The Experts below are selected from a list of 17691 Experts worldwide ranked by ideXlab platform
Guillet Martin - One of the best experts on this subject based on the ideXlab platform.
-
Prise en compte des calculs de cœur via l'utilisation de DONJON5 pour le code de scénario nucléaire CLASS et évaluation de l'impact sur les inventaires isotopiques
2019Co-Authors: Guillet MartinAbstract:Les codes de scénarios nucléaires sont destinés à donner une visibilité sur les inventaires radiotoxiques, la consommation de ressources et la quantité de matière disponible pour le recyclage en fonction des options technologiques, du nombre de réacteurs opérationnels et de leurs conditions d’opération. Ces codes nécessitent de modéliser les éléments du cycle et leurs interactions, de l’unité fabrication du combustible à celle de retraitement en passant par les réacteurs. La simulation de la physique d’un réacteur nucléaire nécessite de résoudre l’équation du transport des neutrons, opération complexe et extrêmement coûteuse, qui n’est en général pas effectué à l’échelle coeur et, a fortiori, pas plus dans les codes de scénarios où plusieurs centaines de cycles sont simulés. Néanmoins, c’est à cette échelle que les phénomènes de fuites neutroniques et d’hétérogénéités prennent toute leur importance. L’objet de ce travail est de quantifier l’impact des calculs de coeurs à l’échelle du scénario. Une étude comparative est menée sur le code de scénario CLASS. D’une part, des simulations sont réalisées en utilisant des réseaux de neurones qui, à partir de calculs de réseaux infinis, permettent de s’extraire de l’échelle coeur tout en permettant de traiter des combustibles à priori inconnus. D’autre part, sont établis des modèles de réacteurs pour le code de scénario, qui font explicitement appel à des calculs de coeurs en 3D dans l’approximation de la diffusion avec le code DONJON5. Ces derniers nécessitent la création préalable d’un objet dans lequel est stocké l’ensemble des données nucléaires issues d’un calcul de réseau DRAGON5. L’établissement des schémas de calculs de réseau et de coeur a nécessité des simplifications et des optimisations poussées afin d’assurer la création d’une base de données contenant les résultats de milliers de calculs assemblage et des calculs de coeurs suffisamment rapides pour assurer des temps de calcul des scénarios raisonnables. Le défi principal de ce projet a été d’inclure les combustibles MOx. En effet, afin de prendre correctement en compte la physique de ce type de combustible, le calcul en diffusion requiert l’interpolation des différents paramètres sur une grille à huit dimensions (burnup, enrichissement, concentration de bore, abondance isotopique en plutonium 238, 240, 241, 242 et américium 241) au lieu de trois pour le combustible UOx (burnup, enrichissement, proportion de poison soluble), ce qui s’avère être extrêmement coûteux. Les résultats sur des scénarios de référence élémentaires et complexes sont comparés afin d’identifier et quantifier les sources d’erreur de la modélisation par les réseaux de neurones comparativement à des calculs de coeurs.----------Abstract Scenario codes are designed to provide a long-term vision about the production of radio toxic inventories, requirements of resources, and estimation of available resources for reprocessing. These codes require the modeling of each Nuclear fuel cycle element as well as their interactions. They include the models of both the fuel fabrication plants and the reactors. The modeling of Nuclear reactors entails solving the neutron transport equation which is very costly and is typically not done at the scale of a reactor. Implementing full-core transport computations in scenarios is impractical since Nuclear scenarios involve hundreds of irradiation cycles. Additionally, in-core heterogeneous effects such as bundle burnup discrepancies or neutron leakage become significant at this scale. This project assesses the impact of full-core calculations in Nuclear scenario simulations with CLASS. In this dynamic scenario code, neural networks based on infinite assembly calculations are generally used to avoid full-core calculations while dealing with a priori unknown fuel composition. This project creates new reactor models calling explicit 3D diffusion full-core computations with DONJON5. These models require infinite assembly databases previously computed with DRAGON5. Some simplifications and optimizations are necessary before processing hundreds of infinite assembly databases as well as for quick and accurate full-core calculations. The main challenge is to include unknown MOx fuels considering the needed interpolation of every Nuclear Property in an eight dimensional grid (burnup, enrichment, poison boron concentration, the mass proportion of plutonium 238, 240, 241, 242 and americium 241). On the other hand, UOx fuels require only a three-dimensional grid (burnup, enrichment, boron concentration) which is considerably less challenging in terms of computing time. This project compares the scenario simulation results on both elementary and complex situation with different fuel management options to identify and quantify main sources of error in the neural network approach. Ultimately, the thesis makes several suggestions to improve inventories prediction in CLASS while keeping a relatively low computing time
春日井 敦 - One of the best experts on this subject based on the ideXlab platform.
-
Conceptual Design of Test Modules for DEMO Blanket, Diagnostic Device, and RI Production for A-FNS
2019Co-Authors: 太田 雅之, 佐藤 聡, 中村 誠, 権 セロム, 朴 昶虎, 落合 謙太郎, 染谷 洋二, 春日井 敦Abstract:In the conceptual design activity of advanced fusion neutron source A-FNS, a variety of test modules for fusion DEMO reactor are planned. In these modules, progresses of design activities on Blanket Nuclear Property Test Module (BNPTM) and Diagnostic and Control Device Test Module (DCDTM) were reported. The BNPTM is a module in order to evaluate accuracies of Nuclear analyses of the DEMO blanket such as tritium production rate. The influences of the cooling water and test cell wall were evaluated in order to decide its design. The DCDTM is one to achieve irradiation data of functional materials on diagnostic and control devices such as mirror and window. It was clarified from its Nuclear analysis that the neutron fluence obtained in the DCDTM was enough to investigate the accumulated irradiation effects on the materials. In addition to the irradiation tests on the fusion reactor materials, various neutron applications are planned at A-FNS. One of the applications is medical isotope production. We clarified that enough amount of the medical isotope molybdenum-99 could be produced in comparison with the demand in Japan by using Radio-Isotope Production Module (RIPM). The RIPM affects little influence on the fusion reactor material irradiation tests.第14回核融合核技術国際シンポジウム(ISFNT-14
Viktoria Weber - One of the best experts on this subject based on the ideXlab platform.
-
neutronic analysis for the ifmif eveda reference test cell and test facility
Fusion Engineering and Design, 2014Co-Authors: Keitaro Kondo, A Serikov, U Fischer, Frederik Arbeiter, Dennis Grose, V Heinzel, A Klix, Martin Mittwollen, Kuo Tian, Viktoria WeberAbstract:Abstract The IFMIF test cell (TC) design has been further developed and optimized in the EVEDA phase, and finally the reference TC design has been proposed. In order to carry out the detailed neutronic analysis for the reference TC design, a very detailed geometrical model for Monte Carlo neutronic calculations has been prepared directly from engineering CAD data by utilizing the McCad conversion software developed at KIT. The geometrical model includes the detailed descriptions of the lithium target system proposed by Japan, all test modules based on the EVEDA phase design, and the 3-dimesional arrangement of the biological shielding. The Monte Carlo code McDeLicious, which is an enhancement to MCNP5, has been utilized in order to adequately simulate the neutron and photon productions from the 6,7Li(d,xn) reactions in the lithium target. The present analysis is focusing on the Nuclear heating distribution inside the biological shielding, the Nuclear Property of the TC liner, and the biological dose distribution around TC during operation. Some countermeasures for reducing the He production in the liner are discussed.
Marcelino Marcello - One of the best experts on this subject based on the ideXlab platform.
-
The compounding parameter and L2 aquisition/learning
[s.n.], 2021Co-Authors: Marcelino MarcelloAbstract:Orientadores: Mary Aizawa Kato, Ruth LopesTese (doutorado) - Universidade Estadual de Campinas, Instituto de Estudos da LinguagemResumo: Esta tese investiga o Parâmetro de Composição (PC) conforme formulado por Snyder (1995) e seu papel na aquisição/aprendizagem de inglês por aprendizes brasileiros. A definição positiva do PC [+] permite a uma língua marcar livremente qualquer item lexical da classe aberta como [+Afixal] e conseqüentemente, detonar toda uma série de propriedades relacionadas (cluster), a saber, composição nominal (N+N), estruturas resultativas (ER), construções V+partícula (V+PRT), construções com objeto duplo (DOC) e isolamento de preposição (PrepStr), entre outras. Diante da aparente existência, em PB, de algumas estruturas semelhantes às do inglês, proponho-me a investigar as seguintes questões: (i) através de comparação entre o PB e o inglês, é possível trazer evidências de que todas as propriedades resultantes da definição positiva do PC, ou pelo menos parte delas estão ligadas ao mesmo parâmetro? (ii) se houver tal parâmetro, a aquisição do inglês como L2 apresenta semelhanças com a aquisição de L1 em relação ao comportamento relativo ao mesmo parâmetro? Após comparação e análise das cinco estruturas propostas, descobri que nenhuma delas, nem mesmo as com correlato estrutural superficial, resulta, no PB, da definição positiva do PC. N+N em PB exemplifica um tipo de composição nominal não recursivo com rigidez de significado; as ERs encontradas em PB são do tipo semântico e não correspondem às versões sintáticas licenciadas pelo PC [+]; estruturas V+PRT e COD são inexistentes; por fim, PB apresenta estruturas com ausência de preposição, que resultam de um diferente arranjo de itens especificados na numeração, diferentemente de estruturas com isolamento de preposição, que são exemplos de predicados complexos decorrentes de Reanálise. Em relação à aquisição das cinco propriedades do PC, sugiro, após análise dos dados de um experimento, que a aquisição das propriedades do (PC) em L2 não é semelhante à sua aquisição em L1. Em L2, as propriedades Nucleares (resultativas, V+Partícula,) parecem ter sido adquiridas pelos falantes avançados, possivelmente via imersão. As estruturas não Nucleares (COD e PrepStr) juntamente com a estrutura Nuclear N+N foram aprendidas via instrução formal. Essas últimas parecem permanecer disponíveis na forma de conhecimento lingüístico conscienteAbstract: This dissertation investigates the Compounding Parameter (CP) as formulated by Snyder (1995) and its role in the acquisition/learning of English by Brazilian learners. The positive setting of the CP [+] allows a given language to freely mark any open-class lexical item as [+Affixal] and consequently trigger a cluster of related properties, namely N+N compounding (N+N), resultative structures (RS), V+Particle (V+PRT) constructions, double object constructions (DOC) and preposition stranding (PrepStr), among others. Upon the apparent existence of structures in Brazilian Portuguese (BP) that resemble some of those in English, namely RS and PrepStr, I set out to explore the following questions: (i) does the comparison between Brazilian Portuguese and English offer evidence that all the properties attributed to the positive setting of the CP, or part of them, follow from the same parameter? (ii) assuming the existence of this parameter, are there any similarities between L1 and L2 acquisition as far as the CP is concerned? After the comparison and analysis of the five related properties, I found out that none of the structures, even the superficially identical looking ones, resulted from the positive setting of the CP. N+N in BP is a result of non recursive N+N compounding with lexically rigid meaning; the RSs found in BP are semantic resultatives that do not correspond to the syntactic versions of the CP-positively valued ones; the V+Particle and DOC constructions are nonexistent; finally, BP offers prepositionless structures that result from a different array of items specified in the numeration, differently from the Reanalysis of V+Preposition, which allows the preposition to be stranded in English. As for the acquisition of the five CP properties, I suggest, after analyzing the data of an experimental study, that the L2 acquisition processes differ from those of the L1 in that the non-Nuclear properties (DOC and PrepStr) along with the Nuclear Property N+N seem to have been learned via formal instruction (ordered input, explicit positive and negative evidence) and remain available as conscious linguistic knowledge. The Nuclear properties (RS and V+PRT), on the other hand, seem to have been acquired by the highly proficient L2 English speakers, probably via immersionDoutoradoDoutor em Linguístic
太田 雅之 - One of the best experts on this subject based on the ideXlab platform.
-
Conceptual Design of Test Modules for DEMO Blanket, Diagnostic Device, and RI Production for A-FNS
2019Co-Authors: 太田 雅之, 佐藤 聡, 中村 誠, 権 セロム, 朴 昶虎, 落合 謙太郎, 染谷 洋二, 春日井 敦Abstract:In the conceptual design activity of advanced fusion neutron source A-FNS, a variety of test modules for fusion DEMO reactor are planned. In these modules, progresses of design activities on Blanket Nuclear Property Test Module (BNPTM) and Diagnostic and Control Device Test Module (DCDTM) were reported. The BNPTM is a module in order to evaluate accuracies of Nuclear analyses of the DEMO blanket such as tritium production rate. The influences of the cooling water and test cell wall were evaluated in order to decide its design. The DCDTM is one to achieve irradiation data of functional materials on diagnostic and control devices such as mirror and window. It was clarified from its Nuclear analysis that the neutron fluence obtained in the DCDTM was enough to investigate the accumulated irradiation effects on the materials. In addition to the irradiation tests on the fusion reactor materials, various neutron applications are planned at A-FNS. One of the applications is medical isotope production. We clarified that enough amount of the medical isotope molybdenum-99 could be produced in comparison with the demand in Japan by using Radio-Isotope Production Module (RIPM). The RIPM affects little influence on the fusion reactor material irradiation tests.第14回核融合核技術国際シンポジウム(ISFNT-14
