2.1 Steel containment

The experimental facility includes a steel containment, a hot air supply, and an injection system and a data acquisition system. Figure 1 presents the schematic of the facility and the details of the distribution of the measuring points on the containment.

Figure 1.Full-screen View

Schematic of the experimental facility

The cylindrical containment is made of Type 304 stainless steel with the height of 3.34 m, a diameter of 2 m, and a wall thickness of 6 mm. The containment has an elliptical upper dome and its cylindrical lower part is embedded underground. There is a rectangle door made of a 0.3-cm-thick stainless steel plate and it can be removed when necessary. The jet tubing is horizontally inserted into the containment through this wall at its center. Air enters the containment from the jet and leaves through the exit on the bottom inside of the containment.

The air supply system provides steady hot, dry air for the experiment with an air blower, an air heater, a volumetric flow meter, a set of corrugated stainless steel jet tubing, an angle-adjust rack, and a nozzle. There are two platinum resistance temperature sensors on the air heater so the temperature can be controlled by the feedback signal. The jet tubing is wrapped by insulating materials to reduce the heat loss. The diameter of the nozzle is 25 mm and the height of the nozzle is 60 cm from the floor of the containment. The flow rate can be monitored by the volumetric flow meter and controlled by the valves.

Thermocouples are distributed in the test area inside the containment. They are connected to the Agilent multifunction switch/measure unit for data collection, which will be processed by the bundled software Bench Link Data Logger for further analysis. The distribution of thermocouple spots for the thermal stratification measurement is also shown in Figure 1. Details of the distribution are as follows:

(a) There are 7 layers (red lines) in the containment, and each elevation is given in Figure 1.

(b) The thermocouples are placed at the center of the cross section, at the midpoint of the radius, and 5 cm away from the inner wall, including the ones that adhere on the inner wall so there are 13 thermocouples on each layer in total.

(c) There is a thermocouple installed at the nozzle.

(d) The T-type thermocouples are used in the containment space and the K-type thermocouples are applied to obtain the wall temperature.

(e) There are five heat flow meters pasted on the inner wall (five black spots), and on the outer wall there are eight thermocouples (eight green spots), as shown in the sketch.

The parameters of the steel containment are listed in Table 2. It can be considered an even smaller modular reactor than the one listed in Table 1.

The experiments were carried out for non-adiabatic and adiabatic conditions. In non-adiabatic conditions, the steel containment was exposed to the air. In adiabatic conditions, the containment was covered with two layers of sponged rubber insulation sheets. In each condition, the experiments were performed at three different injection temperatures (T_in) with a constant flow rate (q_v) of 15 m³/h, and two different flow rates (q_v) with constant injection temperatures (T_in) of 100℃.Then, the comparison can be made by changing only one parameter to get a better understanding of the influence of each parameter. In the beginning of each test, the inlet temperature at the injection nozzle was controlled by setting the power of the air heater. Each individual experiment will last for more than 6 hours to reach a steady state, and then the data can be collected. Equipment in the containment, including thermal couples mounted on lines or on the wall, a tube and a rack, all occupied very little volume, so their impact on the data was limited. Besides, the data was collected in a steady state, so the influence of equipment in the containment was considered negligible for the stratification phenomena.

2.2 Visualized containment

In order to display the jet mixing and stratification development, the visualization experiments were carried out by using boron silicone glass containment. The experimental setup consists of two parts, the water supply control system and glass containment, as shown in Figure 2. The main function of the water supply control system is to provide and control the dyed water. The water tank is made of boron silicone glass and has 6 built-in heating rods. The dyed water must be heated to less than 80℃. A graduated cylinder is used to collect and buffer the dyed water, which is injected into the containment. The top and bottom of the graduated cylinder is sealed and its nominal capacity is 400～600 mL，with a minimum scale of 5 mL. There are three holes on the side of the graduated cylinder, which are distributed uniformly and can be connected to the flange. The middle hole is an inlet, the bottom hole is an outlet, and the top hole is a spare hole. The inner diameter of the graduated cylinder is 50 mm and the height is 300 mm. There is a joint at 50, 150, and 300 mm on the side of the graduated cylinder, respectively. Control valves are located both upstream and downstream of the graduated cylinder. A glass rotameter is used to control the volume flux.

Figure 2.Full-screen View

Visualized experimental setup.

The transparent containment is made of boron silicone glass, which is full of saline water before conducting the experiments. There are holes on the glass containment body. The one at the top of the dome serves as the air vent to ensure that the injected water can fill with the containment. The other two are backup connectors located at the upper and middle vertical sections of the body. The glass containment body is supported by a steel pallet, and there is an inlet at the center and an outlet located nearby. A stainless steel nozzle is inserted in the inlet of the pallet to provide jet water. The parameters of the experimental setup are listed in Table 3.

3.1 CFD model

CFD simulations were done to provide comparisons with the experimental results. Figure 3 shows the geometry of the CFD simulation. All parameters and boundary conditions in the models are consistent with the experiments. The ceiling dome and over ground cylindrical side wall are considered to be air cooled by natural convection under a constant room temperature. The underground wall and the floor are considered to be a constant temperature corresponding to the experimental conditions. The turbulence k-ε model and steady method were used in the calculations.

Figure 3.Full-screen View

Geometry for CFD simulations.

3.2 Comparison

When equations (1) and (2) are applied in this experiment, H_sf is the inside height of the containment and d_bjo is the diameter of the nozzle. The ambient fluid density, ${\rho }_a$, is calculated using the average temperature in the containment. The source fluid density, ${\rho }_0$, is calculated by the jet temperature at the nozzle. The jet source speed, U₀, can be calculated by the flow rate.

Table 4 lists some key parameters of the tests. The Richardson number and the criteria of eq. (1) are calculated to provide the stratification prediction for the corresponding conditions. All the conditions carried out in the experiments were stratified.

The comparison between the experimental results and the calculation results are given in Figure 4 and Figure 5. The X-axis in the graph is the height in the containment, and the Y-axis represents the average temperature of the cross section in the corresponding height. In CFD simulation, the average temperatures of each layer are taken from the cross sections at 20 cm intervals from the bottom to the top. The curves of the simulation temperatures can be extended to the vertex while the experimental data are only available to a height of 315 cm due to the experimental conditions.

Figure 4.Full-screen View

Temperature distribution in the containment, Non-adiabatic atdifferent initial conditions.

Figure 5.Full-screen View

The temperature distribution in the containment, Adiabatic atdifferent initial conditions.

Figure 4 illustrates the temperature distribution comparisons in the containment both experimentally and numerically under non-adiabatic conditions. In the test conditions shown in (a), the flow rate is 15 m³/h, and inlet temperatures of 50℃, 80℃, and 100℃ are applied, respectively. It can be seen that the temperature distribution demonstrates stronger stratification with a higher injection temperature. This can be demonstrated by the CFD simulations, which reach a good agreement with the experiments. From the floor to some height after the nozzle, the temperatures climb up dramatically along the axial height with a large temperature gradient and, hence, cause severe thermal stratification. Then, the temperature increases slowly and becomes smoother from a height of 100 cm to 300 cm and maintains a relatively high temperature, which implies that most energy is stored in the upper part of the cylindrical body. In the upper part of the containment, the temperature increases again due to the shape of the dome, in which heat can be accumulated.

Figure 4(b) displays the temperature distribution with an inlet temperature of 100℃, and flow rates of 10 m³/h and 18 m³/h, respectively. It shows the comparison between different injection flow rates, where the stratification is more pronounced with a higher flow rate. The reasons will be discussed thoroughly through the dimensionless analysis of Figure 6.

Figure 6.Full-screen View

Comparison of dimensionless temperature distribution with dimensionless height for different conditions at low and high modes of flow rate respectively in experiment.

Figure 5 indicates that the trend of the results under adiabatic conditions is in good agreement with the non-adiabatic conditions, i.e. the temperatures surge up in the area near the nozzle and then slightly ascend from the heights of 100 cm to 300 cm. They continue rising to reach the peak at the vertex of the dome. The simulations, however, show a higher temperature than the experiments as a whole. There are two reasons that may account for this. The main reason may lie in insufficient isolation from the sponged rubber sheet wrapping on the containment. The geometrical shape changes significantly at the connection part of ceiling wall and the cylindrical wall so it impairs the heat isolation of the insulating layer. This can explain why there is a slight mismatch between the CFD simulations and experiment that reach a maximum in the upper portion. In addition, there is an outlet at the bottom inside containment and air would exhaust through it, which may take away some energy. However, the temperature there was relatively low, so there was not supposed to be too much heat loss this way.

In our experiments, the increasing flow rate of the vertical injection causes more severe stratification in both conditions. Theoretically, by increasing flow rate, the thermal stratification will be weakened and the stratification will break down due to the enhancement of circulation flow, according to equations (1) and (2). Such phenomena cannot be observed in our experiments, but it can be demonstrated from Table 4 that Ri and the value of the criteria decrease with the increasing flow rate. This is different from the increasing inlet temperature. This can be explained by the insufficient power provided by the air blower. The flow rate is not large enough to reach the critical value to mix the stratification, but a higher flow rate does inject more energy and, as a result, increasing the inlet temperature causes similar thermal stratification to occur.

This point can be interpreted more clearly through the data in Figure 6, which is taken from the experiments. The results are in terms of the dimensionless temperature, (T-T1)/(T2-T1), with the dimensionless height, z/H, where T1 and T2 being, respectively, the temperatures at the bottom and the top of the containment, and H being the total height of the containment. Besides the slightly higher temperature under adiabatic conditions than under non-adiabatic conditions, the stratifications are all formed near the nozzle, and above the nozzle is the high temperature zone. The experiments carried out in the visualized containment will show such a phenomenon in section 3.3. We can deduce from Figure 4 to Figure 6 that the stratifications are formed because:

(a) The increasing flow rate and inlet temperature share the same principle, the increasing the hot air flow rates are actually increasing the temperature in the upper zone within our experimental conditions;

(b) In the containment, the cylindrical wall underground keeps the lower part a relatively low temperature, especially in adiabatic conditions, since most heat was conducted there, hence, accounting for the stable stratification;

(c) Due to (a) and (b), the stable stratified layers are built up and becomes strong enough to overcome mixing forces in these cases. In other words, the flow rates applied in the experiments cannot mix the temperature in such conditions.

In comparing the same conditions under non-adiabatic and adiabatic conditions, respectively, it can be concluded that the thermal stratification is more pronounced under adiabatic conditions rather than the condition that is exposed to the air. In other words, the wall exposed to the air influences the heat conduction in the containment, while the insulating layer decrease the heat loss shown in the comparison between Figure 4 to Figure 5. The cylindrical wall underground keeps the lower part of containment a relatively low temperature in all cases, hence, accounting for the thermal stratification. Therefore, the thermal stratification is the consequence of heat dissipation through the wall and the heat conduction underground when a LOCA scenario occurs.

3.3 Illustrations by the visualized containment

The experiments carried out by the visualization of the containment provided images of the buoyant jet development, and a transient simulation model using ANSYS CFX was done based on the experimental conditions. In the simulation, the dyed water was simplified to clear water, and the transient process lasted for 101 s. The simulation pictures were outputed every second. The meshing figure and cross-sectional views are shown in Figure 7.

Figure 7.Full-screen View

CFD modeling and simulation for visualized containment.

The initial temperature of the containment is 15℃. The volumetric flow rate of the injected water is 50 L/h and the temperature is 60℃. The density of the saline water, which fills with the containment, is 1010 kg/m³. Therefore, the dyed water in the experiment corresponds to the hot water in the simulation.

The comparisons between the experiment and simulation are illustrated from Figure 8 to Figure 13 and show good agreement. The development of the buoyant jets in the containment can be sufficiently described by three stages. At the initial stage, the hot water came from the buoyant jet with a lower density raised by the momentum at the inlet nozzle, then raised to the vertex rapidly (Figure 8 to Figure 9). Then, in the second stage, called developing stage, where the buoyant jet spreads along the inner wall from the top and forms a mushroom shape due to the features of the dome. It gradually spreads out in the space of the containment, and a circular vortex is formed between the jet water and the wall (Figure 10 to Figure 11). Meanwhile, an interface appeared due to the density difference of the jet water and the ambient water in this stage. As time went on, the interface headed downward gradually until it entered the stable stratified stage (Figure 12 to Figure 13), and the stratified interface was finally maintained slightly above the nozzle (Figure 13).This result, in turn, meets the stratification location of the analysis in Figure 6. The experimental results and the simulation show a good agreement in the first stage (jetting stage) and the third stage (stratified stage). In the second stage, as the developing stage, the dyed water diffused a little bit faster than the simulation, but qualitatively, the visualized experiments combined with the numerical simulation illustrated the development of buoyant jets in the containment.

Figure 8.Full-screen View

The 1^st second of the experiment and the 1^st second of the simulation.

Figure 9.Full-screen View

The 10^thsecond of the experiment and the 10^thsecond of the simulation.

Figure 10.Full-screen View

The 30^th second of the experiment and the 30^th second of the simulation.

Figure 11.Full-screen View

The 50^th second of the experiment and the 50^thsecond of the simulation.

Figure 12.Full-screen View

The 70^th second of the experiment and the 70^thsecond of the simulation.

Figure 13.Full-screen View

The 100^th second of the experiment and the 100^th second of the simulation.