Skip to Main content Skip to Navigation

Développements et validation de calculs à énergie continue pondérés par l'importance

Guillaume Truchet 1, 2
1 LEPH - Laboratoire d'Etudes de PHysique
SPRC - Service de Physique des Réacteurs et du Cycle : DEN/DER
Abstract : A key issue in nowadays Reactor Physics is to propagate input data uncertainties (e.g. nuclear data, manufacturing tolerances, etc.) to nuclear codes final results (e.g. keff, reaction rate, etc.). In order to propagate uncertainties, one typically assumes small variations around a reference and evaluates at first sensitivity profiles. Problem is that nuclear Monte Carlo codes are not -- or were not until very recently -- able to straightforwardly process such sensitivity profiles, even thought they are considered as reference codes.First goal of this PhD thesis is to implement a method to calculate keff-sensitivity profiles to nuclear data or any perturbations in TRIPOLI-4, the CEA Monte Carlo neutrons transport code. To achieve such a goal, a method has first been developed to calculate the adjoint flux using the Iterated Fission Probability (IFP) principle that states that the adjoint flux at a given phase space point is proportional to the neutron importance in a just critical core after several power iterations. Thanks to our developments, it has been made possible, for the fist time, to calculate the continuous adjoint flux for an actual and complete reactor core configuration. From that new feature, we have elaborated a new method able to forwardly apply the exact perturbation theory in Monte Carlo codes. Exact perturbation theory does not rely on small variations which makes possible to calculate very complex experiments.Finally and after a deep analysis of the IFP method, this PhD thesis also reproduces and improves an already used method to calculate adjoint weighted kinetic parameters as well as reference migrations areas.
Complete list of metadatas

Cited literature [126 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Wednesday, November 4, 2015 - 4:12:06 PM
Last modification on : Monday, May 18, 2020 - 7:50:46 PM
Document(s) archivé(s) le : Friday, February 5, 2016 - 11:17:25 AM


Version validated by the jury (STAR)


  • HAL Id : tel-01224452, version 1



Guillaume Truchet. Développements et validation de calculs à énergie continue pondérés par l'importance. Génie des procédés. Université Grenoble Alpes, 2015. Français. ⟨NNT : 2015GREAI037⟩. ⟨tel-01224452⟩



Record views


Files downloads