- Email: [email protected]
Initial relocation behavior of control rod materials in boiling water reactors studied via time-resolved visualization
Nuclear Engineering and Design 333 (2018) 99–114
Contents lists available at ScienceDirect
Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes
Initial relocation behavior of control rod materials in boiling water reactors studied via time-resolved visualization
T
⁎
Shota Uedaa, , Byeongnam Job, Masahiro Kondoc, Nejdet Erkana, Takeshi Yajimad, Koji Okamotoa a
Department of Nuclear Engineering and Management, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan Department of Mechanical Engineering, Ajou University, Worldcup-ro 206, Suwon, Republic of Korea c Nuclear Professional School, The University of Tokyo, 2-22 Shirakata, Tokai-mura, Ibaraki, Japan d Institute for Solid State Physics, The University of Tokyo, 5-1-5 Kashiwa, Chiba, Japan b
A R T I C LE I N FO
A B S T R A C T
Keywords: Visualization Severe accident Control rod Eutectic melting Boiling water reactor
For the Fukushima decommissioning, the distribution of boron species in the fuel debris must be determined to assess the risk of recriticality, the debris hardness and thus complicate its successful retrieval. As a result, the relocation behavior of boron carbide (B4C) control rod materials has attracted significant attention. In this work, the influence of the thickness of its stainless steel (SS) clad on the initial relocation behavior of the control rod was investigated. In particular, the initial relocation behavior of the B4C control rod materials was dynamically visualized using a technique previously developed by the authors. To simulate the control rod, a type-304 SS tube filled with B4C powder containing particles with sizes of approximately 20–30 µm was heated to a temperature exceeding its eutectic point (1473 K). The three relocation modes of the obtained eutectic melt corresponded to film formation, droplet formation without collapse and droplet formation with collapse.
1. Introduction
National Laboratories) (Gauntt et al., 1989; Gauntt and Gasser, 1990) and Phebus FPT-3 (IRSN) (Clément et al., 2003; Girault et al., 2009; Haste et al., 2012) experiments. Thus, the B4C-SS eutectic melting process must be examined in detail. Fundamental studies on the progress of the eutectic reaction in a small specimen (so called a separate effect test) have been performed in the past (Hofmann et al., 1990; Nagase et al., 1997; Steinbrück, 2014; Shibata et al. 2015; Furuya and Morooka, 2015). Thus, Hofmann et al. (1990) found that the Arrhenius kinetics can be used in the mathematical description of the growth of the eutectic reaction layer, and Nagase et al. (1997) confirmed its applicability to a wider temperature range. In addition, the sudden increase in the reaction rate on the corresponding Arrhenius plots before and after the eutectic temperature was observed. Furuya and Morooka (2015) investigated the effect of oxidation during eutectic melting and found that B4C oxidation affects the rheological properties of the melt, which affects convection of the melt, and composition profile of compounds. The ability to describe the growth of the eutectic reaction layer by constructing Arrhenius plots indicates that eutectic melting is a time-dependent phenomenon and that the behavior of the produced eutectic melt and its relocation properties depend on the thickness of the SS clad (the latter was confirmed experimentally in the previous studies conducted using a plateshaped specimen (Ueda et al., 2016)). However, the influence of the SS
It is considered that the early liquefaction and relocation of boron carbide (B4C) control rods in boiling water reactors (BWRs) occur in severe accidents. Hence, it is important to investigate the relocation behavior of these rods as part of the establishment of a better severe accident management because of the effect produced on the subsequent accident progress. The obtained results can be also used for examining the boron distribution in fuel debris, which is of great importance to the Fukushima decommissioning because the presence of boron in the relocated core debris affects the possibility of re-criticality (Ikeuchi et al., 2017) and significantly increases its hardness (Takano et al., 2014; Nagase et al., 2017), thus making the process of debris retrieval very complicated. Therefore, a detailed elucidation of the relocation behavior of B4C control rods must be performed. A typical B4C control rod is fabricated from a stainless steel (SS) clad and B4C filling. During its relocation, various chemical reactions, such as the SS-B4C eutectic melting (Hofmann et al., 1990) and oxidation of B4C (Sepold et al., 2006; Repetto et al., 2010) occur. In particular, it is known that the B4C-SS eutectic melting affects its relocation, further contributes to the degradation of the surrounding fuel rod claddings and modifies the composition and characteristics of the final melt that include fuel materials through integral effect tests; the DF-4 (Sandia ⁎
Corresponding author. E-mail addresses: [email protected] (S. Ueda), [email protected] (B. Jo), [email protected] (M. Kondo), [email protected] (N. Erkan), [email protected] (T. Yajima), [email protected] (K. Okamoto). https://doi.org/10.1016/j.nucengdes.2018.04.006 Received 4 December 2017; Received in revised form 1 April 2018; Accepted 3 April 2018 0029-5493/ © 2018 Elsevier B.V. All rights reserved.
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
degradation due to eutectic melting, the dynamic visualization of the initial relocation behavior of the B4C control rod materials has been performed by varying the thickness of the SS clad (the utilized approach has been developed in a previous study (Ueda et al. 2016)). Unlike other methods (such as a metallurgical approach), this technique allows systematic monitoring of the eutectic relocation reaction in a highly time-resolved manner. To simulate a control rod, a type-304 SS tube filled with B4C powder (containing particles with sizes of approximately 20–30 µm) was used. The specimen was heated to a temperature exceeding 1473 K because Hofmann et al. (1990) reported that the eutectic reaction began at approximately 1473 K. The experiment was conducted in an argon atmosphere to eliminate the influence of B2O3 species (produced during the B4C oxidation) on the relocation behavior of the studied rod. The specimen was analyzed via X-ray diffraction (XRD) after the experiment to confirm the absence of significant amounts of oxygen contaminants in the eutectic melt. Fig. 1. XRD profiles of the B4C powder used in this study.
2. Experimental 2.1. Materials In this study, type 304 SS (Nilaco Co.) and B4C powder (Atom Shield Co., Ltd.) were used. The Rietveld analysis on XRD (Rigaku, SmartLab) data of the B4C powder revealed the powder contains 5.8(4)% of graphite as an impurity. Fig. 1 is the refined XRD profile of the B4C powder, showing observed (red cross), calculated (black line), and difference (blue line). The upper and middle ticks represent the positions of the Bragg reflections of the B4C and the bottom ticks represent those of graphite. The SEM image of the utilized B4C powder (see Fig. 2) suggests that its average particle size is equal to approximately 20–30 μm. 2.2. Experimental setup The resulting melt flow was visualized in a time-resolved manner to characterize the initial relocation behavior of the control rod at super high temperatures using the techniques developed in our previous studies (Ueda et al., 2016, 2017). The schematic of the utilized experimental setup is shown in Fig. 3. The specimen was placed between two tungsten heaters (connected to a power supply through copper electrodes) and then heated by thermal radiation. A traction system was attached to the bottom
Fig. 2. An SEM image of the B4C powder used in this study.
clad thickness on the relocation behavior of the control rod geometry has not been examined yet. In the current study, to investigate the initiation of the control rod
Fig. 3. A schematic of the experimental apparatus used in this study.
100
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 4. Schematics of the specimen design, tungsten heaters, and thermocouple positions during the visualization experiment.
during the experiment, G3 grade argon gas was supplied inside the chamber at a constant flow rate of 15–20 L/min. As a result, the concentration of oxygen species inside the chamber was less than 0.1 vol%, which was measured at the outlet of the test facility by an oxygen monitor (XO-2200, NEW COSMOS ELECTRIC CO., Ltd.). Fig. 4 shows the schematic describing the specimen design, configuration of the tungsten heaters, and positions of the thermocouples during visualization. A pole containing a type-304 SS shell filled with B4C powder was used as a specimen. The powder filled approximately 43% of the SS pole volume (the remaining voids were filled with air). The standard deviation of the filling rate was 5.4. For the filling rate calculation, the weight of the filled powder was divided by the calculated weight of the B4C assuming the SS tube was completely occupied with the B4C. The K-type sheathed thermocouples (T35051, Sakaguchi E.H. VOC CORP.) with diameters of 0.5 mm welded to the specimen surface were used for temperature measurements. The depths, heights, widths of the heaters were 5 or 8 mm (depending on the specimen diameter), 70 mm, and 0.2 mm, respectively. The position of the specimen’s top corresponded to the middle parts of the heaters along the vertical direction. Hofmann et al. (1990) reported that the eutectic reaction began at approximately 1473 K. Thus, in this work, the input power was manually controlled to reach temperatures greater than 1473 K at the specimen top. After the thermocouple failure, the input power was set as constant. The specimens were cooled by the air after the experiment. The following tests were conducted to investigate the initial relocation behavior of the B4C powder/SS eutectic melt. Table 1 lists the dimensions of the specimen (here D is the outer diameter of the SS tube, t is the thickness of the SS tube, and “Height” is the height of the SS tube), weight ratios between the SS clad and B4C powder (SS/B4C), and average temperatures with standard deviations (σ) for each test. The “Period” values provided after the average temperatures indicate the periods, over which temperatures were averaged.
Table 1 Experimental conditions utilized in tests. Test #
D, t [mm]
Height [mm]
SS/B4C (mass)
Average temp. [K] (Period [s])
σ [K]
1-1
4.20, 0.28
25 (70)*
2.45
12.2 5.1
1-2
4.20, 0.28
25 (70)*
2.69
1-3
4.20, 0.28
35 (70)*
2.63
2-1
6.00, 1.00
70
8.46
2-2
6.00, 1.00
70
9.82
2-3
6.00, 1.00
70
8.49
3-1
5.00, 0.50
70
3.64
3-2
5.00, 0.50
70
4.03
3-3
5.00, 0.50
70
3.71
Top: 1528.5 (247–521) Mid: 1394.3 (247–1000)** Top: 1489.7 (547–1267) Mid: 1403.0 (547–1600) Top: 1449.3 (575–947) Mid: 1288.6 (575–947) Top: 1498.0 (500–773) Mid: 1002.5 (500–760) Top: 1528.0 (202–724) Mid: 1075.0 (202–521, 879–1000) Top: 1513.4 (500–856) Mid: 1028.3 (500–814) Top: 1499.5 (600–1169) Mid: 1152.7 (600–1075) Top: 1475.3 (283–744) Mid: 917.0 (283–838) Top: 1511.7 (200–218) Mid: 945.6 (200–600)
19.8 23.0 11.3 13.8 4.0 5.3 6.4 21.3 7.3 20.5 13.1 24.2 6.7 24.6 4.0 3.6
* A ceramic holder was installed at the bottom of the specimen. The height in parentheses denotes the total height of the specimen (including the height of the ceramic holder). ** After 1000 s the temperature at Mid. was gradually increased to 1470.2 K.
electrode to prevent the heaters from buckling. The described setup was placed inside the test chamber fabricated from SS, which contained an observation window on one of its walls. To obtain clear images, a slit and a neutral density filter were installed between the observation window and the camera (FASTCAM SA-X, Photron Co.). A halogen lamp was placed in front of the observation window for specimen illumination. To minimize B4C oxidation and maintain an inert environment
3. Results Fig. 5 is a temperature history in test 1-1. During test 1-1, the thermocouple broke at a time of 521 s due to melting. The temperature measured at the middle position gradually increased to 1469.6 K after
101
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 5. A temperature history during test 1-1. Fig. 7. An observation image of the cutting plane at A-A in test 1-1.
1000 s. Fig. 6 shows time-resolved images of the relocated rod obtained during test 1-1. In test 1-1, the molten material covered the specimen surface like a film without falling down (Fig. 6). The collapse of the specimen or formation of melt droplets was not observed. A relatively weak flow of a small amount of the melt was detected on the specimen surface; however; the melt did not visibly move or became deformed as a whole; on the contrary, the specimen retained its shape during the experiment. The tested rod has not melted at once because there was a vertical temperature gradient along the specimen: the gradient was a temperature decrease with an increase of the distance from the top of the specimen. The melting firstly reached to the surface of SS tube at the top part of the specimen. A major part of the B4C powder has not interacted with the melt and was instead exposed to the atmosphere
(see the image at 1554 s in Fig. 6). The described relocation behavior will be further called “film formation” in this work. In addition to the relocation process, the evaporated tungsten species underwent desublimation at the specimen’s bottom, as shown in Fig. 6. This process did not affect the relocation behavior of the rod because the corresponding amount was very small, and the desublimation occurred at a large distance from the melting position. Fig. 7 is an observation image of the cutting plane at A-A in test 1-1 (see Fig. 6). The solidified melt was thin and fragile so that the cutting edge could not be flattened. Circular shape of the melt, which is similar to the original specimen shape, was retained. The B4C powder remained inside the solidified melt.
Fig. 6. Time-resolved images of the relocated rod obtained during test 1-1 (film formation).
102
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 8. A temperature history during test 1-2.
Fig. 8 is a temperature history in test 1-2. During test 1-2, the thermocouple broke at a time of 1267 s due to melting. Fig. 9. shows time-resolved images of the relocated rod obtained during test 1-2. In test 1-2, the molten material covered the specimen surface like a film without falling down showing the similar behavior to that of test 1-1 as well. The melt did not visibly move or became deformed as a whole; on the contrary, the specimen retained its shape during the experiment. The melting firstly reached to the surface of SS tube at the top part of the specimen (see the image at 1000 s in Fig. 9.). A major part of the B4C powder has not interacted with the melt (see the image at 1500 s in Fig. 9.). Thus, the observed relocation behavior was “film formation”.
Fig. 10. An observation image of the cutting plane at A-A in test 1-2.
Fig. 10 shows an observation image of the cutting plane at A-A in test 12 (see Fig. 9.). Circular shape of the melt, which was similar to the original specimen shape, was retained. The B4C powder was remained inside the solidified melt; however it was dropped off from the specimen during the post-test observation, which is shown as “Vacancy where B4C powder remained” in Fig. 10. Fig. 11 is a temperature history in test 1-3. During test 1-3. Fig. 12
Fig. 9. Time-resolved images of the relocated rod obtained during test 1-2 (film formation).
103
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
formation without collapse” in this work. Fig. 24 shows an observation image of the cutting plane at A-A in test 3-1. The melt partially invaded into the B4C powder layer at the plane close to the droplet. Fig. 25 is a temperature history in test 3-2. During test 3-2, the top thermocouple broke at 744 s. Test 3-2 (Fig. 26) did not exhibit any collapse of the specimen; however, the top part tilted slightly. The formation of a melt droplet was observed (see the images obtained after 800 s in Fig. 26). A fraction of the molten material covered the specimen to form a film on its surface, while the remaining melt formed a large droplet (see the image at 1000 s in Fig. 26). The non-reacted B4C powder was exposed to the atmosphere (see the image at 1100 s in Fig. 26). Thus, this relocation behavior was “droplet formation without collapse”. Fig. 27 shows an observation image on the part of the specimen in test 3-2. Inner surface of the solidified melt film was observed. as indicated with a rectangle frame in Fig. 26. The both inner and outer surface of the solidified melt were rough, which was similar to test 1-1, 1-2 and 1-3 (film formation). Fig. 28 is a temperature history in test 3-3. During test 3-3, the top and middle thermocouple broke at 218 s and 600 s, respectively. Test 33 (Fig. 29) did not exhibit any collapse of the specimen; instead, the formation of a melt droplet was observed (see the images obtained after 480 s in Fig. 29). A fraction of the molten material covered the specimen to form a film on its surface, while the remaining melt formed a droplet (see the image at 600 s in Fig. 29). The non-reacted B4C powder was exposed to the atmosphere (see the image at 600 s in Fig. 29). Thus, this relocation behavior was “droplet formation without collapse”. Fig. 30 shows an observation image of the cutting plane at A-A in test 33. The melt did not invade into the B4C powder layer. In tests 2 series (test 2-1, 2-2 and 2-3) and test 3 series (test 3-1, 3-2 and 3-3), small amounts of the non-reacted B4C powder floated on the droplet surface. The migration of its particles on the melt surface was observed as well, indicating that the droplet was not solidified during the experiment and exhibited some fluidity (see also Fig. 31). Fig. 31 shows the cropped image of the left side of the specimen utilized in test 3-2, which describes the shape of the produced droplet and trajectory of the dark spots determined using the particle tracking velocimetry method of the ImageJ software. It is very likely that these dark spots represent the non-reacted agglomerated B4C particles because the furnace chamber was filled with the inert gas (these particles flew upward at an average speed of 710 μm/s, as shown in the magnified images). The migration of non-reacted B4C particles on the melt surface was observed during the other tests other than test 1 series (test 1-1, 1-2 and 1-3) as well, and their trajectory suggest a relatively high fluidity of the melt produced during tests 2 series (test 2-1, 2-2 and 2-3) and test 3 series (test 3-1, 3-2 and 3-3). On the other hand, no migration of B4C powder occurred in test 1 series (film formation) because the thickness of the melt was not thick enough to allow the flow in itself. In order to confirm that significant amount of B2O3 was not formed during the experiment and it did not affect the relocation behavior of the rod, the molten specimens were analyzed by XRD (Rigaku, SmartLab) after the experiments because B2O3 could increase the apparent viscosity of the eutectic melt and suppress flow (Furuya and Morooka, 2015). Here, the results of the analysis on test 1-1, test-2-1 and test 3-1 are shown in Figs. 32-34 (one test case from each relocation mode). The parts of the specimens shown in Fig. 7, Fig. 16 (A-A) and Fig. 24 were crushed for the analysis. The majority of peaks correspond with those of Fe2B. The peaks of B2O3 were not detected in all tests. Thus, the difference of the melt behavior in the current experiments was not affected by chemical factors other than the eutectic melting. Some of small peaks which is indicated as ∗ in the figures were unable to be identified and their peaks were not identical among the tests.
shows time-resolved images of the relocated rod obtained during test 13. In test 1-3, the molten material covered the specimen surface like a film without falling down showing the similar behavior to that of test 11 as well. The melt did not visibly move or became deformed as a whole; on the contrary, the specimen retained its shape during the experiment. A major part of the B4C powder has not interacted with the melt (see the image at 996 s in Fig. 12). Thus, the observed relocation behavior was “film formation”. Fig. 13 shows an observation image of the part of the specimen which is indicated by a rectangle frame in Fig. 12. The both inner and outer surface of the solidified melt were rough, which was same as test 1-1 and test 1-2. Fig. 14 is a temperature history in test 2-1. During test 2-1, the top and middle thermocouple broke after 773 s, and 760 s, respectively. In test 2-1 (Fig. 15), the collapse of the specimen and formation of a melt droplet were observed. The tested rod has not melted at once because there was a vertical temperature gradient along the specimen: the gradient was a temperature decrease with an increase of the distance from the top of the specimen. The melting firstly reached to the surface of SS tube at the top part of the specimen. A large amount of the B4C powder at the specimen’s top did not interacted with the SS and some of the non-reacted B4C powder involve into the melt. The droplet moved downward due to gravity. During the downward movement, the top part of the specimen collapsed. Hence, this relocation behavior will be further called “droplet formation with collapse”. Fig. 16 shows observation images of the cutting plane at A-A and B-B in test 2-1. The Melt invaded into the B4C powder layer at the plane where droplet was formed. The droplet fell down to the bottom where the temperature was lower than the beginning temperature of the eutectic melting. Thus, the melt did not deeply invade into the B4C powder layer at B-B plane. Fig. 17 is a temperature history in test 2-2. During test 2-2, the top thermocouple broke after 724 s, and the middle thermocouple failed to measure the temperature between 521 and 879 s. In test 2-2 (Fig. 18), the collapse of the specimen and formation of a melt droplet were observed. The melting firstly reached to the surface of SS tube at the top part of the specimen. A large amount of the B4C powder at the specimen’s top did not interacted with the SS and some of the non-reacted B4C powder involve into the melt (see the image at 840 s in Fig. 18). The droplet moved downward due to gravity. During the downward movement, the top part of the specimen rolled down (see the images in the region of 840–1250 s depicted in Fig. 18). Thus, this relocation behavior was “droplet formation with collapse”. Fig. 19 is a temperature history in test 2-3. During test 2-3, the top and middle thermocouple broke after 856 s and 814 s, respectively. In test 2-3 (Fig. 20), the collapse of the specimen and formation of a melt droplet were observed. A large amount of the B4C powder at the specimen’s top did not interacted with the SS and some of the non-reacted B4C powder involve into the melt (see the image at 920 s in Fig. 20). The droplet moved downward due to gravity. During the downward movement, the top part of the specimen rolled down. Thus, this relocation behavior was “droplet formation with collapse”. Fig. 21 is an observation image of the cutting plane at A-A in test 2-3. The Melt invaded into the B4C powder layer at the plane where droplet was formed. Fig. 22 is a temperature history in test 3-1. Test 3-1 (Fig. 23) did not exhibit any collapse of the specimen; instead, the formation of a melt droplet was observed (see the images obtained after 990 s in Fig. 23). Briefly, the observed relocation process corresponded to an intermediate behavior between the “film formation” and “droplet formation with collapse” regimes. A fraction of the molten material covered the specimen to form a film on its surface, while the remaining melt formed a large droplet (see the image at 1200 s in Fig. 23). The tested rod has not melted at once because there was a vertical temperature gradient along the specimen: the gradient was a temperature decrease with an increase of the distance from the top of the specimen. The melting firstly reached to the surface of SS tube at the top part of the specimen. The described relocation behavior will be further called “droplet
4. Discussion In this section, possible mechanisms for the observed relocation modes are considered (see Fig. 35 and Fig. 36). Thus, Fig. 35 describes 104
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 13. An observation image on the part of the specimen in test 1-3. Fig. 11. A temperature history during test 1-3.
thin melt is formed in test 1 series. The thin melt cannot be deeply absorbed into the B4C powder layer because of a small quantity of the melt, which agrees with the observation image after the experiment such as Fig. 7. Thus, the packed B4C powder is not disintegrated and the specimen does not collapse. A thin melt film is formed around the B4C powder because a thin melt cannot flow easily by gravity. Comparing with the previous research (Ueda et al., 2016) where the SS plate (0.1 mm) was thinner than that of test 1 series in the current study, the melt film was observed at the beginning of the experiment; however, the melt was not significantly absorbed. The results from the current study and the past study cannot be directly compared; however, the
the vertical cross-section of the specimen whose SS thickness is thin (test 1 series). At the beginning, a small quantity of the eutectic melt is formed between the SS and B4C powder. The formed melt is absorbed by the voids of B4C powder layer and progressively dissolves the B4C while the slightly B4C enriched melt react with the remaining solid SS. Thus, the melt progresses in both directions to the solid SS and B4C powder. As shown in Fe-B binary phase diagram, boron concentration at the eutectic composition is very small (Fe/B ≅ 24). Therefore, the amount of the melt is mainly affected by the thickness of the SS tube: a
Fig. 12. Time-resolved images of the relocated rod obtained during test 1-3 (film formation).
105
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
In the same way as test 1 series, the melt progresses in both directions to the solid SS and B4C powder. In thick-SS cases, a thick melt is formed because the SS tube is thick. A thick melt is deeply absorbed into the B4C powder layer and interact with the B4C because of a large quantity of the melt, which agrees with the observation image after the experiment such as Fig. 16. Thus, the packed B4C powder is disintegrated and the specimen collapsed. Finally, a droplet is formed because a thick melt can flow by gravity. This description also agrees with previous experimental results. The invasion of the melt into the B4C powder layer was observed in the previous studies where the thickness of the SS was thicker than those of test 2 series (Nagase et al., 1997; Ueda et al., 2016, 2017). Ueda et al. (2016) observed that the formed thick melt was deeply absorbed into the B4C powder layer and interact with the B4C while the melt acquired a spherical shape, which lead to the disintegration of the packed B4C powder and subsequent collapse of the specimen. The thickness of SS in test 3 series is between those of test 1-1 and test 2-1. Thus, the results obtained in test 3 series demonstrated the droplet formation without collapse, which represented intermediate behavior between the “film formation” and “droplet formation with collapse” modes. A simple relationship between the velocity of the downflow and the thickness of the surface melt may be considered analytically assuming Newtonian fluid and steady rectilinear flow. Here, we treat the B4C
Fig. 14. A temperature history during test 2-1.
possible reason for the difference is the difference in the filling rate or the SS thickness because the interface condition between the B4C powder layer and the melt affects more strongly the melt flow when the melt amount is small, which means the effect of the interfacial condition becomes dominant over the body force such as gravity than the case with the thick melt. On the other hand, Fig. 36 describes the vertical cross-section of the specimen, for which the thickness of the solid SS is thick (test 2 series).
Fig. 15. Time-resolved images of the relocated rod obtained in test 2-1 (droplet formation with collapse).
106
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 16. Observation images of the cutting plane at A-A and B-B in test 2-1.
Fig. 17. A temperature history during test 2-2.
Fig. 19. A temperature history during test 2-3.
Fig. 18. Time-resolved images of the relocated rod obtained in test 2-2 (droplet formation with collapse). 107
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 20. Time-resolved images of the relocated rod obtained in test 2-3 (droplet formation with collapse).
Fig. 22. A temperature history during test 3-1.
powder layer as a wall and do not consider the absorption of the melt into it. Fig. 37 shows the cross-section of the melt flow on a vertical wall. We assume that the wall is well wetted and the downflow velocity is 0 at the wall. When we consider a very thin liquid, whose thickness is, between y and, gravitational force and viscos force acting on the both sides of the thin liquid are balanced. The balance of forces acting on a
Fig. 21. Observation of the cutting plane at A-A in test 2-3.
108
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 23. Time-resolved images of the relocated rod obtained in test 3-1 (droplet formation without collapse).
unit cross-section of the thin liquid may be described as
∂vz (y ) ⎞ ⎛ ∂vz (y ) − −η ⎜ = ρgn ∂y y + n ∂y y ⎟ ⎝ ⎠
(1)
where, η, vz, ρ and g are the viscosity of the melt, the velocity of the downflow, the density of the melt and gravity, respectively. is sufficiently thin. Thus, Eq. (1) may be wrote as
−η
∂2vz (y ) = ρg. ∂y 2
(2)
We solve Eq. (2) with the following boundary conditions:
vz (0) = 0 and
∂vz (y ) = 0. ∂y
(3)
We obtain the following horizontal velocity profile of the downflow: Fig. 24. An observation image of the cutting plane at A-A in test 3-1.
vz (y ) =
ρg y (2h−y ). 2η
(4)
Thus, the maximum vertical velocity may be described as
vz (h) =
ρg 2 h. 2η
(5)
Eq. (5) means that the vertical velocity is proportional to the square of the melt thickness. In this study, the thicknesses are 0.28 mm in test 1 series, 0.5 mm in test 2 series and 1.0 mm in test 3 series. It may be calculated from Eq. (5) that the velocity in test with 1.0 mm thickness is 12.8 times faster than that of test with 0.28 mm thickness. In this analysis, the absorption was not considered. The time development of the melt absorption into the voids of B4C powder layer may be simply modeled with Lucas-Washburn equation (Washburn, 1921) as follows:
Wl =
Fig. 25. A temperature history during test 3-2. 109
rγ cosθt 2η
(6)
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 26. Time-resolved images of the relocated rod obtained in test 3-2 (droplet formation without collapse).
Fig. 28. A temperature history during test 3-3.
eutectic composition when we consider that the Fe-B phase diagram can illustrate the behavior of SS-B4C melts and that the maximum temperature of the tests is just 100 K higher that the eutectic temperature. Pore radius of the B4C powder layer are almost same as well because of the same way of filling the B4C powder in all tests. They mean Eq. (6) indicates that the absorption speed is almost same in all tests before the melt is completely absorbed. Therefore, it is considered that ratio of the thickness of the thick melt to that of the thin melt film absorbed into the B4C powder layer becomes larger when the absorption speed is same, thus it can be considered that the difference in the velocities among the
Fig. 27. An observation image on the part of the specimen in test 3-2. Inner surface of the solidified melt was observed.
where, t is the time for a liquid of dynamic viscosity η and surface tension γ to penetrate a distance Wl into the capillary whose pore radius is r. The contact angle between liquid and solid is θ. In all tests in this study, fluid characteristics of the melt are almost same because the melts formed during the tests should have Fe/B ratio close to that of the
110
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 29. Time-resolved images of the relocated rod obtained in test 3-3 (droplet formation without collapse).
5. Conclusions In this study, the effect of the thickness of the SS clad on the initial relocation behavior of a rod-shaped specimen containing B4C species was investigated for the first time via dynamic visualization. Experiments were performed at temperatures exceeding the eutectic temperature of the melt (1473 K). The relocation process was visualized using a camera installed at the specimen’s front. As a result, the following three relocation modes of the eutectic melt were observed: film formation, droplet formation without collapse, and droplet formation with collapse. During the droplet formation, the migration of B4C powder was detected on the melt surface. At the same time, no migration of B4C species occurred during the film formation. The observed modes suggest that the geometry of the control rod contribute to the diversity of its relocation behaviors. The possible mechanism for the different relocation modes observed under the utilized experimental conditions include: 1. At the beginning, a small quantity of the eutectic melt is formed between the SS and B4C powder. 2. The formed melt is absorbed by the voids of B4C powder layer and progressively dissolves the B4C while the slightly B4C enriched melt react with the remaining solid SS. Thus, the melt progresses in both directions to the solid SS and B4C powder. 3. The thickness of the melt is mainly affected by the thickness of the SS tube because boron concentration at the eutectic composition in Fe-B binary phase diagram is much smaller than the initial boron concentration on the current experimetal conditions.
Fig. 30. An observation image of the cutting plane at A-A in test 3-3.
tests with different SS thicknesses was enhanced by the influence of the absorption of the melt in the B4C powder layer. However, the change of the interface between the B4C powder and eutectic melt was not considered in above analytical consideration. Therefore, for the better understanding of the phenomena, the assist of detailed numerical calculation would be necessary in the future.
111
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 31. A trajectory of the non-reacted B4C particles observed in test 3-2.
Fig. 32. XRD profiles of the part of the specimen where the eutectic melt occurred in test 1-1.
Fig. 34. XRD profiles of the part of the specimen where the eutectic melt occurred in test 3-1.
4. A thin melt is not deeply absorbed into the B4C powder layer while A thick melt is deeply absorbed into the B4C powder layer causing the collapse of the specimen. 5. A thin melt film is formed around the B4C powder because a thin melt cannot flow by gravity while a droplet is formed because a thick melt can flow by gravity. The findings obtained in this study suggest that the geometry of the control rod may be one of the dominant factors affecting its relocation behavior. For the comprehension of the relocation behavior of the control blade during the Fukushima accident at the early stage, further experimental works are needed because current study was only conducted under argon atmosphere and at atmospheric pressure. The influence of the oxidation and pressure on the relocation behavior should be understood near future. Furthermore, a better understanding of the observed phenomena (and thus the state of the fuel debris after the accident) can be achieved by performing numerical calculations. In
Fig. 33. XRD profiles of the part of the specimen where the eutectic melt occurred in test 2-1.
112
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 35. A vertical cross-section of the specimen whose SS thickness is thin (test 1-1, 1-2 and 1-3). Eutectic melting proceeds from (a) to (c) as time passes.
Fig. 36. A vertical cross-section of the specimen whose SS thickness is thick (test 2-1, 2-2 and 2-3). Eutectic melting proceeds from (a) to (c) as time passes.
Acknowledgments Funding: This study was partially supported by JSPS KAKENHI Grant Number 17J07634 and “High-accuracy computations of the core structure relocation during severe accidents” project funded by the Initiatives for Atomic Energy Basic and Generic Strategic Research program of the Ministry of Education, Culture, Sports, Science and Technology of Japan. References Clément, B.N., Hanniet-Girault, N., Repetto, G., Jacquemain, D., Jones, A.V., Kissane, M.P., Von der Hardt, P., 2003. LWR severe accident simulation: synthesis of the results and interpretation of the first Phebus FP experiment FPT0. Nucl. Eng. Des. 226, 582. Furuya, M., Morooka, S., 2015. Visual Observation of Natural Convection due to Eutectic Melting of Stainless Steel and Boron Carbide and Three-Dimensional Raman Spectrometry. National Heat Transfer Symposium of Japan, Fukuoka, Japan [in Japanese]. Gauntt, R.O., Gasser, R.D., 1990. Results of the DF-4 BWR control blade-channel box test. In: Prc. Eighteenth Water Reactor Safety Information Meeting, NUREG/CP-0114, vol. 2, pp. 25–40. Gauntt, R.O., Gasser, R.D., Ott, L.J., 1989. The DF-4 Fuel Damage Experiment in ACRR with a BWR Control Blade and Channel Box. Sandia National Laboratory, USA, Report NUREG/CR-4671. Girault, N., Fiche, C., Bujan, A., Dienstbier, J., 2009. Towards a better understanding of iodine chemistry in RCS of nuclear reactors. Nucl. Eng. Des. 239, 1162–1170. Haste, T., Payot, F., Dominguez, C., March, P., Simondi-Teisseire, B., Steinbrück, M., 2012. Study of boron behavior in the primary circuit of water reactors under severe accident conditions: a comparison of Phebus FPT3 results with other recent integral and separate-effects data. Nucl. Eng. Des. 246, 147–156. Hofmann, P., Markiewicz, M.E., Spino, J.L., 1990. Reaction behavior of B4C absorber material with stainless steel and zircaloy in severe light water reactor accidents. Nucl. Technol. 90, 226–244. Ikeuchi, H., Piluso, P., Fouquart, P., Excoffier, E., David, C., Brackx, E., 2017. Study on the
Fig. 37. A cross-section of the melt flow on a vertical wall (B4C powder region).
particular, the mixing ratio between the molten fuel and absorbent materials must be determined for successful Fukushima decommissioning.
113
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Shibata, H., Sakamoto, K., Ouchi, A., Kurata, M., 2015. Chemical interaction between granular B4C and 304L-type stainless steel materials used in BWRs in Japan. J. Nucl. Sci. Technol. 52, 1313–1317. Steinbrück, M., 2014. Influence of boron carbide on core degradation during severe accidents in LWRs. Ann. Nucl. Energy 64, 43–49. Takano, M., Nishi, T., Shirasu, N., 2014. Characterization of solidified melt among materials of UO2 fuel and B4C control blade. J. Nucl. Sci. Technol. 51, 859–875. Ueda, S., Madokoro, H., Jo, B., Kondo, M., Erkan, N., Okamoto, K., 2016. Time-resolved visualization technique for eutectic reaction between boron carbide and stainless steel at high temperatures. Mech. Eng. Lett. 2, 16–00056. Ueda, S., Madokoro, H., Jo, B., Kondo, M., Erkan, N., Okamoto, K., 2017. Dynamic visualization of eutectic reaction between boron carbide and stainless steel. J. Nucl. Sci. Technol. 54, 81–88. Washburn, E.W., 1921. The dynamics of capillary flow. Phys. Rev. 17, 273.
distribution of boron in the in-vessel fuel debris in conditions close to Fukushima Daiichi Nuclear Power Station Unit 2. The 8th European Review Meeting on Severe Accident Research (ERMSAR) 2017, Warsaw, Poland. Nagase, F., Gauntt, R.O., Naito, M., 2017. Overview and outcomes of the OECD/NEA benchmark study of the accident at the Fukushima Daiichi nuclear power station. Nucl. Technol. 196, 499–510. Nagase, F., Uetsuka, H., Otomo, T., 1997. Chemical interactions between B4C and stainless steel at high temperatures. J. Nucl. Mater. 245, 52–59. Repetto, G., De Luze, O., Seiler, N., Trambauer, K., Austregesilo, H., Birchley, J., Ederli, S., Lamy, J.S., Maliverney, B., Drath, T., Hollands, T., 2010. B4C oxidation modelling in severe accident codes: applications to PHEBUS and QUENCH experiments. Prog. Nucl. Energy 52, 37–45. Sepold, L., Schanz, G., Steinbrück, M., Stuckert, J., Miassoedov, A., Palagin, A., Veshchunov, M., 2006. Results of the QUENCH-09 experiment compared to QUENCH-07 with incorporation of B4C absorber. Nucl. Technol. 154, 107–116.
114
Contents lists available at ScienceDirect
Nuclear Engineering and Design journal homepage: www.elsevier.com/locate/nucengdes
Initial relocation behavior of control rod materials in boiling water reactors studied via time-resolved visualization
T
⁎
Shota Uedaa, , Byeongnam Job, Masahiro Kondoc, Nejdet Erkana, Takeshi Yajimad, Koji Okamotoa a
Department of Nuclear Engineering and Management, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, Japan Department of Mechanical Engineering, Ajou University, Worldcup-ro 206, Suwon, Republic of Korea c Nuclear Professional School, The University of Tokyo, 2-22 Shirakata, Tokai-mura, Ibaraki, Japan d Institute for Solid State Physics, The University of Tokyo, 5-1-5 Kashiwa, Chiba, Japan b
A R T I C LE I N FO
A B S T R A C T
Keywords: Visualization Severe accident Control rod Eutectic melting Boiling water reactor
For the Fukushima decommissioning, the distribution of boron species in the fuel debris must be determined to assess the risk of recriticality, the debris hardness and thus complicate its successful retrieval. As a result, the relocation behavior of boron carbide (B4C) control rod materials has attracted significant attention. In this work, the influence of the thickness of its stainless steel (SS) clad on the initial relocation behavior of the control rod was investigated. In particular, the initial relocation behavior of the B4C control rod materials was dynamically visualized using a technique previously developed by the authors. To simulate the control rod, a type-304 SS tube filled with B4C powder containing particles with sizes of approximately 20–30 µm was heated to a temperature exceeding its eutectic point (1473 K). The three relocation modes of the obtained eutectic melt corresponded to film formation, droplet formation without collapse and droplet formation with collapse.
1. Introduction
National Laboratories) (Gauntt et al., 1989; Gauntt and Gasser, 1990) and Phebus FPT-3 (IRSN) (Clément et al., 2003; Girault et al., 2009; Haste et al., 2012) experiments. Thus, the B4C-SS eutectic melting process must be examined in detail. Fundamental studies on the progress of the eutectic reaction in a small specimen (so called a separate effect test) have been performed in the past (Hofmann et al., 1990; Nagase et al., 1997; Steinbrück, 2014; Shibata et al. 2015; Furuya and Morooka, 2015). Thus, Hofmann et al. (1990) found that the Arrhenius kinetics can be used in the mathematical description of the growth of the eutectic reaction layer, and Nagase et al. (1997) confirmed its applicability to a wider temperature range. In addition, the sudden increase in the reaction rate on the corresponding Arrhenius plots before and after the eutectic temperature was observed. Furuya and Morooka (2015) investigated the effect of oxidation during eutectic melting and found that B4C oxidation affects the rheological properties of the melt, which affects convection of the melt, and composition profile of compounds. The ability to describe the growth of the eutectic reaction layer by constructing Arrhenius plots indicates that eutectic melting is a time-dependent phenomenon and that the behavior of the produced eutectic melt and its relocation properties depend on the thickness of the SS clad (the latter was confirmed experimentally in the previous studies conducted using a plateshaped specimen (Ueda et al., 2016)). However, the influence of the SS
It is considered that the early liquefaction and relocation of boron carbide (B4C) control rods in boiling water reactors (BWRs) occur in severe accidents. Hence, it is important to investigate the relocation behavior of these rods as part of the establishment of a better severe accident management because of the effect produced on the subsequent accident progress. The obtained results can be also used for examining the boron distribution in fuel debris, which is of great importance to the Fukushima decommissioning because the presence of boron in the relocated core debris affects the possibility of re-criticality (Ikeuchi et al., 2017) and significantly increases its hardness (Takano et al., 2014; Nagase et al., 2017), thus making the process of debris retrieval very complicated. Therefore, a detailed elucidation of the relocation behavior of B4C control rods must be performed. A typical B4C control rod is fabricated from a stainless steel (SS) clad and B4C filling. During its relocation, various chemical reactions, such as the SS-B4C eutectic melting (Hofmann et al., 1990) and oxidation of B4C (Sepold et al., 2006; Repetto et al., 2010) occur. In particular, it is known that the B4C-SS eutectic melting affects its relocation, further contributes to the degradation of the surrounding fuel rod claddings and modifies the composition and characteristics of the final melt that include fuel materials through integral effect tests; the DF-4 (Sandia ⁎
Corresponding author. E-mail addresses: [email protected] (S. Ueda), [email protected] (B. Jo), [email protected] (M. Kondo), [email protected] (N. Erkan), [email protected] (T. Yajima), [email protected] (K. Okamoto). https://doi.org/10.1016/j.nucengdes.2018.04.006 Received 4 December 2017; Received in revised form 1 April 2018; Accepted 3 April 2018 0029-5493/ © 2018 Elsevier B.V. All rights reserved.
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
degradation due to eutectic melting, the dynamic visualization of the initial relocation behavior of the B4C control rod materials has been performed by varying the thickness of the SS clad (the utilized approach has been developed in a previous study (Ueda et al. 2016)). Unlike other methods (such as a metallurgical approach), this technique allows systematic monitoring of the eutectic relocation reaction in a highly time-resolved manner. To simulate a control rod, a type-304 SS tube filled with B4C powder (containing particles with sizes of approximately 20–30 µm) was used. The specimen was heated to a temperature exceeding 1473 K because Hofmann et al. (1990) reported that the eutectic reaction began at approximately 1473 K. The experiment was conducted in an argon atmosphere to eliminate the influence of B2O3 species (produced during the B4C oxidation) on the relocation behavior of the studied rod. The specimen was analyzed via X-ray diffraction (XRD) after the experiment to confirm the absence of significant amounts of oxygen contaminants in the eutectic melt. Fig. 1. XRD profiles of the B4C powder used in this study.
2. Experimental 2.1. Materials In this study, type 304 SS (Nilaco Co.) and B4C powder (Atom Shield Co., Ltd.) were used. The Rietveld analysis on XRD (Rigaku, SmartLab) data of the B4C powder revealed the powder contains 5.8(4)% of graphite as an impurity. Fig. 1 is the refined XRD profile of the B4C powder, showing observed (red cross), calculated (black line), and difference (blue line). The upper and middle ticks represent the positions of the Bragg reflections of the B4C and the bottom ticks represent those of graphite. The SEM image of the utilized B4C powder (see Fig. 2) suggests that its average particle size is equal to approximately 20–30 μm. 2.2. Experimental setup The resulting melt flow was visualized in a time-resolved manner to characterize the initial relocation behavior of the control rod at super high temperatures using the techniques developed in our previous studies (Ueda et al., 2016, 2017). The schematic of the utilized experimental setup is shown in Fig. 3. The specimen was placed between two tungsten heaters (connected to a power supply through copper electrodes) and then heated by thermal radiation. A traction system was attached to the bottom
Fig. 2. An SEM image of the B4C powder used in this study.
clad thickness on the relocation behavior of the control rod geometry has not been examined yet. In the current study, to investigate the initiation of the control rod
Fig. 3. A schematic of the experimental apparatus used in this study.
100
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 4. Schematics of the specimen design, tungsten heaters, and thermocouple positions during the visualization experiment.
during the experiment, G3 grade argon gas was supplied inside the chamber at a constant flow rate of 15–20 L/min. As a result, the concentration of oxygen species inside the chamber was less than 0.1 vol%, which was measured at the outlet of the test facility by an oxygen monitor (XO-2200, NEW COSMOS ELECTRIC CO., Ltd.). Fig. 4 shows the schematic describing the specimen design, configuration of the tungsten heaters, and positions of the thermocouples during visualization. A pole containing a type-304 SS shell filled with B4C powder was used as a specimen. The powder filled approximately 43% of the SS pole volume (the remaining voids were filled with air). The standard deviation of the filling rate was 5.4. For the filling rate calculation, the weight of the filled powder was divided by the calculated weight of the B4C assuming the SS tube was completely occupied with the B4C. The K-type sheathed thermocouples (T35051, Sakaguchi E.H. VOC CORP.) with diameters of 0.5 mm welded to the specimen surface were used for temperature measurements. The depths, heights, widths of the heaters were 5 or 8 mm (depending on the specimen diameter), 70 mm, and 0.2 mm, respectively. The position of the specimen’s top corresponded to the middle parts of the heaters along the vertical direction. Hofmann et al. (1990) reported that the eutectic reaction began at approximately 1473 K. Thus, in this work, the input power was manually controlled to reach temperatures greater than 1473 K at the specimen top. After the thermocouple failure, the input power was set as constant. The specimens were cooled by the air after the experiment. The following tests were conducted to investigate the initial relocation behavior of the B4C powder/SS eutectic melt. Table 1 lists the dimensions of the specimen (here D is the outer diameter of the SS tube, t is the thickness of the SS tube, and “Height” is the height of the SS tube), weight ratios between the SS clad and B4C powder (SS/B4C), and average temperatures with standard deviations (σ) for each test. The “Period” values provided after the average temperatures indicate the periods, over which temperatures were averaged.
Table 1 Experimental conditions utilized in tests. Test #
D, t [mm]
Height [mm]
SS/B4C (mass)
Average temp. [K] (Period [s])
σ [K]
1-1
4.20, 0.28
25 (70)*
2.45
12.2 5.1
1-2
4.20, 0.28
25 (70)*
2.69
1-3
4.20, 0.28
35 (70)*
2.63
2-1
6.00, 1.00
70
8.46
2-2
6.00, 1.00
70
9.82
2-3
6.00, 1.00
70
8.49
3-1
5.00, 0.50
70
3.64
3-2
5.00, 0.50
70
4.03
3-3
5.00, 0.50
70
3.71
Top: 1528.5 (247–521) Mid: 1394.3 (247–1000)** Top: 1489.7 (547–1267) Mid: 1403.0 (547–1600) Top: 1449.3 (575–947) Mid: 1288.6 (575–947) Top: 1498.0 (500–773) Mid: 1002.5 (500–760) Top: 1528.0 (202–724) Mid: 1075.0 (202–521, 879–1000) Top: 1513.4 (500–856) Mid: 1028.3 (500–814) Top: 1499.5 (600–1169) Mid: 1152.7 (600–1075) Top: 1475.3 (283–744) Mid: 917.0 (283–838) Top: 1511.7 (200–218) Mid: 945.6 (200–600)
19.8 23.0 11.3 13.8 4.0 5.3 6.4 21.3 7.3 20.5 13.1 24.2 6.7 24.6 4.0 3.6
* A ceramic holder was installed at the bottom of the specimen. The height in parentheses denotes the total height of the specimen (including the height of the ceramic holder). ** After 1000 s the temperature at Mid. was gradually increased to 1470.2 K.
electrode to prevent the heaters from buckling. The described setup was placed inside the test chamber fabricated from SS, which contained an observation window on one of its walls. To obtain clear images, a slit and a neutral density filter were installed between the observation window and the camera (FASTCAM SA-X, Photron Co.). A halogen lamp was placed in front of the observation window for specimen illumination. To minimize B4C oxidation and maintain an inert environment
3. Results Fig. 5 is a temperature history in test 1-1. During test 1-1, the thermocouple broke at a time of 521 s due to melting. The temperature measured at the middle position gradually increased to 1469.6 K after
101
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 5. A temperature history during test 1-1. Fig. 7. An observation image of the cutting plane at A-A in test 1-1.
1000 s. Fig. 6 shows time-resolved images of the relocated rod obtained during test 1-1. In test 1-1, the molten material covered the specimen surface like a film without falling down (Fig. 6). The collapse of the specimen or formation of melt droplets was not observed. A relatively weak flow of a small amount of the melt was detected on the specimen surface; however; the melt did not visibly move or became deformed as a whole; on the contrary, the specimen retained its shape during the experiment. The tested rod has not melted at once because there was a vertical temperature gradient along the specimen: the gradient was a temperature decrease with an increase of the distance from the top of the specimen. The melting firstly reached to the surface of SS tube at the top part of the specimen. A major part of the B4C powder has not interacted with the melt and was instead exposed to the atmosphere
(see the image at 1554 s in Fig. 6). The described relocation behavior will be further called “film formation” in this work. In addition to the relocation process, the evaporated tungsten species underwent desublimation at the specimen’s bottom, as shown in Fig. 6. This process did not affect the relocation behavior of the rod because the corresponding amount was very small, and the desublimation occurred at a large distance from the melting position. Fig. 7 is an observation image of the cutting plane at A-A in test 1-1 (see Fig. 6). The solidified melt was thin and fragile so that the cutting edge could not be flattened. Circular shape of the melt, which is similar to the original specimen shape, was retained. The B4C powder remained inside the solidified melt.
Fig. 6. Time-resolved images of the relocated rod obtained during test 1-1 (film formation).
102
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 8. A temperature history during test 1-2.
Fig. 8 is a temperature history in test 1-2. During test 1-2, the thermocouple broke at a time of 1267 s due to melting. Fig. 9. shows time-resolved images of the relocated rod obtained during test 1-2. In test 1-2, the molten material covered the specimen surface like a film without falling down showing the similar behavior to that of test 1-1 as well. The melt did not visibly move or became deformed as a whole; on the contrary, the specimen retained its shape during the experiment. The melting firstly reached to the surface of SS tube at the top part of the specimen (see the image at 1000 s in Fig. 9.). A major part of the B4C powder has not interacted with the melt (see the image at 1500 s in Fig. 9.). Thus, the observed relocation behavior was “film formation”.
Fig. 10. An observation image of the cutting plane at A-A in test 1-2.
Fig. 10 shows an observation image of the cutting plane at A-A in test 12 (see Fig. 9.). Circular shape of the melt, which was similar to the original specimen shape, was retained. The B4C powder was remained inside the solidified melt; however it was dropped off from the specimen during the post-test observation, which is shown as “Vacancy where B4C powder remained” in Fig. 10. Fig. 11 is a temperature history in test 1-3. During test 1-3. Fig. 12
Fig. 9. Time-resolved images of the relocated rod obtained during test 1-2 (film formation).
103
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
formation without collapse” in this work. Fig. 24 shows an observation image of the cutting plane at A-A in test 3-1. The melt partially invaded into the B4C powder layer at the plane close to the droplet. Fig. 25 is a temperature history in test 3-2. During test 3-2, the top thermocouple broke at 744 s. Test 3-2 (Fig. 26) did not exhibit any collapse of the specimen; however, the top part tilted slightly. The formation of a melt droplet was observed (see the images obtained after 800 s in Fig. 26). A fraction of the molten material covered the specimen to form a film on its surface, while the remaining melt formed a large droplet (see the image at 1000 s in Fig. 26). The non-reacted B4C powder was exposed to the atmosphere (see the image at 1100 s in Fig. 26). Thus, this relocation behavior was “droplet formation without collapse”. Fig. 27 shows an observation image on the part of the specimen in test 3-2. Inner surface of the solidified melt film was observed. as indicated with a rectangle frame in Fig. 26. The both inner and outer surface of the solidified melt were rough, which was similar to test 1-1, 1-2 and 1-3 (film formation). Fig. 28 is a temperature history in test 3-3. During test 3-3, the top and middle thermocouple broke at 218 s and 600 s, respectively. Test 33 (Fig. 29) did not exhibit any collapse of the specimen; instead, the formation of a melt droplet was observed (see the images obtained after 480 s in Fig. 29). A fraction of the molten material covered the specimen to form a film on its surface, while the remaining melt formed a droplet (see the image at 600 s in Fig. 29). The non-reacted B4C powder was exposed to the atmosphere (see the image at 600 s in Fig. 29). Thus, this relocation behavior was “droplet formation without collapse”. Fig. 30 shows an observation image of the cutting plane at A-A in test 33. The melt did not invade into the B4C powder layer. In tests 2 series (test 2-1, 2-2 and 2-3) and test 3 series (test 3-1, 3-2 and 3-3), small amounts of the non-reacted B4C powder floated on the droplet surface. The migration of its particles on the melt surface was observed as well, indicating that the droplet was not solidified during the experiment and exhibited some fluidity (see also Fig. 31). Fig. 31 shows the cropped image of the left side of the specimen utilized in test 3-2, which describes the shape of the produced droplet and trajectory of the dark spots determined using the particle tracking velocimetry method of the ImageJ software. It is very likely that these dark spots represent the non-reacted agglomerated B4C particles because the furnace chamber was filled with the inert gas (these particles flew upward at an average speed of 710 μm/s, as shown in the magnified images). The migration of non-reacted B4C particles on the melt surface was observed during the other tests other than test 1 series (test 1-1, 1-2 and 1-3) as well, and their trajectory suggest a relatively high fluidity of the melt produced during tests 2 series (test 2-1, 2-2 and 2-3) and test 3 series (test 3-1, 3-2 and 3-3). On the other hand, no migration of B4C powder occurred in test 1 series (film formation) because the thickness of the melt was not thick enough to allow the flow in itself. In order to confirm that significant amount of B2O3 was not formed during the experiment and it did not affect the relocation behavior of the rod, the molten specimens were analyzed by XRD (Rigaku, SmartLab) after the experiments because B2O3 could increase the apparent viscosity of the eutectic melt and suppress flow (Furuya and Morooka, 2015). Here, the results of the analysis on test 1-1, test-2-1 and test 3-1 are shown in Figs. 32-34 (one test case from each relocation mode). The parts of the specimens shown in Fig. 7, Fig. 16 (A-A) and Fig. 24 were crushed for the analysis. The majority of peaks correspond with those of Fe2B. The peaks of B2O3 were not detected in all tests. Thus, the difference of the melt behavior in the current experiments was not affected by chemical factors other than the eutectic melting. Some of small peaks which is indicated as ∗ in the figures were unable to be identified and their peaks were not identical among the tests.
shows time-resolved images of the relocated rod obtained during test 13. In test 1-3, the molten material covered the specimen surface like a film without falling down showing the similar behavior to that of test 11 as well. The melt did not visibly move or became deformed as a whole; on the contrary, the specimen retained its shape during the experiment. A major part of the B4C powder has not interacted with the melt (see the image at 996 s in Fig. 12). Thus, the observed relocation behavior was “film formation”. Fig. 13 shows an observation image of the part of the specimen which is indicated by a rectangle frame in Fig. 12. The both inner and outer surface of the solidified melt were rough, which was same as test 1-1 and test 1-2. Fig. 14 is a temperature history in test 2-1. During test 2-1, the top and middle thermocouple broke after 773 s, and 760 s, respectively. In test 2-1 (Fig. 15), the collapse of the specimen and formation of a melt droplet were observed. The tested rod has not melted at once because there was a vertical temperature gradient along the specimen: the gradient was a temperature decrease with an increase of the distance from the top of the specimen. The melting firstly reached to the surface of SS tube at the top part of the specimen. A large amount of the B4C powder at the specimen’s top did not interacted with the SS and some of the non-reacted B4C powder involve into the melt. The droplet moved downward due to gravity. During the downward movement, the top part of the specimen collapsed. Hence, this relocation behavior will be further called “droplet formation with collapse”. Fig. 16 shows observation images of the cutting plane at A-A and B-B in test 2-1. The Melt invaded into the B4C powder layer at the plane where droplet was formed. The droplet fell down to the bottom where the temperature was lower than the beginning temperature of the eutectic melting. Thus, the melt did not deeply invade into the B4C powder layer at B-B plane. Fig. 17 is a temperature history in test 2-2. During test 2-2, the top thermocouple broke after 724 s, and the middle thermocouple failed to measure the temperature between 521 and 879 s. In test 2-2 (Fig. 18), the collapse of the specimen and formation of a melt droplet were observed. The melting firstly reached to the surface of SS tube at the top part of the specimen. A large amount of the B4C powder at the specimen’s top did not interacted with the SS and some of the non-reacted B4C powder involve into the melt (see the image at 840 s in Fig. 18). The droplet moved downward due to gravity. During the downward movement, the top part of the specimen rolled down (see the images in the region of 840–1250 s depicted in Fig. 18). Thus, this relocation behavior was “droplet formation with collapse”. Fig. 19 is a temperature history in test 2-3. During test 2-3, the top and middle thermocouple broke after 856 s and 814 s, respectively. In test 2-3 (Fig. 20), the collapse of the specimen and formation of a melt droplet were observed. A large amount of the B4C powder at the specimen’s top did not interacted with the SS and some of the non-reacted B4C powder involve into the melt (see the image at 920 s in Fig. 20). The droplet moved downward due to gravity. During the downward movement, the top part of the specimen rolled down. Thus, this relocation behavior was “droplet formation with collapse”. Fig. 21 is an observation image of the cutting plane at A-A in test 2-3. The Melt invaded into the B4C powder layer at the plane where droplet was formed. Fig. 22 is a temperature history in test 3-1. Test 3-1 (Fig. 23) did not exhibit any collapse of the specimen; instead, the formation of a melt droplet was observed (see the images obtained after 990 s in Fig. 23). Briefly, the observed relocation process corresponded to an intermediate behavior between the “film formation” and “droplet formation with collapse” regimes. A fraction of the molten material covered the specimen to form a film on its surface, while the remaining melt formed a large droplet (see the image at 1200 s in Fig. 23). The tested rod has not melted at once because there was a vertical temperature gradient along the specimen: the gradient was a temperature decrease with an increase of the distance from the top of the specimen. The melting firstly reached to the surface of SS tube at the top part of the specimen. The described relocation behavior will be further called “droplet
4. Discussion In this section, possible mechanisms for the observed relocation modes are considered (see Fig. 35 and Fig. 36). Thus, Fig. 35 describes 104
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 13. An observation image on the part of the specimen in test 1-3. Fig. 11. A temperature history during test 1-3.
thin melt is formed in test 1 series. The thin melt cannot be deeply absorbed into the B4C powder layer because of a small quantity of the melt, which agrees with the observation image after the experiment such as Fig. 7. Thus, the packed B4C powder is not disintegrated and the specimen does not collapse. A thin melt film is formed around the B4C powder because a thin melt cannot flow easily by gravity. Comparing with the previous research (Ueda et al., 2016) where the SS plate (0.1 mm) was thinner than that of test 1 series in the current study, the melt film was observed at the beginning of the experiment; however, the melt was not significantly absorbed. The results from the current study and the past study cannot be directly compared; however, the
the vertical cross-section of the specimen whose SS thickness is thin (test 1 series). At the beginning, a small quantity of the eutectic melt is formed between the SS and B4C powder. The formed melt is absorbed by the voids of B4C powder layer and progressively dissolves the B4C while the slightly B4C enriched melt react with the remaining solid SS. Thus, the melt progresses in both directions to the solid SS and B4C powder. As shown in Fe-B binary phase diagram, boron concentration at the eutectic composition is very small (Fe/B ≅ 24). Therefore, the amount of the melt is mainly affected by the thickness of the SS tube: a
Fig. 12. Time-resolved images of the relocated rod obtained during test 1-3 (film formation).
105
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
In the same way as test 1 series, the melt progresses in both directions to the solid SS and B4C powder. In thick-SS cases, a thick melt is formed because the SS tube is thick. A thick melt is deeply absorbed into the B4C powder layer and interact with the B4C because of a large quantity of the melt, which agrees with the observation image after the experiment such as Fig. 16. Thus, the packed B4C powder is disintegrated and the specimen collapsed. Finally, a droplet is formed because a thick melt can flow by gravity. This description also agrees with previous experimental results. The invasion of the melt into the B4C powder layer was observed in the previous studies where the thickness of the SS was thicker than those of test 2 series (Nagase et al., 1997; Ueda et al., 2016, 2017). Ueda et al. (2016) observed that the formed thick melt was deeply absorbed into the B4C powder layer and interact with the B4C while the melt acquired a spherical shape, which lead to the disintegration of the packed B4C powder and subsequent collapse of the specimen. The thickness of SS in test 3 series is between those of test 1-1 and test 2-1. Thus, the results obtained in test 3 series demonstrated the droplet formation without collapse, which represented intermediate behavior between the “film formation” and “droplet formation with collapse” modes. A simple relationship between the velocity of the downflow and the thickness of the surface melt may be considered analytically assuming Newtonian fluid and steady rectilinear flow. Here, we treat the B4C
Fig. 14. A temperature history during test 2-1.
possible reason for the difference is the difference in the filling rate or the SS thickness because the interface condition between the B4C powder layer and the melt affects more strongly the melt flow when the melt amount is small, which means the effect of the interfacial condition becomes dominant over the body force such as gravity than the case with the thick melt. On the other hand, Fig. 36 describes the vertical cross-section of the specimen, for which the thickness of the solid SS is thick (test 2 series).
Fig. 15. Time-resolved images of the relocated rod obtained in test 2-1 (droplet formation with collapse).
106
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 16. Observation images of the cutting plane at A-A and B-B in test 2-1.
Fig. 17. A temperature history during test 2-2.
Fig. 19. A temperature history during test 2-3.
Fig. 18. Time-resolved images of the relocated rod obtained in test 2-2 (droplet formation with collapse). 107
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 20. Time-resolved images of the relocated rod obtained in test 2-3 (droplet formation with collapse).
Fig. 22. A temperature history during test 3-1.
powder layer as a wall and do not consider the absorption of the melt into it. Fig. 37 shows the cross-section of the melt flow on a vertical wall. We assume that the wall is well wetted and the downflow velocity is 0 at the wall. When we consider a very thin liquid, whose thickness is, between y and, gravitational force and viscos force acting on the both sides of the thin liquid are balanced. The balance of forces acting on a
Fig. 21. Observation of the cutting plane at A-A in test 2-3.
108
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 23. Time-resolved images of the relocated rod obtained in test 3-1 (droplet formation without collapse).
unit cross-section of the thin liquid may be described as
∂vz (y ) ⎞ ⎛ ∂vz (y ) − −η ⎜ = ρgn ∂y y + n ∂y y ⎟ ⎝ ⎠
(1)
where, η, vz, ρ and g are the viscosity of the melt, the velocity of the downflow, the density of the melt and gravity, respectively. is sufficiently thin. Thus, Eq. (1) may be wrote as
−η
∂2vz (y ) = ρg. ∂y 2
(2)
We solve Eq. (2) with the following boundary conditions:
vz (0) = 0 and
∂vz (y ) = 0. ∂y
(3)
We obtain the following horizontal velocity profile of the downflow: Fig. 24. An observation image of the cutting plane at A-A in test 3-1.
vz (y ) =
ρg y (2h−y ). 2η
(4)
Thus, the maximum vertical velocity may be described as
vz (h) =
ρg 2 h. 2η
(5)
Eq. (5) means that the vertical velocity is proportional to the square of the melt thickness. In this study, the thicknesses are 0.28 mm in test 1 series, 0.5 mm in test 2 series and 1.0 mm in test 3 series. It may be calculated from Eq. (5) that the velocity in test with 1.0 mm thickness is 12.8 times faster than that of test with 0.28 mm thickness. In this analysis, the absorption was not considered. The time development of the melt absorption into the voids of B4C powder layer may be simply modeled with Lucas-Washburn equation (Washburn, 1921) as follows:
Wl =
Fig. 25. A temperature history during test 3-2. 109
rγ cosθt 2η
(6)
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 26. Time-resolved images of the relocated rod obtained in test 3-2 (droplet formation without collapse).
Fig. 28. A temperature history during test 3-3.
eutectic composition when we consider that the Fe-B phase diagram can illustrate the behavior of SS-B4C melts and that the maximum temperature of the tests is just 100 K higher that the eutectic temperature. Pore radius of the B4C powder layer are almost same as well because of the same way of filling the B4C powder in all tests. They mean Eq. (6) indicates that the absorption speed is almost same in all tests before the melt is completely absorbed. Therefore, it is considered that ratio of the thickness of the thick melt to that of the thin melt film absorbed into the B4C powder layer becomes larger when the absorption speed is same, thus it can be considered that the difference in the velocities among the
Fig. 27. An observation image on the part of the specimen in test 3-2. Inner surface of the solidified melt was observed.
where, t is the time for a liquid of dynamic viscosity η and surface tension γ to penetrate a distance Wl into the capillary whose pore radius is r. The contact angle between liquid and solid is θ. In all tests in this study, fluid characteristics of the melt are almost same because the melts formed during the tests should have Fe/B ratio close to that of the
110
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 29. Time-resolved images of the relocated rod obtained in test 3-3 (droplet formation without collapse).
5. Conclusions In this study, the effect of the thickness of the SS clad on the initial relocation behavior of a rod-shaped specimen containing B4C species was investigated for the first time via dynamic visualization. Experiments were performed at temperatures exceeding the eutectic temperature of the melt (1473 K). The relocation process was visualized using a camera installed at the specimen’s front. As a result, the following three relocation modes of the eutectic melt were observed: film formation, droplet formation without collapse, and droplet formation with collapse. During the droplet formation, the migration of B4C powder was detected on the melt surface. At the same time, no migration of B4C species occurred during the film formation. The observed modes suggest that the geometry of the control rod contribute to the diversity of its relocation behaviors. The possible mechanism for the different relocation modes observed under the utilized experimental conditions include: 1. At the beginning, a small quantity of the eutectic melt is formed between the SS and B4C powder. 2. The formed melt is absorbed by the voids of B4C powder layer and progressively dissolves the B4C while the slightly B4C enriched melt react with the remaining solid SS. Thus, the melt progresses in both directions to the solid SS and B4C powder. 3. The thickness of the melt is mainly affected by the thickness of the SS tube because boron concentration at the eutectic composition in Fe-B binary phase diagram is much smaller than the initial boron concentration on the current experimetal conditions.
Fig. 30. An observation image of the cutting plane at A-A in test 3-3.
tests with different SS thicknesses was enhanced by the influence of the absorption of the melt in the B4C powder layer. However, the change of the interface between the B4C powder and eutectic melt was not considered in above analytical consideration. Therefore, for the better understanding of the phenomena, the assist of detailed numerical calculation would be necessary in the future.
111
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 31. A trajectory of the non-reacted B4C particles observed in test 3-2.
Fig. 32. XRD profiles of the part of the specimen where the eutectic melt occurred in test 1-1.
Fig. 34. XRD profiles of the part of the specimen where the eutectic melt occurred in test 3-1.
4. A thin melt is not deeply absorbed into the B4C powder layer while A thick melt is deeply absorbed into the B4C powder layer causing the collapse of the specimen. 5. A thin melt film is formed around the B4C powder because a thin melt cannot flow by gravity while a droplet is formed because a thick melt can flow by gravity. The findings obtained in this study suggest that the geometry of the control rod may be one of the dominant factors affecting its relocation behavior. For the comprehension of the relocation behavior of the control blade during the Fukushima accident at the early stage, further experimental works are needed because current study was only conducted under argon atmosphere and at atmospheric pressure. The influence of the oxidation and pressure on the relocation behavior should be understood near future. Furthermore, a better understanding of the observed phenomena (and thus the state of the fuel debris after the accident) can be achieved by performing numerical calculations. In
Fig. 33. XRD profiles of the part of the specimen where the eutectic melt occurred in test 2-1.
112
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Fig. 35. A vertical cross-section of the specimen whose SS thickness is thin (test 1-1, 1-2 and 1-3). Eutectic melting proceeds from (a) to (c) as time passes.
Fig. 36. A vertical cross-section of the specimen whose SS thickness is thick (test 2-1, 2-2 and 2-3). Eutectic melting proceeds from (a) to (c) as time passes.
Acknowledgments Funding: This study was partially supported by JSPS KAKENHI Grant Number 17J07634 and “High-accuracy computations of the core structure relocation during severe accidents” project funded by the Initiatives for Atomic Energy Basic and Generic Strategic Research program of the Ministry of Education, Culture, Sports, Science and Technology of Japan. References Clément, B.N., Hanniet-Girault, N., Repetto, G., Jacquemain, D., Jones, A.V., Kissane, M.P., Von der Hardt, P., 2003. LWR severe accident simulation: synthesis of the results and interpretation of the first Phebus FP experiment FPT0. Nucl. Eng. Des. 226, 582. Furuya, M., Morooka, S., 2015. Visual Observation of Natural Convection due to Eutectic Melting of Stainless Steel and Boron Carbide and Three-Dimensional Raman Spectrometry. National Heat Transfer Symposium of Japan, Fukuoka, Japan [in Japanese]. Gauntt, R.O., Gasser, R.D., 1990. Results of the DF-4 BWR control blade-channel box test. In: Prc. Eighteenth Water Reactor Safety Information Meeting, NUREG/CP-0114, vol. 2, pp. 25–40. Gauntt, R.O., Gasser, R.D., Ott, L.J., 1989. The DF-4 Fuel Damage Experiment in ACRR with a BWR Control Blade and Channel Box. Sandia National Laboratory, USA, Report NUREG/CR-4671. Girault, N., Fiche, C., Bujan, A., Dienstbier, J., 2009. Towards a better understanding of iodine chemistry in RCS of nuclear reactors. Nucl. Eng. Des. 239, 1162–1170. Haste, T., Payot, F., Dominguez, C., March, P., Simondi-Teisseire, B., Steinbrück, M., 2012. Study of boron behavior in the primary circuit of water reactors under severe accident conditions: a comparison of Phebus FPT3 results with other recent integral and separate-effects data. Nucl. Eng. Des. 246, 147–156. Hofmann, P., Markiewicz, M.E., Spino, J.L., 1990. Reaction behavior of B4C absorber material with stainless steel and zircaloy in severe light water reactor accidents. Nucl. Technol. 90, 226–244. Ikeuchi, H., Piluso, P., Fouquart, P., Excoffier, E., David, C., Brackx, E., 2017. Study on the
Fig. 37. A cross-section of the melt flow on a vertical wall (B4C powder region).
particular, the mixing ratio between the molten fuel and absorbent materials must be determined for successful Fukushima decommissioning.
113
Nuclear Engineering and Design 333 (2018) 99–114
S. Ueda et al.
Shibata, H., Sakamoto, K., Ouchi, A., Kurata, M., 2015. Chemical interaction between granular B4C and 304L-type stainless steel materials used in BWRs in Japan. J. Nucl. Sci. Technol. 52, 1313–1317. Steinbrück, M., 2014. Influence of boron carbide on core degradation during severe accidents in LWRs. Ann. Nucl. Energy 64, 43–49. Takano, M., Nishi, T., Shirasu, N., 2014. Characterization of solidified melt among materials of UO2 fuel and B4C control blade. J. Nucl. Sci. Technol. 51, 859–875. Ueda, S., Madokoro, H., Jo, B., Kondo, M., Erkan, N., Okamoto, K., 2016. Time-resolved visualization technique for eutectic reaction between boron carbide and stainless steel at high temperatures. Mech. Eng. Lett. 2, 16–00056. Ueda, S., Madokoro, H., Jo, B., Kondo, M., Erkan, N., Okamoto, K., 2017. Dynamic visualization of eutectic reaction between boron carbide and stainless steel. J. Nucl. Sci. Technol. 54, 81–88. Washburn, E.W., 1921. The dynamics of capillary flow. Phys. Rev. 17, 273.
distribution of boron in the in-vessel fuel debris in conditions close to Fukushima Daiichi Nuclear Power Station Unit 2. The 8th European Review Meeting on Severe Accident Research (ERMSAR) 2017, Warsaw, Poland. Nagase, F., Gauntt, R.O., Naito, M., 2017. Overview and outcomes of the OECD/NEA benchmark study of the accident at the Fukushima Daiichi nuclear power station. Nucl. Technol. 196, 499–510. Nagase, F., Uetsuka, H., Otomo, T., 1997. Chemical interactions between B4C and stainless steel at high temperatures. J. Nucl. Mater. 245, 52–59. Repetto, G., De Luze, O., Seiler, N., Trambauer, K., Austregesilo, H., Birchley, J., Ederli, S., Lamy, J.S., Maliverney, B., Drath, T., Hollands, T., 2010. B4C oxidation modelling in severe accident codes: applications to PHEBUS and QUENCH experiments. Prog. Nucl. Energy 52, 37–45. Sepold, L., Schanz, G., Steinbrück, M., Stuckert, J., Miassoedov, A., Palagin, A., Veshchunov, M., 2006. Results of the QUENCH-09 experiment compared to QUENCH-07 with incorporation of B4C absorber. Nucl. Technol. 154, 107–116.
114