Figure - available via license: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
Content may be subject to copyright.
Source publication
A numerical calculation with the commercial computational fluid dynamics code CFX-14.0 was conducted for a test facility simulating the Canadian deuterium uranium moderator thermal-hydraulics. Two kinds of moderator thermal-hydraulic tests at Stern Laboratories Inc. were performed in the full geometric configuration of the Canadian deuterium uraniu...
Similar publications
The Canadian Supercritical Water-cooled Reactor concept, as an evolution of the CANada Deuterium Uranium (CANDU) reactor, includes both pressure tubes and a low temperature heavy water moderator. The current Pressure Tube type SCWR (PT-SCWR) concept features 64-element fuel assemblies placed within High Efficiency Re-entrant Channels (HERCs) that c...
The objective of present study is to predict and analyze the flow through intake manifold and inlet port system using Computational Fluid Dynamics (CFD) and to validate the prediction by experimental data. Three-dimensional model of air intake system was analyzed by using the commercially available FLUENT software. The mesh was generated using the...
Citations
... Issa and Oliveira, 1994;Costa et al., 2006;Wang et al., 2008;Kuczaj et al., 2010;Santos and Kawaji, 2010;Liu and Li, 2011;Sakowitz et al., 2013;Athulyaa and Cherian, 2016;Liu et al., 2016;Arias and Montlaur, 2017;Georgiou and Papalexandris, 2017), mostly simulating turbulence with the use of RANS-based turbulence models or large eddy simulations (LES), despite evidence that RANS models have limited accuracy (Westin et al., 2008;Walker et al., 2010) and the understanding that the proper application of LES to large and complex geometries is not computationally feasible, as it requires an extremely fine mesh near walls, where turbulence length scales are typically very small. A small number of two-phase CFD analyses have also been performed for header/feeder systems (An et al., 2000;Gulshani, 2006;Yazdani et al., 2012;Gandhi et al., 2012;Lee and Jeong, 2019) and other nuclear reactor vessels (Pietralik and Smith, 2006;Kang and Jo, 2008;Kim et al., 2006;Kim and Chang, 2015). ...
Three-dimensional, time-dependent numerical simulations of gas-liquid flow in a multi-branch manifold (an idealised model of the header/feeder system of a CANDU nuclear reactor) were carried out using the volume of fluid approach to separate the phases and the detached eddy simulation model to simulate turbulence. Interest was focusse n the distribution of the liquid to different feeders. Numerical results for a header connected to a small number of vertical feeders were in fair agreement with in-house experimental data. The rates of change of liquid discharge following changes of inlet conditions were estimated and found to depend on feeder location. The interface between an upper region of the header that was occupied by mostly air and a lower region that was occupied by an air-water mixture was found to rise with an increase in either air or liquid inlet flow rate. Simulations were performed to quantify the effects of differences in the values of fluid properties of air-water mixtures at approximately standard atmospheric pressure and temperature from those of saturated steam-water mixtures at significantly higher pressures (up to 10 MPa) and temperatures (up to 310 °C). The liquid flow rate through a feeder that was directly below the header inlet was found to decrease markedly with increasing mixture pressure and temperature. Pilot simulations of air-water flows in a header connected to multiple feeders with inlets at different elevations and orientations revealed that, under certain conditions, some of these feeders discharged little liquid.
... Teyssedou et al. [12] conducted FLUENT code simulation of moderator flow around calandria tubes of CANDU-6 and showed that the standard k-e model is appropriate for turbulence model to perform this kind of simulation. Application of FLUENT and CFX code is successfully performed for the reduced-scale CFD models for various thermal hydraulics problems in nuclear engineering also by the authors [13,14]. ...
... In the computation using OpenFOAM, SIMPLE algorithm, a kind of finite volume method (FVM) is applied for the iteration until the steady state for Equations (1) and (2). In this method, the pressure gradient term in Equation (2) is isolated, and sub-iterations should be performed between predictor and corrector [14]. The PIMPLE method is used for unsteady time marching, which is specified as no under-relaxation and multiple corrector steps in the calculation of momentum. ...
... Among various codes such as ANSYS-CFX and COMSOL, the open source code OpenFOAM displayed similar or better level of coincidence for all kinds of turbulence models, and k-ε model was the best result. The modeling of two-dimensional heat flow can predict the temperature with a maximum local error of 3.5 • C, which can be a reduced model of CANDU-6 moderator [14]. ...
Three-dimensional moderator flow in the calandria tank of CANDU-6 pressurized heavy water reactor (PHWR) is computed with Open Field Operation and Manipulation (OpenFOAM), an open-source computational fluid dynamics (CFD) code. In this study, numerical analysis is performed on the real geometry model including 380 fuel rods in the calandria tank with the heat-source distribution to remove uncertainty of the previous analysis models simplified by the porous media approach. Realizable k-ε turbulence model is applied, and the buoyancy due to temperature variation is considered by Boussinesq approximation for the incompressible single-phase Navier-Stokes equations. The calculation results show that the flow is highly unsteady in the moderator. The computational flow visualization shows a circulation of flow driven by buoyancy and asymmetric oscillation at the pseudo-steady state. There is no region where the local temperature rises continuously due to slow circulating flow and its convection heat transfer.
... En otro trabajo (Kim, 2015) utilizando CFX-14.0 se determinó la distribución de temperatura utilizando datos de una experiencia de un tanque de calandria, que no tenía estrictamente una relación de escala pero con algunas similitudes a un CANDU-6. En dicha experiencia se obtuvo una distribución no simétrica y que está gobernada por la combinación de fuerzas de momento (debido al distribuidor) y fuerzas boyantes (por el calor aplicado). ...
En este trabajo se presentan los cálculos de las distribuciones de temperatura y velocidad del fluido en el recipiente del moderador (calandria de ahora en más) de un reactor tipo CANDU-6, mediante simulaciones en estado estacionario con el software ANSYS CFX 15.0, versión académica. El recipiente de la calandria contiene 380 tubos de calandria que, a su vez, incluyen los tubos de presión y el combustible. Adicionalmente, dentro de la calandria se encuentran los internos del reactor que, para esta etapa de cálculo, no fueron tenidos en cuenta. La geometría se representó a partir de los planos correspondientes al recipiente del moderador y el de sus internos, y como condiciones de contorno se establecieron el caudal de entrada, la presión de salida, y la potencia por moderación y por conducción distribuida radial y axialmente. Finalmente, los resultados obtenidos permitieron demostrar que el modelo en CFX del tanque de calandria posibilitó identificar y representar patrones de flujo determinados experimentalmente y por otros modelos en fluidodinámica computacional (CFD por sus siglas en ingles). Además se determinó que para las condiciones de operación del reactor CANDU-6 el patrón de flujo que domina la distribución de flujo dentro del tanque de calandria es de tipo flujo mixto caracterizado por corrientes de flujo dominadas por fuerzas de momento y por fuerzas boyantes.
... The use of shape function used in FEM for each cell or the elementbased FVM makes it possible to construct a vertex-centered scheme, which is contrast to ANSYS-FLUENT using a cellcentered FVM [7]. Figures 4(a) and 4(b) show an example of full-scale computation for the 2D geometry model with a buoyancy term in momentum equation, (2) where the total number of meshes is 672,912 [13]. + value is defined for Δ , and the normal length of grid at the first one neighboring the wall is ...
... A benchmark test for the performance of each code is proposed for the well-known STERN laboratory experiment [8]. This problem is often used for the comparison with CFD results [10,13,14] and uses a reduced-scale CANDU-6 moderator model: see Figure 8(a). The central part of tube bundles is isolated with flat plates, and the pressure drop is precisely measured from pressure taps located in the three stations, from PT1 to PT3 in Figure 8 ...
... STERN laboratory experiment: (a) experimental configuration and (b) the measurement of pressure[13,15] ...
The moderator system of CANDU, a prototype of PHWR (pressurized heavy-water reactor), has been modeled in multidimension for the computation based on CFD (computational fluid dynamics) technique. Three CFD codes are tested in modeled hydrothermal systems of heavy-water reactors. Commercial codes, COMSOL Multiphysics and ANSYS-CFX with OpenFOAM, an open-source code, are introduced for the various simplified and practical problems. All the implemented computational codes are tested for a benchmark problem of STERN laboratory experiment with a precise modeling of tubes, compared with each other as well as the measured data and a porous model based on the experimental correlation of pressure drop. Also the effect of turbulence model is discussed for these low Reynolds number flows. As a result, they are shown to be successful for the analysis of three-dimensional numerical models related to the calandria system of CANDU reactors.
... The effectiveness of the moderator heat sink to ensure fuel channel integrity at the onset of this ballooning contact is based on a series of contact boiling experiments [1], which derived the moderator subcooling requirements [2] to preclude a sustained CT dryout by the minimum available moderator subcooling and the PT/CT contact temperature. The moderator subcooling or the local moderator temperature is calculated by computational fluid dynamics tools developed from the results of the numerous experimental and numerical studies [3,4]. The International Collaborative Standard Problem (ICSP) on heavy water reactor (HWR) moderator subcooling requirements [5] has been organized by the International Atomic Energy Agency (IAEA) to facilitate the development and validation of computer codes for the analysis of fuel channel integrity. ...
For the blind calculation of the International Collaborative Standard Problem (ICSP) experiment on heavy water reactor moderator subcooling requirements, the COMSOL Multiphysics code is used to simulate plastic deformation of a pressure tube (PT) as a result of the interaction of stress and temperature. It is shown that the thermal stress model of COMSOL is compatible to simulate the multiple heat transfers (including the radiation heat transfer and heat conduction) and stress strain in the simplified two-dimensional problem. The benchmark test result for radiation heat transfer is in good agreement with the analytical solution for the concentric configuration of PT and calandria tube (CT). Since the original strain model of COMSOL only considers an elastic deformation with thermal expansion coefficient, the PT/CT contact cannot be predicted in the ICSP. Therefore, the plastic deformation model by the Shewfelt and Godin, widely used in the fuel channel analysis of CANadian Deuterium Uranium (CANDU) reactor, is implemented to the strain equation of COMSOL. The heat-up of PT, the strain rate, and the contact time of the PT/CT are calculated with the boundary conditions (BCs) given for blind calculation of the ICSP experiment.
The result shows a sudden expansion of the inner concentric PT within a few milliseconds. This unsteady simulation should be helpful for the conceptual design of experiment as well as for the understanding of multiphysics inside the fuel channels of the CANDU reactor.
The passive moderator cooling system (PMCS) for the Advanced Heavy Water Reactor is one of the passive systems incorporated to enhance safety of the reactor. The PMCS comprises of a coupled natural circulation loop, with a Gravity-driven-water pool as the ultimate sink and a shell and tube heat exchanger as the coupling component. Previously (Pal et al., 2016), a scaled experimental facility for PMCS was setup. In this work, validity of scaling theory was verified using 1D thermal-hydraulic code (RELAP5/MOD3.2) and simulations were compared with experimental data. Finally, PMCS was simulated for various geometrical parameters, e.g. elevation difference and pipe diameters, in an attempt to optimize the design. A good agreement with experimental data (within ± 8%), except for initial 2 h, was observed in RELAP/MOD3.2 code simulation to scaled experimental facility because of internal recirculations within Calandria being more prominent than overall circulatory flows in the passive loops.
The distribution of the fluid temperature and mass density of the moderator flow in CANDU-6 nuclear power reactors may affect the reactivity coefficient. For this reason, any possible moderator flow configuration and consequently the corresponding temperature distributions must be studied. In particular, the variations of the reactivity may result in major safety issues. For instance, excessive temperature excursions in the vicinity of the calandria tubes nearby local flow stagnation zones, may bring about partial boiling. Moreover, steady-state simulations have shown that for operating condition, intense buoyancy forces may be dominant, which can trigger a thermal stratification. Therefore, the numerical study of the time-dependent flow transition to such a condition, is of fundamental safety concern. Within this framework, this paper presents detailed time-dependent numerical simulations of CANDU-6 moderator flow for a wide range of flow conditions. To get a better insight of the thermal-hydraulic phenomena, the simulations were performed by covering long physical-time periods using an open-source code (Code_Saturne V3) developed by Électricité de France. The results show not only a region where the flow is characterized by coherent structures of flow fluctuations but also the existence of two limit cases where fluid oscillations disappear almost completely.
