Analysis of flow distribution in a PWR fuel rod bundle model containing a blockage - Part 2. A non-coplanar blockage

An experimental investigation, covering a Reynolds number range from 1900 to 9800, was conducted to study the influence of a non-coplanar blockage on the velocity and turbulence intensity distributions in an unheated 7 X 7 rod bundle. Using the blockage sleeves from a previous 61% coplanar blockage study, the non-coplanarity was obtained by axially staggering these sleeves in a prescribed manner. The results showed that the introduction of non-coplanarity did not result in significant changes from the overall bundle flow behaviour with a coplanar blockage. The effect on the flow immediately upstream and downstream of the blockage and within the blockage was less pronounced, thereby resulting in a smaller degree of flow diversion. The blockage zone, despite being effectively longer than the coplanar geometry, did not seem to adversely influence the downstream flow recovery process. Indeed the recovery to an undisturbed flow profile was more rapidly established. Complete flow recovery was attained for both the non-coplanar and coplanar blockage geometries at the same axial location in the rod bundle. Predictions from the COBRA subchannel computer code again agreed reasonably with the experimental data.     

Analysis of flow distribution a PWR fuel rod bundle model containng A 90% blockage

An experimental investigation, covering a Reynolds number range from 2 × 10³ to 3.5 × 10⁴, was conducted to study the velocity and turbulence intensity distributions due to the presence of a blockage in an unheated 7 × 7 rod bundle. The blockage configuration, consisting of a 4 × 4 rod array, created a maximum flow area reduction of 90% in the central nine subchannels. The blockage sleeve length was 38.3 × rod diameter and the 90% blockage zone length extended for 16.4 × rod diameter. The results showed that upstream of the blockage, the flow was not influenced by the blockage until it reached the location where the inlet taper section of the swelling started. At the downstream end, the flow disturbance was extensive and persisted over a distance of about 83 rod diameters. Compared to the downstream velocity profiles, the turbulence intensity measurements however showed a faster recovery from the blockage influence. At the higher Reynolds number, velocity profiles calculated using the COBRA subchannel computer code compared consistently with the experimental data. The general flow behaviour of the various subchannels was reasonably well predicted. However, at low Reynolds number, due mainly to the frictional form loss calculation scheme in COBRA and uncertainty in the flow transition, the flow diversion due to the blockage to the surrounding unblocked subchannels was overestimated. The influence of the degree of recovery from the rod swelling on the flow was also studied using COBRA.     

Boiling and dryout behind local blockages in sodium cooled rod bundles

Much attention has been given in LMFBR safety analysis to cooling disturbances caused by local blockages within a fuel subassembly. Such blockages are generally considered to be more probable in gridded fuel pin clusters which present the possibility for solid particles in the coolant to be trapped at grids to form a radially extending flow obstruction. The temperature distribution produced in the region of impaired cooling has been studied in water and sodium experiments in pin bundles of various sizes. The experimental work at KfK on local cooling disturbances culminated in two local blockage experiments in the KNS sodium loop simulating LMFBR fuel elements with a 49% central and a 21% corner blockage. In the frame of this work pin cooling in the wake of the blockage was investigated in single-phase conditions, in boiling conditions up to dryout and in conditions simulating gas release from failed pins. The general aims of the studies were to demonstrate that the consequences of a local blockage do not lead to rapid propagation of damage within a pin bundle and to obtain data for validation of theoretical models.     

Flow recovery from a single subchannel blockage in a square-pitched rod array

The flow recovery downstream of a sharp edged plate blockage of one subchannel for axial flow through a square-pitched, six-rod array has been investigated for a rod pitch to rod diameter ratio of 1.107. Wall shear stress, local axial velocities, secondary flow velocities, and all terms of the Reynolds stress tensor were measured. Mean and turbulent components of the air flow were measured by Pitot-static and hot-wire anemometer probes at several axial locations downstream of the blockage for a Reynolds number of 48 × 10³. Flow recovery was incomplete even at the maximum downstream distance ( 90 hydraulic diameters).     

Model for formation and growth of local blockages in grid spaced fast sodium cooled breeder fuel elements

The studies of the formation of local cooling flow disturbances in fast sodium cooled breeder reactors are intended to furnish information, which can be used for safety assessments, about the attainable extent of blockages and the time required for their growth. The formation of major local blockages in normal fuel element geometries will mainly be caused by fuel expelled after a cladding tube failure. A model of the screening process is developed considering the results of inpile experiments, thermohydraulic studies and simulation experiments of the growth of blockages. The model is limited to the initial phase in which the largest free cross sections in a subchannel with a spacer grid geometry must be plugged up to constitute a so-called “fundamental blockage”.Parameter studies on the model show that events constituting the formation of a blockage occur slowly. As a result of the amount of material required, fundamental blockages can be generated in only a small number of subchannels around the point of defect, compared with the familiar release data from in-pile experiments. Other constraints on the basis of detection criteria indicate that the formation of local cooling channel blockages of a size which allows them to be detected by thermocouples is highly improbable. However, even smaller blockages are safely detected by delayed neutron detection because of the very large free surfaces of the fuel particles.     

Results of the temperature noise measurements in subchannels of sodium cooled electrically heated 28 rod bundle behind a blockage

For LMFBR safety studies a 28 rod bundle has been built at Petten (cooperation of GfK and ECN), representing a 60-degrees section of an SNR-300 fuel element having a 70% flat type central blockage. The aims of the temperature noise measurements were to determine the subchannel coolant velocities behind the blockage to study the mixing of coolant in subchannels of different temperature from behind the blockage to the outlet and to study the temperature noise due to boiling in a subchannel. The temperature noise measurements were carried out in parallel to the other measurements (temperature distribution, etc.), using signals of fourteen subchannel thermocouples placed in five measuring planes behind the blockage. The single phase measurements were made with several heat fluxes (5 W/cm² to 120 W/cm²), inlet flows (0.25 to 3 m/s) and inlet temperatures (250°C to 600°C). Two phase flow is initiated and sustained either by a slow and continuous pressure reduction or by stepwise reduction of the main flow. The temperature noise signals were amplified and recorded in analog form. Later the signals were digitized and analysed by digital computers. Part of the signals was also processed by a hardware correlator. The experimental results of the temperature noise measurements will be shown for the different conditions of the loop. Measurements clearly show the following effects:

• - the recirculation flow pattern due to the vortex in the wake behind the blockage;

• - the dependence of r.m.s. value of the noise on the heat flux and the coolant flow;

• - the increase in noise and change in power spectra when going from single phase to the boiling condition.

     

Analysis of temperature noise propagation in a large fuel subassembly of FBR

We have developed a method to calculate the three-dimensional distribution of root-mean-square (RMS) values of temperature noise in the single phase flow in a fast reactor fuel subassembly with a local flow blockage. Employed are the subchannel method in a pin bundle region and the finite difference method in the region downstream of the bundle. We have compared the calculated RMS values of temperature noise with experimental data for a sodium loop test using a wire-spacered 91-pin-bundle fuel sub-assembly with a local blockage. We have investigated the possibility of detection of the blockage by temperature noise by taking into account the influence of structures in the upper part of the subassembly.     

稠密栅元不同子通道内湍流流动的RANS和URANS模拟

本工作采用RANS和非稳态雷诺平均纳维斯托克斯模拟（URANS）方法对稠密栅元内典型子通道——中心通道和壁面通道内的湍流流动进行CFD模拟。研究分析了稠密栅元子通道内的不同周向角度的主流速度、壁面剪应力、湍动能等参数。将模拟计算结果和实验测量结果进行对比,结果表明：RANS模拟在采用各向异性的湍流模型的情况下能较好地模拟P/D较大的稠密栅元通道,但对于P/D较小（P/D<1.1）的稠密栅元通道,CFD结果和实验数据存在较大差距。相比之下,URANS方法可模拟紧密栅元子通道间隙区的大尺度、准周期的流动振动,从而和实验数据拟合良好。推荐采用雷诺应力湍流模型(SSG,ORS)进行RANS模拟,而采用SAS湍流模型进行URANS模拟。     

Analytical studies on local flow blockages in LMFBR subassemblies, using the UZU code

Extensive experimental and analytical studies on local flow blockages in LMFBR subassemblies have been performed in several countries. However, analytical tools have not yet been established. In the present study, therefore, a three-dimensional analytical code, UZU, has been developed for analysing important factors which affect the flow and temperature fields in locally blocked subassemblies.In the UZU code a subchannel approach was employed and the finite difference equations, consisting of momentum, mass and energy balances, were solved by the SOR method. The special features of this code are: (1) fluid properties are functions of local fluid temperature, consequently buoyancy is included in the momentum equations, (2) fluid flow properties, such as turbulent friction factors and eddy diffusivities, are expressed as functions of local fluid velocities, geometries and physical properties, (3) peripheral subchannels are treated separately from inner ones, (4) thickness and porosity of a blockage plate can be considered and (5) a blockage at the upstream end of a grid spacer can be dealt with.The validity of the code was proven by comparison with the measured velocity, pressure and temperature distributions in simulation experiments conducted in parallel with the code development.     

PIV and CFD Study of Turbulent Flow Behavior in Rod Bundle

At the downstream of the spacer grid in a PWR fuel assembly, local disturbance damps out along the flow direction and the flow returns to stable eventually. The turbulent flow and mixing behavior of the coolant are key factors affecting the economy and safety of a nuclear reactor, and need in-depth investigations. In the present paper, the turbulent flow of water in a 3×3 rod bundle was studied using PIV (particle image velocimetry) and CFD. First-order mean velocity and second-order turbulent statistics were obtained. It is found that the velocity in the central subchannel is higher than that in the gap region, but the streamwise root-mean-square velocity behaves inversely. Large-scale flow pulsation induced by the strong streamwise velocity gradient between adjacent subchannels, is observed in the rod bundle, and the wave length increases with Reynolds numbers. In addition, the measured turbulent mixing coefficient is 10% higher than that predicted by the Castellana correlation for PWRs, but this deviation reduces with the increase of Reynolds numbers.     

棒束内湍流流动特性的PIV与CFD研究

在压水堆燃料组件的定位格架下游,局部扰动沿流动方向逐渐衰减,流场最终趋于稳定。光滑棒束区冷却剂的湍流流动和交混特性是影响反应堆经济性和安全性的重要因素,有必要进行深入研究。本文采用粒子图像测速（PIV）与数值模拟(CFD)相结合的方法,对3×3小规模棒束内水的流动特性进行研究,得到了一阶平均流速以及二阶湍流统计信息。结果表明,中心子通道的速度明显高于棒间隙区,但轴向均方根速度呈现出相反的变化趋势。在相邻子通道横向速度梯度的作用下,棒束内出现了大尺度的流量脉动现象,且脉动波长随雷诺数的增加而增大。此外,实验得到的湍流交混系数较压水堆采用的Castellana公式预测值偏高10%左右,这一偏差随雷诺数的增加有减小的趋势。     

Flow Transition Characteristics in 5×5 Rod Bundle Channel

The special geometric structure of the rod bundle channel can induce complicated flow transition of the coolant, and investigation on the flow transition rules is sufficiently important. In the current study, experimental and numerical study on the flow transition characteristics in the 5×5 rod bundle channel was carried out. Experiments were performed to obtain the variation characteristics of the resistance coefficient and CFD simulation was performed using different turbulence models in ANSYS Fluent. The results show that the simulation with SST k-ω turbulence model agrees well with the experimental data. The simulated turbulence intensity and resistance coefficient at different measurement locations and in different flow conditions were compared. For different subchannels, the turbulence intensity and the resistance coefficient are higher in the center subchannel than those in the edge subchannel. For the same subchannel, the turbulence intensity and the shear stress in the subchannel center are higher than those in the subchannel edge. This result indicates that the turbulence intensity, shear stress and resistance coefficient in the rod bundle are not uniform due to the influence of the wall surface. This non-uniform spatial interaction makes the transition point obscure.     

5×5棒束通道内流动转捩特性研究

棒束通道的特殊结构导致其内部流动转捩情况较为复杂,探究其内部流动转捩规律具有重要意义。本文针对棒束通道内的流动转捩特性开展实验与CFD模拟研究,通过实验获得了棒束通道内沿程阻力系数的变化规律;采用不同湍流模型进行了数值模拟。结果表明,SST k-ω模型能较好地反映实验结果。进一步对比了不同雷诺数工况下通道内不同位置的沿程阻力系数与湍流强度,发现对于不同子通道,中心子通道湍流强度与沿程阻力系数高于边角子通道;对于同一子通道,子通道中心处湍流强度与壁面切应力高于子通道边缘处。这一结果说明,受壁面影响,棒束内湍流强度、壁面切应力、阻力特性具有不均匀性,这些空间上的不均匀性相互作用会引起总体上棒束转捩点不明显。     

Coolant mixing and subchannel velocities in an LMFBR fuel assembly

Experimental mixing studies were conducted using water with a 91-pin wire-wrapped fuel assembly to establish a data base for calibrating thermal-hydraulic codes and to supplement and clarify existing mixing data. An electrolytic tracer was employed in conjunction with an isokinetic sampling technique which permitted mixed mean subchannel concentrations and flowrates to be measured without incurring errors resulting from local subchannel gradients. Emphasis was placed upon the behavior of peripheral subchannels.

A complete set of mixing data was obtained which can be used to calibrate thermal-hydraulic codes. Isokinetic velocity measurements revealed that the pitch-average corner, edge, and central subchannel velocities were nearly equal to the bundle average velocity. It was shown that the average peripheral swirl velocity was approximately 1.2 times greater than predicted assuming the fluid follows the wire-wraps. Also the magnitude of the average swirl flow decreases with increasing bundle size. Parametric studies revealed that neither the velocity nor concentration profiles were sensitive to Reynolds number in the range Re = 9000–24000.     

Cladding tube deformation and core emergency cooling in a loss of coolant accident of a pressurized water reactor

The paper summarizes the dominant effects which finally ensure the core coolability of a pressurized water reactor in a loss-of-coolant accident (LOCA).The main results are summarized as follows:

&#x02022; — The cooling effect of the two-phase mixture which is intensified during reflooding increases temperature differences on the cladding tube circumference and thus limits the mean circumferential burst strains to values of about 50%.

&#x02022; — An unidirected flow through the fuel rod bundle during the refill and reflooding phases causes maximum cooling channel blockage of about 70%.

&#x02022; — The coolability of deformed fuel elements can be maintained up to flow blockages of about 90%.

All effects investigated indicate that in a LOCA no impairment of core coolability and public safety has to be expected.     

堵塞条件下紧密栅湍流交混特性研究

开展堵塞工况下紧密栅内流体子通道间隙湍流交混研究,对事故工况下燃料组件热工水力行为的预测具有重要意义。本文采用CFD方法对紧密栅内堵塞工况的流体流动现象进行了模拟,模拟结果与相关文献结果吻合较好。进一步对比分析了不同堵塞工况下,堵塞段及堵塞下游的速度场、涡结构以及湍流交混系数分布。所得不同堵塞工况下的横向与轴向湍流交混系数变化规律,可为子通道分析程序的参数设置提供参考。     

Void distribution and bubble motion in bubbly flows in a 4×4 rod bundle. Part I: Experiments

Lack of local void fraction data in a rod bundle makes it difficult to validate a numerical method for predicting gas–liquid two-phase flow in the bundle. Distributions of local void fraction and bubble velocity in each subchannel in a 4×4 rod bundle were, therefore, measured using a double-sensor conductivity probe. Liquid velocity in the subchannel was also measured using laser Doppler velocimetry (LDV) to obtain relative velocity between bubbles and the liquid phase. The size and pitch of rods were 10 and 12.5 mm, respectively. Air and water at atmospheric pressure and room temperature were used for the gas and liquid phases, respectively. The volume fluxes of gas and liquid phases ranged from 0.06 to 0.15 m/s and from 0.9 to 1.5 m/s, respectively. Experimental results showed that the distributions of void fraction in inner and side subchannels depend not only on lift force acting on bubbles but also on geometrical constraints on bubble dynamics, i.e. the effects of rod walls on bubble shape and rise velocity. The relative velocity between bubbles and the liquid phase in the subchannel forms a non-uniform distribution over the cross-section, and the relative velocity becomes smaller as bubbles approach the wall due to the wall effects.     

Simulation of turbulent flow inside different subchannels in tight lattice bundle

In this paper, both steady and unsteady Reynolds Averaged Navier Stokes (RANS and URANS) methodology are applied to the prediction of turbulent flow inside different subchannels in tight lattice bundles.Two typical configurations of subchannels (i.e., wall subchannel and center subchannel) are chosen to be investigated. In this work the application of different turbulence models implemented in the commercial code CFX v12 is shown. The validity of the methodology is assessed by comparing computational results of axial velocity, wall shear stress and turbulent intensity distributions with the experimental data (Krauss, 1996; Krauss and Meyer, 1998). This study shows that RANS simulation with anisotropic turbulent model produces excellent agreement with experiment, whereas it failed to predict the flow behavior accurately in the case of tightly packed geometries (P/D < 1.1). On the other hand, the URANS simulation is in good agreement with the results in tightly packed geometries with flow oscillation in the gap region. The effects of the Reynolds number and the bundle geometry on the flow oscillation are investigated in details.     

Multi-dimensional analysis of two-phase flow inside subchannels of nuclear fuel assembly

In subchannel analysis, the conservation equations are solved for each channel in a complex fuel bundle, where the effects of fluid exchange between each subchannel are considered. The fluid exchange is commonly referred to as that caused by cross flow. Void drift is considered to be phenomenon resulting from attaining a hydrodynamic equilibrium state. Its mechanism has not been clarified, and the transport due to void drift is therefore estimated through empirical models in conventional subchannel analyses. Therefore, mechanistic model for the void drift phenomenon is required to apply the subchannel analysis to a variety of fuel bundle geometry. In this study, multi-dimensional analysis using two-fluid model was applied to two-phase flow inside a geometry simulating fuel bundle subchannels, for the purpose of clarifying the void drift mechanism. The comparison between the results of the numerical analysis and the experiment confirmed that the reliability of the numerical method used in this study. In this paper, a mechanistic model based on the Stanton number, which expresses the void diffusion coefficient based on the Lahey's proposal, was proposed.     

Method of Critical Power Prediction Based on Film Flow Model Coupled with Subchannel Analysis

A new method was developed to predict critical powers for a wide variety of BWR fuel bundle designs. This method couples subchannel analysis with a liquid film flow model, instead of taking the conventional way which couples subchannel analysis with critical heat flux correlations. Flow and quality distributions in a bundle are estimated by the subchannel analysis. Using these distributions, film flow rates along fuel rods are then calculated with the film flow model. Dryout is assumed to occur where one of the film flows disappears. This method is expected to give much better adaptability to variations in geometry, heat flux, flow rate and quality distributions than the conventional methods.

In order to verify the method, critical power data under BWR conditions were analyzed. Measured and calculated critical powers agreed to within ±7%. Furthermore critical power data for a tight-latticed bundle obtained by LeTourneau et al. were compared with critical powers calculated by the present method and two conventional methods, CISE correlation and subchannel analysis coupled with the CISE correlation. It was confirmed that the present method can predict critical powers more accurately than the conventional methods.