bias potentials, however the limiting performance of hyperdynamics is still unknown. In this work, a nearly ``ideal'' bias potential is designed to study the limiting performing of hyperdynamics. This bias potential is constructed based on the minimum energy pathways (MEP) of all the pathways out of the current state. We apply this MEP-based hyperdynamics (MEP-HD) to several metallic surface diffusion systems. By using proper parameters for constructing such ``ideal'' bias potential, both the Kramers recrossings and the branch ratios of different transitions can be reproduced. Since such MEP-based bias potential is directly built on reaction coordinates, in most cases it gives boost factors that are orders of magnitude larger than the best existing bias potentials. Such impressive performance of MEP-HD is believed to be very close to the limiting performance of hyperdynamics, and shows that further development of hyperdynamics could have a significant payoff. [Preview Abstract] |
Wednesday, March 5, 2014 1:51PM - 2:03PM |
M27.00012: Discrete Event-based Performance Prediction for Temperature Accelerated Dynamics Christoph Junghans, Susan Mniszewski, Arthur Voter, Danny Perez, Stephan Eidenbenz We present an example of a new class of tools that we call {\em application simulators}, parameterized fast-running proxies of large-scale scientific applications using parallel discrete event simulation (PDES). We demonstrate our approach with a TADSim {\em application simulator} that models the Temperature Accelerated Dynamics (TAD) method, which is an algorithmically complex member of the Accelerated Molecular Dynamics (AMD) family. The essence of the TAD application is captured without the computational expense and resource usage of the full code. We use TADSim to quickly characterize the runtime performance and algorithmic behavior for the otherwise long-running simulation code. We further extend TADSim to model algorithm extensions to standard TAD, such as speculative spawning of the compute-bound stages of the algorithm, and predict performance improvements without having to implement such a method. Focused parameter scans have allowed us to study algorithm parameter choices over far more scenarios than would be possible with the actual simulation. This has led to interesting performance--related insights into the TAD algorithm behavior and suggested extensions to the TAD method. [Preview Abstract] |
Wednesday, March 5, 2014 2:03PM - 2:15PM |
M27.00013: Beyond the Harmonic Approximation: Lattice Dynamics and Thermal Conductivity on Massively Parallel Heterogenous Systems Weston Nielson, Fei Zhou, Vidvuds Ozolins The theory of lattice dynamics provides the mathematical foundation necessary to solve, exactly, the potential energy and interatomic forces of a system on a lattice, which can easily be written as functions of atomic displacements and so-called force constants. We have used this formalism to develop a highly-parallel algorithm that is capable of calculating the potential energy and interatomic forces across large computational clusters and on graphics processing units (GPUs). The necessary force constants are calculated well beyond the simple harmonic approximation (up to n-order), via a compressive sensing-based algorithm, which are then used in molecular dynamics simulations to study lattice thermal conductivity in a variety of crystal systems. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700