Full article: Development of microbubble generator for suppression of pressure waves in mercury target of spallation source

Abstract
Formulae display:?Mathematical formulae have been encoded as MathML and are displayed in this HTML version using MathJax in order to improve their display. Uncheck the box to turn MathJax off. This feature requires Javascript. Click on a formula to zoom.

A MW-class mercury target for the spallation neutron source is subjected to the pressure waves and cavitation erosion induced by high-intense pulsed-proton beam bombardment. Helium-gas microbubbles injection into mercury is one of the effective techniques to suppress the pressure waves. The microbubble injection technique was developed. The selection test of bubble generators indicated that the bubble generator utilizing swirl flow of liquid (swirl-type bubble-generator) will be suitable from the viewpoint of the produced bubble size. However, when single swirl-type bubble-generator was used in flowing mercury, swirl flow of mercury remains at downstream of the generator. The remaining swirl flow causes the coalescence of bubbles which results in ineffective suppression of pressure waves. To solve this concern, a multi-swirl type bubble-generator, which consists of several single swirl-type bubble-generators arraying in the plane perpendicular to mercury flow direction, was invented. The multi-swirl type bubble-generator was tested in mercury and the geometry was optimized to generate small bubble with low flow resistance based on the test results. It is estimated to generate the microbubbles of 65 μm in radius under the operational condition of the Japanese Spallation Neutron Source mercury target, which is the sufficient size to suppress the pressure waves.

Keywords:

mercury target spallation neutron source pressure wave cavitation erosion microbubbles (multi) swirl-type bubble generator

1. Introduction

A mercury target is installed at the Japanese Spallation Neutron Source (JSNS) of the Materials and Life Science Experimental Facility (MLF) in the Japan Proton Accelerator Research Complex (J-PARC). It is bombarded by proton beams with energy of 3 GeV, repetition rate of 25 Hz, pulse duration of about 1 μs and power of 1 MW (expected at the final stage) to produce high intense neutrons for promoting innovative sciences [1].

Figure 1(a)–1(c) shows the schematics of JSNS, JSNS mercury target system and mercury target vessel, respectively. A mercury target vessel is a container of mercury producing the neutron by the spallation reaction. The proton beams bombard mercury passing through the front of the target vessel which is called the beam window and the mercury is rapidly heated by the proton beam bombardment. Flow guides set in the target vessel make flow distribution of mercury as matching the generated heat distribution in the mercury target. In Figure 1(b), conceptual drawing of the gas supply system is also shown. The gas supply system is set on a mercury circulation loop in the mercury target system which has been already operated in the MLF for the heat removal in the mercury target, and a bubble generator is set into a mercury target vessel.

Figure 1. Schematic views of (a) Japanese Spallation Neutron Source (JSNS), (b) mercury target system and (c) mercury target vessel.

Display full size

In the mercury target vessel, pressure waves are generated by the proton bombardment and propagate in mercury. They impose stresses on a target vessel. The pressure waves shorten a lifetime of the target vessel not only by fatigue due to a cyclic stress but also by the cavitation erosion, so-called pitting damage resulting from the pressure wave propagations [2, 3]. Pitting damage occurs due to a microjet and shock wave which is generated when the cavitation bubble collapses. Since the cavitation bubble is generated by negative pressure in mercury which is caused by expansion of the target vessel due to the pressure wave propagation, it is necessary to suppress the pressure waves to mitigate the pitting damage. Then suppression techniques on the pressure waves should be developed to realize the high power operation of JSNS.

Microbubbles of helium gas which is noncondensable is effective to suppress the pressure waves in mercury, because the microbubbles absorb thermal expansion of mercury due to proton beam bombardment, and kinetic energy of the pressure wave is changed to thermal energy by oscillation of the gas microbubbles. Since pressure rising is very fast, e.g. the maximum pressure reaches to 40 MPa at 1 μs after the proton beam bombardment under the 1 MW beam condition, large bubbles of which resonant frequency is low cannot absorb the thermal expansion of mercury due to the proton beam bombardment [4]. It is reported that the injection of microbubbles less than 100 μm in radius with sufficient gas volume fraction into the mercury, 0.1%, will suppress the pressure waves [4] and the pitting damages on the target vessels [5]. Then, the bubble generator should generate the bubbles less than 100 μm in radius.

There are some measures to make microbubbles into liquid such as a pressurized dissolution method [6] and the method utilizing the force induced by liquid flow such as shear force and pressure change. In the pressurized dissolution method, gas which is confined in a chamber with liquid is dissolved into liquid by pressurizing inside of the chamber; after that, the pressure in the chamber is rapidly decreased and then the microbubbles are produced. For the pressurized dissolution method in JSNS, new components are required, such as the chamber, a mercury feed line from the mercury circulation loop to the chamber and return line of bubble mixed mercury, and they have to connect to the mercury circulation loop although the loop has been already operated and activated. It is also reported that the solubility of the noble gases in mercury are very low [7]. The pressurized dissolution method was dismissed to apply for the JSNS.

On the other hand, the method utilizing the force induced by the liquid flow is suitable in the JSNS mercury target because mercury flows in the target vessel and installation of the bubble generator is possible when the mercury target vessel is exchanged. However, installing the bubble generator increases the flow resistance in the mercury target vessel. Mercury flow rate should keep 41 m³/h to remove the heat in mercury target vessel; however, the maximum discharged pressure of the pump in JSNS mercury circulation loop is 0.3 MPa at 41 m³/h [8]. The flow resistance of the mercury target system without the bubble generator was 0.1 MPa when the mercury flow rate was 41 m³/h. Therefore, the flow resistance of the bubble generator has to be less than 0.2 MPa at 41 m³/h in mercury flow rate for cooling the mercury target vessel.

To suppress the pressure waves in the mercury target, the bubble generator should generate the microbubbles less than 100 μm in radius and the flow resistance in it should be less than 0.2 MPa at 41 m³/h in mercury flow rate. The development of the bubble generator was carried out to achieve the design goal as mentioned above. Furthermore, since spatial distribution of the bubble in the target is important, measurement of bubble distribution was carried out by using a model simulating the target vessel installing the developed bubble generator.

2. Development of multi-swirl type bubble-generator

2.1. Selection of candidate bubble generator

Three kinds of bubble-generators were screened from the viewpoints of the generated bubble size and flow resistance of the bubble generator. Figure 2 shows the schematic views of three kinds of bubble generators. Figure 2(a) shows the needle-type bubble-generator. It has a needle with 200 μm in outer diameter and 100 μm in inner diameter. The gas passes through the inside of the needle and sheared by flowing mercury at the outside of the needle to make small bubbles. Figure 2(b) shows the porous-metal bubble-generator combined with Venturi. The porous metal is set in the throat of the Venturi. Diameters of the Venturi at inlet/outlet and throat are 22 mm and 6 mm, respectively. Length of throat is 5 mm. The average diameter of the porous is 50 μm and thickness of the porous metal is 2 mm. The gas passing through the porous metal is sheared by mercury flow whose speed is higher than the case of needle type. Furthermore, since the bubbles generated at the throat are exposed in pressure increase between throat part and outlet of the throat, it is expected that the smaller bubbles are generated at higher pressure field of the outlet of the throat. Figure 2(c) and 2(d) shows a cut view of the swirl-type bubble-generator and its detailed structure, respectively. A static swirler fastened in an outer cylinder whose diameter is 20 mm makes swirl flow of liquid at the downstream of the swirler. The swirler consists of a centre column and guide vanes whose angle, θ_f is 68 degrees. Gas is injected from the centre of the swirler to make gas column as shown in Figure 2(c). The gas column is bended at the outlet of the bubble generator by the Coanda effect. The bended column is broken down to the microbubbles due to the shear force induced by the vortex-breakdown at the outlet of the bubble generator and the pressure distribution as shown in Figure 3 which are screenshots at the outlet of the swirl-type bubble-generator observed by a high speed camera.

Figure 2. Schematic views of three kinds of bubble generators; (a) needle-type bubble-generator, (b) porous-metal bubble-generator combined with Venturi, (c) and (d) swirl-type bubble-generator.

Display full size

Figure 3. Screenshots of microbubble generation at the outlet of the swirl-type bubble-generator.

Display full size

Figure 4 shows the bubble size distribution generated by three bubble generators in experimental mercury loop having the test section with a diameter of 22 mm. Here, the bubble size was calculated from the projection radius in contacting with the transparent acrylic window [5]. The swirl-type bubble-generator generated the smallest bubbles. Not only generating small bubbles but also low flow resistance is important for installing the bubble generator in the mercury target. The flow resistance was measured by the pressure difference between the inlet and outlet of the bubble generator. Regarding the needle-type bubble-generator, the flow resistance is almost null and flow velocity around the bubble-generator was about 0.7 m/s. However, it is reported an interfacial tension force between gas and nozzle in mercury is strong because of the large surface tension force of mercury (σ = 0.47N/m which is about 7 times larger than water) and bad wettability of mercury to the nozzle material. To generate the bubble less than 100 μm in radius in mercury, the required flow velocity around the nozzle is estimated to be about 5 m/s [9]. Since such high flow velocity induces another problem such as the flow-induced vibration, the nozzle-type bubble-generator is not suitable for the real target. In the tests with the porous-metal bubble-generator with the Venturi or the swirl-type bubble-generator, the flow resistance and flow velocity were 0.015 MPa and 0.3 m/s. The flow resistance is estimated to be 0.17 MPa if the flow velocity passing through the bubble generator increases to 1.0 m/s which is comparable with the flow velocity in the real target. The swirl-type bubble-generator generated smaller bubbles than the porous-metal bubbler in the same flow resistance condition. Since the swirl-type bubble-generator had possibility to generate small bubbles with suitable flow resistance for the real target system, we selected the swirl-type bubble-generator as a candidate bubble generator for the mercury target.

Figure 4. Bubble size distribution generated from the various bubble generator in mercury.

Display full size

2.2. Concern on swirl-type bubble-generator

The swirl-type bubble-generator has been applied for purifying the water in the pond, water tank, etc., that is, the bubble generator has been used under the condition of its outlet facing to a stagnant wide area. On the other hand, in the mercury target, the outlet area will be limited and it will be installed in the flowing condition.

Figure 5(a) shows the bubble behaviour at the outlet of the single swirl-type bubble-generator in the pipe where the water was flowing. Outer cylinder of the swirl-type bubble-generator is 70 mm in diameter. In this case, the swirl flow remains at the downstream of the bubble generator and gathers the microbubbles at the centre of the swirl to make the gas column again. To avoid the swirl at the outlet of the bubble generator, a swirl stopper was installed at 50 mm downstream of the outlet of the swirl-type bubble-generator. Figure 5(b) shows the effect of the swirl stopper. It is shown that gas column was not observed and microbubbles were distributed homogeneously at downstream of the swirl stopper in comparison with the case without the swirl stopper. Although the swirl stopper was effective to suppress the swirl at downstream of the bubble generator, it was difficult to install the swirl stopper because it had an unacceptable demerit that the swirl stopper increased the flow resistance.

Figure 5. Gas behaviours from swirl-type bubble-generator (a) without and (b) with swirl stopper.

Display full size

2.3. Multi-swirl type bubble-generator

To suppress the swirl flow at downstream of the swirl-type bubble-generator, we focused on the relationship between the geometries of the mercury target vessel and the bubble generator. Figure 6 shows a schematic drawing of the cross section of the target vessel at the position where the bubble-generator is going to be installed. The cross section of the mercury target at the bubble-generator position is almost rectangular. On the other hand, the outer view of the swirl-type bubble-generator is cylindrical as shown in Figure 2(c) and 2(d). When the large single bubble generator is set as shown in Figure 6(a), there is an area where mercury cannot flow around the bubble generator. When a diameter of the bubble generator is made small, several small single bubble-generators can be arrayed as shown in Figure 6(b). When several bubble generators are arrayed so that the swirl directions of the bubble generators are alternative, which is named the multi-swirl type bubble-generator, it is expected that the swirl flow at the downstream of the bubble generator is diminished by the interferences due to each swirl flow from each single generator. Arraying the small bubble-generators can increase the cross sectional area for mercury flow, which decreases the flow resistance in the bubble generator. For example, the diameter of the bubble generator is 76 mm in the maximum when the single swirl-type bubble-generator is installed. On the other hand, seven single generators of 30 mm in diameter can be installed as multi-swirl type bubble-generator, in which the cross sectional area for mercury flow is 10% wider than the single generator of 76 mm. The characteristics of the multi-swirl type bubble-generator were investigated.

Figure 6. Cross sectional view of the mercury target at bubble generator position under conditions of (a) single swirl-type bubble-generator and (b) multi-swirl type bubble-generator installation.

Display full size

3. Measurement of characteristics of multi-swirl type bubble-generator in JSNS target

3.1. Target model

To investigate the bubble behaviour which is generated by the swirl-type bubble-generator, a target model was fabricated. Figure 7 shows the target model to investigate the bubble distribution in the target and the flow resistance in the bubble generator. The model sizes were 961 mm in length, 545 mm in width and 80 mm in height. They were 10% smaller in horizontal and the same in the height of actual target vessel. The top of the model was made of acrylic resin to be transparent, and so the bubble population and motion could be observed. Reinforcement ribs made by the stainless steel were set on the transparent wall to secure an operational pressure up to 1 MPa. Other parts of the vessel were made of the stainless steel. The bubble generator was set at the inlet of the model as shown in Figure 7. We investigate the effect of the bubble generator type on bubble production, flow resistance and flow pattern at the downstream of the bubble generator by changing the bubble generators (single swirl-type bubble-generator with/without swirl stopper and multi-swirl type bubble-generator). Since bubble distribution could not be observed in mercury except for the top surface of the target model, the tests were carried out in both water and mercury to estimate the bubble distribution inside mercury from the result of test in water. The water loop in JAEA and mercury loop, TTF (the Target Test Facility) in ORNL (the Oak Ridge National Laboratory: USA) were utilized for water and mercury tests, respectively. The flow rate varied up to 27 m³/h (7.5 L/s) in both tests. The bubble size distribution was measured by using a captured image taken by a digital still camera. In the case of the water test, bubble size was measured at different heights by varying focus of the camera. In the mercury, on the other hand, it was measured only the projection size of bubble attaching on the acrylic top wall. The projection size was converted to free bubble size by estimating the volume of the bubble in mercury taking account of the contact angle of mercury on the acrylic wall. The pressure sensors were mounted on the inlet and outlet of the bubble-generator to measure the flow resistance of the bubble-generator.

Figure 7. Test model to investigate the bubble distribution in the target and flow resistance in the bubble generator.

Display full size

3.2. Results of flow resistance

The flow resistance, ΔP, is expressed as follows: (1) $Δ P = \frac{1}{2} ρ (C_{D} + f) V_{i n}^{2}$ (1) where, ρ is the liquid density; C_D, the flow resistance coefficient depending on the geometry of the bubble-generator independent of the sort of liquid; f, the coefficient by the effect of boundary between liquid and the bubble-generator material such as the friction and the wettability; V_in, the flow velocity at the bubble generator inlet obtained by the following relation, (2) $V_{i n} = \frac{Q}{n \times π D^{2} / 4}$ (2) where, n is the number of the single bubble-generator arrayed in the multi-swirl type bubble-generator; Q, the flow rate of liquid; D, the inner diameter of the bubble generator inlet as shown in Figure 2(d).

The dependency of the liquid density on flow resistance can be cancelled due to dividing the flow resistance by liquid density. Figure 8 shows the flow resistance divided by the liquid density against the flow velocity at the bubble generator inlet. The results of the flow resistance in water and mercury are coloured in black and grey, respectively in Figure 8. The flow resistances in water and mercury divided by the liquid density are same independent of sorts of liquid. It means, the flow resistance by the effect of boundary in the swirl-type bubble-generator is negligibly low compared with that by the geometry of the bubble generator. The flow resistance in mercury can be estimated by using the results obtained in water test.

Figure 8. Flow resistance in bubble generator divided by the liquid density against the flow velocity.

Display full size

The flow resistance of the multi-swirl type bubble-generator without the swirl stopper reduced to 1/3 of that of the single bubble generator with the swirl stopper at the same flow velocity while the generated bubble sizes are almost same between them as described in Section 3.3.

3.3. Results of produced bubble size

Figure 9 compares bubble size distributions at the outlet of the bubble generator in water among single bubble generators with and without the swirl stopper and multi-bubble-generator without swirl stopper. The bubble sizes from the multi-bubble generator without the swirl stopper are as same as those from the single bubble-generator with the swirl stopper. The peak count appears at about 100 μm in radius. In the single swirl-type bubble-generator without the swirl stopper, although the peak appeared at about 100 μm, the number of counts is smaller compared to other two generators because coalescence of bubbles occurred and bubbles over 500 μm in radius were generated.

Figure 9. Comparison of bubble size distribution in water at the outlet of the swirl-type bubble-generators with and without swirl stopper (Flow rate of water: 5L/s, Injected gas rate: 30 cm³/min).

Display full size

The bubble distribution near the beam window is important to suppress the pressure wave. Figure 10 shows the bubble distribution in water at the position A shown in Figure 7 in the case that the flow rate is 5 L/s. In Figure 10, the change of the size distribution depending on the height in the model is also shown. The peak count of the bubble size appeared at around 100 μm in radius independent of the height from the target bottom surface, z. The number of the bubbles increases with the height since the generated bubble rises up with water flow by buoyancy.

Figure 10. Change of the bubble size distribution depending on height in water at the position A in the model shown in Figure 7 (Flow rate of water : 5L/s, Injected gas rate: 300 cm³/min).

Display full size

Figure 11 shows the bubble size distribution in mercury at the position A in the case that the flow rate is 7.5 L/s. Only bubbles attached on the top wall newly are counted. The peak count appeared at 40 μm in bubble radius. The generated bubble radius in mercury is smaller than that in water as shown in Figure 10 although liquid flow rate is higher in the mercury test. The difference of the bubble radius between mercury and water is discussed in Section 4.2.

Figure 11. Bubble size distribution generated by multi-bubble generator without swirl stopper in mercury at position A (Flow rate of mercury: 7.5 L/s, Injected gas flow rate: 450 cm³/min).

Display full size

3.4. Gas accumulation in the target

Figure 12(a) and 12(b) shows the top view of the model in the case that the mercury flow rates were 5.5 and 7.0 L/s, respectively. It is observed that huge amount of gas accumulated at the downstream of the flow guide vane in the case of 5.5 L/s, which was not observed in the water test. However, the gas accumulation phenomenon was never observed over the flow rate of 6.5 L/s as shown in Figure 12(b). Since the operational flow rate is 11.4 L/s in the running mercury target, this gas accumulation will never occur in the real system.

Figure 12. Gas accumulation behaviour in the mercury target model in the cases of mercury flow rate of (a) 5.5 L/s and (b) 7.0 L/s.

Display full size

4. Discussions

4.1. Flow resistance

As shown in Figure 8, flow resistance in mercury can be estimated by water test. To apply the bubble-generator to the real target, the flow resistance of the bubble generator is required to be less than 0.2 MPa. The flow resistance depends on the coefficient, C_D, and flow velocity of mercury as shown in Equation (1). Estimation of the coefficient of flow resistance, C_D, is important to design the bubble generator geometry, such as the vane angle of the swirler, diameter of the single bubble-generator, the number of the single bubble-generator, etc. for the real target. The experiment was carried out in water by varying the vane angle and the diameter of the bubble generator. The coefficient, C_D, is derived from the experimental results and Equation (1). Figure 13 shows the coefficient of the flow resistance, C_D, as the function of the vane angle of the swirler and the diameter of the single bubble-generator. The coefficient, C_D, is expressed as the following equation: (3) $C_{D} = \frac{C_{1} \times {tan}^{C_{2}} θ_{f}}{D^{C_{3}}}$ (3) where, θ_f is the vane angle as shown in Figure 6(a); D, the inner diameter of the bubble generator; C₁, C₂ and C₃, the constant. Based on the Equations (1), (2) and (3), the geometry of the bubble generator was designed for the target system to make flow resistance low. Actually, the number of the bubble generator, n, was restricted to 5 because a part of the cross sectional area where the bubble generator is installed should be used for the gas supply line connection, and manufacturing process should be considered. The diameter of the single generator was decided to be as large as possible in the cross sectional area. Then, the vane angle was decided so that the flow resistance could be less than 0.2 MPa. After that, the generated bubble size was estimated by Equation (4) which will be described in Section 4.2. Finally, the geometry of the swirl-type bubble-generator was decided as θ_f = 63°, D = 31 mm and n = 5. With this multi-swirl type bubble- generator, the flow resistance will be 0.194 MP in the real target.

Figure 13. Coefficient of flow resistance in swirl-type bubble-generator against the vane angle of the swirler.

Display full size

4.2. Bubble size and distribution

Under the turbulent condition between liquid and gas as well as the condition at the swirl-type bubble-generator, the generated bubble radius, R_b, is represented as the following equation [10]: (4) $R_{b} = C_{4} {(\frac{σ^{3}}{ρ^{3} ϵ^{2}})}^{1 / 5}$ (4) where, σ is the surface tension of liquid; ρ, the density of the liquid; ϵ, the energy dissipation rate which depends on the flow velocity and shape of the bubble generator; C₄, the constant. This equation reproduces the generated bubble radius at the peak frequency as shown in Figures 10 and 11. In the tests by changing the flow rates of water and mercury, the generated bubble radii were similar to the results calculated by Equation (4) within 50% error. Based on Equation (4), the peak bubble radius is estimated to be 65 μm in the real target by installing fine bubble generator with the vane angle and the inner diameter mentioned in Section 4.1.

The bubble size in mercury as shown in Figure 11 became smaller than that in water as shown in Figure 10. This is because σ/ρ of mercury is 1.9 times smaller than those of water since the surface tension, σ, and the density, ρ, of mercury are 7 times and 13 times larger than those of water, respectively. The bubble size generated in mercury is smaller than that in water when ϵ in mercury and water are same value, which means the same geometry of the bubble generator and same flow velocity.

By installing the multi-swirl type bubble-generator into the target, the small bubble size and low flow resistance will be achieved. However, the volume fraction near the beam window is also important to suppress the pressure waves in the mercury target. Figure 14 shows the volume fraction distribution in water as a function of the height, which was obtained by the bubble volume in the image divided by the image volume (the image area multiplied by a focus depth of the camera; 10 mm). In this test, the injected gas volume rate was 0.1% to the water flow rate of 5 L/s. The volume fraction increases with the measured height because of the effect of the buoyancy force. The main bubble radius was 100 μm under the water flow condition of 5 L/s which corresponds to 0.5 m/s in flow velocity in the target model. On the other hand, the bubble radius is estimated to be 65 μm in the real target in which the flow velocity of mercury is ca. 1.0 m/s in the target. The rising velocity of bubble of 100 μm in radius in water and 65 μm in radius in mercury are 0.020 and 0.022 m/s, respectively [11]. The volume fraction at the half height of the target is 0.025% near the beam window in the 0.5 m/s flowing water as shown in Figure 14. It means that the volume fraction at the half height of the target near the beam window is reduced to about 1/4 of the injected gas fraction which was 0.1% by the buoyancy force. Based on the rising velocity of bubble, it is estimated that the bubble flows to 1.0 and 1.8 m in horizontal direction for the bubble rising to 0.04 m (half height of the target) in 0.5 m/s flowing water and 1.0 m/s flowing mercury, respectively. This indicates that more bubbles will reach near the window of the real target than the case of the water test and the volume fraction near the beam window will be more than 0.025% if the injected gas fraction is 0.1% and the bubble keeps its radius within 65 μm. From this discussion, the volume fraction over 0.1% will be realized by injecting the gas over 0.4% in flow rate to the mercury flow rate.

Figure 14. Distribuition of volume fraction depending on height in water (Flow rate of water: 5 L/s, Injected gas rate: 300 cm³/min).

Display full size

5. Conclusions

The swirl-type bubble-generator was developed to suppress the pressure waves in the JSNS mercury target. The developed bubble genrator can generate bubble less than 100 μm in radius with flow resistance of less than 0.2 MPa at 41 m³/h in mercury flow rate. Through the developement, the following knowledges were obtained:

-	The swirl-type bubble-generator is promising for the JSNS mercury target system from the viewpoint of both the produced bubble size and the flow resistance comparing with the needle type and porous material with Venturi bubble generators.
-	The developed multi-swirl type bubble-generator is effective on reduction of both of flow resistance and suppression of the swirl flow at downstream of the bubble generator.
-	The coefficient of the flow resistance of the swirl-type bubble-generator could be expressed as the function of the vane angle and the diameter of the swirl-type bubble generator.
-	Based on this result, the multi-swirl type bubble-generator was designed; the number of the bubble generator n = 5, the vane angle θ_f = 63° and the diameter of the single generator D = 31 mm, in which flow resistance will be less than 0.2 MPa at 41 m³/h in mercury flow rate and the generated bubble size will be 65 μm in radius in the real target.
-	Since the generated bubbles are rising in the mercury target by buoyancy force, the volume fraction at the half height of the target vessel near the beam window will be reduced to about 1/4 of the injected gas fraction. The bubble fraction near the beam window will be improved by increasing the injected gas fraction.
-	Gas accumulation occurred in the mercury target model at low mercury flow rate while it disappeared by increasing the mercury flow rate, more than 6.5 L/s which is less than the operational flow rate for the real target system.

Acknowledgements

We deeply appreciate Mr. Bernei Riemer, Mr. Mark Wendel and Mr. David Felde of the Oak Ridge National Laboratory for their kind help to carry out the bubble generator tests in mercury at the Target Test Facility (TTF). This work was partly supported by the Japan Society for the Promotion of Science through a Grant-in-Aid for Scientific Research [grant numbers 20360090 and 23360088].

References

http://J-parc.jp/MatLife/en/index. [Google Scholar]
Futakawa M, Naoe T, Kogawa H, Tsai CC, Ikeda Y. Pitting damage formation up to over 10 million cycles off-line test by MIMTM. J Nucl Sci Technol. 2003;40:895–904. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]
Futakawa M, Naoe T, Tsai CC, Kogawa H, Ishikura S, Ikeda Y, Soyama H, Date H. Pitting damage by pressure waves in a mercury target. J Nucl Mat. 2005;343:70–80. [Crossref], [Web of Science ®], [Google Scholar]
Okita K, Takagi S, Matsumoto Y. Propagation of pressure waves caused by a thermal shock in liquid metals containing gas bubbles. J Fluid Sci Technol. 2008;3:116–128. [Crossref], [Google Scholar]
Naoe T, Ida M, Futakawa M. Cavitation damage reduction by microbubble injection. Nucl Instrum Methods Phys Res Sec. A. 2008;586:382–386. [Crossref], [Web of Science ®], [Google Scholar]
Hosokawa S, Tanaka K, Tomiyama A, Maeda Y, Yamaguchi S, Ito Y. Measurement of micro bubbles generated by a pressurized dissolution method. J. Phys.: Conf. Series. 2009;147:012016. [Crossref], [Google Scholar]
Hasegawa S, Naoe T, Futakawa M. Solubility of helium in mercury for bubbling technology of the spallation neutron mercury target. J Nucl Mat. 2010;398:189–192. [Crossref], [Web of Science ®], [Google Scholar]
Kogawa H, Haga K, Wakui T, Futakawa M. Development on mercury pump for JSNS. Nucl Inst Meth Phys Res A. 2009;600:97–99. [Crossref], [Web of Science ®], [Google Scholar]
Kogawa H, Shobu T, Futakawa M, Bucheeri A, Haga K, Naoe T. Effect of wettability on bubble formation at gas nozzle under stagnant condition. J Nucl Mat. 2008;377:189–194. [Crossref], [Web of Science ®], [Google Scholar]
Martinez-Bazán C, Montañés LJ, Lasheras CJ. On the break-up of an air bubble injected into a fully developed turbulent flow, Part I: break-up frequency. J Fluid Mech. 1999;401:157–182. [Crossref], [Web of Science ®], [Google Scholar]
Peebles NF, Garber JH. Studies on the motion of gas bubbles in liquids. Chem Eng Progress. 1953;49:88. [Web of Science ®], [Google Scholar]

Article

Development of microbubble generator for suppression of pressure waves in mercury target of spallation source

Article

Development of microbubble generator for suppression of pressure waves in mercury target of spallation source

Abstract
Formulae display:?Mathematical formulae have been encoded as MathML and are displayed in this HTML version using MathJax in order to improve their display. Uncheck the box to turn MathJax off. This feature requires Javascript. Click on a formula to zoom.

1. Introduction

Published online:

2. Development of multi-swirl type bubble-generator

2.1. Selection of candidate bubble generator

Published online:

Published online:

Published online: