|
|
2009 International Conference on Advances in Mathematics, Computational Methods, and Reactor Physics
|
Several Companies, Laboratories, and Universities will be presenting tutorials on a wide variety of computer codes and other topics at M&C2009. Sessions covering Monte Carlo and Discrete Ordinates codes, as well as Analytic Benchmarks and Computational Toolkits will be available for the Conference Attendees. Many of these tutorials will provide the opportunity for hands-on experience with the subject computer codes.
Please consider specifying a preference for sessions you may attend when you register for the conference. Specifying a preference for specific tutorials is not required for you to attend any specific sessions but will help us to plan for adequate facilities. Please note that seating will be on a first-come-first serve basis for all of the tutorials.
As with past meetings, these tutorials are provided free of charge to registered Conference Attendees.
Please click on any of the tutorial topics in the table below for a brief course synopsis.
Check this page later for final room assignments for the tutorials.
**************************************************Schedule of Tutorial Sessions
|
Date/Time |
Session 1 |
Session 2 |
Session 3 |
Session 4 |
Sunday, May 3 8:30 am - 12:00 pm |
Topic: MCBEND/MONK Room: |
Topic: SCALE 6/MAVRIC Room: |
|
|
Sunday, May 3 1:30 pm - 5:00 pm |
Topic: TRIPOLI Room: |
Topic: Attila Room: |
|
|
Thursday, May 7 1:30 am - 5:00 pm |
Topic: Advances in Monte Carlo |
Topic: TRILINOS Room: |
Topic: Analytic Benchmarks Room: |
Topic: PENTRAN Room: |
Topic: MCBEND/MONK Instructor(s): Adam Bird Affiliation: SERCO, ANSWERS Software Service Date: Sunday, May 3 Time: 8:30 am - 12:00 pm Room: Summary: This tutorial is aimed at providing a broad understanding of the capabilities of the codes and is a compression of two 4 day courses "Introduction to MCBEND" and "Introduction to MONK". The tutorial comprises a mixture of lectures and demonstrations. Demonstrations will include constructing MCBEND & MONK models for specified problems, checking the input specifications, displaying the geometry model and performing full calculations on a PC. An overview of the ANSWERS code RANKERN and WIMS will be provided at the end of the tutorial. Attendees will be provided with a 1 month trial license of the MCBEND and MONK software. Those bringing their own computers may install the software before the tutorial and run the examples themselves. MCBEND is a computer program written to solve problems of radiation transport in sub-critical systems using the Monte Carlo method. MONK is a UK regulator recognised Monte Carlo neutronics code written to assist in the study of both criticality safety and reactor physics depletion problems. MONK and MCBEND are distributed and actively supported in use by Serco in UK as part of its ANSWERS Software Service, with the code development itself being managed by a collaboration comprising Serco and Sellafield Limited. RANKERN is a point kernel package for the assessment of complex gamma-ray shielding. RANKERN is distributed and actively supported in use by Serco in UK as part of its ANSWERS Software Service, with the code development itself being managed by a collaboration comprising Serco and Sellafield Limited. WIMS is an internationally-known modular reactor physics software package for neutronics calculations. WIMS is developed, distributed and actively supported in use by Serco in the UK as part of its ANSWERS Software Service. Main Objectives: * Provide an overview of the MONK & MCBEND codes * Familiarisation with the input manual, input style, output files and method of execution. * Introduce the geometry modelling package * Introduce the HOLE geometry modelling package * Demonstrate the graphics package for modifying, displaying, checking and running models. * Explore the source options in MCBEND & MONK * Demonstrate the principal variance reduction techniques of MCBEND. * Demonstrate the use of superhistory powering in MONK. * Provide an overview of the features of WIMS and RANKERN A more detailed description of the codes and topics covered in this tuturial (in MS WORD format) can be seen by clicking here.
Topic: New Shielding Methods in SCALE 6 Instructor(s): Douglas E. Peplow Affiliation: Oak Ridge National Laboratory Date: Sunday, May 3 Time: 8:30 am - 12:00 pm Room: Summary: This half-day demonstration/tutorial will highlight the automated variance reduction capabilities of the MAVRIC sequence using several simple example shielding problems. To optimize a given tally, MAVRIC first computes an importance map and biased source distribution based on the results of approximate discrete ordinates calculations using the new Denovo SN code. The importance map and biased source are then used by the Monte Carlo functional module Monaco to compute that tally much more efficiently than an analog calculation. Examples will include: * Calculating dose near a spent fuel cask * Calculating a dose contour map from an array of storage casks * Calculating the doses at the detectors of a criticality accident alarm system. (This last example uses a fission distribution computed by KENO-VI as the source term for MAVRIC.) Registered SCALE 6 users are welcome to bring their own laptop and follow along.
Topic: TRIPOLI Instructor(s): Jean-Christophe Trama, Yi-Kang Lee, Eric Dumonteil, Stephane Bourganel Affiliation: CEA/SERMA Date: Sunday, May 3 Time: 1:30 pm - 5:00 pm Room: Summary: TRIPOLI-4 is the fourth generation of the TRIPOLI Monte Carlo code, originally developed in the 1960's by CEA. Totally rewritten in C and C++, this 3D full pointwise code is dedicated to radiation shielding, criticality safety and reactor physics. It is used as a reference tool by CEA, EDF, IRSN and other industrial partners, as well as in the European nuclear reactor simulation platform (NURESIM). It is distributed by the NEA Databank and by RSICC in the US. This 3-hour tutorial is a short version of the 4 day "OECD/NEA TRIPOLI-4 training course". After a general introduction of TRIPOLI (scope, semantic, v&v), the people attending the workshop will have the opportunity to run TRIPOLI-4 on the following test cases: * Keff calculation to determine the neutron spectra, leakage current and fission rate distribution * Fuel lattice construction to simulate a MOX fuel assembly * Reactor pressure vessel fluence calculation to demonstrate the powerful TRIPOLI-4 variance reduction, the automatic creation of importance maps and the migration of the neutron collision sites on projection maps. Six laptops will be provided by the training team. Up to 3 people per laptop may actively participate in the workshop. Neither the TRIPOLI-4 code nor the test cases will be provided for other laptops.
Topic: Attila Instructor(s): Gregory Failla Affiliation: Transpire, Inc. Date: Sunday, May 3 Time: 1:30 pm - 5:00 pm Room: Summary: Attila is a CAD based deterministic ratiation transport software system which is being used for a broad range of radiation protection and shielding applications. Attila combines an intuitive, process based graphical user interface (GUI) with a leading edge solver and insightful post processing. Attila calculates the angular and energy dependent flux for all solved particles everywhere in the computational domain. Response functions such as dose, reaction rates and user defined quantities can all be calculated through the GUI as post processing operations. Since Attila is acurate and efficient through large attenuations, and can import arbitrary CAD geometries, it is well suited for shielding design applications. Additionally, Attila can automatically calculate and export optimized weight windows in a format readable by MCNP/MCNPX, enabling solutions to be rapidly verified through two independent first principles solution methods. The tutorial will begin with a half hour presentation introducing Attila, followed by a hands-on workshop where participants will set-up, analyze, and post process representative shielding applications. Attendees wishing to participate in the hands-on workshop should bring their own laptops. Please contact Greg Failla (greg@transpireinc.com)with any questions prior to the workshop.
Advances in Monte Carlo Criticality Calculations
Topic: Advances in Monte Carlo Criticality Calculations Instructor(s): Forrest Brown (LANL), Brian Kiedrowski (U. of Wisconsin/LANL) William Martin (U. of Michigan), Gokhan YesilYurt (U. of Michigan) Affiliation: Los Alamos National Laboratory, U. of Wisconsin, U. of Michigan Date: Thursday, May 7 Time: 1:30 pm - 5:00 pm Room: Summary: Monte Carlo criticality calculations are performed routinely on large, complex models for reactor physics and criticality safety applications. This tutorial includes a thorough review of best practices for calculations, along with in-depth coverage of several important R&D areas. It should benefit both Monte Carlo practitioners and developers. The following topics will be covered in the tutorial: I. Best Practices for Monte Carlo Criticality Calculations A review of the theory and practice of Monte Carlo criticality calculations, including best practices for assuring convergence, avoiding bias in Keff and tallies, assessing bias in confidence intervals, and taking advantage of symmetry. Includes numerous practical examples with MCNP and recent advances. II. Adjoints, Importance, and Reactor Kinetics Parameters A review of adjoint calculations and the need for importance weighting in computing reactor kinetics parameters. The iterated fission probability and its use in Monte Carlo calculations are discussed at length. Numerous examples are presented, along with an overview of current efforts to develop a continuous-energy importance weighting method for MCNP. III. Temperature Dependence For realistic, detailed reactor calculations, Monte Carlo codes are part of a multiphysics simulation involving thermal-hydraulic feedback to adjust temperatures and densities. This process can result in 1000s of material temperatures for which broadened cross-sections are needed. Existing codes (eg, MCNP) were not designed to accommodate this need. This tutorial reviews the broadening temperature dependence and discusses a novel new on-the-fly broadening scheme that would permit an unlimited number of temperatures for only a modest computing cost.
Topic: Trilinos Solvers Instructor(s): Roger Pawlowski Affiliation: Sandia National Laboratory Date: Thursday, May 7 Time: 1:30 pm - 5:00 pm Room: Summary: The Trilinos Project is an effort to facilitate the design, development, integration and ongoing support of mathematical software libraries within an object-oriented framework for the solution of large-scale, complex multi-physics engineering and scientific problems. Trilinos addresses two fundamental issues of developing software for these problems: (i) Providing a streamlined process and set of tools for development of new algorithmic implementations and (ii) promoting interoperability of independently developed software. Trilinos uses a two-level software structure designed around collections of packages. A Trilinos package is an integral unit usually developed by a small team of experts in a particular algorithms area such as algebraic preconditioners, nonlinear solvers, etc. Packages exist underneath the Trilinos top level, which provides a common look-and-feel, including configuration, documentation, licensing, and bug-tracking. Here we present an overview of Trilinos capabilities and the overall Trilinos design, integration of Trilinos into an application and use of its linear and nonlinear solver capabilities. The approximate schedule is: * Intro to Trilinos and its framework (30 minutes) * The Epetra and Teuchos packages: Integrating Trilinos into an application (60 minutes) * Solving linear systems using Trilinos (90 minutes) * Solving nonlinear problems using Trilinos (30 minutes)
Topic: Analytic Benchmarks Instructor(s): Barry Ganapol Affiliation: University of Arizona/University of Tennessee Date: Thursday, May 7 Time: 1:30 pm - 5:00 pm Room: Summary: The study of the neutron transport equation is a delicate blend of theoretical mathematics, numerical methods and computational strategies describing the interaction of neutrons and nuclei. In this short course, "Case Studies in Neutron Transport Theory", we shall concentrate on transforming theoretical solution representations of the neutron transport equation into numerically useable forms. In this 3 and 1/2 hour course, we will study reactor physics from neutron slowing down to multidimensional multigroup theory and criticality. Though the backdrop is reactor physics, our emphasis will be on analytical manipulations of the transport equation and the numerical realization of its solutions. The main objective is to provide a basis for understanding the fundamental concepts of evaluating the solutions of neutron transport theory. This will include recent theoretical as well as numerical advances in analytical benchmarking. The course text will be "The Analytical Benchmark Library for Nuclear Engineering" by Dr. Barry Ganapol, OECD/NEA Press, Paris, 2007. Free hard copies of the text will be available to the first 20 course participants. All Participants will receive electronic copies of the book and benchmark library. Several demonstrations of the analytical benchmark library that accompanies the text as well as analytical benchmarking practices will be included in the tutorial session. Course attendees will become familiar with: * Some analytical forms of the transport equation * Various analytical methods of solution * Numerical evaluation of analytical representations * Semi-analytical benchmarking techniques. Participants should be familiar with reactor physics and the operation of nuclear systems. In addition, some familiarity with mathematics through vector calculus and linear algebra is helpful. The participant should also be familiar with elementary numerical methods and come with an open mind awaiting new information. For best results, participants are encouraged to bring a laptop equipped with a FORTRAN compiler and plotting package. We consider reactor physics concepts in three distinct physical categories, energy dependence, space dependence and coupled energy and space dependence. The following topics will be covered: 1. The Neutron Transport Equation (1/2 Hour) 2. Neutron Slowing Down (1/2 Hour) 3. Monoenergetic and Multigroup Transport Theory (1 Hour) 4. Multidimensional Transport Theory (1/2 Hour) 5. The future of Analytical Benchmarking (1/2 Hour) A more detailed description of the short course presented by Dr. Ganapol (in MS WORD format) can be seen by clicking here.
Topic: PENTRAN Instructor(s): Glenn Sjoden Affiliation: University of Florida Date: Thursday, May 7 Time: 1:30 pm - 5:00 pm Room: Summary: The PENTRAN code system was first developed by Sjoden and Haghighat in 1996, and can be used for 3-D multigroup forward or adjoint discrete ordinates (SN) simulations on parallel (or serial) computers. The PENTRAN system is actually a suite of codes that allow one to readily generate mesh geometries, solve 3-D transport models, and automatically collate parallel data. PENTRAN is a multi-group, anisotropic SN code for 3-D Cartesian geometries; it has been specifically designed for distributed memory, scalable parallel computer architectures using the Message Passing Interface library. Automatic domain decomposition of the phase space among the angular, energy, and/or spatial variables with an adaptive differencing algorithm and other numerical enhancements make PENTRAN an extremely robust solver, with >0.975 parallel code fraction (based on Amdahl's law). Numerous simulations have been performed using the PENTRAN code system, including many international benchmark computations. The many advanced numerical features in PENTRAN, including adaptive differencing with a two-level parallel angular memory structure in a scalable architecture, are such that it is well-suited for deterministic work; at present, it is likely that PENTRAN is the only deterministic code to be directly capable of rendering a solution to extremely large-scale transport problems in a rapid time using parallel computing. PENTRAN has demonstrated excellent agreement with both Monte Carlo and experimental flux measurement in a variety of problems in reactor physics, detection, and medical physics applications. The code is available free to colleges and universities through RSICC, and is commercially available for a fee. The workshop will highlight code features, supporting tools, and results from problems of interest. Course attendees with wi-fi and SSH enabled on their laptops, who also wish to have hands-on experience with the PENTRAN code during the tutorial, should send their name, email, and contact info to: glenn@hswtech.com prior to the meeting so access to the code can be set up.