Robert S. Sinkovits

The interaction of shocks and defects in Lennard-Jones crystals

The authors examine, using computational molecular dynamics, shocks launched in two-dimensional crystals by a flying plate. The interaction of the shock with various lattice defects is observed, and is seen to create sites of rapidly growing, thermalized, hot fluid-like phases included within the crystal lattice. They hypothesize that these fluid-like regions are the sites of the initial chemical reactions leading to detonation in energetic materials, and that crystallographic defects therefore control the sensitivity of single-crystal high explosives to shock-initiation. The computations are carried out on the massively parallel CM-200 using a parellelized version of the MLG algorithm.

Parallelization of direct simulation Monte Carlo method combined with monotonic Lagrangian grid

The monotonic Lagrangian grid (MLG) and the direct simulation Monte Carlo (DSMC) methodology were combined on the Thinking Machines CM-5 to create a fast DSMC-MLG code with automatic grid adaptation based on local number densities. The MLG is a data structure in which particles that are close in physical space are also close in computer memory. Using the MLG data structure, physical space is divided into a number of templates (cells), each containing the same number of particles. An MLG-regularization method, stochastic grid restructuring, is implemented to minimize the occurrence of highly skewed cells. Parallelization of the DSMC-MLG is achieved by two different mapping techniques. First, simulated particles are mapped onto the parallel processors for the particle-oriented processes, such as convection, boundary interactions, and MLG sorting. Second, particle templates are mapped onto the processors for computing the macroscopic quantities (i.e., pressure, velocity, density, and temperature) and statistical sampling. In both levels of mapping, the code logic focuses on the structured and fast communications on the CM-5 architecture. The computing time required by the parallel DSMC-MLG code was significantly decreased compared with other parallel efforts and its parallel efficiency on 512 processors achieved approximately 80% for simulation involving one-half million particles.

Scaling relations for the slippery ballistic growth model

A slippery ballistic growth model is introduced in which deposited particles stick with probability ps to the sides of existing columns. Numerical simulations in 1 + 1 dimensions show that while this model belongs to the KPZ universality class, it also exhibits a short-time scaling regime during which the width of the growth region obeys the power law w ∼ t12 before crossing over to KPZ behavior. The crossover times between the three scaling regimes both scale with sticking probability as ps−1. The interface width at saturation ws is found to scale with ps, in addition to its scaling with substrate size.

Simulations of High Knudsen Number Flows in a Channel-Wedge Configuration

A rarefied hypersonic flow through a channel containing a wedge is simulated using the combined direct simulation Monte Carlo and monotonic Lagrangian grid methodologies on the massively parallel Connection Machine. Numerical issues related to the effects of resolution in terms of the simulated to actual particle ratio and the use of time averaging to obtain a statistically converged solution are discussed. Because the flow is rarefied, important aerodynamic features are diffuse compared to what is expected in a continuum flow. For example, the reattachment shock behind the trailing edge of the wedge degenerates into a diffuse viscous layer. Other low-density effects include the velocity slip, which peaks near the leading edges where the density is low, and the temperature jump of the gas adjacent to the solid surfaces, which is highest at the entrance of the channel and decreases farther downstream. The calculated skin friction and heat transfer agree well with the Reynolds analogy for boundary-layer flow.

Direct simulation Monte Carlo study of H/H2 and H/H2/CO mixtures for diamond chemical vapor deposition

One‐dimensional direct simulation Monte Carlo calculations have been carried out on H/H2 and H/H2/CO mixtures under operating conditions typical of diffusion‐dominated diamond chemical vapor deposition processes. Mechanisms have been included in the model for the adsorption and recombination of hydrogen atoms on the diamond surface and the dissociation of molecular hydrogen at the interior of the reactor. Hydrogen atom fluxes and recombinative and conductive heat fluxes to the diamond surface are calculated as a function of pressure, gas composition, hydrogen dissociation and surface reaction probabilities, reactor temperature, and distance between the activating source and substrate. The numerical calculations are shown to be in excellent agreement with analytical results in the limiting regimes of free‐streaming particles at low pressures and continuum hydrodynamics at high pressures.

Evaluation of I/O technologies on a flash-based I/O sub-system for HPC

To meet the growing demand for high performance computing systems that are capable of processing large datasets, the San Diego Supercomputer Center is deploying Gordon. This system was specifically designed for data intensive workloads and uses flash memory to fill the large latency gap in the memory hierarchy between DRAM and hard disk.n In preparation for the deployment of Gordon, we evaluated the performance of multiple remote storage technologies and file systems for use with the flash memory. We find that OCFS and XFS are both superior to PVFS at delivering fast random access to flash. In addition, our tests indicate that the Linux SCSI target framework (TGT) can export flash storage devices with minimal overhead and achieve a large fraction of the theoretical peak I/O performance.n Despite the difficulties in fairly comparing I/O solutions due to the many differences in file systems and service implementations, we conclude that OCFS on TGT is a viable option for our system as it provides both excellent performance and a user-friendly shared file system interface. In those instances where a parallel file system is not required, XFS on TGT is a better alternative.

Subset removal on massive data with Dash

Ongoing efforts by the Large Synoptic Survey Telescope (LSST) involve the study of asteroid search algorithms and their performance on both real and simulated data. Images of the night sky reveal large numbers of events caused by the reflection of sunlight from asteroids. Detections from consecutive nights can then be grouped together into tracks that potentially represent small portions of the asteroids sky-plane motion. The analysis of these tracks is extremely time consuming and there is strong interest in the development of techniques that can eliminate unnecessary tracks, thereby rendering the problem more manageable. One such approach is to collectively examine sets of tracks and discard those that are subsets of others. Our implementation of a subset removal algorithm has proven to be fast and accurate on modest sized collections of tracks, but unfortunately has extremely large memory requirements for realistic data sets and cannot effectively use conventional high performance computing resources. We report our experience running the subset removal algorithm on the TeraGrid Appro Dash system, which uses the vSMP software developed by ScaleMP to aggregate memory from across multiple compute nodes to provide access to a large, logical shared memory space. Our results show that Dash is ideally suited for this algorithm and has performance comparable to or superior to that obtained on specialized, heavily demanded, large-memory systems such as the SGI Altix UV.

Kinetic effects in the chemistry of diamond CVD source gases and implications for diamond growth

Abstract Perfectly stirred reactor calculations have been carried out on six different gas mixtures containing ethanol, water, methane, acetylene, oxygen and hydrogen at conditions typical of low-pressure (about 1 Torr) diamond chemical vapor deposition (CVD) reactors. For reactors operating under conditions in which the flow or diffusive time-scales are short compared with the chemical time-scales, the selection of viable source gas mixtures for diamond growth based solely on the overall C-H-O stoichiometry becomes impossible and kinetic effects must be taken into account. Numerical results are presented which show the dependence of the maximum theoretical diamond deposition rate and diamond quality as a function of mass flow rate.

An analysis of gas phase ethanol-water chemistry for diamond CVD

Abstract : Chemical kinetics calculations have been carried out on ethanol (EtOH)water mixtures at a range of temperatures, pressures snd EtOH/H2O ratios. The results show that the mole fractions of many chemical species depend very sensitively on the initial EtOH/H20 ratio and diverge sharply after an induction time which depends both on pressure and temperature. We explain this divergence primarily in terms of the stability of carbon monoxide and show that the rate-limiting steps that dictate the induction time are the molecular hydrogen dissociation reactions. Our calculations also predict the dependence of the concentrations of potential growth (C, CH3, C2H2) and etching (O, OH, O2) species for diamond chemical vapor deposition (CVD) on operating conditions. jg

Computer simulation of random sequential adsorption of two interacting species on a lattice

A computer-simulation model is introduced to study the variation in the coverage and porosity in a binary system by random sequential adsorption on a periodic square lattice. We study the effects of the range of the repulsive interaction between unlike species and of the probability of deposition of each particle type. For all choices of the interaction range there is a minimum in the total coverage of the lattice which occurs for equal deposition probability of the two species. The saturation coverage decreases on increasing the range of the interaction. For proper choices of the parameters of the model, regimes exist in which either pores or particles of one type form an infinite percolating network.