The Center for Simulating Dynamic Response of Materials, a DOE ASCI Center of Excellence at the California Institute of Technology, is constructing a virtual shock physics facility in which the full three-dimensional response of a variety of target materials can be computed for a wide range of compressive, tensional, and shear loadings including those loadings produced by detonation of energetic materials. The goals are to \beginitemize \item Facilitate computation of a variety of experiments in which strong shock and detonation waves are made to impinge on targets consisting of various combinations of materials, \item Compute the subsequent dynamic response of the target materials, \item Validate these computations against experimental data. \enditemize The research is centered on the three primary stages required to conduct a ``virtual experiment'' in this facility: detonation of high explosives, interaction of shock waves with materials, and shock-induced compressible turbulence and mixing. The modeling requirements are addressed through five integrated research initiatives that guide the key disciplinary activities: \beginitemize \item Modeling and simulation of fundamental processes in detonation, \item Modeling dynamic response of solids, \item First principles computation of materials properties, \item Compressible turbulence and mixing, and \item Computational and computer science infrastructure. \enditemize Of critical importance in performing a simulation of this type is the ability to bridge widely disparate length scales which realte to the dynamic response of both the explosive and the target materials. In this presentation the speaker will provide an overview of the multi-scale modeling research in progress at the center and the computational sciecne research which facilitates the simulations. Additional information on our work can be obtained by visiting our web page http://www.cacr.caltech.edu/ASAP.
[LC41.02] Connecting Astrophysics to Laboratory Experiments: Astrophysical Thermonuclear Flashes
Robert Rosner (The University of Chicago)
Our understanding of highly dynamic and nonlinear astrophysical phenomena such as transient thermonuclear burning within, or on the surface of, evolved compact stars such as white dwarfs is strongly dependent on the versimilitude of numerical simulations. Because the dynamic range of the spatial scales of such phenomena are far beyond our ability to carry out direct numerical simulations, we must rely upon modeling of the sub-gridscale physics; similarly, we need to develop approximation schemes that allow us to overcome the limitations imposed by the enormous dynamic range of the temporal behavior. It is then extremely important to test the versimilitude of our calculations, independent of the astrophysical observations we seek to understand; this has led us to use laboratory fluid dynamical experiments as test beds for our calculations.
[LC41.03] Large-Scale Quantum Monte Carlo Simulations of Correlated Electron Systems
Carey Huscroft (University of Cincinnati)
Quantum Monte Carlo (QMC) is an exact numerical technique for calculating quantities such as charge and spin correlations, densities of states, and thermodynamic properties in quantum systems. QMC has been used successfully to gain insight into the physics of many systems, but its great computational expense has limited its applicability. Using recently available large-scale computational resources, we can now address formerly intractable problems with QMC. An example is the f-electron correlation driven phase transition under pressure, seen in a number of rare-earth metals and characterized by large (5--15%) volume changes. In the high-pressure regime, approximate techniques give a good description of the material structure. However, exact treatments of correlations are needed to reproduce the transition and the low-pressure, strongly correlated phase. While exact, full-orbital approaches are still beyond reach, using large-scale QMC one can obtain exact solutions of appropriate models, such as the two-band Periodic Anderson Model, and obtain considerable insight into the nature of such transitions.
[LC41.04] Large Scale Ab Initio Molecular Dynamics Simulations in Materials Science
Francois Gygi (Lawrence Livermore National Laboratory, Livermore CA 94551.)
Ab initio molecular dynamics has a history of applications to numerous problems in solid state physics, chemistry and, more recently, biochemistry. The advent of large scale computing facilities is now extending the domain of application of this technique to ever more complex systems. The ability to extend simulation sizes to several hundred atoms, for durations of tens of picoseconds allows to study new problems such as the behaviour of complex liquids under pressure, the solvation properties of biomolecules, or microfracture in complex solids. We will discuss the results of ab initio MD simulations carried out on the recently installed ASCI platforms at LLNL. The new challenges implied by the unprecendented scale of these calculations will also be described.