Vous êtes ici : GDR - Version française > Annonces > Thèses et Post-docs
-
Partager cette page
Appel à candidatures
|
Recherche, Emploi
PhD on multi-scale nuclear reactor core simulation
Du 1 septembre 2025 au 31 août 2028
Cadarache (France)
Contacts : daniele.vivaldi@asnr.fr
Contacts : daniele.vivaldi@asnr.fr
Fluid-structure interactions (FSI) in a nuclear reactor core involve thermalhydraulics mechanisms at different scales. FSI simulations at the local (CFD) scale are feasible only for reduced domains. Such domains must however be representative of the actual core conditions. A multi-scale approach is therefore required to understand how to consistently simulate the core thermalhydraulics and the related FSI phenomena.
PhD topic context:
A nuclear reactor core is made of thousands of cylinders containing Uranium oxide and subjected to an axial water flow. The flow is highly turbulent and can induce vibration of the cylinders that may result in fretting wear [1]. This fluid structure interaction (FSI) issue is driven by the specific flow around the structures.
Two main flow scales can be identified: first, a local scale, represented by the fluid flowing around the fuel rod, that induces the forces on the solid walls and, in turn, is at some extent modified by the structure displacement. Then, a component scale, represented by the global coolant flow distribution inside the core, that can feature regions with higher velocity, asymmetries, periodic flow behavior, local increased cross-flow, etc. The local scale is characterized by a certain degree of coupling between the fluid flow and the structure displacement: a consistent numerical simulation of this scale requires CFD scale approaches and fluid-structure coupling. On the other hand, the impact of the structure deformation and displacement on the component scale can be neglected: the global flow distribution inside a core/RPV can be modeled by averaged approaches, such as porous media simulations.
PhD goals:
Coupled FSI simulations at CFD scale have become feasible in the recent year, see for example [2,3,4]. However, their applicability is limited to small domains (one or a few cylinders). Such reduced scale of application requires an assumption of the local flow boundary conditions to impose: an evaluation of their representativity for the application to an arbitrary core configuration is, therefore, mandatory. The goal of this PhD is the comprehension of the physical coupling degree between the different scales, the definition of the sizes of representative domains and the impact of boundary conditions.
To do that, the PhD candidate will have to study how different boundary conditions (periodicity, imposed velocity and turbulence, etc.) imposed at the CFD scale influence parameters such as pressure, velocity and temperature and how they propagate within the fuel assembly. The importance of explicitly simulating local discontinuities such a spacer grids will have to be accessed.
The goal is to determine what are reduced-size domains representative (compared to the real geometry) in terms of thermal-hydraulic and FSI behaviors.
The study will focus on the interpretation of CFD results based on LES and/or hybrid URANS/LES approaches et on porous media for the component scale. To validate the methodology, CFD simulations will be compared to available experimental results (average and rms velocity profiles and force spectra), obtained on simplified fuel assembly configurations [5]. Once a methodology developed for the simplified experimental case, the study of a PWR core configuration at prototypical scale will be realized. The porous media simulation will be used to determine the coolant flow patterns in different regions of the core, as a function, for example, of different supposed flow rate distributions at the core inlet.
To realize the numerical simulations (both for the porous media and CFD scales), the CFD codes code_Saturne or Ansys Fluent will be used. The focus of the PhD is on the fluid dynamics of the flow, i.e., without considering the coupling with the structure displacements. Nevertheless, once this step finalized, fluid-structure interaction simulations will be possible following a coupling approach proposed implemented inside code_Saturne [6].
Profile:
Master Degree in Engineering (Nuclear, Mechanical, Energy).
Knowledge of fluid dynamics, numerical simulation, CFD.
References:
[1] I. A. E. Agency, Review of Fuel Failures in Water Cooled Reactors., Technical Report, IAEA-NF-T-2, 2010.
[2] Hofstede, E., Shams, A., and van Zuijlen, A., 2015, “Numerical prediction of flow induced vi-brations in nuclear reactor applications”, Conference: 16th International Topical Meeting on Nuclear Reactor Thermal Hydraulics (NURETH-16)At: Chicago, USA, 09.
[3] Brockmeyer, L., Merzari, E., Solberg, J., Karazis,K., and Hassan, Y., 2019, “High fidelity simulation and validation of crossflow through a tube bundle and the onset of vibration”, International Journal of Non-Linear Mechanics, 117, p. 103231.
[4] D. Vivaldi, 2024. An assessment of CFD-scale fluid–structure interaction simulations through comprehensive experimental data in cross-flow. Computers & Fluids, 278, p. 106303
[5] N. Turankok, et al., 2020. Unsteady pressure and velocity measurements in 5 × 5 rods bundle using grids with and without mixing vanes. Nuclear Engineering and Design 364 110687.
[6] Vivaldi, D. & Ricciardi, G., 2024. Optimizing Coupled Fluid-Structure Simulations for Nuclear-Relevant Geometries. ASME J. of Pressure Vessel Tech.
A nuclear reactor core is made of thousands of cylinders containing Uranium oxide and subjected to an axial water flow. The flow is highly turbulent and can induce vibration of the cylinders that may result in fretting wear [1]. This fluid structure interaction (FSI) issue is driven by the specific flow around the structures.
Two main flow scales can be identified: first, a local scale, represented by the fluid flowing around the fuel rod, that induces the forces on the solid walls and, in turn, is at some extent modified by the structure displacement. Then, a component scale, represented by the global coolant flow distribution inside the core, that can feature regions with higher velocity, asymmetries, periodic flow behavior, local increased cross-flow, etc. The local scale is characterized by a certain degree of coupling between the fluid flow and the structure displacement: a consistent numerical simulation of this scale requires CFD scale approaches and fluid-structure coupling. On the other hand, the impact of the structure deformation and displacement on the component scale can be neglected: the global flow distribution inside a core/RPV can be modeled by averaged approaches, such as porous media simulations.
PhD goals:
Coupled FSI simulations at CFD scale have become feasible in the recent year, see for example [2,3,4]. However, their applicability is limited to small domains (one or a few cylinders). Such reduced scale of application requires an assumption of the local flow boundary conditions to impose: an evaluation of their representativity for the application to an arbitrary core configuration is, therefore, mandatory. The goal of this PhD is the comprehension of the physical coupling degree between the different scales, the definition of the sizes of representative domains and the impact of boundary conditions.
To do that, the PhD candidate will have to study how different boundary conditions (periodicity, imposed velocity and turbulence, etc.) imposed at the CFD scale influence parameters such as pressure, velocity and temperature and how they propagate within the fuel assembly. The importance of explicitly simulating local discontinuities such a spacer grids will have to be accessed.
The goal is to determine what are reduced-size domains representative (compared to the real geometry) in terms of thermal-hydraulic and FSI behaviors.
The study will focus on the interpretation of CFD results based on LES and/or hybrid URANS/LES approaches et on porous media for the component scale. To validate the methodology, CFD simulations will be compared to available experimental results (average and rms velocity profiles and force spectra), obtained on simplified fuel assembly configurations [5]. Once a methodology developed for the simplified experimental case, the study of a PWR core configuration at prototypical scale will be realized. The porous media simulation will be used to determine the coolant flow patterns in different regions of the core, as a function, for example, of different supposed flow rate distributions at the core inlet.
To realize the numerical simulations (both for the porous media and CFD scales), the CFD codes code_Saturne or Ansys Fluent will be used. The focus of the PhD is on the fluid dynamics of the flow, i.e., without considering the coupling with the structure displacements. Nevertheless, once this step finalized, fluid-structure interaction simulations will be possible following a coupling approach proposed implemented inside code_Saturne [6].
Profile:
Master Degree in Engineering (Nuclear, Mechanical, Energy).
Knowledge of fluid dynamics, numerical simulation, CFD.
References:
[1] I. A. E. Agency, Review of Fuel Failures in Water Cooled Reactors., Technical Report, IAEA-NF-T-2, 2010.
[2] Hofstede, E., Shams, A., and van Zuijlen, A., 2015, “Numerical prediction of flow induced vi-brations in nuclear reactor applications”, Conference: 16th International Topical Meeting on Nuclear Reactor Thermal Hydraulics (NURETH-16)At: Chicago, USA, 09.
[3] Brockmeyer, L., Merzari, E., Solberg, J., Karazis,K., and Hassan, Y., 2019, “High fidelity simulation and validation of crossflow through a tube bundle and the onset of vibration”, International Journal of Non-Linear Mechanics, 117, p. 103231.
[4] D. Vivaldi, 2024. An assessment of CFD-scale fluid–structure interaction simulations through comprehensive experimental data in cross-flow. Computers & Fluids, 278, p. 106303
[5] N. Turankok, et al., 2020. Unsteady pressure and velocity measurements in 5 × 5 rods bundle using grids with and without mixing vanes. Nuclear Engineering and Design 364 110687.
[6] Vivaldi, D. & Ricciardi, G., 2024. Optimizing Coupled Fluid-Structure Simulations for Nuclear-Relevant Geometries. ASME J. of Pressure Vessel Tech.