Application d'un modèle de fuites simplifié pour le calcul d'assemblage de réacteurs nucléaires en environnement

Elouan Madic

Master's thesis (2026)

Access this document
Available
Open access to the full text in PolyPublie
Text Text • 6MB •

Abstract

In nuclear physics, the calculation of a reactor core is classically performed in two steps. The first one consists in solving the transport equation on an elementary part of the reactor, usu-ally a fuel assembly. During this step, one generates multi-parametric libraries containing the quantities such as cross sections or diffusion coefficients. The second step of the calculation relies on these libraries and interpolates them to perform a fast and precise calculation of the complete reactor core. During the first step of the calculation, one usually assumes that the assembly studied is repeated and paves the complete space, thus forming an infinite lattice. This assumption is however a strong one, since the reactor core is not uniform and is composed of different assemblies with different neutron behavior. Therefore, one often choses to introduce in the transport equation a neutron leakage term, which models the neutron leakage within the reactor and the exchanges between the different assemblies. Usually, a homogeneous leakage model is used, which provides an isotropic and homogeneous leakage probability on the whole assembly. Heterogeneous leakage models also exist and allow to model space-dependent leak-age on the assembly at the cost of a longer calculation time. The reactor core is composed of various assemblies with different neutron behavior. Those assemblies can vary in terms of the fuel that is used (Uranium Oxide, also called UOX, or Mixed Oxide, also called MOX), on their irradiation received in the core, or on the presence of control rods or neutron poisons. The assumption made of an infinite lattice calculation is therefore questionable. Indeed, during a complete core calculation, the calculated assemblies can face environments very different from the one in which they were calculated (themselves, in an infinite lattice calculation). The work proposed in this thesis is to calculate these assem-blies in an environment closer to the one they might face during a complete core calculation. To do so we use the lattice calculation code Dragon5 to calculate these assemblies with an environment. We do so by placing the assembly of interest in the middle of a 3 × 3 assem-blies cluster. However, by adding an environment, the use of an homogeneous leakage model becomes questionable, since it would provide a diffusion coefficient for the whole cluster and not for the calculated assembly. To palliate this, we chose to use a simplified B1 heteroge-neous leakage model present in Dragon5. This model allows the user to calculate leakage coefficients that are homogenous by region, thus allowing to decouple the leaks in the central assembly and in the rest of the cluster.

Program:
Génie énergétique
Academic/Research Directors :
Institution:
Polytechnique Montréal
OAI:
oai:publications.polymtl.ca:76780
ORCID
Date Deposited:
15 Jul 2026 13:15
Last Modified:
13 Oct 2026 15:26
Cite in APA 7:
Madic, E. (2026). Application d'un modèle de fuites simplifié pour le calcul d'assemblage de réacteurs nucléaires en environnement [Master's thesis, Polytechnique Montréal]. PolyPublie. https://publications.polymtl.ca/76780/

Statistics

Total downloads

Downloads per month in the last year

Origin of downloads

Repository Staff Only

View Item
View Item