- Email: [email protected]
Joint resistance measurements of pancake and terminal joints for JT-60SA EF coils
Fusion Engineering and Design 88 (2013) 2773–2776
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Joint resistance measurements of pancake and terminal joints for JT-60SA EF coils Tetsuhiro Obana a,∗ , Kazuya Takahata a , Shinji Hamaguchi a , Toshiyuki Mito a , Shinsaku Imagawa a , Kaname Kizu b , Haruyuki Murakami b , Kiyoshi Yoshida b a b
National Institute for Fusion Science, 322-6 Oroshi, Toki, Gifu 509-5292, Japan Japan Atomic Energy Agency, 801-1 Mukoyama, Naka, Ibaraki 311-0193, Japan
h i g h l i g h t s • • • •
To evaluate the joint fabrication technology for the JT-60SA EF coils, joint resistance measurements were conducted with a joint sample. The joint sample was composed of pancake and terminal joints. The measurements demonstrated that both joints fulfilled the design requirement. Considering the measurements, the characteristics of both joints were investigated using an analytical model that represents the joints.
a r t i c l e
i n f o
Article history: Received 18 September 2012 Received in revised form 8 April 2013 Accepted 8 April 2013 Available online 17 May 2013 Keywords: Cable-in-conduit conductor JT-60SA NbTi Joint resistance Shake-hands lap joint
a b s t r a c t To evaluate the joint fabrication technology for the JT-60SA EF coils, joint resistance measurements were conducted using a sample consisting of pancake and terminal joints. Both joints are shake-hands lap joints composed of cable-in-conduit conductors and a pure copper saddle-shaped spacer. The measurements demonstrated that both joints fulfilled the design requirement. Considering these measurements, the characteristics of both joints were investigated using analytical models that represent the joints. The analyses indicated that the characteristics of the conductors used in the joints affect the characteristics of the joints. © 2013 Elsevier B.V. All rights reserved.
1. Introduction The magnet system in the JT-60 Super Advanced (JT-60SA) fusion experiment is composed of 18 toroidal field coils, 4 stacks of central solenoid (CS) coils, and 6 plasma equilibrium field (EF) coils [1,2]. The procurement of the CS and EF coils is being undertaken by the Japan Atomic Energy Agency. In the EF coils, two types of cable-in-conduit (CIC) conductors are utilized because of the difference of the maximum magnetic field in the coils. The EF coil for high field (EF-H) conductor is composed of NbTi strands, and the EF coil for low field (EF-L) conductor is composed of NbTi strands and copper wires. The joint between the EF conductors is a one box-type joint that is suitable for a NbTi joint without using a bonding plate of copper and stainless joint. A joint can be assembled using a 60Sn-40Pb solder and covered by a simple case
∗ Corresponding author. Tel.: +81 572 58 2137; fax: +81 572 58 2616. E-mail address: [email protected] (T. Obana). 0920-3796/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fusengdes.2013.04.022
with closure welding. The EF coil has a “pancake joint”, which is the joint between pancake coils, and a “terminal joint” which is the joint between the pancake coil and the current feeder [3]. The sample, including the pancake and terminal joints, was developed to evaluate the fabrication technology of the joints. Using the sample, joint resistance tests were conducted at the National Institute for Fusion Science (NIFS) test facility [4,5]. In this paper, the results of the joint resistance measurement are described, and the characteristics of the pancake and terminal joints are discussed. 2. Sample of the pancake and terminal joints Fig. 1 shows the configuration of the joint sample. It has a racket shape and is 300 mm in diameter at the circular section. The pancake joint is composed of the shake-hands lap joint between the EF-H coil conductors, and the terminal joint is composed of the shake-hands lap joint between the EF-H and EF-L coil conductors. The EF-H and EF-L coil conductors are CIC conductors equipped
2774
T. Obana et al. / Fusion Engineering and Design 88 (2013) 2773–2776
Fig. 1. Schematic view of the sample. V2–V9 indicate the position of the voltage taps.
is 160 mm in the longitudinal direction. This length is the same as the final pitch of the conductors. The conductors are compacted in the joint, and the void fraction of the conductors is 25%. There is a liquid helium (LHe) inlet located at the circular section of the joint sample. 3. Experimental setup
Fig. 2. Cross-section of the EF-H and EF-L coil conductors.
with a central spiral. The EF-H conductor’s cable is composed of only 450 NbTi strands, and the EF-L conductor’s cable is composed of 216 NbTi strands and 108 copper wires. The NbTi strands are plated with Ni. Fig. 2 shows the cross-sectional view of the EF-H and EF-L coil conductors. Specifications for the conductors are described in Refs. [1,2]. As shown in Fig. 3, a saddle-shaped spacer of pure copper (C1100) is located between the conductors in the joints of the sample, removing a conduit and the Ni plating of the conductor surface. To reduce AC loss, the saddle-shaped spacer is divided into seven sections in the longitudinal direction using 1-mm-thick polyimide sheets. The spacer and conductors are electrically connected with a 60Sn-40Pb solder and clamped with SUS304. Additionally, the central spiral is replaced with a stainless tube. The connected length
Fig. 3. Cross-section of the sample at the joint.
The joint resistance of the sample was tested at the superconducting test facility of NIFS. The test facility can accommodate the testing of superconductors cooled by LHe, under an external field generated by a superconducting split coil. Details of the test facility are described in Refs. [4,5]. The joint sample was installed into the gap of the split coil so as to fit the center of the joint sample with that of the split coil. The joint sample was subsequently immersed in LHe. As illustrated in Fig. 1, the joint sample was equipped with voltage taps attached to the conduit. To measure the joint resistance properly, two pairs of the voltage taps were used for each joint such as the pairs (V2–V5, V3–V4) for the pancake joint and those (V6–V9, V7–V8) for the terminal joint. 4. Measurements Electrical resistances were measured at the joints of the sample. For these measurements, the sample was energized to 20 kA at a ramp rate of 100 A/s, and was maintained for 500 s. The sample was then degaussed.
Fig. 4. Measurement result of the joint resistance.
T. Obana et al. / Fusion Engineering and Design 88 (2013) 2773–2776
2775
Fig. 7. Analytical model for the joint resistance between voltage taps.
From the crossing to the degauss, the decay time constants of the pancake and terminal joints were 49 s and 62 s, respectively. 5. Discussion Fig. 5. Voltage profiles of the joint during the measurement without the external field.
Fig. 4 shows the resistances for the pancake and terminal joints. The resistances of both joints are proportional to the external field strength, and the resistance of the pancake joint is slightly higher than that of the terminal joint. The resistances of the pancake and terminal joints are 1.85 n and 2.1 n respectively, at the external field of 3 T. This fulfilled the 5 n at 3 T design requirement. Figs. 5 and 6 show the voltage profiles generated by both joints at the external fields of 0 T and 3 T, respectively. At the external field of 0 T, the voltage profiles of both joints were consistent during excitation. After the end of excitation, both profiles decreased rapidly from 38 V to 30 V. Then, the difference between the voltage profiles occurred. Until the sample was degaussed, the voltage of the pancake joint decreased significantly, in comparison with that of the terminal joint. During this period, the decay time constants () of each joint were as follows: was 54 s for the pancake joint and 71 s for the terminal joint. When the external field was 3 T, the voltage of the pancake joint was higher than that of the terminal joint during excitation. At the end of excitation, the difference in the voltages reached approximately 10 V. Subsequently, both voltage profiles crossed at 43 V, and the voltage of the pancake joint was lower than that of the terminal joint until degaussing occurred.
5.1. Voltage profiles of the pancake and terminal joints Operating current of the EF coils is significantly changed to provide the position equilibrium of plasma current and the plasma vertical stability. Hence, understanding the characteristics of the joints under transitional condition is necessary. As shown in Figs. 5 and 6, the voltage profiles of both joints were changed by the external field. This phenomenon is related to the magnetic resistances of the saddle-shaped spacer, copper wires, and solder. With the increase from 0 T to 3 T, the magnetic resistivity of the saddle-shaped spacer and copper wires increased approximately three-fold. The solder maintains a superconducting state blow approximately 0.1 T, and some parts of the solder in the joint will be in a superconducting state under the external field of 0 T, even if the self-field is considered. Under the same external field, a difference between voltage profiles at each joint occurred. This is because the makeup of the conductors is different at each joint; the pancake joint is composed of the EF-H conductors using only non-plated NbTi strands, and the terminal joint is composed of the EF-H and EF-L conductors using NbTi strands and copper wires. 5.2. Estimation of contact resistance between the conductor and saddle-shaped spacer The contact resistance between the CIC conductor and saddleshaped spacer was estimated on the basis of the measurements and theoretical assumption. Fig. 7 illustrates the electric circuit between two voltage taps attached to the conduit. The electric resistances of the pancake and terminal joints are respectively described as follows: Rpancake = 2Rcontact@EF-H + Rsaddle Rterminal = Rcontact@EF-H + Rcontact@EF-L + Rsaddle where Rpancake and Rterminal are the resistances of the pancake and terminal joints, Rcontact@EF-H is the contact resistance between the EF-H conductor and spacer, Rcontact@EF-L is the contact resistance between the EF-L conductor and spacer, and Rsaddle is the resistance of the copper spacer. Rsaddle is given as follows [6]: Rsaddle =
Fig. 6. Voltage profiles of the joint during the measurement under the external field of 3 T.
( · ı) (a · l)
where is the resistivity, ı is the thickness, a is the width of the conductor being jointed, and l is the length. The Rsaddle values of each joint were assumed to be 0.56 n
2776
T. Obana et al. / Fusion Engineering and Design 88 (2013) 2773–2776
(=(0.189 nm·10.4 mm)/(21.8 mm·160 mm)) for the pancake joint, and 0.69 n (=(0.189 nm·12.65 mm)/(21.8 mm·160 mm)) for the terminal joint at the external field of 3 T. From the measurements at 3 T, Rpancake was 1.85 n and Rterminal was 2.1 n. As a result, Rcontact@EF-H and Rcontact@EF-L were estimated to be 0.65 n and 0.77 n, respectively. Provided that the assembly quality is equal at each joint, the contact area between NbTi strands and the saddleshaped spacer is considered to be the cause of the difference of the contact resistance. According to Refs. [1,2], the EF-H conductor is composed of 450 NbTi strands and the EF-L conductor is composed of 216 NbTi strands and 108 copper wires. The final twist pitch of both conductors is 160 mm, which is the connected length to the spacer. Relative to the contact area between the NbTi strands and the spacer, the EF-H conductor is larger than the EF-L conductor. Therefore, the contact resistance of the EF-H conductor will be lower than that of the EF-L conductor. 6. Conclusion To evaluate joint fabrication technology, resistance measurements were conducted using a sample consisting of pancake and terminal joints for the JT-60SA EF coils. Both joints fulfilled the design requirement of 5 n at the external field of 3 T. The electrical resistance of the pancake joint was slightly lower than that of the terminal joint. Analyses indicated that the characteristics of the conductors used in the joints affect those of the joints. The
presence or absence of copper wires in the conductor is one factor that determines the characteristics of the joints. Acknowledgments The authors would like to thank Mr. S. Moriuchi, Dr. T. Kobuchi, and Mr. H. Noguchi of the NIFS Department of Engineering and Technical Services for their technical support. The authors also thank Mr. K. Ueda of Taiyo Nippon Sanso Co. Ltd. and the operating staff of Hitachi Co. Ltd. for their technical support. References [1] K. Yoshida, K. Tsuchiya, K. Kizu, H. Murakami, K. Kamiya, T. Obana, et al., Development of JT-60SA superconducting magnet system, Physica C: Superconductivity and its Applications 470 (20) (2010) 1727–1733. [2] K. Kizu, Y. Kashiwa, H. Murakami, T. Obana, K. Takahata, K. Tsuchiya, et al., Fabrication and tests of EF conductors for JT-60SA, Fusion Engineering and Design 86 (2011) 1432–1435. [3] T. Takao, K. Nakamura, T. Takagi, N. Tanoue, H. Murakami, K. Yoshida, et al., Influence of external magnetic field AC loss at EF coil joints of JT-60SA, IEEE Transactions on Applied Superconductivity 22 (3) (2012) 4704604. [4] T. Obana, K. Takahata, S. Hamaguchi, N. Yanagi, T. Mito, S. Imagawa, et al., Upgrading the NIFS superconductor test facility for JT-60SA cable-in-conduit conductors, Fusion Engineering and Design 84 (2009) 1442–1445. [5] J. Yamamoto, T. Mito, K. Takahata, S. Yamada, N. Yanagi, I. Ohtake, et al., Superconducting test facility of NIFS for the Large Helical Device, Fusion Engineering and Design 20 (1993) 147–151. [6] Y. Iwasa, Case Studies in Superconducting Magnets – Design and Operational Issues, 2nd ed., Springer, New York, NY, USA, 2009.
Contents lists available at ScienceDirect
Fusion Engineering and Design journal homepage: www.elsevier.com/locate/fusengdes
Joint resistance measurements of pancake and terminal joints for JT-60SA EF coils Tetsuhiro Obana a,∗ , Kazuya Takahata a , Shinji Hamaguchi a , Toshiyuki Mito a , Shinsaku Imagawa a , Kaname Kizu b , Haruyuki Murakami b , Kiyoshi Yoshida b a b
National Institute for Fusion Science, 322-6 Oroshi, Toki, Gifu 509-5292, Japan Japan Atomic Energy Agency, 801-1 Mukoyama, Naka, Ibaraki 311-0193, Japan
h i g h l i g h t s • • • •
To evaluate the joint fabrication technology for the JT-60SA EF coils, joint resistance measurements were conducted with a joint sample. The joint sample was composed of pancake and terminal joints. The measurements demonstrated that both joints fulfilled the design requirement. Considering the measurements, the characteristics of both joints were investigated using an analytical model that represents the joints.
a r t i c l e
i n f o
Article history: Received 18 September 2012 Received in revised form 8 April 2013 Accepted 8 April 2013 Available online 17 May 2013 Keywords: Cable-in-conduit conductor JT-60SA NbTi Joint resistance Shake-hands lap joint
a b s t r a c t To evaluate the joint fabrication technology for the JT-60SA EF coils, joint resistance measurements were conducted using a sample consisting of pancake and terminal joints. Both joints are shake-hands lap joints composed of cable-in-conduit conductors and a pure copper saddle-shaped spacer. The measurements demonstrated that both joints fulfilled the design requirement. Considering these measurements, the characteristics of both joints were investigated using analytical models that represent the joints. The analyses indicated that the characteristics of the conductors used in the joints affect the characteristics of the joints. © 2013 Elsevier B.V. All rights reserved.
1. Introduction The magnet system in the JT-60 Super Advanced (JT-60SA) fusion experiment is composed of 18 toroidal field coils, 4 stacks of central solenoid (CS) coils, and 6 plasma equilibrium field (EF) coils [1,2]. The procurement of the CS and EF coils is being undertaken by the Japan Atomic Energy Agency. In the EF coils, two types of cable-in-conduit (CIC) conductors are utilized because of the difference of the maximum magnetic field in the coils. The EF coil for high field (EF-H) conductor is composed of NbTi strands, and the EF coil for low field (EF-L) conductor is composed of NbTi strands and copper wires. The joint between the EF conductors is a one box-type joint that is suitable for a NbTi joint without using a bonding plate of copper and stainless joint. A joint can be assembled using a 60Sn-40Pb solder and covered by a simple case
∗ Corresponding author. Tel.: +81 572 58 2137; fax: +81 572 58 2616. E-mail address: [email protected] (T. Obana). 0920-3796/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fusengdes.2013.04.022
with closure welding. The EF coil has a “pancake joint”, which is the joint between pancake coils, and a “terminal joint” which is the joint between the pancake coil and the current feeder [3]. The sample, including the pancake and terminal joints, was developed to evaluate the fabrication technology of the joints. Using the sample, joint resistance tests were conducted at the National Institute for Fusion Science (NIFS) test facility [4,5]. In this paper, the results of the joint resistance measurement are described, and the characteristics of the pancake and terminal joints are discussed. 2. Sample of the pancake and terminal joints Fig. 1 shows the configuration of the joint sample. It has a racket shape and is 300 mm in diameter at the circular section. The pancake joint is composed of the shake-hands lap joint between the EF-H coil conductors, and the terminal joint is composed of the shake-hands lap joint between the EF-H and EF-L coil conductors. The EF-H and EF-L coil conductors are CIC conductors equipped
2774
T. Obana et al. / Fusion Engineering and Design 88 (2013) 2773–2776
Fig. 1. Schematic view of the sample. V2–V9 indicate the position of the voltage taps.
is 160 mm in the longitudinal direction. This length is the same as the final pitch of the conductors. The conductors are compacted in the joint, and the void fraction of the conductors is 25%. There is a liquid helium (LHe) inlet located at the circular section of the joint sample. 3. Experimental setup
Fig. 2. Cross-section of the EF-H and EF-L coil conductors.
with a central spiral. The EF-H conductor’s cable is composed of only 450 NbTi strands, and the EF-L conductor’s cable is composed of 216 NbTi strands and 108 copper wires. The NbTi strands are plated with Ni. Fig. 2 shows the cross-sectional view of the EF-H and EF-L coil conductors. Specifications for the conductors are described in Refs. [1,2]. As shown in Fig. 3, a saddle-shaped spacer of pure copper (C1100) is located between the conductors in the joints of the sample, removing a conduit and the Ni plating of the conductor surface. To reduce AC loss, the saddle-shaped spacer is divided into seven sections in the longitudinal direction using 1-mm-thick polyimide sheets. The spacer and conductors are electrically connected with a 60Sn-40Pb solder and clamped with SUS304. Additionally, the central spiral is replaced with a stainless tube. The connected length
Fig. 3. Cross-section of the sample at the joint.
The joint resistance of the sample was tested at the superconducting test facility of NIFS. The test facility can accommodate the testing of superconductors cooled by LHe, under an external field generated by a superconducting split coil. Details of the test facility are described in Refs. [4,5]. The joint sample was installed into the gap of the split coil so as to fit the center of the joint sample with that of the split coil. The joint sample was subsequently immersed in LHe. As illustrated in Fig. 1, the joint sample was equipped with voltage taps attached to the conduit. To measure the joint resistance properly, two pairs of the voltage taps were used for each joint such as the pairs (V2–V5, V3–V4) for the pancake joint and those (V6–V9, V7–V8) for the terminal joint. 4. Measurements Electrical resistances were measured at the joints of the sample. For these measurements, the sample was energized to 20 kA at a ramp rate of 100 A/s, and was maintained for 500 s. The sample was then degaussed.
Fig. 4. Measurement result of the joint resistance.
T. Obana et al. / Fusion Engineering and Design 88 (2013) 2773–2776
2775
Fig. 7. Analytical model for the joint resistance between voltage taps.
From the crossing to the degauss, the decay time constants of the pancake and terminal joints were 49 s and 62 s, respectively. 5. Discussion Fig. 5. Voltage profiles of the joint during the measurement without the external field.
Fig. 4 shows the resistances for the pancake and terminal joints. The resistances of both joints are proportional to the external field strength, and the resistance of the pancake joint is slightly higher than that of the terminal joint. The resistances of the pancake and terminal joints are 1.85 n and 2.1 n respectively, at the external field of 3 T. This fulfilled the 5 n at 3 T design requirement. Figs. 5 and 6 show the voltage profiles generated by both joints at the external fields of 0 T and 3 T, respectively. At the external field of 0 T, the voltage profiles of both joints were consistent during excitation. After the end of excitation, both profiles decreased rapidly from 38 V to 30 V. Then, the difference between the voltage profiles occurred. Until the sample was degaussed, the voltage of the pancake joint decreased significantly, in comparison with that of the terminal joint. During this period, the decay time constants () of each joint were as follows: was 54 s for the pancake joint and 71 s for the terminal joint. When the external field was 3 T, the voltage of the pancake joint was higher than that of the terminal joint during excitation. At the end of excitation, the difference in the voltages reached approximately 10 V. Subsequently, both voltage profiles crossed at 43 V, and the voltage of the pancake joint was lower than that of the terminal joint until degaussing occurred.
5.1. Voltage profiles of the pancake and terminal joints Operating current of the EF coils is significantly changed to provide the position equilibrium of plasma current and the plasma vertical stability. Hence, understanding the characteristics of the joints under transitional condition is necessary. As shown in Figs. 5 and 6, the voltage profiles of both joints were changed by the external field. This phenomenon is related to the magnetic resistances of the saddle-shaped spacer, copper wires, and solder. With the increase from 0 T to 3 T, the magnetic resistivity of the saddle-shaped spacer and copper wires increased approximately three-fold. The solder maintains a superconducting state blow approximately 0.1 T, and some parts of the solder in the joint will be in a superconducting state under the external field of 0 T, even if the self-field is considered. Under the same external field, a difference between voltage profiles at each joint occurred. This is because the makeup of the conductors is different at each joint; the pancake joint is composed of the EF-H conductors using only non-plated NbTi strands, and the terminal joint is composed of the EF-H and EF-L conductors using NbTi strands and copper wires. 5.2. Estimation of contact resistance between the conductor and saddle-shaped spacer The contact resistance between the CIC conductor and saddleshaped spacer was estimated on the basis of the measurements and theoretical assumption. Fig. 7 illustrates the electric circuit between two voltage taps attached to the conduit. The electric resistances of the pancake and terminal joints are respectively described as follows: Rpancake = 2Rcontact@EF-H + Rsaddle Rterminal = Rcontact@EF-H + Rcontact@EF-L + Rsaddle where Rpancake and Rterminal are the resistances of the pancake and terminal joints, Rcontact@EF-H is the contact resistance between the EF-H conductor and spacer, Rcontact@EF-L is the contact resistance between the EF-L conductor and spacer, and Rsaddle is the resistance of the copper spacer. Rsaddle is given as follows [6]: Rsaddle =
Fig. 6. Voltage profiles of the joint during the measurement under the external field of 3 T.
( · ı) (a · l)
where is the resistivity, ı is the thickness, a is the width of the conductor being jointed, and l is the length. The Rsaddle values of each joint were assumed to be 0.56 n
2776
T. Obana et al. / Fusion Engineering and Design 88 (2013) 2773–2776
(=(0.189 nm·10.4 mm)/(21.8 mm·160 mm)) for the pancake joint, and 0.69 n (=(0.189 nm·12.65 mm)/(21.8 mm·160 mm)) for the terminal joint at the external field of 3 T. From the measurements at 3 T, Rpancake was 1.85 n and Rterminal was 2.1 n. As a result, Rcontact@EF-H and Rcontact@EF-L were estimated to be 0.65 n and 0.77 n, respectively. Provided that the assembly quality is equal at each joint, the contact area between NbTi strands and the saddleshaped spacer is considered to be the cause of the difference of the contact resistance. According to Refs. [1,2], the EF-H conductor is composed of 450 NbTi strands and the EF-L conductor is composed of 216 NbTi strands and 108 copper wires. The final twist pitch of both conductors is 160 mm, which is the connected length to the spacer. Relative to the contact area between the NbTi strands and the spacer, the EF-H conductor is larger than the EF-L conductor. Therefore, the contact resistance of the EF-H conductor will be lower than that of the EF-L conductor. 6. Conclusion To evaluate joint fabrication technology, resistance measurements were conducted using a sample consisting of pancake and terminal joints for the JT-60SA EF coils. Both joints fulfilled the design requirement of 5 n at the external field of 3 T. The electrical resistance of the pancake joint was slightly lower than that of the terminal joint. Analyses indicated that the characteristics of the conductors used in the joints affect those of the joints. The
presence or absence of copper wires in the conductor is one factor that determines the characteristics of the joints. Acknowledgments The authors would like to thank Mr. S. Moriuchi, Dr. T. Kobuchi, and Mr. H. Noguchi of the NIFS Department of Engineering and Technical Services for their technical support. The authors also thank Mr. K. Ueda of Taiyo Nippon Sanso Co. Ltd. and the operating staff of Hitachi Co. Ltd. for their technical support. References [1] K. Yoshida, K. Tsuchiya, K. Kizu, H. Murakami, K. Kamiya, T. Obana, et al., Development of JT-60SA superconducting magnet system, Physica C: Superconductivity and its Applications 470 (20) (2010) 1727–1733. [2] K. Kizu, Y. Kashiwa, H. Murakami, T. Obana, K. Takahata, K. Tsuchiya, et al., Fabrication and tests of EF conductors for JT-60SA, Fusion Engineering and Design 86 (2011) 1432–1435. [3] T. Takao, K. Nakamura, T. Takagi, N. Tanoue, H. Murakami, K. Yoshida, et al., Influence of external magnetic field AC loss at EF coil joints of JT-60SA, IEEE Transactions on Applied Superconductivity 22 (3) (2012) 4704604. [4] T. Obana, K. Takahata, S. Hamaguchi, N. Yanagi, T. Mito, S. Imagawa, et al., Upgrading the NIFS superconductor test facility for JT-60SA cable-in-conduit conductors, Fusion Engineering and Design 84 (2009) 1442–1445. [5] J. Yamamoto, T. Mito, K. Takahata, S. Yamada, N. Yanagi, I. Ohtake, et al., Superconducting test facility of NIFS for the Large Helical Device, Fusion Engineering and Design 20 (1993) 147–151. [6] Y. Iwasa, Case Studies in Superconducting Magnets – Design and Operational Issues, 2nd ed., Springer, New York, NY, USA, 2009.