Unsteady Elbow Pipe Flow to Develop a Flow-Induced Vibration Evaluation Methodology for Japan Sodium-Cooled Fast Reactor

This paper describes the current status of flow-induced vibration evaluation methodology development for primary cooling pipes in the Japan sodium-cooled fast reactor (JSFR), with particular emphasis on recent research and development activities that investigate unsteady elbow pipe flow. Experimental efforts have been made using 1/3-scale and 1/10-scale single-elbow test sections for the hot-leg pipe. The 1/10- scale experiment simulating the hot-leg pipe indicated no effect of pipe scale in comparison with the 1/3- scale experiment under inlet-rectified-flow conditions. The next experiment using the 1/3-scale test section was performed to investigate the effect of swirl flow at the inlet. Although the flow separation region was deflected at the downstream from the elbow, the experiment clarified a less significant effect of swirl flow on pressure fluctuation onto the pipe wall. An additional experiment was intended to study the effect of elbow curvature. The experiments with water revealed no clear flow separation in a larger curvature elbow case than that of the JSFR. Since the interference of multiple elbows should be investigated to understand turbulent flow in the cold-leg pipe geometry, 1/15-scale experiments with double elbows were carried out to clarify that flow in the first elbow influenced a flow separation behavior in the second elbow. Simulation activities include Unsteady Reynolds Averaged Navier Stokes equation (URANS) approach with a Reynolds stress model using a commercial computational fluid dynamics (CFD) code and Large Eddy Simulation (LES) approach using an in-house code. A hybrid approach that combined with RANS and LES was also applied using a CFD code. Several numerical results appear in this paper, focusing on its applicability to the hot-leg pipe experiments. These simulations reasonably agreed with the experimental data using the 1/3-scale test section.     

Experimental identification for flow distribution inside APR+ reactor vessel and direction of internal structure design improvement

A test program to quantify the reactor flow distribution has been performed using a test facility, named ACOP, having a 1/5 length scale referring to the APR+ reactor design. The flow characteristics of the prototype plant could be preserved by designing the test facility by adopting a linear reduced scaling principle. An Euler number is considered as a primary dimensionless parameter, which was preserved with a 1/41.0 Reynolds number scaling ratio based on the balanced flow conditions. The important measuring parameters are the core inlet flow, outlet pressure distribution, and sectional pressure drops along the major flow path inside the reactor vessel as well as static pressure and temperature at the vessel and boundary legs. The reactor flow distribution is identified by a series of three reactor flow balancing conditions: (1) balanced cold leg flow condition (2) 5% unbalanced cold leg flow condition, and (3) extreme unbalanced flow condition under the assumption of a single pump failure. This paper describes the design features for the test facility and the measuring method, and summarizes the reactor flow and pressure characteristics by ensemble averaging for each group of tests.     

Model experimentation and analysis of flow-induced vibrations of PWR internals

In a PWR the reactor coolant flow that goes through the reactor internals and the fuel assemblies is characterized by high turbulence and this flow is able to induce some structural vibration. A few years ago, some nuclear power plants were obliged to shut down for many months, due to the heavy damage caused by vibration. The design of reactors must be carefully checked taking into account the possible interaction between hydraulic excitation and reactor structure response. The reactor assembly of a PWR consists of: (1) a reactor vessel which withstands the internal pressure of the primary fluid and maintains the reactor core; (2) reactor internals which maintain fuel assemblies, guide the control rods and wear a thermal shield in order to reduce the fast neutron exposure of the reactor vessel wall; and (3) fuel assemblies and control rods.The SAFRAN test loop consists of a reduced-scale ( ) model of a reactor vessel, reactor internals, dummies representing fuel assemblies and a system of three loops including pumps and damping tanks connected to the reactor vessel, the purpose of which is to simulate the flow distribution of a three-loop PWR. The scaling laws for designing the model and the test loop are: same geometry and attachment conditions; same flow velocity: V model = V reactor; same Cauchy number, i.e. same ratio of inertia forces to stiffness forces; and same Euler number, i.e. same ratio of inertia forces to pressure forces. Nevertheless, it is not possible to use the same Reynolds number. The ratio between the Reynolds number of the reactor and the Reynolds number of the model, for the same fluid velocity, is 70. This is mainly due to scale ratio and to the viscosity of the fluid in the hot condition. But in most cases, we are above the critical values of Reynolds number where there is a variation of the Strouhal number S = ƒD/V. The measured frequencies in the model will be eight times the frequencies occurring in the reactor. In general, the construction technology used for the model is the same as that used for the reactor. All the structures in contact with the fluid are made of stainless steel. The instrumentation used on the SAFRAN test loop consists of accelerometers, pressure sensors and relative displacement sensors.Vibration phenomena are studied using two different approaches. In the first approach, the vibration properties of the structure are measured by means of tests performed in air and water to obtain, in both cases, frequencies, modes, damping and stiffness values. The hydraulic excitation sources are measured by tests on the loop: frequencies, Δp values, direct- and cross-correlation lengths. During these tests, structures are stiffened in order to prevent their motion. By means of a computer program based on the POWELL method, the structural response can be calculated according to the density of Δp distributed around the structure. The second approach consists of measuring directly the structural response to hydraulic excitations. Comparison of the results given by these two approaches shows: (a) the system non-linearities and (b) the coupling between the fluid and the structure. By using two different approaches a better knowledge of complex phenomena can be gained.     

The effects of break location on PWR small break LOCA: Experimental study at the ROSA-IV LSTF

This paper discusses the effect of break location on the break flow rate and break flow quality transitions during a small-break loss-of-coolant accident (LOCA) in a pressurized water reactor (PWR). Results from five experiments conducted at the ROSA-IV Large Scale Test Facility (LSTF) are compared for this purpose. These experiments simulated a 2-inch break at the lower plenum, upper head, pressurizer top, cold leg, and hot leg, respectively. The controlling phenomena for the break flow quality transitions in cold-leg and hot-leg break experiments are described.     

Towards an elastic-plastic fracture mechanics predictive capability for reactor piping

Intergranular stress corrosion cracks have been discovered in the recirculation bypass piping and core spray lines of several boiling water reactor (BWR) plants. These cracks initiate in heat-affected zones of girth welds and grow circumferentially by combined stress corrosion and fatigue. Reactor piping is mainly type 304 stainless steel, a material which exhibits high ductility and toughness. A test program described in this paper demonstrates that catastrophic crack growth in these materials is preceded by considerable amounts of stable crack growth accompanied by large plastic deformation. Thus, conventional linear elastic fracture mechanics, which only applies to the initiation of crack growth in materials behaving in a predominantly linear elastic fashion, is inadequate for a failure analysis of reactor piping.This paper is based upon research initiated by a need to develop a realistic failure prediction and a way to delineate leak-before-break conditions for reactor piping. An effective engineering solution for the type of cracks that have been discovered in BWR plants was first developed. This was based upon a simple net section flow stress criterion. Subsequent work to develop an elastic-plastic fracture mechanics methodology has also been pursued. A survey of progress being made is described in this paper. This work is based on the use of finite element models together with experimental results to identify criteria appropriate for the onset of crack extension and for stable crack growth. A number of criteria have been evaluated. However, the optimum fracture criterion has not yet been determined, even for conditions which do not include all of the complications involved in reactor piping.     

Characteristics of two-phase natural circulation in freon-113 boiling loop

A freon-113 flow visualization loop for simulating the hot-leg U-bend natural circulation flow has been constructed and hot-leg two-phase flow behavior has been studied experimentally. From the present experiments, an understanding of the basic mechanisms of the two-phase natural circulation and flow termination were obtained. The power input, loop friction and the liquid level in the simulated steam generator played key roles in the overall flow behavior. Experimental results show that the flow behavior strongly depends on phase changes and coupling between hydrodynamic and heat transfer phenomena. Non-equilibrium phase-change phenomena such as flashing create unstable hydrodynamic conditions which lead to cyclic or oscillatory flow behaviors.     

Benchmark Simulation of Turbulent Flow through a Staggered Tube Bundle to Support CFD as a Reactor Design Tool. Part II: URANS CFD Simulation

In Part II, we described the unsteady flow simulation and proposed a modification of a traditional turbulence flow model. Computational fluid dynamics (CFD) simulations of an isothermal, fully periodic flow across a tube bundle using unsteady Reynolds averaged Navier-Stokes (URANS) equations, with turbulence models such as the Reynolds stress model (RSM) were investigated at a Reynolds number of 1.8 × 10⁴, based on the tube diameter and inlet velocity. As noted in Part I, CFD simulation and experimental results were compared at five positions along (x; y) coordinates. The steady RANS simulation showed that four diverse turbulence models were efficient for predicting the Reynolds stresses, and generally, SRANS results were marginal to poor, using a consistent evaluation terminology. In the URANS simulation, we modeled the turbulent flow field in a manner similar to the approach used for large eddy simulation (LES). The time-dependent URANS results showed that the simulation reproduces the dynamic stability as characterized by transverse oscillatory flow structures in the near-wake region. In particular, the inclusion of terms accounting for the time scales associated with the production range and dissipation rate of turbulence generates unsteady statistics of the mean and fluctuation flow. In spite of this, the model implemented produces better agreement with a benchmark data set and is thus recommended.     

Experimental investigation on heat transfer characteristics in a safety injection nozzle component in PWR

Thermal fatigue is a potentially significant degradation mechanism in Nuclear Power Plants (NPP). For the fatigue analysis, the thermal load information about components must be determined firstly. In this paper, an experimental study was carried out to obtain local fluid temperatures and local heat transfer coefficients for the safety injection nozzle component in reactor coolant system (RCS). In this mixing tee component a hot jet issues into a cold cross-flow stream from an oblique pipe and the turbulent mixing of two fluids induces local cycling stresses on the adjacent piping wall. Experiments were performed using a special-made heat fluxmeter, which can measure the mixed fluid temperature close to the wall and the heat transfer coefficient between the fluid and the wall. Plexiglass and metallic 1/9-scale mockups were manufactured for flow visualization and heat transfer tests, respectively. All tests were conducted at range of 0–40 for the jet-to-cross-flow velocity ratio. The flow visualization test has obtained general pattern of the flow and identified sensitive zones in the component where the jet and cross-flow interact intensively to cause thermal fatigue more possibly. In the heat transfer test, heat fluxmeters were positioned in the wall at these sensitive zones. The measurement results of temperatures and heat transfer coefficients have been discussed in detail in the paper. These experimental results allow us improving the state of knowledge of the thermal load to be used in the industrial mixing tees in operating for long lifetime assessment and for the design in the basic Nuclear Power Plants.     

Seismic stops vs. snubbers, a reliable alternative

The Seismic Stops methodology has been developed to provide a reliable alternative for providing seismic support to nuclear power plant piping. The concept is based on using rigid passive supports with large clearances. These gaps permit unrestrained thermal expansion while limiting excessive seismic displacements. This type of restraint has performed successfully in fossil fueled power plants.A simplified production analysis tool has been developed which evaluates the nonlinear piping response including the effect of the gapped supports. The methodology utilizes the response spectrum approach and has been incorporated into a piping analysis computer program RLCA-GAP.Full scale shake table tests of piping specimens were performed to provide test correlation with the developed methodology. Analyses using RLCA-GAP were in good agreement with test results. A sample piping system was evaluated using the Seismic Stops methodology to replace the existing snubbers with passive gapped supports. To provide further correlation data, the sample system was also evaluated using nonlinear time history analysis. The correlation comparisons showed RLCA-GAP to be a viable methodology and a reliable alternative for snubber optimization and elimination.     

A study of the sensitivity of loca heat transfer analysis for a water-cooled reactor system

Results are given of computer calculations, using the reactor thermal analysis code THETA1-B, to determine the significance and relative importance of various heat transfer regimes in predicting maximum fuel cladding temperature for the blowdown phase of a postulated loss-of-coolant accident (LOCA) in a pressurized water reactor system. The factors considered include the choice of heat transfer correlation for a particular heat transfer regime, the method of delineating the boundaries between regimes, and core inlet coolant flow conditions.For a hot-leg rupture, the maximum surface temperature is sensitive to a number of factors, including choices of critical heat flux correlation, flow boiling transition heat transfer correlation, and in particular, stable film flow boiling correlation. However, for a LOCA resulting from a double-ended rupture of an inlet feeder, these factors have only marginal effects on maximum cladding temperature. In this case the importance of heat transfer to dry steam coolant at low net flow rate conditions is demonstrated, indicating a need for further information.     

Technical note Leak detection in nuclear piping outside containment

The results of an experimental investigation performed at Wyle Laboratories to evaluate various methods for detecting small leaks in high energy piping system is described. These experiments were designed to support the leak-before-break methodology currently being employed by the United States nuclear industry. This methodology requires that:

1. (1) the flow rate through a hypothetical leak be accurately predicted, and

2. (2) the lower limits of detecting the flow rate through such a leak be established.

The research described in this work was designed to establish experimentally this limit of detectability.The experiments performed covered a range of leakage flow rates between 0.1 and 5 gpm (378–18925 cm³/min through small penetrations in a 6-in. diameter carbon steel pipe. Insulation, typical of the types found in nuclear power plants, covered the test pipe.The key observation made is that leak rates down to 0.1 gpm (378 cm³/min) are easily detectable.     

Matched refractive-index PIV visualization of complex flow structure in a three-dimentionally connected dual elbow

Flow structure in a three-dimensionally connected dual elbow is visualized using a 1/15-scale experimental apparatus simulating the 1st and 2nd elbows of JSFR cold-leg piping. A matched refractive-index PIV measurement clarifies that a low-velocity region formed on the inner wall side of the 1st elbow develops toward the 2nd elbow. This low-velocity region consists of the following two ones: a flow separation region accompanied mainly with the generation and disappearance of transverse vortices, and a velocity recovery region that has longitudinal vortices with strong unsteadiness. These longitudinal vortices exist as twin vortices in the time-averaged flow field, and their dynamic characteristics highly depend on high-velocity creeping flows generated in the 1st elbow that flow into the velocity recovery region through the side walls. Since the velocity recovery region reaches the 2nd elbow, the geometry of the 2nd elbow has a significant impact on the characteristics of the vortex shedding in the velocity recovery region. On the other hand, obvious flow separation is not observed in the 2nd elbow, whereas high-velocity flow with intense velocity fluctuation is confirmed on the inner wall side. Furthermore, the unsteady vortices shed from the velocity recovery region are transferred to the central area of the 2nd elbow while growing significantly. The visualization of the secondary elbow shortly after the 2nd elbow clarifies that a strong swirling flow is formed in the 2nd elbow. These flow structures are due to the distorted flow formed in the 1st elbow and the shape effect of the 2nd elbow.     

Comparison of Reynolds average Navier-Stokes turbulence models in numerical simulations of the DC arc plasma torch

Five turbulence models of Reynolds average Navier-Stokes(RANS),including the standard k-ω model,the RNG k-e model taking into account the low Reynolds number effect,the realizable k-ω model,the SST k-ω model,and the Reynolds stress model(RSM),are employed in the numerical simulations of direct current(DC)arc plasma torches in the range of arc current from 80 A to 240 A and air gas flow rate from 10 m^3 h^-1 to 50 m^3 h^-1.The calculated voltage,electric field intensity,and the heat loss in the arc chamber are compared with the experiments.The results indicate that the arc voltage,the electric field,and the heat loss in the arc chamber calculated by using the standard k-ω model,the RNG k-ωmodel taking into account the low Reynolds number effect,and the realizable k-ω model are much larger than those in the experiments.The RSM predicts relatively close results to the experiments,but fails in the trend of heat loss varying with the gas flow rate.The calculated results of the SST k-ω model are in the best agreement with the experiments,which may be attributed to the reasonable predictions of the turbulence as well as its distribution.     

Simulation of flow pulsations in a twin rectangular sub-channel geometry using unsteady Reynolds Averaged Navier-Stokes modelling

An unsteady Reynolds Averaged Navier-Stokes (URANS) based turbulence model, the Spalart-Allmaras (SA) model, was used to investigate the flow pulsation phenomena in compound rectangular channels for isothermal flows. The studied geometry was composed of two rectangular sub-channels connected by a gap, on which experiments were conducted by Meyer and Rehme (1994) and were used for the validation of numerical results. Two case studies were selected to study the effect of the advection scheme. The results from the first order upwind advection scheme had clear symmetry and periodicity. The frequency of flow pulsations was under predicted by almost a factor of two. Due to inevitable numerical diffusion of the first order upwind scheme, a second order accurate in space advection scheme was also considered. The span-wise velocity contours, velocity vector plots, and time traces of the velocity components showed the expected cross-flow mixing between the sub-channels through the gap. The predicted kinetic energy in the unsteady velocity fluctuations showed two clear peaks at the edges of the gap. The dynamics of the flow pulsations were quantitatively described through temporal auto-correlations and power spectral functions. The numerical predictions were in agreement with the experiments. Studies on the effect of the Reynolds number and the computational length of the domain were also performed. The numerical results reproduced the relationship between the Reynolds number and the frequency of the flow pulsations. The impact of the channel length was tested by simulating a longer channel with respect to the base case. It was found that the channel length did not significantly affect the numerical predictions. Simulations were also performed using the standard k-? model. While the flow pulsations were predicted with this model, the frequency of the pulsation was in poor agreement with the experimentally measured value.     

Two-group interfacial area transport equations at bubbly-to-slug flow transition

In relation to nuclear reactor accident and safety studies, experiments on hot-leg U-bend two-phase natural circulation in a loop with a relatively large diameter pipe (10.2 cm inner diameter) were performed for understanding the two-phase natural circulation and flow termination during a small break loss of coolant accident in LWRs. The loop design was based on the scaling criteria developed under this program and the loop was operated either in a natural circulation mode or in a forced circulation mode using nitrogen gas and water. Various tests were carried out to establish the basic mechanism of the flow termination as well as to obtain essential information on scale effects of various parameters such as the loop frictional resistance, thermal center and pipe diameter. The void distribution in a hot-leg, flow regime and natural circulation rate were measured in detail for various conditions. The termination of the natural circulation occurred when there was insufficient hydrostatic head in the downcomer side. The superficial gas velocity at the flow termination could be predicted well by the simple model derived from a force balance between the frictional pressure drop along the loop and the hydrostatic head difference. The bubbly-to-slug flow transition was found to be dependent on axial locations. It turned out that the formation of cap bubbles in the large diameter pipe caused the increased drift velocity, which would affect the prediction of the void fraction in the hot leg.     

Numerical Simulation of Stall Flow Control Using a DBD Plasma Actuator in Pulse Mode

A numerical simulation method is employed to investigate the effects of the unsteady plasma body force over the stalled NACA 0015 airfoil at low Reynolds number flow conditions.The plasma body force created by a dielectric barrier discharge actuator is modeled with a phenomenological method for plasma simulation coupled with the compressible Navier-Stokes equations.The governing equations are solved using an efficient implicit finitevolume method.The responses of the separated flow field to the effects of an unsteady body force in various interpulses and duty cycles as well as different locations and magnitudes are studied.It is shown that the duty cycle and inter-pulse are key parameters for flow separation control.Additionally,it is concluded that the body force is able to attach the flow and can affect boundary layer grow that Mach number 0.1 and Reynolds number of 45000.     

Influence of elbow curvature on flow structure at elbow outlet under high Reynolds number condition

In the design of Japan Sodium-cooled Fast Reactor (JSFR), coolant velocity is beyond 9 m/s in the primary hot leg pipe of 1.27 m diameter. The Reynolds number in the piping reaches 4.2 × 10⁷. Moreover, a short-elbow is adopted in the hot leg pipe in order to achieve compact plant layout and to reduce plant construction cost. Therefore, the flow-induced vibration (FIV) arising from the piping geometry may occur in the short-elbow pipe. The FIV is due to the excitation source which is caused by the pressure fluctuation in the pipe. The pressure fluctuation in the pipe is closely related with the velocity fluctuation. As the first step of clarification of the FIV mechanism, it is important to grasp the mechanism of flow fluctuation in the elbow. In this study, water experiments with two types of elbows with different curvature ratios were conducted in order to investigate the interaction between flow separation and the secondary flow due to the elbow curvature. The experiments were conducted with the short-elbow and the long-elbow under Re = 1.8 × 10⁵ and 5.4 × 10⁵ conditions. The velocity fields in the elbows were measured using a high-speed Particle Image Velocimetry (PIV). The time-series of axial velocity fields and the cross-section velocity fields obtained by the high-speed PIV measurements revealed the unsteady and complex flow structure in the elbow. The flow separation always occurred in the short-elbow while the flow separation occurred intermittently in the long-elbow case. The circumferential secondary flows in clockwise and counterclockwise directions flowed forward downstream of reattachment point alternately in both elbows.     

Experimental investigation and CFD validation of countercurrent flow limitation (CCFL) in a large-diameter PWR hot-leg geometry

ABSTRACT

Countercurrent flow limitation (CCFL) is a phenomenon that consists of several flow patterns occurring simultaneously which produces a complex gas/liquid interface and interfacial momentum transfer, thus making it one of the most challenging two-phase flow configurations for computational fluid dynamics (CFD) validation. Numerous experimental investigations have been carried out in recent years regarding this, but most of those investigations were performed in small-diameter pipes or in non-pipe geometries (rectangular cross sections). A review of these experimental investigations has shown that the scale and geometry of the test section has a large impact upon the onset and characteristics of the CCFL. In order to provide a better understanding of this phenomenon in an actual pressurized water reactor (PWR) hot-leg geometry at a relatively large-diameter and scale, a test facility with a ～1/3.9 scale and a 190 mm inner diameter was constructed. Experiments were carried out at atmospheric pressure using water and air. High-speed recording was used to acquire high-quality images of the air/water interface. CCFL mechanisms, flow patterns, and the limits of the onset of CCFL and deflooding were experimentally identified. CFD simulations of two representative cases were carried out and assessed against experimental results. The analysis of the CFD simulations has provided insights into the improvements necessary for the accurate simulation of CCFL in large-scale geometries.     

Pressurized water reactor (PWR) hot-leg streaming: Part 1: Computational fluid dynamics (CFD) simulations

A thermo-hydraulic analysis model was developed to analyze thermal stratification phenomena observed in the hot-legs of pressurized water reactors (PWR). The model uses VIPREW code to determine the flow field and temperature distribution in the reactor fuel region. The temperature readings from the thermal couples located at the exit of the reactor core were used to compare with the VIPREW computed results. The predicted values agree well with the measurements. The VIPREW results are then used as the boundary conditions for the CFD analysis. The CFD computational domain includes the upper plenum and hot-legs and the fifty two (52) control rod guiding tubes to properly include the additional obstructions imposed to the fluid. Different fuel loading patterns were studied to investigate the effects of different power distribution and fuel channel exit water temperature on hot-leg thermal stratification magnitude. The analysis results show that the 52 control rod guide tubes have major contribution to the mixing effect in the upper plenum. The sudden expansion of the cross sectional area in the upper plenum leads to the formation of recirculation vortex that prolongs the duration of coolant in the reactor vessel. The hotter coolant from the center portion tends to flow upwards to the top before exiting at the upper portion of the hot-leg pipes. It leads to higher temperature in the upper portion of the hot-legs. Water from the cooler outer fuel channels tends to trap in the recirculation region before exiting from the lower portion of the hot-legs.     

加减速流条件下棒束通道阻力特性研究

压水反应堆的启动、停堆、事故等多种工况会使反应堆堆芯内部冷却剂产生加减速流,影响反应堆的热工水力特性。在雷诺数（Re）700相似文献