A study of bubble behaviors in the volumetrically heated packed bed

A study of bubble behaviors in the volumetrically heated packed bed

Progress in Nuclear Energy 76 (2014) 100e105 Contents lists available at ScienceDirect Progress in Nuclear Energy journal homepage: www.elsevier.com...

1MB Sizes 0 Downloads 7 Views

Progress in Nuclear Energy 76 (2014) 100e105

Contents lists available at ScienceDirect

Progress in Nuclear Energy journal homepage: www.elsevier.com/locate/pnucene

A study of bubble behaviors in the volumetrically heated packed bed Guangzhan Xu a, Zhongning Sun a, *, Xianke Meng b, Xiaoning Zhang a a b

Fundamental Science on Nuclear Safety and Simulation Technology Laboratory, Harbin Engineering University, Harbin 150001, China State Nuclear Power Research Institute, Beijing 100029, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 December 2013 Received in revised form 20 March 2014 Accepted 8 May 2014 Available online

The volumetrically heated packed bed has been widely utilized in modern industry, however, no research on the bubble behaviors in forced convection subcooled boiling was studied. To study the bubble behaviors in the volumetrically heated packed bed, here electromagnetic induction heating method was used to heat oxidized carbon steel balls adopted to stack packed bed, while water was utilized as the refrigerant in the experiment. Bubble behaviors were observed by a high speed camera for particle diameter varying from 8 mm to 12 mm, mass flux varying from 29.3 kg m2 s1 to 84.2 kg m2 s1, heat flux varying from 14.5 kW m2 to 50 kW m2, inlet pressure varying from 0.116 MPa to 0.125 MPa, inlet subcooling varying from 7 k to 9.2 k and porosity ¼ 0.39. Obtained flow visualization images were analyzed. The experimental results indicated that the bubbles were blocked by steel balls and easily attached to the surface of balls, then slipped along the surface of steel balls. There was “regrowth phenomenon” in the packed bed and generated bubbles repeated growth several times in the lifetime. The nucleate boiling was firstly observed in the contact surface. Structures of contact surface had great impacts on the bubble shapes, departure diameter and frequency. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Bubble behaviors Forced convection subcooled boiling Packed bed

1. Introduction Water-cooled packed bed reactor utilizes the light water reactor technology but uses the new type spherical coated particles as fuel elements. It is a new small type reactor with high economy and inherent safety, and has a great potential in special industries. For example, AFPR-100 (Senor et al., 2007) consists of spherical fuel elements and cooled by single-phase water flow within the particle bed. The concept reactor requires more than 20-year core life. Sümer and Farhang (2008) descript the fixed bed nuclear reactor, fuel element, and criticality calculations. The preliminary neutronics calculations show that the core lifetime can be as long as 17 years. At present, the researches on water-cooled packed bed reactor are in the concept design in the world. So many scientific issues should be studied. Bubble behaviors in forced convective subcooled boiling flow are of considerable interest to nuclear reactors. Subcooled flow boiling heat transfer characteristics are closely related to the bubble behaviors in forced convection subcooled boiling. A visual study of bubble behaviors is helpful to study the boiling heat transfer mechanisms in the packed bed.

* Corresponding author. Tel./fax: þ86 451 82569655. E-mail addresses: [email protected] (G. Xu), [email protected] edu.cn (Z. Sun). http://dx.doi.org/10.1016/j.pnucene.2014.05.005 0149-1970/© 2014 Elsevier Ltd. All rights reserved.

Most researches were concentrated in heat transfer enhancement. The packed bed was heated on one side rather than volumetrically heated. For example, Izadpanah et al. (1998) and Jamialahmadi et al. (2005) found that in the channels filled with metallic or non-metallic particles whose working fluid was water, the heat transfer coefficient can be increased by 5e10 times compared with empty channels. Dryout experiments are based on the issues when the nuclear reactor severely damaged without water. It was investigated by several researchers. For example, Sch€ afer et al. (2006) studied the pressure distribution inside a boiling particle bed. The crucible had an inner diameter of 125 mm and was filled with oxidized stainless steel balls of 6 or 3 mm diameter. The particles were heated via a two-winding induction coil. The performed boiling and dryout experiment showed clearly that present models without the explicit consideration of the interfacial drag cannot predicate pressure distribution inside a boiling particle bed, not even qualitatively. Atkhen and Berthoud (2006) studied the coolability of a debris bed in multidimensional configurations. Particles were simulated by steel beads with diameters ranging from 2 to 7 mm and were heated by an induction coil. The results showed that the steady temperature overheat up to 200  C with the increase of power, while the bed was still cooling. At present, the visual researches of packed bed channels were more concentrated in flow patterns. For example, Ford (1960) was

G. Xu et al. / Progress in Nuclear Energy 76 (2014) 100e105

101

Fig. 1. The schematic diagram of experimental setup.

the first to research the flow patterns in packed bed. He observed gaseliquid flow in the packed bed filled with small particles. Based on the two-phase flow distribution in the packed bed, they divided the flow patterns into the single-phase pore flow and the twophase pore flow. When the mass flux of gas is low, the interspaces were occupied by liquid phase, so the flow pattern was called the single-phase pore flow. When the mass flux of gas was high, the interspaces were occupied by gaseliquid two phases, so the flow pattern was called the two-phase pore flow. Turpin and Huntington (1967), Mazzarino et al. (1987) Based on macroscopic features of the packed bed channel, and divided the flow patterns into bubble flow, pulse flow, spray flow. When the mass flux of gas was low, the bubbles were nearly spherical and dispersed in the continuous liquid phase. The bubble could freely move in the packed bed. With the increase of mass flux of gas, bubbles were elongated and deformed. This flow pattern was called the bubble flow. When mass flux of liquid was medium and mass flux of gas was high, the interaction of two phases was enhanced. Fluids containing more liquid and more gas go through the packed bed in turn. The fluid density was changed in turn and the pulse flow was formed. With the increase of mass flux of gas, the pulse frequency was increased and the density difference between the fluid containing more liquid and the fluid containing more gas was no longer apparent. Further increase the gas mass flux, the spray flow appeared. The spray flow was a continuous gas flow, liquid was suspended or dispersed in the gas. Zhang et al. (2009) was the first to investigate the boiling flow pattern in the packed bed. The experimental section was a rectangular channel filled with 4, 6, 8 mm diameter glass balls. In order to observe, the one side of the channel was plexiglass. In order to provide the heat flux, another side of the channel was a metal plate with many hemispherical projections. The metal plate generated heat by providing the current. They observed the bubble flow, bubble-slug flow, slug flow, slug-annular flow. Wang et al. (2002) was the first to research the bubble behaviors in the pool boiling in the packed bed. The test section was a glass container filled with glass balls and was heated at the bottom. The experimental results showed that the dynamic bubble behavior was significantly affected by the packed bed structure. It was difficult to appear dryout phenomena due to the behavior of the replenished liquid caused by special pore geometry.

No data has been found in the open literature on bubble behaviors in the subcooled flow boiling in the volumetrically heated packed bed, so an experimental setup was built to study it. 2. Experimental setup The experimental setup is shown in Fig. 1. The flow direction in the test section was vertical upward. The water was stored in the tank with 18 heaters. The dissolved gases were removed by heating water to reach saturated. Saturated water flowed through the heat exchanger, flowmeter, pre-heater and the test section by the driving of the pump, and then absorbed the heat generated by the packed bed to change to vapor and liquid mixtures. The mixtures entered the steam separator and was divided into water and vapor separately. The vapor was condensed into water in the condenser. Finally the water condensed from vapor and divided in steam separator flowed back to the tank. As shown in Fig. 2, the test section was consisted of an annular channel filled balls. The annular channel had an outer diameter of

Fig. 2. Schematic of packed bed.

102

G. Xu et al. / Progress in Nuclear Energy 76 (2014) 100e105

72 mm, inner diameter of 45 mm, and length of 980 mm. The outside of the annular channel was quartz glass tube and the inside of the annular channel was Teflon bar. The middle part of the channel was filled with oxidized carbon steel balls of 8 mm or 12 mm diameter. The upper and lower ends were filled with glass balls. Oxidized steel balls have high magnetic conductivity, and can change the electrical energy to heat easily. The steel balls were heated by the induction coil wrapped around the test section and the heat flux was determined by the lumped capacity method which has been used by Catton and Jakobsson (1987) with an accuracy of about 15%. The flow rate was measured via a turbine flow meter with an accuracy of ± 0.0084 m3. The inlet and outlet water temperature were measured by the copper-constantan thermocouples with an accuracy of ±0.1  C, The inlet pressure was measured with an accuracy of ±0.6 kPa and the outlet pressure was measured with an accuracy of ±0.25 kPa. All measurement signals were acquired by NI system and recorded by the computer. The high-speed camera was fixed at a two-dimensional slideway and the height of the camera and distance between test section and the camera could be continuously adjusted. The camera frame rate was set as 2000 frames per second. Because the packed bed was filled with steel balls, so it was not transparent. The lights must be in the front of the test section rather than behind the test section. In order to avoid the glisten of quartz glass, two 500 W halogen lights were fixed at the upper and bottom of the camera and formed a certain angle with observation window. 3. Results and discussion 3.1. Bubble slipping Bubble sliding process is shown in Fig. 3. T ¼ 0 ms, the bubble located at the bottom of the ball. The buoyancy, drag force, surface tension acting on the bubble could be decomposed into two forces. One force was the pressure whose direction was vertical to the

heated wall. Another force was driving force driving bubble motion along with the heated wall. Direction of driving force was parallel to the wall. At the effects of these two forces, the bubble could attach to the wall and slide along the wall. T ¼ 22.5 ms, Bubble slipped from the bottom to the upper of the ball. The effects of buoyancy and drag force would promote the bubble to detach from the wall. T ¼ 33 ms, Bubble moved to the almost peak of the ball, then left the wall. Fig. 3 results indicated that the bubble slipping started from the bottom of the ball, and left at the almost peak. Slipping distance was long and almost equal to the half of the ball circumference. The bubble away from the ball did not end the slipping, but also would reattach to the upper balls and go on slipping. Bubbles undergo a circular motion process which contained the slipping, departure, reattachment. 3.2. “Regrowth phenomenon” In subcooled flow boiling region, lots of bubbles repeated growth several times in the lifetime. The flow in the packed bed was complex, so “regrowth phenomenon” was divided into two cases. 1. Bubble left from the heated wall and began to shrink at the effects of subcooled liquid. The bubble should condense until it disappeared, but it would reattach the heat wall before the disappearing and then obtained the regrowth opportunity. The results were shown in Fig. 4a. T ¼ 0 ms, bubble attached on the heated wall and slipped along the wall. T ¼ 15 ms, at the influences of buoyancy and drag force, bubble departed from the heated wall and was condensed. T ¼ 37.5 ms, the bubble reattached to the upper balls and began a new growth. As shown in Fig. 4a, although there was no nucleate boiling in the upper ball, but it would capture the bubbles which was not condensation off by subcooled liquid. The upper ball would provide

Fig. 3. Bubbles sliding process.

G. Xu et al. / Progress in Nuclear Energy 76 (2014) 100e105

103

Fig. 4. “Regrowth phenomenon”.

the heat flux for bubble growth. The single-phase flow changed into two-phase flow around the upper ball. The flow field disturbance was enhanced, so the heat transfer would be improved. 2. The bubble did not left from the heated wall, but the heat flux was low and could not provide enough energy for bubble growth. The volume of bubble should be reduced or keep constant, however, there were a large number of contact surface in the packed bed and bubble could attach to the contact surface easily. The bubble located at the corner would break the original growth rate and then grow fast. The results were shown in Fig. 4b. From 0 ms to 24 ms, bubble growth rate was very slow and volume of bubble almost remains constant. T ¼ 24 ms, the bubble located at contact surface and was elongated and deformed, so the contact area between the balls and bubble was enlarged. The more energy could provide for bubble. Water subcooled in contact surface also was lower than other place. So the bubble growth was accelerated.

3.3. Bubble behaviors in the contact surface There were many different structures of contact surface in packed bed. The bubble firstly appeared in the contact surface and the structure of contact surface had great impacts on the bubble behaviors. The structures of contact surface were multiple and the effects on the bubble behaviors were complex. It was very difficult to study all the structures of contact surface. So we only discussed two simple but important structures. They were called structure A and structure B separately (as shown in Fig. 5). The direction of

Flow direction

The reasons of bubble slipping and regrowth in the volumetrically heated packed bed were depended on the structure of the channel. One reason was many balls existing in the channel. For another reason was the balls were heated. So the bubbles behaviors were not different from the conventional channels.

Flow direction

Structure A

Structure B Fig. 5. Structures of contact surface.

104

G. Xu et al. / Progress in Nuclear Energy 76 (2014) 100e105

connection of two ball centers was parallel to the direction of flow in structure A. The direction of connection of two ball centers was normal to the direction of flow in structure B. 3.3.1. Bubble behaviors in the structure A In structure A, the bubble growth process is shown in Fig. 6. T ¼ 0 ms, bubble generated in the contact surface. From T ¼ 0 ms to 10 ms, bubble growth rate was fast. The space was narrow in the contact surface, so the bubble was elongated and deformed. The bubble was almost an ellipsoid. The bubble volume was small and the effects of wall adhesion were great. So the bubble stability was stronger. The left and right side of bubble were symmetrical. From T ¼ 10 ms to 20 ms, Bubble volume continues to increase, Buoyancy and drag forces were increased also. The forces on the left and right sides of a bubble were not equivalent. The bubble stability was reduced. The bubble started slipping from one side. The left and right sides of the bubble were not symmetrical. T ¼ 22.5 ms, bubble left the contact surface and its shape was almost a spherical at the effects of surface tension. 3.3.2. Bubble behaviors in the structure B In the structure B, the bubble growth frequency was higher and the volume was smaller than in the structure A. We could not provide all the images in every time due to experimental conditions. So we only provided a bubble image in the structure B (as shown in Fig. 7). The bubble volume was small, so space in the structure B was enough for bubble growth. The bubble was not deformed and was nearly spherical. 3.3.3. The effects of structure A and structure B on the bubble behaviors As shown in Section 3.3.1 and 3.3.2, the bubble shapes in the different structures were different. In structure A, the bubble was elongated and deformed to an ellipsoid. In structure B, the bubble was a spherical. Figs. 8 and 9 were the bubble departure diameter and frequency change versus the heat flux, respectively. As shown in figures, the bubble departure frequency and diameter were very different. Bubble departure diameter in structure A was four times in structure A. The frequency was 1/45e1/3 times. Because in the structure A, the contact surface located the back flow region. The flow velocity was reduced by resisting of the lower ball, so the drag force on the bubble was decreased. The upper ball

Fig. 7. Bubble behaviors in the structure B.

hindered bubble to departure. At the effects of upper and lower balls, the bubble stability was enhanced in structure A and not easily to departure. In the structure B, the bubble faced fluid, the drag force on bubble was larger than in structure B, and no balls hindered bubble to departure. The bubble was easily departed from heated wall. So the bubble departure frequency was lower in structure A than in structure B, and the bubble growth time was longer in structure A than in structure B. The bubble had enough time to grow larger in structure A. Except the bubble shapes, departure diameter and frequency were different in the structure A and B, the forms of bubble left the contact surface were also different. The bubble left the contact surface by slipping in structure A, and the bubble left the contact surface by lift-off in structure B. Because the bubble volume was larger in structure A and the upper balls hindered bubble to left. So the bubble left contact surface only by slipping. In structure B the bubble volume was smaller and no ball hindered bubble to left. The drag force was also lager. All these factors promoted the bubble left corner by lift-off.

Fig. 6. Bubble behaviors in the structure A.

G. Xu et al. / Progress in Nuclear Energy 76 (2014) 100e105

105

bubble departure diameter and frequency. Maybe the ranges of inlet subcooled and mass flux are small in the experiment. 4. Conclusion 1. The bubbles were not easily to left wall due to the hindering of balls and undergo a circular motion process which contained the slipping, departure, reattachment. 2. Lots of bubbles would grow many times. Although there was no nucleate boiling in the upper ball, but it would capture the bubbles. The single-phase flow could change into two-phase flow around the upper ball. The flow field disturbance was enhanced. 3. The bubbles firstly appeared in the contact surface. The bubble shape, departure diameter and frequency, the forms of left the contact surface were also different in structure A and structure B. Fig. 8. Bubble departure diameter change versus the heat flux.

Acknowledgments The financial support of this work by the National Natural Science Foundation of China (Grant No. 11075042) is gratefully appreciated. References

Fig. 9. Bubble departure frequency change versus the heat flux.

From Figs. 8 and 9, the results shown that Bubble departure diameter and frequency almost increase with heat flux. Because the growth force increases with heat flux and prevent the bubble to departure, so bubble diameter is larger. The growth rate increases with heat flux, so the growth time decrease and the frequency is increased. The mass flux and subcooled have less effects on the

Atkhen, K., Berthoud, G., 2006. SILFIDE experiment: coolability in a volumetrically heated debris bed. Nucl. Eng. Des. 236 (19), 2126e2134. Catton, I., Jakobsson, J., 1987. The effect of pressure on dryout of a saturated bed of heat-generating particles. J. Heat Transf. (Transcations of the ASME (American Society of Mechanical Engineers), Series C); (United States) 109 (1), 185e195. Ford, L.H., 1960. Multiphase Flow Through Porous Media with Special Reference to Turbulent Region. Imperial College London. PhD thesis. Izadpanah, M., Müller-Steinhagen, H., Jamialahmadi, M., 1998. Experimental and theoretical studies of convective heat transfer in a cylindrical porous medium. Int. J. Heat Fluid Flow 19 (6), 629e635. Jamialahmadi, M., Müller-Steinhagen, H., Izadpanah, M., 2005. Pressure drop, gas hold-up and heat transfer during single and two-phase flow through porous media. Int. J. Heat Fluid Flow 26 (1), 156e172. Mazzarino, Sicardi., S., Baldi, G., 1987. Hydrodynamics and solideliquid contacting effectiveness in an upflow multiphase reactor. Chem. Eng. J. 36 (3), 151e160. €fer, P., Groll, M., Kulenovic, R., 2006. Basic investigations on debris cooling. Scha Nucl. Eng. Des. 236 (19), 2104e2116. Senor, D.J., Painter, C.L., Geelhood, K.J., 2007. An Innovative Spherical Cermet Nuclear Fuel Element to Achieve an Ultra-long Core Life for Use in Gridappropriate LWRs. PNNL-16647. Pacific Northwest National Laboratory. Sümer, S¸., Farhang, S., 2008. The fixed bed nuclear reactor concept. Energy Convers. Manage. 49 (7), 1902e1909. Turpin, Jim L., Huntington, R.L., 1967. Prediction of pressure drop for two-phase, two-component concurrent flow in packed beds. AIChE J. 13 (6), 1196e1202. Wang., Z., Peng., X., Wang., B., Liu., T., 2002. Experimental investigation of nucleate boiling phenomena in bead-packed structures. J. Eng. Thermophys. 23 (6), 742e744. Zhang, X., Su, W., Bai, B., Yan, X., 2009. Flow regime classification and transition of flow boiling through porous channel. J. Eng. Thermophys. 30 (6), 967e970.