Thermal Stability of Yttrium Based Superconducting Coil for Accelerator Application

Transcription

1 Thermal Stability of Yttrium Based Superconducting Coil for Accelerator Application Xudong Wang, 1 Kiyosumi Tsuchiya, 1 Shinji Fujita, Shogo Muto, Masanori Daibo, 3 Yasuhiro Iijima and Kunihiro Naoe Yttrium(Y)-based coated conductors are expected for miniaturization and higher performance of superconducting devices by applying to the coils, because of their high critical current characteristics in a high magnetic fields and relatively high temperatures above K. As an application study of Y-based superconducting coils for a sextupole magnet considered by the High Energy Accelerator Research Organization (KEK), we evaluated the thermal stability of the Y-based superconducting coil. 1. Introduction Superconductivity is a phenomenon that the electrical resistance of a substance becomes zero when cooled below a certain temperature. The substance with this kind of property is called superconductor. The temperature of the zero electrical resistance is called the critical temperature (Tc). Based on Tc, superconductors are broadly divided into two groups; low critical temperature superconductors (LTS) and high critical temperature superconductors (HTS). Materials of the LTS, such as NbTi and Nb3Sn, are industrialized and the LTS wires are applied to medical magnetic resonance imaging (MRI) magnets and other purposes. Materials of the HTS include bismuth (Bi)- based superconductors (BiSrCaCuO8+d (Bi1) and BiSrCaCu3O1+d (Bi3)) and Yttrium (Y)- based superconductors (REBaCu3O7-d, RE; rare earth elements such as Y and Gd). These HTS materials are actively developed and commercially available as wires. The LTS wires are usually cooled by costly liquid helium (temperature of 4 K = -69 C) because of its low Tc. On the other hand, the HTS materials can be cooled by costless liquid nitrogen (temperature of 77 K = -196 C) or by using small refrigerators. Y- based coated conductors have been developed particularly for magnet coil applications by virtue of their high mechanical strength and high critical current (Ic) in a magnetic field 1)-3). We have developed Y-based coated conductors over many years, and now we can supply long wires having high performances 4). We have also conducted basic studies for coil applications 5)-7). Based on these studies, we succeeded in developing the world s largest Y-based high temperature superconducting magnet having stored 46 kj using our 1 Accelerator Laboratory of High Energy Accelerator Research Organization Energy Technology Research Department of Advanced Technology Laboratory 3 Superconductor Business Development Division of New Business Development Center Y-based coated conductors of about 7. km 8)-1). Important issues for stable operation of superconducting magnets are the quench detection and the quench protection. In case of a LTS magnet, the quench will be triggered by even small energy of the order of several mj. One of the reasons is that the specific heat of the LTS magnet is extremely small because of its low operating temperature. Another reason is its small difference between the magnet s Tc and operating temperature (temperature margin). When the magnet quenches, a large voltage would appear in the coil because the normal zone propagation velocity (NZPV) in the coil is fast (from several dozens to several hundreds of meters per second). Therefore, the detection of quench is relatively easy. On the other hand, the HTS magnet would be operated at a higher temperature. The specific heat of the HTS magnet is two to three orders of magnitude larger than that of the LTS magnet and its temperature margin is also larger than that of the LTS magnet. Therefore, the HTS magnet has extremely high thermal stability. However, quench detection of the HTS magnet is difficult because the NZPV of the magnet is extraordinarily slow (from several to several dozens of milimeters per second) and a generated voltage due to the normal transition is small. Once the HTS magnet quenches, its temperature is increased locally (called hotspot), resulting in a local break of the coil 11). The High Energy Accelerator Research Organization (KEK) is currently constructing an electron-positron colliding accelerator, called SuperKEKB. It is considering to install the magnets using Y-based superconducting coils for a part of the accelerator 1)-13). The reason for the consideration is that this application will bring about advantages in terms of economic efficiency by not using liquid helium, and high stability against the beam loss in the accelerator. Development of a quench detection and a protection method is the 1

2 Panel 1. Abbreviations, Acronyms, and Terms. Accelerator A device that accelerates electrons, protons, and other particles close to the speed of light to create a high energy state. KEK has a large circular accelerator (KEKB) with a circumference of 3 km. It uses this accelerator to study subatomic physics and other basic sciences by colliding electrons and positrons. KEKB demonstrated the Kobayashi- Masukawa theory, contributing Mr. Kobayashi and Mr. Masukawa earning the Nobel Prize in Physics in 8. Critical temperature, Tc Maximum temperature that can keep a superconducting state. Critical current, Ic Critical current (Ic) is the maximum current value that can flow in a superconducting state. This value depends on the temperatures and magnetic fields. Y-based coated conductor An oxide superconducting wire whose superconducting layer contains yttrium (Y), gadolinium (Gd), and other rare earth elements. Compared to the other HTSs, this conductor has high Ic in a magnetic field at a relatively high temperature of 3 K or higher. Quench A phenomenon causing thermal runaway in a coil that occurs when a part of the coil is changed to a normal conducting state (called normal transition) due to some reasons; such as increase of the temperature and/or magnetic instabilities, after which this normal conducting region propagates and spreads to the entire coil. Cryostat A vacuum vessel with a heat-insulating layer to keep coils and other components at an extremely low temperature. n-value V-I characteristics of a superconducting wire around the critical current are expressed as V=Vc(l/lc) n (Vc: Voltage criterion for critical current). This expression is referred to as an n-value model and the index is called n-value. This index is apparently decreased if a part of the superconducting wire is degraded. For this reason, this model is used as an index of soundness when manufacturing coils. essential issue for the real applications of Y-based superconducting coils. Although several studies on the quench of Y-based coated conductors and Y-based superconducting coils have been reported 14)-16), the quench behavior depends on the structure of the conductor and other factors. Therefore, we investigated the quench behavior of a small coil, which was fabricated by using our Y-based coated conductors 17)-18). The following reports the overview of this investigation result. Y. Special Sextupole Magnet for KEKB Accelerator A special sextupole magnet as shown in Fig. 1 is currently considered by KEK. It consists of six coils placed inside of the magnet ( normal coil ) and six coils located outside ( skew coil ). This small magnet will be used to correct the beam divergence due to momentum dispersion, which is referred to as chromaticity. A total of 16 magnets would be dispersedly installed in the region of 3 to 1 m away from the point where accelerated particles collide. The specifications of each coil are listed in Table 1. Table 1. Specifications of normal coils and skew coils. 5 1 Y 5 5X Z normal coil 5 skew coil Fig. 1. Schematic view of a sextpole magnet. Item Normal coil Dimensions Number of turns of conductor Inductance Stored energy Operating current Skew coil Dimensions Number of turns of conductor Inductance Stored energy Operating current Specifications mm 83.6 mm 9. mm 11 turns 6.6 mh. kj 57.6 A mm 94. mm 4.5 mm 43 turns 3.9 mh.1 kj 59.5 A Fujikura Technical Review, 16 13

3 3. Quench Behavior of Small Coil 3.1 Fabrication of Small Coil The dimensions and the structure of the Y-based coated conductor used for the small coil fabrication are shown in Fig.. The structure of the conductor is the same as that would be used for the special sextupole magnet. This conductor was wound around a GFRP bobbin with a diameter of 5 mm and impregnated with epoxy resin by a method called vacuum impregnation. The specifications of this coil are listed in Table. This coil consists of two layers of pancake coils. An NiCr heater was installed into the innermost turn of the first layer and voltage taps were attached on the coil around the heater (Fig. 3). The fabricated coil was installed in a cryostat and cooled by a small refrigerator. The temperature of the coil could be controlled at any temperatures by a heater installed inside. 3. Electric Circuit Used for Tests As illustrated in Fig. 4, a voltage detection system using a quench detector with a bridge circuit was ap- Insulation [Polyimide tape] 5 μm Stabilizer [Electroplate Cu] μm Protection Layer [Ag] μm Superconducting Layer [GdBa Cu 3 O x ] μm Buffer Layer [MgO, etc.] ~.7 μm Substrate [Hastelloy ] 75 μm Width 5 mm Fig.. Schematic view of Y-based Coated Conductor. Coil 1st Layer Coil nd Layer Heater (Innermost Turn) Voltage Taps (Copper foil) (a) (c) V1 (cm) V1 (3cm) V11 (cm) V6(1cm) V7(1cm) V5(1cm) V4(1cm) V3(1cm) V(1cm) V8(1cm) V1 (3cm) V9(1cm) NiCr Heater (4 8 mm) K 3rd Turn nd Turn Bobbin 1st Turn Coil 1st Layer (b) (splice between layers) Fig. 3. Schematic of a small test coil, (a)photograph (side view) (b)photograph (top view) (c)location of voltage taps and a heater. Table. Specifications of a small test coil. Trigger Item Specifications Inside diameter 5 mm Outside diameter 69.7 mm Height 11.7 mm Number of turns of conductor 5 Length of conductor to be used 1 m Critical current (77 K, self field (s. f.)) 11.1 A (1-7 V/cm criterion) n-value (77 K, s. f.) 3.8 ( V/cm) Power Supply Shunt Resistor Dump Resistor Superconducting Coil V Quench Detector Detect Voltage Vd Detect Time t d Fig. 4. Quench detect and protect circuit. 14

4 Voltage [mv] Quench detection Current shut down 5 3 transport current voltage of 18 1st-layer coil 1 heater voltage voltage of 1 nd-layer coil 6.7 sec.75 sec.8 sec.85 sec.9 sec time [sec] Current [A] Fig. 5. An example of quench by heater. plied to a quench detection and a protection circuit. This detection system is a basic system used most frequently for quench detection. It can detect a small change of the coil voltage during the ramping of the coil current by cancelling the inductive voltage in the bridge circuit. When the voltage becomes larger than a preset detection voltage (VQD) and the voltage elapses for a time longer than a preset time (tqd), the quench detector regards it as quench and opens a cutoff switch of the coil current. Then, the current is decreased with a time constant determined by a dump resistor and the coil inductance. 3.3 Test Methods The test coil was kept at a certain temperature with a constant transport current (It; the ratio of It to Ic of the coil, It/Ic is called load factor). The first layer of the coil was forcibly induced a quench by a heater with gradually increasing output time (Fig. 5). Fig. 6 shows an example of the voltage generated at each voltage tap on the coil during the quenching. A voltage (normal conduction region) was first appeared at the part where the heater was installed (hotspot), and then the region was spread. NZPV was calculated from a time difference between the generation of mv voltage in a voltage tap section and that in the adjacent section. The hotspot temperature was estimated by converting from the generated voltage using electrical resistancetemperature characteristics of the conductor measured beforehand (Fig. 7). Voltage [mv] V5 6 6 time delay V6 NZPV 4 V4 4 Heater other taps time [sec] V5 Resistance [W] Fig. 6. An example of quench voltage generation Coil Temperature [K] Fig. 7. Temperature dependence of electrical resistance at V5 tap. Heater Power [W] 3.4 Test Results 5 The test was performed at a transport coil current whose load factor was from.3 to.6 and in the temperature range of K to 5 K, which was the expected range when operating the magnet. As shown in Fig. 8, the energy required to induce a quench was from 1 to 5 J, which was extraordinarily larger than that of the LTS coil. Fig. 9 shows the measured NZPV. The velocity range of to 8 mm/s, which is very slow and nearly the same to the reported value 14). Fig. 1 indicates the hotspot temperature of the coil under each Quench Energy [J] / I t I C Fig. 8. Quench energy at 5~ K. 5 K 4 K 3 K K Fujikura Technical Review, 16 15

5 quench detection condition. It reveals that the temperature is largely dependent not on the initial coil temperature but also on the coil current, and that the current dependence varies with the detection time. In the range of VQD less than 1 mv and tqd less than 1 ms, the hotspot temperature was 5 K or less. Since the characteristics of the coil were not degraded after the test, it is confirmed that the quench detection and protection system is useful in this range. 4. Numerical Analysis 4.1 Conditions for Analysis Considering that the voltage was generated only in the innermost turn of the first layer coil during the experiment, the first turn and its surroundings of the first layer coil shown in Fig. 11 were modelled and Equations (1) and () were used to formulate the modelled part by the finite element method (FEM). In order to express superconductive properties, n-value model was adopted. The Ic and the n-value actually measured were used. Assuming that refrigerator cooling has little effect for several seconds during the quench, the surface of the model was regarded as an adiabatic boundary. In Equation (1), f is a scalar potential and s is the electric conductivity of the stabilizing layer or superconducting layer. In Equation (), T is temperature, C is heat capacity, k is heat conductivity, Qj is Joule heat generation calculated from Equation (1), and Qheater is heat input to the heater. The temperature dependence of each physical property value was also considered. C T t s ( f) = (1) = ( k T) + Qj + Qheater () 4. Analysis Results Fig. 1 shows the results of the experiment and analysis when the coil temperature was 5 K and Ic was 175 A (load factor of.5). In both of the results, NZPV [mm/s] / I t I C 5 K 4 K 3 K K V1 V3 V4 V V1 V1 V11 V5 V6 V7 V8 V9 3rd Turn nd Turn 1st Turn Fig. 9. NZPV at 5~ K. Symmetrical thermal boundary Simulation model 3 5 K Fig. 11. Schematic of simulation model. Hotspot Temperature [K] K 3 K K 1 mv 4 mv V QD Coil Current [A] 1 ms 5 ms 1 ms 1 ms 5 ms 1 ms t QD Voltge [mv], Temperature [K] Doted line : Experiment Solid line : Simulation Heater.5 Vcoil V5 V6.5 V Time [sec] T5 1 Heaer Power [W] Fig. 1. Hotspot temperature of the coil after quenching for certain detection conditions. Fig. 1. Experimental and simulation results of quench test at temperature of 5 K and with coil current of 175 A. 16

6 the voltage was increased first in the voltage section V5 where the heater was installed, and then this increase propagated into the adjacent voltage section (V6). When the voltage of the entire coil (Vcoill) reached 1 mv, the temperature of V5 (T5) was about 145 K. Fig. 13 shows the analysis result of the hotspot temperature with respect to VQD of 4, 6, and 1 mv and tqd of 1, 5, and 1 ms at the coil temperatures of 3 K, 4 K, and 5 K. The analyzed trend and absolute value agree well with the experiment result illustrated in Fig. 1. The temperature of the hotspot is increased in proportion to VQD and tqd. In the case of tqd of 1 ms, the temperature is decreased with the increased operating current. The reason of this is that the hotspot temperature when the voltage reaches VQD is low with the larger transport current. The experiment and analysis revealed that the hotspot temperature during the quench was independent of the operating temperature and dependent on VQD, tqd and the operating current. Analysis also showed that the generation of the voltage and the hotspot temperature during the quench could be predicted. Hotspot Temperature [K] K 4 K 3 K K 1 mv 4 mv V QD 1 ms 5 ms 1 ms 1 ms 5 ms 1 ms t QD Coil Current [A] Fig. 13. Simulation results of hotspot temperature of the coil after quenching for certain detection conditions. 5. Conclusion As a first step to study the applicability of the Y- based superconducting coil to the special sextupole magnet currently considered by KEK, a small test coil was fabricated using our Y-based coated conductor and investigated the quench behavior of the coil. This study revealed that the energy required for quenching the coil was several J, which was far larger than that for the LTS coil, and NZPV was very slow. The hotspot temperature of the coil during the quench was lower than the room temperature (3 K) under some quench detection conditions; the coil was not degraded at the temperature. Numerical analysis showed that the simulation results agreed well with the experimental results. This analysis showed that the generation of the voltage and the hotspot temperature in the coil during quenching could be predicted and quench detection and protection could be used for small coils. Next step, we will fabricate prototype coils having the size of the actual magnet, and investigate the quench characteristics and the field quality of the coils in order to confirm the feasibility of the special sextupole magnet. In parallel to this study, we will proceed with the development of Y-based superconducting coils for the various application fields. Acknowledgment This study was supported by JSPS KAKENHI Grant Number 15H3667 from the Japan Society for the Promotion of Science (JSPS). References 1) S. Fujita et al.: Evaluation of Rare-earth-based Coated Conductors: Mechanical, Delamination and In-field Critical Current Properties, TEION KOGAKU, Vol. 48, No. 4, pp , 13 ) S. Fujita et al.: In-field critical current property of IBAD/ PLD coated conductors, J. Phys.: Conf. Ser. vol. 57, 7, 14 3) S. Fujita et al.: Characteristics of REBCO coated conductors for 5T cryogen-free superconducting magnet, IEEE Trans. Appl. Supercond. Vol.5, No.3, 8434, 15 4) Y. Iijima, et al.: Development for mass production of homogeneous RE13 coated conductors by hot-wall PLD process on IBAD template technique, IEEE Trans. Appl. Supercond., vol. 5, no, 3, 66414, 15 5) M. Daibo, et al.: Characteristics of cryocooled racetrack magnet fabricated using REBCO coated conductor, Physica C 471, pp , 11 6) M. Daibo, et al.: Characteristics of Impregnated Pancake Coils Fabricated using REBCO Coated Conductors, IEEE Trans. Appl. Supercond. Vol., No.3, 394, 1 7) M. Daibo, et al: Evaluation of thermal stability of conduction-cooled REBCO coil with.3-mm-thick stabilizer, Proceedings of ICEC 4-ICMC 1, p.57-51, 13 8) M. Daibo, et al: Development of a 5 T Rare-earth-based High Temperature Superconducting Magnet with a -cmdiameter Room Temperature Bore, TEION KOGAKU, Vol. 48, No. 5, pp. 6-3, 13 9) M. Daibo, et al.: World s Largest 5 T Yttrium-based High Temperature Superconducting Magnet with a -cm-diameter Room Temperature Bore, Fujikura Technical Review, No. 14, pp.37-45, 13 1) M. Daibo, et al.: Evaluation of a 46 kj cryocooled magnet and a model magnet with REBCO coated conductors, IEEE Trans. Appl. Supercond. Vol. 4, N. 3, June, 14 11) A. Ishiyama, et al.: Stability Criterion and Quench Detection/Protection Methods for HTS Coils Cryocooler-cooled YBCO Coil for SMES, TEION KOGAKU, Vol. 48, No. 4, pp , 13 Fujikura Technical Review, 16 17

7 1) K. Tsuchiya, et al.: Development of HTS sextupole magnet for SuperKEKB interaction region, IEEE Trans. Appl. Supercond. Vol.6, No.4, 4194, ) K. Tsuchiya, et al.: Development of HTS accelerator magnet (4-1) Design of HTS chromaticity correction sextupole for SuperKEKB interaction region, 9th CSSJ Conference, Extended Abstract, p. 194, 15 14) X. Wang, et al.: Near-adiabatic quench experiments on short YBCu3O7-d coated conductors, J. Appl. Phys. 11, 5394, 7 15) M. Daibo, et al.: Evaluation of normal-zone propagation characteristics of REBCO coated conductor with laminated Cu tape, IEEE Trans. Appl. Supercond. Vol.1, No.3, pp , 11 16) M. Daibo, et al.: Study of quench behavior of REBCO impregnated pancake coil with a 75-µm-thick copper stabilizer under conduction-cooled conditions, Physics Procedia 67, pp , 15 17) S. Fujita et al.: Development of HTS accelerator magnet (4- ) Measurement of the quench characteristics of REBCO impregnated coils, 9th CSSJ Conference, Extended Abstract, p. 195, 15 18) X. Wang, et al.: Development of HTS accelerator magnet (4-3) Numerical analysis on quench characteristics of REBCO impregnated coil, 9th CSSJ Conference, Extended Abstract, p. 196, 15 18