**1. Introduction**

46 Nuclear Reactors

Zuber, N.; Staub, F. W. & Bijwaard, G. (1966). Vapour void fraction in subcooled boiling

24-38, Chicago.

systems, Proceeding of the third International heat transfer conference, Vol. 5, pp.

Buoyancy-driven exchange flows of helium-air were investigated through horizontal and inclined small openings. Exchange flows may occur following a window opening as ventilation, fire in the room, over the escalator in the underground shopping center as well as a pipe rupture accident in a modular high temperature gas-cooled nuclear reactor. Fuel loading pipe is located in the inclined position in the pebble bed reactor such as Modular reactor (Fumizawa, 2005, Kiso, 1999) and AVR(El-Wakil, 1982, Juni-1965, 1965).

In safety studies of High Temperature Gas-Cooled Reactor (HTGR), a failure of a standpipe at the top of the reactor vessel or a fuel loading pipe may be one of the most critical designbase accidents. Once the pipe rupture accident occurs, helium blows up through the breach immediately. After the pressure between the inside and outside of he pressure vessel has balanced, helium flows upward and air flows downward through he breach into the pressure vessel. This means that buoyancy-driven exchange flow occurs through the breach, caused by density difference of the gases in the unstably stratified field. Since an air stream corrodes graphite structures in the reactor, it is important to evaluate and reduce the air ingress flow rate during the standpipe rupture accident.

Some studies have been performed so far on the exchange flow of two fluids with different densities through vertical and inclined short tubes. Epstein(Epstein, 1988) experimentally and theoretically studied the exchange flow of water and brine through the various vertical tubes. Mercer et al. (Mercer, 1975) experimentally studied an exchange flow through inclined tubes with water and brine. He performed the experiment in the range of 3.5 <L/D < 18 and 0 deg < θ < 90 deg, and pointed out that the length-to-diameter ratio L/D, and the inclination angle θof the tube are the important parameter for the exchange flow rate. Most of these studies were performed on the exchange flow with a relatively small difference of the densities of the two fluids (up to 10 per cent). However, in the case of HTGR standpipe rupture accident, the density of the outside gas is at least three times larger than that of the gas inside the pressure vessel. Few studies have been performed so far in such a large range of density difference. Kang et al. (Kang, 1992) studied experimentally the exchange flow through a round tube with a partition plate. Although we may think that the partitioned plate, a kind of obstacle in the tube, decrease the exchange flow rate, he found that the exchange flow rate is increased by the partition plate because of separation of an upward and downward flow.

Herium-Air Exchange Flow Rate Measurement Through a Narrow Flow Path 49

"S-shape smoke line

Exchange flow path

Air

Fig. 3. The visualized data listed along the elapsed time (L/D=18.3, flame rate=200 f/s,

The typical exchange flow in the tube was visualized in the figure 2. The visualized exchange flow resembles to the S-shape. The flames are detected 200 and 500 flames in a second by the high-speed camera. The visualized data listed along the elapsed time in the figures 3 and figure 4, respectively. It is clearly visualized that helium up-flow along left hand side and air down-flow along right hand side. Figure 5 shows time history of the upward flow peak velocity measured by PTV. In the case of L/D=18.3, the average velocity value *U0* is evaluated as 0.315 m/s. In the case of L/d=10, the average velocity value *U0* is

(a) Ignition: 0 sec (b)Elapsed time:0.050 sec

Fig. 2. Mechanism of exchange flow

Heliumas

θ=30 deg)

**2.2 Results** 

The objectives of the present study are to investigate the behavior of the exchange flow, i.e., exchange flow through the round long tube by several flow visualization method, then to evaluate the exchange flow rate by the PTV and PIV methods and mass increment with helium-air system. Therefore the following methods are investigated in the present study.


## **2. Smoke wire method**

#### **2.1 Experimental apparatus and procedure**

The smoke wire method was used for the present investigations. Figure 1 shows a typical sketch of the apparatus. It consists of a smoke pulse generator, thin Nichrome wire with oil and a test chamber. This figure also shows the high-speed camera system, and it transfers the visual digital data to the personal computer for data acquisition. The experimental procedure is as follows. The test chamber is filled with pure helium. By removing the cover plate placed on the top of the tube, exchange flow, i.e., exchange flow of helium and air is initiated. At such condition, the smoke pulse generator ignites the high voltage. Immediately a smoke appeared and visualized the helium up flow and the air down flow in the flow path in the long tube. Test chamber diameter and height are 350 mm, the long tube on the test chamber diameter(=D) are 17.4 and 20mm, length of it (=L) are 200 and 319mm. They denote L/D=10 and L/D=18.3 respectively. It simulates a typical long tube. The inclination angle θis 30 deg. The smoke wire conditions are as follows. The voltage is around 250 (V), current duration is 30 m sec, and the oil of thin wire is CRC-556. The high-speed camera system using D-file records the visual data up to 1600 flames in a second. Upward flows peak velocity measured by PTV method.

Fig. 1. Exchange Flow Apparatus of Smoke Wire Method High speed camera system

Fig. 2. Mechanism of exchange flow

Fig. 3. The visualized data listed along the elapsed time (L/D=18.3, flame rate=200 f/s, θ=30 deg)

### **2.2 Results**

48 Nuclear Reactors

The objectives of the present study are to investigate the behavior of the exchange flow, i.e., exchange flow through the round long tube by several flow visualization method, then to evaluate the exchange flow rate by the PTV and PIV methods and mass increment with helium-air system. Therefore the following methods are investigated in the present study.

The smoke wire method was used for the present investigations. Figure 1 shows a typical sketch of the apparatus. It consists of a smoke pulse generator, thin Nichrome wire with oil and a test chamber. This figure also shows the high-speed camera system, and it transfers the visual digital data to the personal computer for data acquisition. The experimental procedure is as follows. The test chamber is filled with pure helium. By removing the cover plate placed on the top of the tube, exchange flow, i.e., exchange flow of helium and air is initiated. At such condition, the smoke pulse generator ignites the high voltage. Immediately a smoke appeared and visualized the helium up flow and the air down flow in the flow path in the long tube. Test chamber diameter and height are 350 mm, the long tube on the test chamber diameter(=D) are 17.4 and 20mm, length of it (=L) are 200 and 319mm. They denote L/D=10 and L/D=18.3 respectively. It simulates a typical long tube. The inclination angle θis 30 deg. The smoke wire conditions are as follows. The voltage is around 250 (V), current duration is 30 m sec, and the oil of thin wire is CRC-556. The high-speed camera system using D-file records the visual data up to 1600 flames in a second. Upward

> High speed camera VCC-H1600C

Air

**Control computer**

D-file

L:length of flow path D: diameter of flow path

1. Smoke wire method

**2. Smoke wire method** 

3. The method of the mass increment

**2.1 Experimental apparatus and procedure** 

flows peak velocity measured by PTV method.

Smoke pulse generator

> Helium gas

θ

D

Fig. 1. Exchange Flow Apparatus of Smoke Wire Method High speed camera system

L

Flow path

Test chamber

**Smoke wire**

2. The optical system of the Mach-Zehnder interferometer

The typical exchange flow in the tube was visualized in the figure 2. The visualized exchange flow resembles to the S-shape. The flames are detected 200 and 500 flames in a second by the high-speed camera. The visualized data listed along the elapsed time in the figures 3 and figure 4, respectively. It is clearly visualized that helium up-flow along left hand side and air down-flow along right hand side. Figure 5 shows time history of the upward flow peak velocity measured by PTV. In the case of L/D=18.3, the average velocity value *U0* is evaluated as 0.315 m/s. In the case of L/d=10, the average velocity value *U0* is

Herium-Air Exchange Flow Rate Measurement Through a Narrow Flow Path 51

Fig. 6. The visualization data of PIV measurement of 0.050sec after ignition (L/D=18.3 and

In PTV method, the exchange flow rate Q is calculated as 1.47x10-5 m3/s under the conditon of . L/D=18.3. Therefore, the densimetric Froude number Fr is evaluated as 0.202, in this condition Reynords number Re is 79.2. In the condition of L/D=10, the densimetric Froude

Air

**Exchange flow path**

Computer

Computer

**High speed camera**

D

Cross section of Exchange flow path

Inclined angle

**Mach-Zehnder Interferometer**

L

**Helium Gas Mixture**

Fig. 7. Experimental apparatus of optical system and mass increment

**Test chamber**

**Electronic balance**

flame rate=200 f/s, θ=30 deg )

number Fr is evaluated as 0.287.

evaluated as 0.662 m/s. It means that high exchange velocity detects in low L/D ratio. Figue.6 shows the visualization data of PIV measurement of 0.050sec after ignition (L/d=18.3 and flame rate=200 f/s ). The adopted mesh size is 40x10 points in the flow area. The upward peak velocity was measured to 0.517 m/s. It means the value is higher than average velocity in this case. The exchange flow rate *Q* is derived as follows assuming parabolic flow profile, where *r* is the radius of flow path of the horizontal direction.

$$Q = \frac{2\pi l L\_0 r^2}{9} \tag{1}$$

(a) Elapsed time 0.062 sec (b) Elapsed time 0.082 sec

Fig. 4. The visualized data listed along the elapsed time (L/D=10, flame rate=500 f/s, θ=30 deg)

Fig. 5. Upward flow peak velocity measured by PTV method (L/D=18.3)

evaluated as 0.662 m/s. It means that high exchange velocity detects in low L/D ratio. Figue.6 shows the visualization data of PIV measurement of 0.050sec after ignition (L/d=18.3 and flame rate=200 f/s ). The adopted mesh size is 40x10 points in the flow area. The upward peak velocity was measured to 0.517 m/s. It means the value is higher than average velocity in this case. The exchange flow rate *Q* is derived as follows assuming

> 2 <sup>0</sup> 2 9 *U r <sup>Q</sup>*

(1)

parabolic flow profile, where *r* is the radius of flow path of the horizontal direction.

(a) Elapsed time 0.062 sec (b) Elapsed time 0.082 sec Fig. 4. The visualized data listed along the elapsed time (L/D=10, flame rate=500 f/s, θ=30

> 0 0.1 0.2 0.3 Elapsed time (sec)

θ=30゜

Fig. 5. Upward flow peak velocity measured by PTV method (L/D=18.3)

**Upward flow peak velocity**

**(mm/s)**

deg)

Fig. 6. The visualization data of PIV measurement of 0.050sec after ignition (L/D=18.3 and flame rate=200 f/s, θ=30 deg )

In PTV method, the exchange flow rate Q is calculated as 1.47x10-5 m3/s under the conditon of . L/D=18.3. Therefore, the densimetric Froude number Fr is evaluated as 0.202, in this condition Reynords number Re is 79.2. In the condition of L/D=10, the densimetric Froude number Fr is evaluated as 0.287.

Fig. 7. Experimental apparatus of optical system and mass increment

Herium-Air Exchange Flow Rate Measurement Through a Narrow Flow Path 53

If the density of the test section is homogeneous, the interference fringes are parallel and equidistant (Keulegan, 1958). If it is not homogeneous, the interference fringes are curved. An inhomogeneity in the test section produces a certain disturbance of the non-flow fringe pattern. The digital camera and high-speed camera using D-file is able to attach to the

Figure 9 shows the typical interference fringes for the inclined long round tube (L/D=5). The curved interference fringes indicate that the lighter helium flows in the upper passage of the tube. The straight fringes indicate that the heavier air flows in the bottom of the tube. It is clearly visualized that the exchange flows take place smoothly and stable in the separated passages of the tube. This leads to less resistance for the exchange flow in the inclined tubes compared to he vertical ones. In the case of 30 deg, the curvature of the interference fringes is larger than that at other angles, indicating that the exchange flow rate

Fig. 10. The relationship between Fr and inclination angle θ with L/D as a parameter

The method of the mass increment was used for the investigations. Figure 7 shows a rough sketch of the apparatus. It consists of a test chamber, an electronic balance and a personal computer for data acquisition. The experimental procedure is mentioned in Sec. 2.1. Air enters the test chamber and the mass of the gas mixture in the test chamber increases.

The mass increment m is automatically measured by the electronic balance with high accuracy. From mass increment data, the density increment of the gas mixture ρL = m/V is calculated. The density increment means the difference of densities of the gas mixture from the density of pure helium in the test chamber. Then, volumetric exchange

and the densimetric Froude number are the largest at 30 deg.

**4. Method of mass increment** 

**4.1 Experimental apparatus and procedure** 

flow rate is evaluated by the following equation:

L/D

interferometer.

**3.2 Results** 

Fig. 8. The mechanism of optical system of the Mach-Zehnder interferometer, MZC-60S

Fig. 9. The typical interference fringes for the inclined long round tube (L/D=5)

### **3. Optical system of mach-zehnderinterferometer**

#### **3.1 Experimental apparatus and procedure**

The optical system of the Mach-Zehnder interferometer, MZC-60S is shown in figure 7 and figure 8 to visualize the exchange flow. After being rejoined behind the splitter, the test and reference laser beams interfere, and the pattern of interference fringes appears on the screen. If the density of the test section is homogeneous, the interference fringes are parallel and equidistant (Keulegan, 1958). If it is not homogeneous, the interference fringes are curved. An inhomogeneity in the test section produces a certain disturbance of the non-flow fringe pattern. The digital camera and high-speed camera using D-file is able to attach to the interferometer.

## **3.2 Results**

52 Nuclear Reactors

Fig. 8. The mechanism of optical system of the Mach-Zehnder interferometer, MZC-60S

**(a) (b) (c)**

**(d) (e) (f)** Fig. 9. The typical interference fringes for the inclined long round tube (L/D=5)

The optical system of the Mach-Zehnder interferometer, MZC-60S is shown in figure 7 and figure 8 to visualize the exchange flow. After being rejoined behind the splitter, the test and reference laser beams interfere, and the pattern of interference fringes appears on the screen.

**3. Optical system of mach-zehnderinterferometer** 

**3.1 Experimental apparatus and procedure** 

Figure 9 shows the typical interference fringes for the inclined long round tube (L/D=5). The curved interference fringes indicate that the lighter helium flows in the upper passage of the tube. The straight fringes indicate that the heavier air flows in the bottom of the tube. It is clearly visualized that the exchange flows take place smoothly and stable in the separated passages of the tube. This leads to less resistance for the exchange flow in the inclined tubes compared to he vertical ones. In the case of 30 deg, the curvature of the interference fringes is larger than that at other angles, indicating that the exchange flow rate and the densimetric Froude number are the largest at 30 deg.

Fig. 10. The relationship between Fr and inclination angle θ with L/D as a parameter
