Simulation of two-phase flows in vertical tubes with the CTFD code FLUBOX

The computational two-fluid dynamics (CTFD) code FLUBOX is developed at GRS for the multidimensional simulation of two-phase flows. The single-pressure two-fluid model is used as basis of the simulation. A basic mathematical property of the two-fluid model of FLUBOX is the hyperbolic character of the advection. The numerical solution methods of FLUBOX make explicit use of the hyperbolic structure of the coefficient matrices. The simulation of two-phase flow phenomena needs, apart from the conservation equations for each phase, an additional transport equation for the interfacial area concentration. The concentration of the interfacial area is one of the key parameters for the modeling of interfacial friction forces and interfacial transfer terms. A new transport equation for the interfacial area concentration is in development. It describes the dynamic change of the interfacial area concentration due to mass exchange and a force balance at the phase boundary. Results from FLUBOX calculations for different experiments of two-phase flows in vertical tubes are presented as part of the validation.     

Two fluid model for two-phase turbulent jets

Eulerian two-fluid models are widely used in nuclear reactor safety and CFD. In these models turbulent diffusion of a dispersed phase must be formulated in terms of the fluctuating interfacial force and the Reynolds stresses. The interfacial force is obtained using the probability distribution function approach by Reeks (1992). This paper is the first application of this force to a case of engineering interest outside homogeneous turbulence. An Eulerian multidimensional two-fluid model for a cylindrical two-phase dispersed particle jet is proposed and compared with experimental data. The averaged conservation equations of mass and momentum are solved for each phase and the turbulent kinetic energy equation is solved for the continuous phase. The turbulent diffusion force and the Reynolds stresses are constituted within the context of the k- model of turbulence. A dissipation term has been added to the k- model for the turbulence modulation by the particles. Once the constitutive relations have been defined, the two-fluid model is implemented in a computational fluid dynamics code. It is shown that when the particles are very small the model is consistent with a convection-diffusion equation for particle transport where the diffusivity is defined according to Taylor's model (Taylor, G.I., 1921. Diffusion by continuous movements. Proc. London Math. Society, A20, pp. 196–211). The two-fluid model is also compared against two experimental data sets. Good agreement between the model and the data is obtained. The sensitivity of the results to various turbulent mechanisms is discussed.     

Development of one-dimensional two-fluid model with consideration of void fraction covariance effect

Accurate evaluation of gas-liquid two-phase flow behavior within rod bundle geometry is crucial for the safety assessment of the nuclear power plants. In safety assessment codes, two-phase flow in rod bundle geometry has been treated as a one-dimensional flow. In order to obtain the reliable one-dimensional two-fluid model, it is essential to utilize proper area-averaged models for governing equations and constitutive relations. The area-averaged interfacial drag term utilized to evaluate two-phase interfacial drag force is typically given by the drift-flux parameters which consider the velocity profile in two-phase flow fields. However, in a rigorous sense, the covariance due to void fraction profile is ignored in traditional formulations. In this paper, the rigorous formulation of one-dimensional momentum equation was derived by taking consideration of void fraction covariance, and a new set of one-dimensional momentum equation and constitutive relations for interfacial drag was proposed. The newly obtained set of formulations was embedded into TRAC-BF1 code and numerical simulation was performed to compare against the traditional model without covariance. It was found that effect of covariance was almost negligible for steady-state adiabatic conditions, but for high void fraction condition with added perturbation, the traditional model underpredicted the damping ratio at around 8%.     

相界面浓度输运方程在一维两流体模型中的应用研究

为解决一维两流体模型核电厂系统分析程序中使用流型图所带来的缺陷,提高系统分析程序计算的准确性,探索在一维两流体模型中应用相界面浓度输运方程（IATE）对两相流动进行预测。采用FORTRAN语言开发耦合了IATE的一维两流体模型求解器（Solver-IATE）,并对其进行验证。基于SolverIATE对小直径绝热圆管内向上泡状流进行了数值模拟,并与采用流型图的计算结果进行了对比。研究结果表明：采用IATE计算的相界面浓度结果比采用流型图的计算结果更接近实验值。因此,在一维两流体模型中使用IATE可以提高其计算相界面浓度的准确性,进而提高一维两流体模型核电厂系统分析程序计算两相间相互作用项的准确性,能更准确预测反应堆的瞬态响应特性。     

Effects of non-uniform inlet boundary conditions and lift force on prediction of phase distribution in upward bubbly flows with Fluent-IATE

This study investigates the profile effects of the boundary conditions in two-phase flows, such as the inlet void fraction, interfacial area concentration, and phase velocity, on the predictions of flow behaviors downstream. Simulations are performed for upward air-water bubbly flows in a 48.3-mm inner diameter pipe by employing Fluent's two-fluid model together with an interfacial area transport equation (IATE) model. The IATE was developed in the literature to model the interfacial area concentration by taking into account the bubble coalescence and disintegration, and phase change effects.In this study, two types of inlet boundary conditions are considered, one being a uniform-profile boundary condition in the radial direction with area-averaged experimentally measured values while the other being a non-uniform profile condition based on the actual measured profiles at the inlet. The numerical predictions of downstream profiles of the phase distributions indicate that the two types of boundary conditions yield similar results for the downstream flow behaviors for the bubbly flow conditions investigated. In addition, the results with and without the lift force demonstrated that the lift force is essential to obtain accurate lateral phase distribution.     

Numerical simulation of hyperbolic two-phase flow models using a Roe-type solver

The application of the generalized Roe scheme to the numerical simulation of two-phase flow models requires a fast and robust computation of the absolute value of the system matrix. In several models such as the two-fluid model or a general multifield model, this matrix has a non-trivial eigenstructure and the eigendecomposition is often ill conditioned. We give a general algorithm avoiding the diagonalization process. It is based on an iterative approach, but turns out to be an exact computation when the eigenvalues are real. The knowledge of the characteristic polynomial gives us an easy access to the eigenvalues but however, the iterative scheme can be used with only estimates of the eigenvalues, using for example Gershgorin’s disk localization. We finally show some numerical results of two-fluid model simulations involving interfacial pressure and a virtual mass force model.     

Code performance with improved two-group interfacial area concentration correlation for one-dimensional forced convective two-phase flow simulation

In gas–liquid two-phase flow simulation for reactor safety analysis, interfacial momentum transfer in two-fluid model plays an important role in predicting void fraction. Depending on flow conditions, a shape of the two-phase interface complicatedly evolves. One of the proposed approaches is to quantify the gas–liquid interface information using interfacial area transport equation. On the other hand, a more simplified and robust approach is to classify bubbles into two-groups based on their transport characteristics and utilize constitutive equations for interfacial area concentration for each group. In this paper, interfacial drag model based on the two-group interfacial area concentration correlations is implemented into system analysis code, and void fractions were calculated for the evaluation of numerical behaviors. The present analysis includes (1) comparison of one-group and two-group relative velocity models, (2) comparison with separate effect test database, (3) uncertainty evaluation of drag coefficient, (4) numerical stability assessment in flow regime transition, and (5) transient analysis for simulating the prototypic condition. Results showed that utilization of interfacial drag force term using constitutive equations of two-group interfacial area concentration yields satisfactory void fraction calculation results. The proposed solution technique is practical and advantageous in view of reducing the computational cost and simplifying the solution scheme.     

Numerical simulations of basic interfacial instabilities with incompressible two-fluid model

The incompressible two-fluid model for stratified flow was improved. The interface of the stratified two-phase flow was successfully recognized and sharpened within the two-fluid model. After the advection step of volume fraction the numerical diffusion of the interface was reduced in such a way that the thickness of the interface is kept constant during the simulation. The surface tension force was implemented in the system of the two-fluid model equations. The two basic instabilities of stratified flows: Rayleigh-Taylor and Kelvin-Helmholtz instability were used to validate the proposed two-fluid model. The proposed two-fluid model with interface sharpening presents a step towards the simulations of mixed flows, where locally dispersed flow or stratified flow will be simulated with appropriated submodels within the two-fluid model.     

Numerical simulation using CFD software of countercurrent gas-liquid flow in a PWR hot leg under reflux condensation

In order to improve the countercurrent flow model of a transient analysis code, countercurrent air-water tests were previously conducted using a 1/15 scale model of the PWR hot leg and numerical simulations of the tests were carried out using the two-fluid model implemented in the CFD software FLUENT 6.3.26. The predicted flow patterns and CCFL characteristics agreed well with the experimental data. However, the validation of the interfacial drag correlation used in the two-fluid model was still insufficient, especially regarding the applicability to actual PWR conditions. In this study, we measured water levels and wave heights in the 1/15 scale setup to understand the characteristics of the interfacial drag, and we considered a relationship between the wave height and the interfacial drag coefficient. Numerical simulations to examine the effects of cell size and interfacial drag correlations on numerical predictions were conducted under PWR plant conditions. Wave heights strongly related with the water level and interfacial drag coefficient, which indicates that the interfacial drag force mainly consists of form drag. The cell size affected the gas velocity at the onset of flooding in the process of increasing gas flow rate. The gas volumetric fluxes at CCFL predicted using fine cells were higher than those using normal cells. On the other hand, the cell size did not have a significant influence on the process of decreasing gas flow rate. The predictions for the PWR condition using a reference set of interfacial drag correlations agreed well with the Upper Plenum Test Facility data of the PWR scale experiment in the region of medium gas volumetric fluxes. The reference interfacial drag correlations employed in this study can be applied to the PWR conditions.     

Prediction of interfacial area transport in a coupled two-fluid model computation

A computer code has been written to predict interfacial area transport within the framework of the two-fluid model. The suitability of various constitutive models was evaluated from a scientific and numerical standpoint, and selected models were used to close the two-fluid model. The resulting system was then used to optimize the empirical constants in the interfacial area transport equation for large diameter pipes. The optimized model was evaluated based on comparison with the data of Shen et al. and Schlegel et al. The optimization shows agreement with previous research conducted by Dave et al. and Talley et al. using TRACE-T, and reduced the RMS error in the interfacial area concentration prediction for the large diameter pipe data from 52.3% to 34.9%. The results also highlight a need for additional high-resolution data at multiple axial locations to provide a more detailed picture of the axial development of the flow. The results also indicate a need for improved modeling of the interfacial drag, especially for Taylor cap bubbles under relatively low void fraction conditions.     

New analytical solutions to the two-phase water faucet problem

The one-dimensional water faucet problem is one of the classical benchmark problems originally proposed by Ransom to study the two-fluid two-phase flow model. With simplifications, such as massless gas phase and no wall and interfacial frictions, analytical solutions had been previously obtained for the transient liquid velocity and void fraction distribution. The water faucet problem and its analytical solutions have been widely used for code assessment, benchmark, and numerical verification. In this work, we present a new set of analytical solutions to the water faucet problem at the steady-state condition, with the gas-phase density’s effect on pressure distribution considered. This new set of analytical solutions is used in a rigorous numerical verification process from which the anticipated second-order spatial accuracy is achieved for a second-order spatial discretization scheme. On the contrary, the same anticipated order of accuracy could not be obtained using the Ransom solutions as the reference. In addition, extended Ransom transient solutions for the gas-phase velocity and pressure are derived with the assumption of decoupled liquid and gas pressures. Numerical benchmark on the extended Ransom solutions is also presented.     

Flooding correlation based on the concept of hyperbolicity breaking in a vertical annular flow

Flooding is classified into exit flooding and entrance flooding depending on the positions at which it is initiated. It is postulated that flooding may result from the instabilities of roll waves and the stationary wave generated at the entrance and the exit respectively. Based on the two-fluid model, the neutral stability condition for wave instabilities is derived by studying hyperbolicity breaking near a singular point in two-phase flow. It turns out to be a type of onset condition of Helmholtz instability; the critical relative velocity depends on the void fraction derivative of the interfacial pressure force as well as on the void fraction and density ratio. In order to obtain information on the interfacial pressure force, a Korteweg-de Vries solitary wave is studied with the assumption that its wavenumber is the same as that of the fastest-growing sine wave. Predictions by the correlations for entrance flooding and exit flooding are in good agreement with experimental data. Also, the present model is able to consider the effect of the test section length on flooding.     

Modified two-fluid model for the two-group interfacial area transport equation

This paper presents a modified two-fluid model that is ready to be applied in the approach of the two-group interfacial area transport equation. The two-group interfacial area transport equation was developed to provide a mechanistic constitutive relation for the interfacial area concentration in the two-fluid model. In the two-group transport equation, bubbles are categorized into two groups: spherical/distorted bubbles as Group 1 while cap/slug/churn-turbulent bubbles as Group 2. Therefore, this transport equation can be employed in the flow regimes spanning from bubbly, cap bubbly, slug to churn-turbulent flows. However, the introduction of the two groups of bubbles requires two gas velocity fields. Yet it is not practical to solve two momentum equations for the gas phase alone. In the current modified two-fluid model, a simplified approach is proposed. The momentum equation for the averaged velocity of both Group-1 and Group-2 bubbles is retained. By doing so, the velocity difference between Group-1 and Group-2 bubbles needs to be determined. This may be made either based on simplified momentum equations for both Group-1 and Group-2 bubbles or by a modified drift-flux model.     

Dynamic model for the interfacial shear as a closure law in two-fluid models

The particular structure of the interfacial shear as a closure law in two-fluid models ought to bridge the gap between microscale phenomena and the macro-averaged representation of the flow. Conventionally, quasi-steady models of the interfacial shear are adopted in relation to the local phase holdup and velocities. The main attempts to account for the complicated interfacial interactions have been focused on improving the friction factor modelling. The present study suggests incorporation of an explicit functional dependence of the interfacial shear on the interface slope due to interfacial waviness. The implementation of the proposed model as a closure law in the stability analysis of stratified flows reveals the crucial role of the dynamic term in determining the stability characteristics. It is shown that with the inclusion of the newly proposed dynamic term of the interfacial shear the stratified smooth-stratified wavy transitional boundary is satisfactorily predicted for a wide range of two-fluid horizontal and inclined systems.     

Implementation and evaluation of one-group interfacial area transport equation in TRACE

The present study implements the one-dimensional interfacial area transport equation into the TRACE code, being developed by the U.S. Nuclear Regulatory Commission. The interfacial area transport equation replaces the conventional flow regime dependent correlations and the regime transition criteria for furnishing the interfacial area concentration in the two-fluid model. This approach allows dynamic tracking of the interfacial area concentration by mechanistically modeling bubble coalescence and disintegration mechanisms. Thus, it eliminates potential artificial bifurcations or numerical oscillations stemming from the use of conventional static correlations. To implement the interfacial area transport equation, a three-field version of TRACE is utilized, which is capable of tracking both the continuous liquid and gas fields as well as a dispersed gas field. To demonstrate the feasibility of the present approach, the steady-state one-group interfacial area transport equation applicable to adiabatic air-water bubbly two-phase flow is first tested in the present study. Data obtained in 18 different flow conditions from two vertical co-current upward air-water bubbly two-phase flow experiments performed in round pipes (25.4 mm and 48.3 mm) are used to help evaluate the implementation. Results obtained from TRACE with the interfacial area transport equation (TRACE-T) and those from TRACE without the transport equation (TRACE-NT) are compared to demonstrate the enhancement in prediction accuracy. The predictions made by TRACE-T agree well with the data with an average percent difference of approximately ±8%. It is also evident from the results that while TRACE-T accounts for dynamic interaction of bubbles along the flow field, the predictions made by TRACE-NT are attributed primarily to the pressure change.     

Capability of the RELAP5 code to simulate natural circulation behavior in test facilities

Many advanced reactor designs incorporate passive systems mainly to enhance the operational safety and possible elimination of severe accident condition. Some reactors are even designed to remove the nominal fission heat passively by natural circulation without using mechanical pumps e.g. ESBWR, AHWR, CHTR, CAREM, etc. while in most other new reactor concepts, the decay heat is removed passively by natural circulation following the pump trip conditions. The design and safety analysis of these reactors are carried out using the best estimate codes such as RELAP5, TRAC and CATHARE, etc. These best estimate codes have been developed for pumped circulation systems and it is not proven about their adequacy or applicability for natural circulation systems wherein the driving mechanism is completely different. Some of the key phenomena which are difficult to model but are significantly important to assess the natural circulation system performances are – low flow natural circulation mainly because the flow is not fully developed and can be multi-dimensional in nature; flow instabilities; critical heat flux under oscillatory condition; flow stratification particularly in large diameter vessel; thermal stratification in large pools; effect of non-condensable gases on condensation, etc. Though, these best estimate codes use a six equation two-fluid model formulation for the thermal-hydraulic calculation which is considered to be the best representative of two-phase flows, but their accuracies depend on the accuracies of the models for interfacial relationships for mass, energy and momentum transfer which are semi-empirical in nature. The other problem with two-fluid models is the effect of ill-posedness which may cause numerical instability. Besides, the numerical diffusion associated due to truncation of higher order terms can affect the prediction of flow instabilities. All these effects may lead to inability to capture the important physical instability in natural circulation systems and instability characteristics i.e. amplitude and frequency of flow oscillation. In view of this, it is essential to test the capability of these codes to simulate natural circulation behavior under single and two-phase flow conditions before applying them to the future reactor concepts.In the present study, one of the extensively used best estimate code RELAP5 has been used for simulation of steady state, transient and stability behavior of natural circulation based experimental facilities, such as the High-Pressure Natural Circulation Loop (HPNCL) and the Parallel Channel Loop (PCL) installed and operating at BARC. The test data have been generated for a range of pressure, power and subcooling conditions. The computer code RELAP5/MOD3.2 was applied to predict the transient natural circulation characteristics under single-phase and two-phase conditions, thresholds of flow instability, amplitude and frequency of flow oscillations for different operating conditions of the loops. This paper presents the effect of nodalisation in prediction of natural circulation behavior in test facilities and a comparison of experimental data in with that of code predictions. The errors associated with the predictions are also characterized.     

CFD prediction of flow and phase distribution in fuel assemblies with spacers

This paper is concerned with the modeling and computation of multi-dimensional two-phase flows in BWR fuel assemblies. The modeling principles are presented based on using a two-fluid model in which lateral interfacial effects are accounted for. This model has been used to evaluate the velocity fields of both vapor and liquid phases, as well as phase distribution, between fuel elements in geometries similar to BWR fuel bundles. Furthermore, this model has been used to predict, in a detailed mechanistic manner, the effects of spacers on flow and phase distribution between, and pressure drop along, fuel elements. The related numerical simulations have been performed using a CFD computer code, CFDS-FLOW3D.     

New Interfacial Drag Force Model Including Effect of Bubble Wake, (I) Model Development for Steam-Water Bubbly Flow in Large-Diameter Pipes

Several experimental results show that bubbles can easily be captured in the wake formed by leading bubbles when multiple bubbles are rising in a liquid. It is suggested from this experimental result that the effect of bubble wake should be included in the constitutive relationships representing the interfacial drag force. In the present study, steam-water bubbly flow experiments were performed to develop a new interfacial drag force model including the effect of bubble wake. Since the validity of the existing constitutive equations have been tested mainly for two-phase flow in small-diameter pipes, our study focused on two-phase flow in a large-diameter pipe. Using a one-dimensional two-fluid model, the applicability of the new interfacial drag force model to our experimental conditions was investigated. As a result, it was shown that the present model markedly improves the accuracy of the predicted results. It was therefore demonstrated that the present bubble wake model is effective at least for the conditions which were used for model development. Its applicability to different conditions will be discussed in a subsequent study.     

Void fraction estimation within rod bundles based on three-fluid model and comparison with x-ray CT void data

An interfacial shear stress equation in the dispersed-annular two-phase flow regime has been developed, which is based on a three-fluid model consisting of a liquid film on a rod, vapor and entrained liquid associated with a vapor flow. It is an extension of J.G.M. Andersen's procedure that provides a two-fluid interfacial shear stress equation using the drift flux parameters C₀ and V_gj. This interfacial shear stress equation can take into account a phase and velocity distribution through an equivalence between the drift flux parameters and the interfacial shear stress.

Using the three-fluid subchannel analysis code TEMPO with the three-fluid interfacial shear stress model, the capability of a three-fluid calculation using the drift flux parameters C₀ and V_gj that reproduce a measured void fraction is demonstrated. A comparison was made with advanced X-ray computed tomography (CT) void fraction data within a 4×4 rod bundle in diabatic 1 MPa pressure conditions. The three-fluid velocity field was estimated to be in good agreement with the experimental result of a void fraction.     

Modeling and validation of interfacial area transport equation in subcooled boiling flow

The first comprehensive validation of the interfacial area transport equation in subcooled boiling is presented and shown to perform exceptionally when compared with experimental data. The formulation and closure of the bubble layer averaged interfacial area transport equation is reviewed along with the treatment of the two-fluid model in subcooled boiling. Interfacial area concentration source and sink terms in subcooled boiling are presented including the bubble interaction mechanisms (random collision and turbulent impact), as well as phase change terms (wall nucleation and condensation). Additionally, the volume source terms from phase change are described and discussed in terms of their significance to the interfacial area transport equation. The validation of the interfacial area transport equation with a recently proposed wall nucleation source term is shown to have excellent prediction at low and elevated pressure, as well as a wide range of mass flux. With new confidence in the wall nucleation source term, the interfacial area concentration in subcooled boiling can be accurately predicted. Due to its strong dependence in the modeling of active nucleation site density, bubble departure frequency, and departure diameter, the calculation is shown to be very sensitive to wall temperature.