The European Journal of Physics N (EPJ-N)
Not a member yet
448 research outputs found
Sort by
Representativity studies of GEN-III large cores to ZPR experiments with respect to nuclear data
No full textUncertainty quantification plays a crucial role in demonstrating the safety of nuclear reactors by assessing and accounting for the various sources of uncertainty in reactor performance predictions. This process helps establish safety margins, which are essential for ensuring that the reactor operates safely under a wide range of conditions. For existing reactors, it is mainly based on comparisons between calculations and measurements. However, the lack of experimental data in some cases (new reactor concepts, accidental conditions,…) has made the so-called “transposition”, at the very least, a complement to the latter. The most commonly used methods for this purpose rely on bayesian inference and requires a high degree of similarity between the integral parameters of the different configurations, also called representativity. This paper presents the methodology and some results of evaluated representativity factors between ZPR experiments and a Gen-III+ target core issued from the UAM benchmark at different scales and their evolution throughout the fuel cycle life, using the industrial state-of-the-art code COCAGNE. The goal is to study the relevance of such approach in an industrial context. The paper focuses on the effective multiplication factor and the center over periphery fission rate ratio. Standard (SPT) and generalized (GPT) perturbation theories are employed to determine sensitivities with respect to nuclear data and their uncertainties are propagate to the outputs through the sandwich rule with covariance data collapsed from a fine to a coarse energy mesh
Doppler broadening of neutron elastic scattering cross section using
No full textIn nuclear data processing codes, Free Gas and Crystal Lattice Models are used to account for temperature effects in neutron elastic scattering cross sections. A somewhat arbitrary thermal cutoff energy is introduced to define the energy ranges of application of the two models, given that the Crystal Lattice Model in processing codes can only be applied to constant cross sections. Despite works performed on 238U, no algorithm reported in the literature provides a unified Crystal Lattice Model valid below and above the thermal cutoff energy. This paper presents two new Doppler models, namely α0 and α′ models. The α0 model is able to account for the Doppler broadening effects in the resonance range of the neutron cross sections, while the α′ model provides a unified description of the Doppler effect over the entire energy range of interest for nuclear reactor applications. The performances of the models are illustrated with 238U, 240Pu and 237Np
Multi-output Gaussian processes for the reconstruction of homogenized cross-sections
No full textDeterministic nuclear reactor neutronics codes employing the prevalent two-step scheme often generate a substantial amount of intermediate data at the interface of their two subcodes, which can impede the overall performance of the software. The bulk of this data comprises “few-groups homogenized cross-sections” or HXS, which are stored as tabulated multivariate functions and interpolated inside the core code. A number of mathematical tools have been studied for this interpolation purpose over the years, but few meet all the challenging requirements of neutronics computation chains: extreme accuracy, low memory footprint, fast predictions. We here present a new framework to tackle this task, based on multi-output Gaussian processes (MOGP). These smooth and tunable bayesian regressors are able to model several quantities at once, and to capture the correlations between them – a key asset in the modeling of HXS’s, which we show to be highly similar from one another. Several models of this family are discussed, compared, adapted to the case of very numerous HXS’s, and their possible modeling choices are experimented on. These machine learning models enable us to interpolate HXS’s with improved accuracy compared to the current multilinear standard, using only a fraction of its training data – meaning that the amount of required precomputation is reduced by a factor of several dozens. They also necessitate an even smaller fraction of its storage requirements, preserve its reconstruction speed, and unlock new functionalities such as adaptive sampling and facilitated uncertainty quantification. We demonstrate the efficiency of this approach on a rich test case reproducing the VERA benchmark, proving in particular its scalability to datasets of millions of HXS’s
Supporting trans-national access to key nuclear research infrastructures – OFFERR and JRC OASIS: two complementary projects – One goal
No full textThe OFFER and OASIS projects both provide transnational access to key nuclear research infrastructures. Since September 2022, the OFFERR project, funded by Horizon Europe/Euratom, supports the SNETP Association by providing R&D experts access to key nuclear research infrastructures across Europe. It addresses financial and logistical barriers that hinder nuclear research by offering a platform for financial support and access to more than 230 experimental facilities. Researchers can submit applications through the OFFERR Call Platform, ensuring eligible projects receive necessary funding and access. This initiative aims to accelerate innovation in nuclear energy by bridging the gap between research ideas and advanced facilities, adhering to EU regulations and fostering international collaboration. The OASIS project is funded by an Administrative Arrangement between the Directorate General for Research and Innovation (DG RTD) and JRC since February 2020 and aims at enhancing open access to JRC's nuclear facilities (11 out of 16 are opened) and the associated technical support. So in the OASIS project, JRC makes available its nuclear research infrastructures to external users free of charge while DG RTD provides their financial support to eligible users to cover their travel and subsistence costs. This allows an optimal use of JRC's unique facilities and nuclear materials not available to European scientists at their home institutions and results in scientific excellence in research that could otherwise not be performed. With a large participation of students and young scientists, OASIS also contributes to the training of the next generation of European scientists in various nuclear fields. Whereas the two projects have one goal, they follow complementary approaches
ASVAD
No full textDuring a Long-Term Station Blackout (LTSBO) accident like the Fukushima accident, the main way to remove decay heat from Pressurized Water Reactors (PWRs) is the Steam Generators (SGs). During these accidents, a Loss of Coolant Accident (LOCA) can also occur, as the main pump seals degrade quickly without proper cooling. To cope with LOCA, all current PWRs are equipped with 3 or more accumulators. Each accumulator consists of a tank filled with subcooled borated water. These tanks are pressurized with nitrogen to a pressure of around 4.5 MPa. When the primary pressure falls below the initial accumulator pressure, its water is pushed by the pressurized nitrogen into the primary system, increasing the mass inventory and cooling the reactor. This system has the advantage of being fully passive. However, nitrogen may flow into the primary system once the water has been depleted. To avoid gas intrusion, the isolation valve must be closed in time. However, this is very difficult to do when there is no power. If the valve isn’t closed in time and successfully, nitrogen will enter the primary system as soon as the accumulator is depleted. This nitrogen soon reaches and accumulates in the SG tubes. Here, the gas will significantly decrease steam condensation, which is the main method of cooling the core. This leads to a sudden increase in primary pressure and a strong decrease in natural circulation, threatening core cooling. To avoid all these complications, a special valve has been designed. The ASVAD valve is fully passive and automatic. It automatically vents the nitrogen at the correct moment, when the accumulator empties, preventing nitrogen from reaching the SG tubes. Being fully passive and automatic guarantees its proper operation without any operator action even during LTSBO scenarios. With ASVAD, operators are relieved from managing nitrogen injection and can remain focused on recovery tasks. By allowing further primary depressurization, the valve facilitates accident recovery, gives a longer coping time. The overall safety of the PWR reactors can be improved by installing ASVAD on each accumulator
Transposition studies with a hybrid experimental database combining ZPR and PWR measurements
No full textTransposition offers the possibility of extending the conclusions of the validation of a scientific calculation tool to a wider field of use by integrating the experimental information in the calculation process. It generally results in bias adjustment and posterior uncertainty reduction. The aim of this paper is to apply a transposition method on main neutronics parameters to a Gen-III PWR benchmark core, based on a combination of measurements from both critical mock-ups and industrial reactors, thus distinguishing ourselves from most of the studies carried out on the subject. The studies carried out focus mainly on highlighting the main nuclear data requirements for reducing the uncertainties propagated in the EDF and Framatome industrial core code COCAGNE. Two observables are analyzed in this context, one is global: the critical boron concentration, rarely studied in the literature. The second is local: the center/periphery fission rate ratio. It is an indicative measure of the center/periphery power bulge inside a reactor core and has received particular attention. A general overview of the advantages of using transposition for integral quantities requiring significant improvements in nuclear data is also given. As a result, it is shown that the hybrid experimental database enables a wide range of sensitivity profiles to be covered and thus a very large number of nuclear data to be constrained, which ultimately leads to significant reductions in the uncertainties after assimilation (about 70% in the case of the fission rate ratio and 80% in the case of the critical boron concentration). The main nuclear data contributors are identified in each case and some recommendations are given to improve them
TRUST: the HPC open-source CFD platform – from CPU to GPU
No full textSince 1993, the CEA has developed TRUS
Data-driven reduced order modelling with malfunctioning sensors recovery applied to the Molten Salt Reactor case
No full textThis work presents the use of two Data-Driven Reduced Order Modelling techniques in predicting the transient response of a Molten Salt Fast Reactor when one or more sensors fail and, thus, provide wrong information; Supervised Machine Learning techniques are used to compensate for the failed sensors. Data-Driven Reduced Order Modelling integrate the physical knowledge contained in high-fidelity mathematical models with that coming from data measured on the actual system. This enables refining and updating the mathematical model, and address the challenges related to local-only observations, allowing for global state estimation. These methods are of interest when both sources of information are present, albeit incomplete, as is the case of the Molten Salt Fast Reactor. In these designs, typically operating in the fast neutron spectrum, the fuel is liquid, and no solid structures are foreseen in the core, thus making sensing and monitoring of safety-critical parameters and quantities quite challenging. Additionally, most literature studies on Data-Driven Reduced Order Modelling take the experimental observations as (noisy) ground-truth: very few works consider the case in which sensor fail or malfunction, and how this affect the state estimation
ARCHER-a Monte Carlo code for multi-particle radiotherapy through GPU-accelerated simulation and DL-based denoising
No full textThe ARCHER project was initiated about 14 years ago to explore the use of emerging GPU technologies for fast Monte Carlo (MC) calculations. This paper presents the latest work to integrate the newly developed deep conventional neural network (dCNN) based MC denoising method with GPU-based MC multi-particle radiation transport simulation method to demonstrate a real-time dose computing capability for clinically realistic radiotherapy examples. The computing process involves GPU-based dose calculations that is followed by dCNN-based denoising. The dCNN-based dose denoiser is designed and employed to reduce the statistical uncertainty in dose distributions in patient anatomy defined by 3D computed tomography (CT) images. The training data include a range of dose distributions covering low-count/high-noise (DoseLCHN) and high-count/low-noise (DoseHCLN). The extremely large DoseLCHN and DoseHCLN dataset was generated from ARCHER. The DoseLCHN dataset is input into the trained model to output a predicted DoseHCLN dataset. For the evaluation, the DoseHCLN dataset produced by ARCHER is considered to be the ground truth. Experimental results show that the dose distributions generated from newly proposed method agreed consistently with the DoseHCLN produced from ARCHER. For hundreds of patient radiation treatment cases involving photons and protons, the average running time for one patient (GPU-based dose simulation followed by dCNN-based denoising) is about 200 ms. These preliminary results have demonstrated the feasibility of real-time Monte Carlo dose computing using an integrated dCNN-based denoising and GPU-based dose calculational approach. On-going studies involving more radiation types and clinical procedures are expected to facilitate the use of real-time MC dose planning and verification in the clinical workflow
The MCNP
No full textAfter several years of effort involved in merging the Los Alamos National Laboratory MCNP5 and MCNPX codes, in 2013 the first production release of version 6 of the Monte Carlo N-Particle®, or MCNP®, code MCNP6.1 was distributed publicly. Since then, three significant releases have been issued: MCNP6.1.1beta in 2014, MCNP6.2 in 2018, and MCNP6.3 in 2023. While each release always contains new features, code enhancements, and bug fixes, each version has had a different primary focus, ranging from improved calculational efficiency to new powerful utilities and tools, to software modernization of the code base. With all that has been learned over the first decade of the MCNP6 code, continuous progress is being made toward a modernized, general-purpose Monte Carlo radiation transport code that remains a trusted resource for the global community of practitioners. This paper describes these first 10+ years of the MCNP6 code and its continually improving data libraries, and gives some insight into how the next decade is expected to unfold