Canada Spends Logo

Explorez notre analyse du budget 2026

Grant: Discovery Grants Program - Individual

École Polytechnique de Montréal

$27,000.00

Nuclear reactor full-core deterministic computational schemes in transport theory

EF 2011-12

Résumé
We are focusing on full-core reactor physics codes required for simulation of operation, fuel management and safety analyses of third- and fourth-generation reactors. Neutron diffusion theory is currently used for this purpose, but reference calculations require a more accurate representation of transport phenomena. Based on recent advances related to the MOC and to the evolution of computer clusters, we can now propose the implementation of a family of rigorous full-core solution technique relying on the discrete ordinates method (for regular geometries) and on the MOC (for unstructured geometries). The MOC we propose to develop is an original contribution based on: 1. linear sources. Scattering and fission sources are assumed to vary linearly along each track segment crossing the domain. A parabolic order diamond differencing scheme is used to obtain the flux moments in each volume. Most current implementations are based on flat sources approximation and step-characteristic differencing. 2. scattering anisotropy. With the MOC solution, we propose to implement a variable-order Legendre expansion model for scattering anisotropy. 3. synthetic acceleration. The MOC already implemented in DRAGON4 code is accelerated using a second-order algebraic collapsing (AC) method limited to flat sources and isotropic scattering. We propose to replace the second-order AC method with a first-order AC method and to generalize its scattering representation to higher orders. 4. parallelism. We propose to use a capability of Fortran-2008 ISO standard named ""coarray"" to implement a massively-parallel solution technique adapted to the MOC and to the AC method. The DRAGON4 implementation of the MOC is currently not parallelized. Using the Unified Portable C (UPC) is an alternative massively-parallel solution technique for algorithms written in C language.
Programme
Discovery Grants Program - Individual
Emplacement
Quebec, Canada
Type de bénéficiaire
Academia
Type
Grant
Source
NSERC's Awards Data

Tel que publié

Application id
107847-2011
Name nom
Hébert, Alain
Department département
Génie mécanique
Organization id
64
Institution établissement
École Polytechnique de Montréal
Province en
Québec
Province fr
Québec
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2011
Competition year année de concours
2011
Award amount
27000
Program id
RGPIN
Program nane en
Discovery Grants Program - Individual
Program name fr
Programme de subventions à la découverte - individuelles
Group en
Discovery grants programs
Group fr
Programmes de subventions à la découverte
Committee code
1512
Committee name en
Mechanical Engineering
Committee name fr
Génie mécanique
Area of application code
300
Area of application group en
Energy resources
Area of application group fr
Ressources d'énergie
Area of application en
Energy resources (including production, exploration, processing, distribution and use)
Area of application fr
Ressources d'énergie (production, exploration, traitement, distribution et utilisation de l'énergie)
Research subject code
2401
Research subject group en
Nuclear engineering
Research subject group fr
Génie nucléaire
Research subject en
Reactor design and operation
Research subject fr
Conception et fonctionnement des réacteurs
Partie
1
Nb partie
5
Application title
Nuclear reactor full-core deterministic computational schemes in transport theory
Keyword
NSERC
Application summary
We are focusing on full-core reactor physics codes required for simulation of operation, fuel management and safety analyses of third- and fourth-generation reactors. Neutron diffusion theory is currently used for this purpose, but reference calculations require a more accurate representation of transport phenomena. Based on recent advances related to the MOC and to the evolution of computer clusters, we can now propose the implementation of a family of rigorous full-core solution technique relying on the discrete ordinates method (for regular geometries) and on the MOC (for unstructured geometries). The MOC we propose to develop is an original contribution based on: 1. linear sources. Scattering and fission sources are assumed to vary linearly along each track segment crossing the domain. A parabolic order diamond differencing scheme is used to obtain the flux moments in each volume. Most current implementations are based on flat sources approximation and step-characteristic differencing. 2. scattering anisotropy. With the MOC solution, we propose to implement a variable-order Legendre expansion model for scattering anisotropy. 3. synthetic acceleration. The MOC already implemented in DRAGON4 code is accelerated using a second-order algebraic collapsing (AC) method limited to flat sources and isotropic scattering. We propose to replace the second-order AC method with a first-order AC method and to generalize its scattering representation to higher orders. 4. parallelism. We propose to use a capability of Fortran-2008 ISO standard named ""coarray"" to implement a massively-parallel solution technique adapted to the MOC and to the AC method. The DRAGON4 implementation of the MOC is currently not parallelized. Using the Unified Portable C (UPC) is an alternative massively-parallel solution technique for algorithms written in C language.