Experimental investigation on bubble characteristics entrained by surface vortex

The cover gas entrainment at the free surface of sodium coolant becomes one of the significant issues according to the compact sizing of reactor vessel in the latest reactor design. In the present study, some basic experiments for the gas entrainment due to the surface vortex were performed in order to obtain the fundamental knowledge about the entrained bubble size. Distributions of entrained bubble diameters in several experimental conditions were obtained from bubble images using an image processing technique. Velocity fields around vortices and surface dimple shapes (gas cores) due to surface vortices were measured to grasp those influences on bubble shapes. The result showed that mean equivalent diameters of bubbles were varied from 1.3 to 2.1 mm in the range of present experimental conditions. The bubble sizes were influenced by the thickness of gas core.     

Energetics of simulated HCDA bubble expansions: Some potential attenuation mechanisms

Experiments conducted to increase our understanding of the dynamics and thermodynamics of expanding bubbles similar to the core disruptive accident (CDA) bubble in liquid metal fast breeder reactors (LMFBR) are described. The experiments were conducted in a transparent model of a typical demonstration-size loop-type LMFBR in which water at room temperature simulated the sodium coolant. Nitrogen gas (1450 psia) and flashing water (1160 psia) qualitatively simulated sodium vapor and molten fuel expansions. Three physical mechanisms that may result in attenuation of the work potential of a hypothetical CDA (HCDA) were revealed by the experiments: (1) the pressure gradient existing between the lower core and the bubble within the pool, (2) the hydrodynamic effects of vessel internal structures, and (3) the nonequilibrium flashing process occurring in the lower core. These three mechanisms combine to result in a coolant axial slug kinetic energy that is only 14% of the work potential of the ideal quasi-static nitrogen expansion and only 5% of the work potential of the ideal quasi-static flashing water expansion.     

Study on thermal-hydraulic behavior during molten material and coolant interaction

In a core disruptive accident (CDA) of a Fast Breeder Reactor, the post accident heat removal (PAHR) is crucial for the accident mitigation. The molten core material should be solidified in the sodium coolant in the reactor vessel. The material, being fragmented while solidification and forming debris bed, will be cooled in the coolant.

In the experiment, molten material jet is injected into water to experimentally obtain the visualized information of the fragmentation and boiling phenomena during PAHR in CDA. The experiment shows that the break up of the molten material into fine fragments is observed at the front, side and middle part of the jet during very short time interval. The distributed particle behavior of the molten material jet is observed with high-speed video camera. And the visual data is analyzed with Particle Imaging Velocimetry (PIV).

The experimental results are compared with the existing theories. Consequently, the marginal wavelength on the surface of a water jet is close to the value estimated based on the Rayleigh–Taylor instability. Moreover, the fragmented droplet diameter obtained from the interaction of molten material and water is close to the value estimated based on the Kelvin–Helmholtz instability.     

Annular flow entrainment rate experiment in a small vertical pipe

Two-fluid model predictions of film dryout in annular flow, leading to nuclear reactor fuel failure, are limited by the uncertainties in the constitutive relations for the entrainment rate of droplets from the liquid film. The main cause of these uncertainties is the lack of separate-effects experimental data in the range of the operating conditions in nuclear power reactors. An air–water experiment has been performed to measure the entrainment rate in a small pipe. The current data extend the available database in the literature to higher gas and liquid flows and also to higher pressures. The measurements were made with the film extraction technique. A mechanistic model was obtained based on Kelvin–Helmholtz' instability theory. The dimensionless model includes the Weber number of the gas and the liquid film Reynolds number. Kataoka and Ishii's correlation (Kataoka, I., Ishii, M., 1982. NUREG/CR-2885, ANL-82-44) is modified based on this model and the new data. The new correlation collapses the present air–water data and Cousins and Hewitt's data (Cousins, L.B., Hewitt, G.F., 1968. UKAEA Report AERE-R5657) The effects of pressure and surface tension were considered in the derivation so it may be applied for boiling water reactor operating conditions.     

Analysis of gas–liquid metal two-phase flows using a reactor safety analysis code SIMMER-III

SIMMER-III, a safety analysis code for liquid-metal fast reactors (LMFRs), includes a momentum exchange model based on conventional correlations for ordinary gas–liquid flows, such as an air–water system. From the viewpoint of safety evaluation of core disruptive accidents (CDAs) in LMFRs, we need to confirm that the code can predict the two-phase flow behaviors with high liquid-to-gas density ratios formed during a CDA. In the present study, the momentum exchange model of SIMMER-III was assessed and improved using experimental data of two-phase flows containing liquid metal, on which fundamental information, such as bubble shapes, void fractions and velocity fields, has been lacking.

It was found that the original SIMMER-III can suitably represent high liquid-to-gas density ratio flows including ellipsoidal bubbles as seen in lower gas fluxes. In addition, the employment of Kataoka–Ishii’s correlation has improved the accuracy of SIMMER-III for gas–liquid metal flows with cap-shape bubbles as identified in higher gas fluxes. Moreover, a new procedure, in which an appropriate drag coefficient can be automatically selected according to bubble shape, was developed.

Through this work, the reliability and the precision of SIMMER-III have been much raised with regard to bubbly flows for various liquid-to-gas density ratios.     

A review of modeling concepts for sodium-concrete reactions and a model for liquid sodium transport to the unreacted concrete surface

A review of existing modeling concepts and studies of sodium-concrete reactions is presented. Consistent with experimental observations, the current modeling study being conducted at Hanford Engineering Development Laboratory assumes for hydrated concretes the presence of liquid layer of reaction products intervening between sodium pool and concrete surface. Primary liquid component in this layer is NaOH which has a low melting point. This liquid component dissolves the reaction products such as silicates, aluminates and forms a very viscous liquid more dense than sodium. As this layer assumes a significant thickness, the only mechanism available for transport of sodium to fresh concrete surface is the motion and agitation induced by gas bubbles consisting of hydrogen, water vapor, CO₂ and sodium vapor. However, to date there exists no satisfactory model that describes this transport mechanism. To rectify this shortcoming, we propose a mass “iffusion” model for sodium transport. The model reduces the sodium transport process by bubble motion to a single unknown parameter which has the appearance of a diffusion coefficient and consequently can be determined by solving an inverse problem in conjunction with measured “concentration” distributions in simulant material experiments.     

气泡破裂产生膜液滴现象可视化实验研究

气泡破裂产生液滴的现象普遍存在于核电厂蒸汽发生器中，由此产生的液滴是蒸汽流夹带液滴的主要来源之一。本工作利用可视化装置及高速摄像技术拍摄气泡在自由液面处破裂产生液滴的实验现象。结果表明，在实验气泡尺寸范围内，气泡液帽破裂产生膜液滴形式都是单点破裂，首先液膜进行排液，其后液帽顶部或底部产生孔隙，然后孔隙迅速扩大、液膜卷曲形成液体环，最终发生不稳定射流形成细小的膜滴。对实验数据进行拟合，得到气泡等效半径在3~25 mm范围内膜液滴数量同气泡尺寸关系式，实验结果与前人结果变化趋势相似。     

Gas entrainment allowance level at free surface and gas dynamic behavior of sodium-cooled fast reactor

In a sodium-cooled fast reactor (SFR), inert gases exist in the primary coolant system either in a state of dissolved gas or free gas bubbles. The sources of the gas bubbles are entrainment and dissolution of the reactor cover gas (argon) at the vessel free surface and emission of the helium gas that is produced as a result of disintegration of B₄C control rod material. The gas in the primary system may cause disturbance in reactivity, nucleation site for boiling, etc. Therefore, it is a key issue from the design and safety viewpoint and the allowance level is necessary regarding the gas entrainment at the free surface and the gas bubble concentration in the primary system. In the present study, a gas entrainment allowance level at the free surface is discussed and rationalized for the Japanese SFR (JSFR) design. The influence of the gas entrainment is evaluated using the void fraction at the core inlet. Design criteria for the acceptable level of the gas entrainment and gas concentration are proposed in consideration of the background level of gasses in the coolant. For the purpose, a plant dynamics code VIBUL has been developed to apply to the JSFR design to evaluate the concentration distribution of the dissolved gas and the free gas bubble in the JSFR system. Using the plant dynamics code for the bubble behavior, the background level of the free gas (void fraction at the core inlet) has been obtained. Assuming that the total void fraction should be kept below 105% of the background level, a preliminary design allowance level of gas entrainment is proposed as the map in terms of the entrainment rate and the entrained bubble radius. Furthermore, the possibility of bubble removal and design requirement of the device is investigated to satisfy the allowance level. It is noted that the background level is already very low in comparison with the induced void reactivity by the void passing the reactor core.     

Two-dimensional numerical simulation of single bubble rising behavior in liquid metal using moving particle semi-implicit method

Gas-lift pump in liquid metal cooling fast reactor (LMFR) is an innovative conceptual design to enhance the natural circulation ability of reactor core. The two phase flow characteristics of gas–liquid metal make significant improvement of the natural circulation capacity and reactor safety. It is important to study bubble flow in liquid metal. In present study, the rising behaviors of a single nitrogen bubble in 5 kinds of common stagnant liquid metals (lead bismuth alloy (LBE), liquid kalium (K), sodium (Na), potassium sodium alloy (Na–K) and lithium lead alloy (Li–Pb)) and in flowing lead bismuth alloy have been numerically simulated using two-dimensional moving particle semi-implicit (MPS) method. The whole bubble rising process in liquid was captured. The bubble shape, rising velocity and aspect ratio during rising process of single nitrogen bubble were studied. The computational results show that, in the stagnant liquid metals, the bubble rising shape can be described by the Grace's diagram, the terminal velocity is not beyond 0.3 m/s, the terminal aspect ratio is between 0.5 and 0.6. In the flowing lead bismuth alloy, as the liquid velocity increases, both the bubble aspect ratio and terminal velocity increase as well. This work is the fundamental research of two phase flow and will be important to the study of the natural circulation capability of Accelerator Driven System (ADS) by using gas-lift pump.     

Comparison of various models for the simulation of a Core Disruptive Accident in the MARA 10 mock-up

During a Hypothetical Core Disruptive Accident in a Liquid Metal Fast Breeder Reactor, it is assumed that the core of the nuclear reactor melts down partially and that the interaction between hot molten fuel and relatively cold liquid sodium creates a high-pressure gas bubble in the core. The violent expansion of this bubble loads and deforms the reactor vessel and the internal structures, thus endangering the safety of the nuclear plant.The MARA 10 experimental test simulates a Core Disruptive Accident in a 1/30-scale mock-up schematising a reactor block. In the mock-up, the liquid sodium cooling the reactor core is replaced by water and the argon blanket laying below the reactor roof is simulated by an air blanket. The explosion is triggered by an explosive charge.This paper presents some models available within the EUROPLEXUS code to simulate a Core Disruptive Accident and an analysis of the computed results. In particular, results are compared with experimental measurements and previous numerical simulations carried out with the codes SIRIUS and CASTEM-PLEXUS.     

Numerical interpretation of the 8 experiment simulating a hypothetical core disruptive accident

In the case of a hypothetical core disruptive accident (HCDA) in a liquid metal fast breeder reactor (LMFBR), it is assumed that the core of the nuclear reactor has melted partially and that the chemical interaction between the molten fuel and the liquid sodium has created a high-pressure gas bubble in the core. The violent expansion of this bubble loads and deforms the reactor vessel, thus endangering the safety of the nuclear plant. The experimental test 8 simulates the explosive phenomenon in a mock-up included in a flexible vessel with a flexible roof. This paper presents a numerical simulation of the test and a comparison of the computed results with the experimental results and previous numerical ones.     

Numerical simulation of a Hypothetical Core Disruptive Accident in a small-scale model of a nuclear reactor

In the case of a Hypothetical Core Disruptive Accident (HCDA) in a Liquid Metal Fast Breeder Reactor, it is assumed that the core of the nuclear reactor has melted partially and that the chemical interaction between molten fuel and liquid sodium has created a high-pressure gas bubble in the core. The violent expansion of this bubble loads and deforms the reactor vessel and the internal structures, thus endangering the safety of the nuclear plant.The MARA 10 experimental test simulates a HCDA in a 1/30-scale mock-up schematising a reactor block. In the mock-up, the liquid sodium cooling the reactor core is replaced by water and the argon blanket laying below the reactor roof is simulated by an air blanket. The explosion is triggered by an explosive charge.This paper presents a numerical simulation of the test with the EUROPLEXUS code and an analysis of the computed results. In particular, the evolution of the fluid flows and the deformations of the internal and external structures are analysed in detail. Finally, the current computed results are compared with the experimental ones and with previous numerical results computed with the SIRIUS and CASTEM-PLEXUS codes.     

AP1000 ADS-4夹带卸压试验模拟分析

第4级自动降压系统（ADS-4）是AP1000极为重要的非能动安全设施。ADS-4能在AP1000小破口失水事故中为反应堆系统提供可控卸压。然而,大量的冷却剂可通过卸压过程中ADS-4夹带和上腔室夹带被带到安全壳中,从而引发堆芯裸露和堆芯熔化事故。为研究小破口事故中的ADS-4夹带卸压和上腔室夹带过程,在以AP1000为原型、按直径/高度比1∶5.6设计建造的ADS-4喷放卸压试验回路（ADETEL）中,研究了不同初始压力、压力容器混合液位和加热功率下的夹带和卸压行为,以及反应堆内部构件的夹带沉积效应。试验数据表明,大量的水在短时间内迅速通过ADS-4支管被夹带出来。液体的夹带率和压力容器混合液位的降低速率随系统初始压力的增加而增大。值得注意的是,在本试验特定工况下,初始压力为0.5 MPa时出现堆芯裸露。堆内构件对夹带量和压力容器混合液位无显著影响。     

Hydrodynamic characteristics of a gas blowdown into a two-dimensional liquid pool

Hydrodynamic experiments were performed on transient gas jet discharge into liquid pool. Liquid entrainment, jet development, and the possibility of liquid vaporization by a noncondensible gas were investigated.Liquid entrainment during jet expansion in the plenum was seen to increase with increasing pressure. However, the fraction of the expansion volume which is liquid was not affected by gas pressure. Higher initial entrainment was seen for a more dense fluid. A Taylor instability mechanism for entrainment seems to fairly predict the rate of entrainment. Another entrainment mechanism besides that of Taylor instabilities may exist at higher pressures.     

Computation of a core disruptive accident in the MARS mock-up

A hypothetical core disruptive accident in a liquid metal fast breeder reactor (LMFBR) results from the interaction between molten fuel and liquid sodium, which creates a high-pressure bubble of gas in the core. The violent expansion of this bubble loads and deforms the vessel and the internal structures. The MARS experimental test simulates a HCDA in a small-scale mock-up containing all the significant internal components of a fast breeder reactor. The mock-up is filled with water, topped by an argon blanket, and the explosion is generated by an explosive charge.This paper presents a numerical simulation of the test with the EUROPLEXUS code. The top closure is represented by massive structures and the main internal structures are described by shells. The current numerical results are described and compared with the experimental ones, and previous computations with the CASTEM-PLEXUS code.     

Hydraulic design of sparger arrangement for long guide tubes of a fast breeder reactor

In liquid metal cooled fast reactors, the core is submerged in sodium pool by ∼5 m below sodium free surface. This necessitates the control and shutdown of reactor be achieved by long overhanging mechanisms housed inside a control plug. These mechanisms are protected by porous guide tubes with a sparger type arrangement for the sodium flow through them. Comprehensive knowledge of flow distribution of sodium through these guide tubes is essential to assess the risks of flow induced vibration of thin thermowell tubes that pass close to these shroud tubes and entrainment of cover gas due to high free surface velocities. Three dimensional hydraulic analysis of single isolated shroud tube and integrated assembly of shroud tubes have been carried out using CFD tools to acquire this knowledge. The predictions of the CFD models have been validated against experimental predictions. These studies have provided important information regarding critical design parameters. Size of holes in the shroud tube, location of holes in the control plug shell and arrangement for breaking sodium jets emanating from shroud tubes have been optimized to reduce free surface velocity.     

Onset Criterion for Liquid Entrainment in Reflooding Phase of LOCA

Reflooding tests were conducted in a rod bundle geometry at the maximum pres- sure of 12 MPa to investigate thermal-hydraulic behavior during a small break loss-of-coolant accident (SBLOCA) in a nuclear reactor. The test conditions ranged 0.6 ～ 12 MPa for pressure, up to 920 K for rod surface temperature, up to 20 cm/s for bundle inlet flow velocity and up to 2 kW/m for linear power input. The principal objective of this paper is to investigate the onset condition for liquid entrainment by steam flow in the relatively high pressure reflooding phase. Experimental results showed a tendency that the liquid entrainment decreased with increase in pressure when the other parameters such as an inlet flow rate and rod temperature were fixed. A new correlation for the onset criterion for liquid entrainment was derived from the experimental results and an analysis of a force balance for a liquid droplet. Effects of pressure on liquid entrainment in the reflooding phase were made clear from the experimental and analytical results.     

液态金属高温热管传热极限研究

热管作为一种高效可靠、可进行长距离传热的非能动设备,在核能领域有着广泛的应用。本文针对工质为钠、充液量为158 g与208 g的毛细驱动热管,对其传热极限开展实验和理论研究。实验方面,设计搭建了高温热管传热极限测试分析实验平台,研究了液态金属高温热管在不同水平倾角和不同加热功率下传热功率的变化。理论方面,验证了连续流动极限与夹带极限理论模型的正确性,总结了两种极限的发生规律。研究发现,热管连续流动极限影响热管的启动;由于水平夹角较大时转变温度较高,因此大角度下的热管更容易发生连续流动极限,小角度下经验模型的预测误差在6.58%以内,大角度下误差超过28%。夹带极限发生时热管蒸发段温度骤升且冷凝段温度出现波动,热管倾角越大夹带极限越容易发生,经验模型在不同角度下均存在误差,大角度下误差超过100%。本文总结了连续流动极限与夹带极限的发生规律,为先进核反应堆系统中热管的设计提供参考。     

A fundamental approach to specify thermal and pressure loadings on containment buildings of sodium cooled fast reactors during a core disruptive accident

Reactor Containment Building (RCB) is the ultimate barrier to the environment against activity release in any nuclear power plant. It has to be designed to withstand both positive and negative pressures that are credible. Core Disruptive Accident (CDA) is an important event that specifies the design basis for RCB in sodium cooled fast reactors. In this paper, a fundamental approach towards quantification of thermal and pressure loadings on RCB during a CDA, has been described. Mathematical models have been derived from fundamental conservation principles towards determination of sodium release during a CDA, subsequent sodium fire inside RCB, building up of positive and negative pressures inside RCB, potential of in-vessel sodium fire due to failed seals and temperature evolution in RCB walls during extended period of containment isolation. Various heating sources for RCB air and RCB wall and their potential have been identified. Scaling laws for conducting CDA experiments in small-scale water models by chemical explosives and the rule for extrapolation of water leak to quantify sodium leak in reactor are proposed. Validation of the proposed models and experimental simulation rules has been demonstrated by applying them to Indian prototype fast breeder reactor. Finally, it is demonstrated that in-vessel sodium fire potential is very weak and no special containment cooling system is essential.