ETD

• DF
• Diffusion
• Fuel Doppler Temperature
• Full 3D equivalence
• GET
• Hi2Low
• NEM
• Serpent

This PhD research is part of a broader initiative led by NINE (Nuclear and Industrial Engineering) to develop an advanced multi-physics, multi-scale simulation platform aimed at enhancing nuclear power plant safety analysis. The increasing complexity of nuclear systems, combined with the need for improved predictive accuracy in safety
and design modeling, has driven international research efforts towards the development of high-fidelity multi-physics tools.
The core objective of this research is to advance the reactor physics module of the multiphysics platform, particularly focusing on neutron behavior modeling, cross-section generation, and diffusion solutions. A key contribution of this work is the implementation of the full 3D equivalence between the Serpent 2 Monte Carlo code and the Nodal Expansion Method (NEM) diffusion code. This approach enhances the accuracy of coarse-mesh calculations, making them comparable to high-fidelity Monte Carlo simulations while significantly reducing computational costs.
The research is structured around two primary methodologies:
• High-Fidelity (HiFi) Multi-Physics Modeling: Traditional reactor physics tools typically employ low-fidelity models with simplified coupling between neutron transport, thermal-hydraulics, and fuel performance. In contrast, novel highfidelity tools integrate these phenomena in real-time, capturing local interactions
such as the Doppler effect and coolant flow variations with higher precision.
However, these HiFi tools are computationally demanding, restricting their application to reference calculations and targeted studies.
• Hi2Lo Approach for Enhanced Computational Efficiency: To bridge the gap between high-fidelity and traditional methods, this research leverages a High-toLow (Hi2Lo) approach to transfer high-fidelity Monte Carlo-calculated nuclear data into a computationally efficient diffusion model. The methodology ensures that reactor safety analyses maintain Monte Carlo accuracy while achieving feasible runtimes.
A critical aspect of this work is the development of a full 3D equivalence framework between Serpent 2 and NEM, allowing for reliable homogenization of nuclear data. A novel subroutine for side-dependent discontinuity factors (DFs) was introduced, significantly enhancing power distribution predictions, particularly in axial and radial flux calculations. Extensive code-to-code validation was conducted using the C5G7 mini-core benchmark, demonstrating the effectiveness of the proposed methodology in achieving accurate power distribution and reactivity predictions.
Another major contribution of this research is the development of a refined methodology for modeling fuel Doppler temperature, a critical factor influencing reactor kinetics. Traditional approaches using a single effective temperature have shown nonconservative behavior under certain conditions. The new methodology improves accuracy by refining temperature distribution modeling, leading to better reactivity estimates in both fresh fuel and burnup scenarios. This approach was validated through depletion analyses of UO2 fuel assemblies and full-core C5G7 benchmark simulations, confirming that the effective temperature assumption underestimates core reactivity.
Future Perspectives:
While the research presents significant advancements in multi-physics safety analysis, computational cost remains a challenge. The next steps involve:
• Optimization of Monte Carlo Simulations: Implementing variance reduction techniques and acceleration methods to improve computational efficiency.
• Thermal-Hydraulic Data Extrapolation: Developing interpolation schemes for homogenized nuclear data to minimize the need for frequent Monte Carlo runs during transient scenarios.
• Integration into NINE’s Multi-Physics Platform: The developed methodologies will be incorporated into the broader NINE simulation framework, enhancing its capability to perform best-estimate safety analyses.
This PhD research aligns with ongoing global efforts to improve nuclear safety and operational flexibility through advanced simulation tools. The work contributes to the international community by offering a validated methodology for high-accuracy, efficient reactor modeling, ensuring that nuclear power plants can operate safely under diverse and challenging conditions.