This PhD work was motivated by the CEA R&D program to provide solid technological basis for the use of Brayton power conversion system in Sodium-cooled Fast nuclear Reactors (SFRs). Multi-channel compact heat exchangers are necessary for the present application because of the low heat transfer capacity of the gas foreseen. In ASTRID project, a minimum size of Na channels section is required to avoid the plugging risk. However, this induces very low pressure losses in the bundle. Considering an additional inlet flow condition, a real risk of bad flow distribution remains. As a result, the thermal performance and thermal loading of the heat exchanger degrades due to it. The main goal of this work was to overcome the flow maldistribution problem by means of an innovative design of sodium distribution system (PATENT FR1657543), the development of a numerical strategy and the construction of an experimental database to validate all theoretical studies. The innovative sodium distribution system consists on an inlet header which tries to guide the evolution of the impinging jet flow while a system of bifurcating pre-distribution channels increases pressure drops in the bundle. Lateral communications between pre-distribution channels are introduced to further homogenize the flow. Two experimental facilities have been conceived to study the flow behavior in bifurcating channels and in the inlet header, respectively. At the same time, their effect on the flow distribution between channels is evaluated. The acquired PIV aerodynamic database allows to validate the numerical models and to prove the design basis for the proposed distribution system. Once having validated the CFD turbulence models and the strategy to study the flow maldistribution in the SGHE module, a decisive and trustworthy optimization of each component of the sodium distribution system has been performed. Finally, an optimal configuration has been proposed for the actual phase of ASTRID project.

Ce travail de thèse s’inscrit dans le cadre du programme R&D du CEA en support au système de conversion d’énergie à gaz du prototype industriel de Réacteur à Neutrons Rapides refroidi au Sodium (RNR-Na). Cette technologie représente une alternative aux cycles Rankine conventionnels à eau/vapeur, ayant pour avantage principal l’élimination du scenario accidentel de réaction sodium-eau. Cependant, la faible capacité de transfert de chaleur du gaz nécessite une technologie d’échangeurs compacts à plaques avec un nombre élevé de canaux à alimenter. Coté sodium, une section minimale de passage est nécessaire pour éviter le risque de bouchage par impureté. Cela induit de très faibles pertes de pression dans le faisceau qui, couplées à une condition de vitesse élevée à l’entrée, génèrent un risque réel de mauvaise distribution du débit. Les performances d’échange thermique et la tenue mécanique du composant sont alors dégradées. L’objectif principal de ce travail de thèse a été de résoudre ce problème de mauvaise distribution, en s’appuyant sur une conception innovante (BREVET FR16 57543), sur une stratégie de calcul numérique et l’établissement d’une base de données expérimentale pour la validation des travaux théoriques. Le nouveau système de distribution sodium se compose d’un collecteur d'entrée dont le design permet de guider la trajectoire du jet et d’un système de bifurcation de canaux qui augmente les pertes de pression dans le faisceau. De plus, des communications latérales entre les canaux sodium aident à homogénéiser davantage le flux. Deux installations expérimentales ont été conçues pour caractériser l'écoulement dans les canaux de bifurcation et dans le collecteur d'entrée. La conception des maquettes a permis de quantifier leur effet sur la distribution du flux entre les canaux. La base de données aérodynamiques PIV acquises a permis de valider les modèles numériques et de prouver l’efficacité du système de distribution proposé. Après avoir validé les modèles de turbulence CFD et la stratégie d'étude de la distribution dans le module SGHE, une optimisation de chaque composant du système de distribution de sodium a été réalisée. Le travail de cette thèse s’achève par la description de la conception optimale retenue pour la phase actuelle du projet ASTRID. / This PhD work was motivated by the CEA R&D program to provide solid technological basis for the use of Brayton power conversion system in Sodium-cooled Fast nuclear Reactors (SFRs). Multi-channel compact heat exchangers are necessary for the present application because of the low heat transfer capacity of the gas foreseen. In ASTRID project, a minimum size of Na channels section is required to avoid the plugging risk. However, this induces very low pressure losses in the bundle. Considering an additional inlet flow condition, a real risk of bad flow distribution remains. As a result, the thermal performance and thermal loading of the heat exchanger degrades due to it. The main goal of this work was to overcome the flow maldistribution problem by means of an innovative design of sodium distribution system (PATENT FR1657543), the development of a numerical strategy and the construction of an experimental database to validate all theoretical studies. The innovative sodium distribution system consists on an inlet header which tries to guide the evolution of the impinging jet flow while a system of bifurcating pre-distribution channels increases pressure drops in the bundle. Lateral communications between pre-distribution channels are introduced to further homogenize the flow. Two experimental facilities have been conceived to study the flow behavior in bifurcating channels and in the inlet header, respectively. At the same time, their effect on the flow distribution between channels is evaluated. The acquired PIV aerodynamic database allows to validate the numerical models and to prove the design basis for the proposed distribution system. Once having validated the CFD turbulence models and the strategy to study the flow maldistribution in the SGHE module, a decisive and trustworthy optimization of each component of the sodium distribution system has been performed. Finally, an optimal configuration has been proposed for the actual phase of ASTRID project.