12 Matching Results

Search Results

Advanced search parameters have been applied.

Material Interface Reconstruction for Monte Carlo Particle Tracking

Description: In this project we implement material interface reconstruction into a large, massively parallel Monte Carlo particle transport code. Here we examine the benefit of resolving a material interface for criticality calculations. Input to the code is a mesh with material and density defined on the mesh. For mesh zones that contain more than one material (mixed zones), the old approximation made in the code is to homogenize the material properties of all the materials in the zone. The neutron mean free path is a function of the material density that the neutron is traveling through, so for mixed zones, we use the average density of the zone, rather than reconstructing a material interface, determining which material within the zone the particle is in and using the correct density based on the position of the particle within the zone. In order to get a better answer, here we implement material interface reconstruction and rather than homogenizing the materials in a mixed zone, we have a material interface divide the zone so we can tell which material the particle is in, based on the particle's position and the location of the material interface.
Date: March 15, 2006
Creator: O'Brien, M J
Partner: UNT Libraries Government Documents Department

Multigrid Methods for Mesh Relaxation

Description: When generating a mesh for the initial conditions for a computer simulation, you want the mesh to be as smooth as possible. A common practice is to use equipotential mesh relaxation to smooth out a distorted computational mesh. Typically a Laplace-like equation is set up for the mesh coordinates and then one or more Jacobi iterations are performed to relax the mesh. As the zone count gets really large, the Jacobi iteration becomes less and less effective and we are stuck with our original unrelaxed mesh. This type of iteration can only damp high frequency errors and the smooth errors remain. When the zone count is large, almost everything looks smooth so relaxation cannot solve the problem. In this paper we examine a multigrid technique which effectively smooths out the mesh, independent of the number of zones.
Date: June 12, 2006
Creator: O'Brien, M J
Partner: UNT Libraries Government Documents Department

PyMercury: Interactive Python for the Mercury Monte Carlo Particle Transport Code

Description: Monte Carlo particle transport applications are often written in low-level languages (C/C++) for optimal performance on clusters and supercomputers. However, this development approach often sacrifices straightforward usability and testing in the interest of fast application performance. To improve usability, some high-performance computing applications employ mixed-language programming with high-level and low-level languages. In this study, we consider the benefits of incorporating an interactive Python interface into a Monte Carlo application. With PyMercury, a new Python extension to the Mercury general-purpose Monte Carlo particle transport code, we improve application usability without diminishing performance. In two case studies, we illustrate how PyMercury improves usability and simplifies testing and validation in a Monte Carlo application. In short, PyMercury demonstrates the value of interactive Python for Monte Carlo particle transport applications. In the future, we expect interactive Python to play an increasingly significant role in Monte Carlo usage and testing.
Date: November 29, 2010
Creator: Iandola, F N; O'Brien, M J & Procassini, R J
Partner: UNT Libraries Government Documents Department

Load Balancing of Parallel Monte Carlo Transport Calculations

Description: The performance of parallel Monte Carlo transport calculations which use both spatial and particle parallelism is increased by dynamically assigning processors to the most worked domains. Since he particle work load varies over the course of the simulation, this algorithm determines each cycle if dynamic load balancing would speed up the calculation. If load balancing is required, a small number of particle communications are initiated in order to achieve load balance. This method has decreased the parallel run time by more than a factor of three for certain criticality calculations.
Date: May 25, 2005
Creator: Procassini, R J; O'Brien, M J & Taylor, J M
Partner: UNT Libraries Government Documents Department

On the Elasto-Plastic Response of a Large-Tow Triaxial Braided Composite

Description: The elastic-plastic response of a large-tow 0{sup o}/{+-}{theta}{sup o} tri-axially braided composite is numerically simulated to determine the elastic coefficients and post-yield behavior. The ratios of extensional to flexural effective Young's moduli vary from 0.30 to 0.52 in the longitudinal direction and 0.90 to 0.95 in the transverse direction. Measurements on a 2-ply 0{sup o}/{+-} 30{sup o} braid support these numerical trends. The onset of macro yield in uniaxial extension coincides with the experimental values in the longitudinal direction while it is nearly twice the experimental values in the transverse direction. In simple shear, matrix plasticity around the undulations facilitates local rotation of the braiders at the onset of macro yield. Under uniaxial flexure, modest stiffening occurs prior to strain softening in both the principal directions.
Date: June 14, 2000
Creator: Zywicz, E.; O'Brien, M.J. & Nguyen, T.
Partner: UNT Libraries Government Documents Department

New Capabilities in Mercury: A Modern, Monte Carlo Particle Transport Code

Description: The new physics, algorithmic and computer science capabilities of the Mercury general-purpose Monte Carlo particle transport code are discussed. The new physics and algorithmic features include in-line energy deposition and isotopic depletion, significant enhancements to the tally and source capabilities, diagnostic ray-traced particles, support for multi-region hybrid (mesh and combinatorial geometry) systems, and a probability of initiation method. Computer science enhancements include a second method of dynamically load-balancing parallel calculations, improved methods for visualizing 3-D combinatorial geometries and initial implementation of an in-line visualization capabilities.
Date: March 8, 2007
Creator: Procassini, R J; Cullen, D E; Greenman, G M; Hagmann, C A; Kramer, K J; McKinley, M S et al.
Partner: UNT Libraries Government Documents Department

Update on the Development and Validation of MERCURY: A Modern, Monte Carlo Particle Transport Code

Description: An update on the development and validation of the MERCURY Monte Carlo particle transport code is presented. MERCURY is a modern, parallel, general-purpose Monte Carlo code being developed at the Lawrence Livermore National Laboratory. During the past year, several major algorithm enhancements have been completed. These include the addition of particle trackers for 3-D combinatorial geometry (CG), 1-D radial meshes, 2-D quadrilateral unstructured meshes, as well as a feature known as templates for defining recursive, repeated structures in CG. New physics capabilities include an elastic-scattering neutron thermalization model, support for continuous energy cross sections and S ({alpha}, {beta}) molecular bound scattering. Each of these new physics features has been validated through code-to-code comparisons with another Monte Carlo transport code. Several important computer science features have been developed, including an extensible input-parameter parser based upon the XML data description language, and a dynamic load-balance methodology for efficient parallel calculations. This paper discusses the recent work in each of these areas, and describes a plan for future extensions that are required to meet the needs of our ever expanding user base.
Date: June 6, 2005
Creator: Procassini, R J; Taylor, J M; McKinley, M S; Greenman, G M; Cullen, D E; O'Brien, M J et al.
Partner: UNT Libraries Government Documents Department