IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 27, NO. 4, JUNE 2017
4003307
First Cold Powering Test of REBCO Roebel Wound
Coil for the EuCARD2 Future Magnet
Development Project
Glyn A. Kirby, Jeroen van Nugteren, H. Bajas, V. Benda, A. Ballarino, M. Bajko, Luca Bottura, K. Broekens,
M. Canale, A. Chiuchiolo, L. Gentini, N. Peray, Juan Carlos Perez, Gijs de Rijk, Adriaan Rijllart, Lucio Rossi,
J. Murtomaeki, Jacky Mazet, Francois-Olivier Pincot, Giovanni Volpini, Maria Durante, P. Fazilleau, Clement Lorin,
A. Stenvall, Wilfried Goldacker, Anna Kario, and Alexander Usoskin
(Invited Paper)
Abstract—EuCARD-2 is a project partly supported by FP7European Commission aiming at exploring accelerator magnet
technology for 20 T dipole operating field. The EuCARD-2 collaboration is liaising with similar programs for high field magnets
in the USA and Japan. EuCARD-2 focuses, through the workpackage 10 “Future magnets,” on the development of a 10 kA-class
superconducting, high current density cable suitable for accelerator magnets, for a 5 T stand-alone dipole of 40 mm bore and about
1 m length. After standalone testing, the magnet will possibly be inserted in a large bore background dipole, to be tested at a peak field
up to 18 T. This paper starts by reporting on a few of the highlight
simulations that demonstrate the progress made in predicting: dynamic current distribution and influence on field quality, complex
quench propagation between tapes, and minimum quench energy
in the multitape cable. The multiphysics output importantly helps
predicting quench signals and guides the development of the novel
early detection systems. Knowing current position within individual tapes of each cable we present stress distribution throughout
the coils. We report on the development of the mechanical component and assembly processes selected for Feather-M2 the 5 T
EuCARD2 magnet. We describe the CERN variable temperature
flowing helium cold gas test system. We describe the parallel integration of the FPGA early quench detection system, using pickup
Manuscript received August 31, 2016; accepted January 6, 2017. Date of
publication January 26, 2017; date of current version March 1, 2017. This
work was supported by the European Commission under the FP7 Research
Infrastructures project EuCARD-2, Grant agreement no. 312453.
G. A. Kirby, J. van Nugteren, H. Bajas, V. Benda, A. Ballarino, M. Bajko,
L. Bottura, K. Broekens, M. Canale, A. Chiuchiolo, L. Gentini, N. Peray, J. C.
Perez, G. de Rijk, A. Rijllart, L. Rossi, J. Murtomaeki, J. Mazet, and F.-O. Pincot are with the CERN, Geneva CH-1211, Switzerland (e-mail: Glyn.Kirby@
cern.ch; jeroen.van.nugteren@cern.ch; luca.bottura@cern.ch; Juan.Carlos.
Perez@cern.ch; Gijs.derijk@cern.ch; Adriaan.Rijllart@cern.ch; Lucio.Rossi@
cern.ch; Jacky.Mazet@cern.ch; Francois-Olivier.Pincot@cern.ch).
G. Volpini is with the Istituto Nazionale di Fisica Nucleare Milano and CERN,
Segrate 20090, Italy (e-mail: giovanni.volpini@mi.infn.it).
M. Durante, P. Fazilleau, and C. Lorin are with the CEA, Gif-surYvette 91191, France (e-mail: maria.durante@cea.fr; philippe.fazilleau@cea.fr;
clement.lorin@gmail.com).
A. Stenvall is with the Tampere University of Technology (TUT) Korkeakoulunkatu 10, FI-33720 Tampere, Finland (e-mail: antti.stenvall@tut.fi).
W. Goldacker and A. Kario are with the Karlsruhe Institute of Technology, Eggenstein-Leopoldshafen 76344, Germany (e-mail: wilfried.Goldacker@
kit.edu; anna.kario@kit.edu).
A. Usoskin is with the Bruker HTS GmbH, Alzenau 63755, Germany (e-mail:
alexander.usoskin@bruker.com).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TASC.2017.2653204
coils and temperature sensors, alongside the standard CERN magnet quench detection system using voltage taps. Finally we report
on the first cold tests of the REBCO 10 kA class Roebel subscale
coil named Feather-M0.
Index Terms—Superconducting magnets, accelerators magnets,
high-temperature-superconductors, EuCARD-2, future magnets,
HTS, Roebel cable, cryogenic systems, metal 3-D printing.
I. INTRODUCTION
OST-LARGE Hadron Collider (LHC) energy-frontier collider development has started. Triggered by the first conceptual study for a high energy LHC (HE-LHC), based on 20
T collider quality dipole magnets [1]–[3], CERN has launched
a collaboration aimed at exploring the use of high temperature
superconductor (HTS) (the only material viable for field in excess of 15-16 T) for accelerator magnets.
EuCARD2 program work package 10 [4] has chosen to develop Roebel cable based on REBCO tape and the first magnets
will be cold tested during 2016-2017. The EuCARD2 project is
achieving its main goal with the development, manufacture and
delivery of long lengths of multi strand HTS Roebel cable, now
under way and being wound into test magnet coils. The aligned
block accelerator type magnet named Feather-M2 [5] (see Fig. 1)
is under construction at CERN Geneva and an alternative, Cosine Theta coil design is under construction at CEA-Saclay Paris.
The detailed design for both magnets are complete and dummy
coil winding with Roebel cable has started for both magnets:
the 5 T magnet Feather-M2, and the Cosine-Theta magnet. Four
short cable lengths approximately 6 m have been delivered and
are planned to be used in a sequence of subscale coil windings
(called Feather-M0) which will be cold tested. The first length
of Robel cable was assembled with tape from Bruker (member
of the EuCARD2 consortium). This cable was wound into a coil.
Initial test results are presented later in this paper. Groundbreaking progress has been made in the areas of: cable production,
magnetic simulation, assembly techniques and processes.
P
II. REBCO CABLE & MAGNETS DESIGN SIMULATIONS
During the course of the EuCARD2 program from 2013 many
papers have been published which cover in detail the computer
1051-8223 © 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
4003307
IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 27, NO. 4, JUNE 2017
Fig. 4. Simulation of current redistribution between tapes in a cable leading to
a full cable quench, approximately 400 ms after the start of current redistribution.
Fig. 1. This is a rendered image of the full three-dimensional Feather-M2 coil
model. That was used during quench and field quality simulations. It contains
the full geometry of the 15 tapes that form the REBCO cable that is wound into
final coil.
Fig. 5. Simulation of quench voltages in multi strand quenching cable. Red
trace shows coil voltage, blue trace shows pick-up array voltage signals.
Fig. 2. Reaction time to 300 K as a function of magnet current and temperature
calculated for the performance expected in the first Bruker tapes assembled into
a Robel cable with the field in Feather-M0.4. A 3 turn coil.
Fig. 6.
Fig. 3.
Simulation showing thermal map of quenching Roebel cable.
code used for the HTS Roebel cable. Simulations and code
descriptions can be found in [6]–[8].
A. Quench Analysis With Simulated Signals
The time of the “pre-quench drift” from nominal conditions
to current sharing is very difficult to predict, because of complex
structure and of strong anisotropy. A quench starts by forcing the
edge of one tape over the critical surface. This spreads across
that tape (see Fig. 3). Current starts to redistribute between
Simulation of quench temperatures in multi strand quenching cable.
adjacent tapes. As the current exits the initial quenching tape,
we see that its temperature is gradually reducing while other
tapes in the cable begin to exceed the superconducting surface
(see Figs. 3 to 6 which refer to the same quenching event). The
voltages developed during this initial stage are extremely small,
approximately 1 mV (see Fig. 5 red trace). At this low voltage
level noise may obscure the quench signals. As the last tapes in
the cable quench, only then does the voltage grow rapidly at the
same time as the temperature, leaving little time to react before
a temperature run-away is initiated (see Figs. 5 and 6).
At high operating temperatures where the maximum available
current density is lower, we have more time to react we therefore
KIRBY et al.: FIRST COLD POWERING TEST OF REBCO ROEBEL WOUND COIL FOR THE EUCARD2 FUTURE MAGNET DEVELOPMENT
4003307
Fig. 7. Dynamic harmonics variation during ramping for Aligned Block design with cable oriented parallel to field lines, 10 units variation during two
current ramps from 0 to 8000 A and back to 0 A, in 13 T background field.
Fig. 9. Minimum quench energy calculations for a range of cables having
different numbers of tapes in the cable at 4.2 K, Bo = 17 T, á= 2◦ .
20 minutes similar to a LHC ramp profile. The light green areas
in coil cross section image contain virtually no current. The red
areas have high current density. The position of these currents
will impose the final field quality, which is plotted in Fig. 7.
The field variation is very low under, 10 units. This calculation
does not take into account iron saturation. The conclusion is that
the variation in field errors due to current distribution within the
coil appears to be controllable. The calculations for the Cosine
Theta (CT) design exhibits higher errors in the range of 40 units
due to the field angle at the magnet mid-plane, however much
better than initially expectations.
Fig. 8. (Left) cable current density looking at cable wide face. (Right) current
density in tapes of cable, looking at a cut through the left side of FeatherM2
coil.
designed a test plan that would start at higher temperatures (see
Fig. 2) and Section V, sub-section C, below.
The pre-quench drift should give an early indication that the
quench is starting, (see Figs. 4 to 6), despite the very low voltages, was the idea behind the advanced quench detection systems which try to use the temperature or fields generated during
the current redistribution within the cable. Temperature sensors
placed throughout the coil proved to be impractical. However
placing temperature sensors adjacent to the superconducting
joints may be effective. Pickup coils are extremely promising
(see Fig. 5 blue trace) and their design is described later in this
paper. The quenching tape, locally redistributes current within
the cable, this new current distribution can spread axially approximately 30 m along the cable. This leads to the possibility
that a set of compact pick-up coil arrays, may be able to detect
quenches up to 30 m from the quench origin.
B. Field Quality
Fig. 7 presents the dynamic field quality that was calculated
for the Feather-M2 coil, when it is ramped in a background field
of 13 T. In Fig. 8, we see the current distribution through the
cable and coil after the current has been ramped up to 8000 A in
C. Power Losses Due to Magnetization
The cable angle in relation to the field vector plays a strong
role in the magnetisation losses. The Aligned Block (AB) design dissipates approximately 2 orders-of-magnitude less power
dissipation than the (CT) design. The AB dissipates a maximum
total power of 0.25 W/m and Average 0.024 W/m. The CT dissipates a maximum total power of 24.4 W/m, with an average
of 8.37 W/m [9].
D. Minimum Quench Energy in HTS Cable
The minimum quench energy (MQE) for the HTS cable is
in the range of Joules, approximately 3 orders-of-magnitude
higher than the Nb3 Sn and Nb-Ti low temperature superconductors. When operating at lower temperature, 4 K, the temperature margin can be as high as 50 K. Therefore quenches
in an operating magnet may only occur due to: heating due to
cryogenic systems fail, catastrophic conductor failure through
fatigue, temperature rise due to resistive joints, beam loss heating, or over-current induced during a fast ramp down of the outsert magnet. Coil movement leading to frictional heating, which
have always been suspected for low temperature superconductor (LTS) quench initiation may not release sufficient energy to
induce a quench. Fig. 9, presents the minimum quench energies
for a selection of cables with varying numbers of tapes, starting
4003307
IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 27, NO. 4, JUNE 2017
Fig. 10. (Left) image of pressure vectors in each tape of each cable in coil.
(Right) horizontal pressure distribution in outermost single cable. Coil image
shows high-pressure >200 MPa at the edges of the tapes. This case is calculated
with a 17 T central field and perfect tape alignment.
with a single tape, moving up to the full 15 tape cable in the full
FeatherM2 coil, in 13 T and self-field.
E. Mechanical Analysis Coil Stress
Knowing the precise position of the current flowing in individual tapes, it is then possible to calculate with finite element
model (FEA) the build-up of stress through the coil pack. As
the current predominantly flows at the edge of the tapes in the
cable, the pressure generated due to the magnetic forces are also
applied at the cable edges (see Fig. 10). The pressure is then calculated with a sophisticated FEA model, taking into account the
full detail of impregnated glass sleeve around cable and sliding
faces between cable tapes/strands. This predicted stresses can
be locally as high as 200 MPa compressive when inside the 13 T
background field [10]. The University of Twente have just repeated their benchmark test, which applied a uniform pressure
during powering of a short length of cable. The cable was insulated with the selected Feather-M2 impregnation system. No
degradation up to 450 MPa was seen [11]. The stress build up
through the coil is diluted by shear stresses which are averaging
out the peak stress. This highlights the need for detailed cable pressure measurement for further investigation and possible
instrumentation during high field testing.
Fig. 11. Coil former produced with additive manufacture in 316L stainless
steel. (A) 7-part Set of 212 mm long printed former sections, (B) former assembly model, (C) hollowed section III mm wall thickness for unloaded volumes,
(D) laser welding test sample.
developed. The 1.3 m long former was too long to be printed
in one piece, so an assembly of seven sections was needed,
see Fig. 11.
The cost of the former is proportional to the volume of material. By hollowing out the former at non-loaded volumes the
cost was reduced by 38%. A magnet former wall thickness
of 2.5 to 3 mm is an optimum for this component 3D printing, build stability. The former design was optimised through
close collaboration with 3T in Newbury UK [12]. A Standard
Material specification for Stainless steel 316L yield stress is
260 MPa. It is interesting to note that the 3-D printed material
has a much improved yield stress of 470 - 530 MPa depending
on the build direction (vertical, horizontal, respectively). Laser
welding would provide minimal distortion, but was not compatible with the micro-structure of the 3D printed surface. So
before welding the contact surface must be machined.
C. Pickup Coil Array
Pickup coil arrays are placed on the magnet former. Then the
insulated Robel cable is wound on to the former which has been
treated with a mold release. Copper spaces pack out the volume
between the coil and the energy extraction copper rings allowing adjustment for a range of cable thicknesses. Instrumentation
is mounted. The coil assembly is then resin impregnated. Finally the two coils that make up the Feather-M2 magnet are
mounted inside a High-strength tube. Joints are soldered and
mechanically/electrically finished.
The idea behind the pickup coil is to detect the flux change as
one tape in the cable quenches and the current in that individual
tape moves to the adjacent tapes (see Fig. 5). The changing flux
position then induces a signal on the pickup coils. The pickup
coils are photographically printed onto both sides of a 50 µm
Kapton sheet. Each coil has 285 full turns. There are 6 coils
on the upper deck, and 24 coils on the lower deck. The copper
coil pick-up tracks are 5 microns thick, 100 microns wide. The
insulation space between tracks are 75 microns. The detailed
design layout is presented in Fig. 12.
Polarity of the coils are chosen to match the twist pitch of
the Robel cable. This amplifies the pickup signal. A flexible
flat printed circuit board is electrically connected to the pick-up
coil array to take the signal wires out of the magnet. During a
full quench, low mutual inductance with the main field and high
circuit impedance limits the pick-up coil currents.
B. 3D Printing Magnet Former
D. Joints
Some features of the magnet former are “impossible” to machine conventionally so an additive manufacture approach was
Two joint types were required: first to connect the 15 tape
Roebel cable to the cryostat current lead, and second, make the
III. “FEATHER-M2” THE FIVE TESLA MAGNET
CONSTRUCTION STATUS HIGHLIGHTS
A. Feather-M2 Magnet Design Key Features
KIRBY et al.: FIRST COLD POWERING TEST OF REBCO ROEBEL WOUND COIL FOR THE EUCARD2 FUTURE MAGNET DEVELOPMENT
4003307
Fig. 12. Pickup coil array design details. (A) Rectangular pickup coil,
(B) Exit flex-PCB trace, (C) full pickup coil array for one pole, (D) Printed
circuit board connections between the five sheets, (E) Central 12 coil set.
connection between two Roebel cables which is the connection between the two poles of Feather-M2. A controlled solder
procedure was performed using: low pressure (6 to 10 MPa), following the eutectic specific temperature profile, with short time
at maximum temperature. The removable connection needed
for the cryostat current lead used the LTS system. Were Roebel
cable is soldered into a 300 mm long block (which is the final
design length of the Roebel cable twist pitch), with a protective
copper cover. This block is then lightly clamped to the current
lead with an indium sheet. This allows the joint to be disassembled without delamination on the tapes in the cable. The 300 mm
joint design will need to be improved as demonstrated when the
joint voltage was exceeded in the powering test, descried later
in Section IV, sub-section C, later.
IV. FEATHER-M0.4: CABLE PERFORMANCE &
INSTRUMENTATION
A. First REBCO Roebel Cable in Coil.
The 12 mm wide Bruker tape was fine-blanked from a tape
width of 5.5 mm and then silver coated, then copper coated
with a 20 µm layer. The 15 strand/tape Roebel cable that was
assembled by KIT, has a transposition pitch of 226 mm. This
early copper coated tape exhibited “dog-bone” edges. Later, the
“dog-boning” was improved. The predicted cable performance,
when wound into the first test coil is presented in [13] & [14].
This cable was then used to wound the first superconducting 3
turn subscale coil, named Feather-M0.4.
B. Feather-M0.4 Instrumentation:
As detailed in Fig. 13 the first Feather-M0.4 test coil contained the following instrumentation: voltage taps, pickup coils,
temperature sensor array, calibrated point temperature sensors,
Hall probes, integrating field coil and three spot heaters. The
cryostat is equipped with thermocouples, CERNOX temperature sensors and many voltage taps.
V. COLD TEST, OF SUBSCALE COIL FEATHER-M0.4
A. Variable Temperature Cooling System
Liquid helium from a liquefier at 4 K is supplied to a 100 l
satellite vessel. Its level is maintained at 50%. 2-300 W heaters
Fig. 13. (Left) Feather-M0.4 instrumented coil. (Right) Connections to test
insert. Key: 1) Coil entry lead. 2) Coil exit lead, 3) Insulation between leads,
seen in both pictures. 4) Part of CCS. Temp. Sensor array. 5) CCS temp sensor.
6) Cu coil former loop. 7) Magnet former support. 8) Pick-up coil sets. 9) Hall
probe (a). 10) Spot heater #1. 11) Spot heater #2. 12) Position of fibre optic
temp sensor in coil, adjacent to (11). 13) Hall probe (b). 14) Integrating field
PCB coil. 15) Cover plate. 16) Spot heater #3. 17) 15 kA current leads. 18)
–ve pole indium clamped joint. 19) +ve pole indium clamped joint. 20) Copper
cryostat current lead to Roebel clamped connection adaptor plate. 21) Roebel
cable clamped joints using indium sheets, Brass bolts. 22) Roebel cable link
between Feather-M0.4 coil joint and cryostat current leads. With copper shunt &
Kapton insulation 23) Coil joint box –ve pole. 24) Coil joint box +ve pole. 25)
Variable temp gas supply pipe to bottom of magnet. 26) Fibre optic connector.
27) CERNOX temp sensor. 28) 4K helium auxiliary supply tube hidden in wires.
29) Kapton sheet ground insulation over Feather-M0.4 assembly that is inside
the magnetic yoke. 30) Magnetic yoke top end plate.
are used to evaporate >3 g/s of cold gas to a second chamber
containing a high performance variable power heater 500 W to 5
kW, this sets the gas temperature in the 20 K to 300 K range with
a stability better than 1 K. A valve system splits the GHe gas
flow and directs gas to the test cryostat and refrigerator return
line. This enables variable mass flow delivery to cryostat with
stable temperature control when low flow rates are required.
A second 4 K helium liquid line is available passing into the
4003307
IEEE TRANSACTIONS ON APPLIED SUPERCONDUCTIVITY, VOL. 27, NO. 4, JUNE 2017
magnet test cryostat for additional cooling of the current leads
or other (see Fig. 13, item 28).
B. HTS Quench Detection-Protection System
Data acquisition is performed by three systems in parallel.
The standard LTS test stand system, cRIO-9068 [15] and Micron
Optics SM130 [16]. The test stand consists of a digital multimeter and a ‘quench DAQ’. The digital multi-meter has the
purpose of continuous temperature acquisition during cryostat
operation with low frequency (1 Hz). The ‘quench DAQ’ stores
its data in a certain time window when triggered (archive mode)
at three different frequencies: LF 10Hz, MF 5 kHz and HF
200 kHz. The cRIO is fitted with 7 NI-9205 [17] modules,
with 32 channels and a maximum aggregate sample frequency
of 250 kHz. Pick up coils and hall probes are measured with
10 kHz, while temperature sensors and system state signals are
measured with 100 Hz. A NI 9401 DO-module [18] is used
for synchronisation, protection triggering and provoked quench
heater control. Synchronisation triggers are send to the two
other DAQ systems for data-fitting. Heaters are switched by
solid state relays with power is delivered by a GEN3300W.
Voltage and current is measured by the cRIO for quench energy
estimations. Quench detection is performed by the PotAim and
the cRIO in parallel. The PotAim processes all quench related
data with respect to the SM18 system and the voltage taps
in Feather-M0. Quench detection of all other signals coming
from the magnet is performed by the cRIO. Both systems use a
comparison of signals with a threshold over a predefined time
window. Protection trigger of the cRIO is directly fed into the
CERN SM18 test station safety matrix.
C. Summary of Initial Cold Testing
Roebel cable to current lead joint testing started in March
2016. The circuit was successfully powered to 5 kA, the temporary room temperature current lead limit. Cold testing of
Feather-M0.4 coil started June 2016. Initially cooling to 4 K
liquid maintaining a 20 K temperature difference over the magnet at 2.4 K/min was achieved. We did not see a sharp transition
to the superconducting state, so it was difficult to say at which
temperature the coil becomes superconducting. This also is true
for the Coil RRR 5. Using a 600 amp power supply low current
powering was performed so that these sensor system noise could
be assessed, leading to some rewiring and removal of unwanted
additional Earth connections. There are three significant joint
types within Feather-M0.4 assembly: the Copper Indium joint
to the bottom of the cryostat current lead (see Fig. 13, items 18
& 19), (approximately 170 nΩ @ 70 K and 100 nΩ @ 50 K),
Roebel cable to copper transition plate (see Fig. 13, item 21),
(<10 nΩ), and the Roebel cable to Roebel cable joints (see
Fig. 13, items 23 & 24), (<10 nΩ). Joint resistances were measured up to 3000 A, at 56 K. Coil inductance was measured to be
6 µH, in line with calculations, an extremely low value. The Test
plan aimed to set high temperatures 80 K to 85 K where energy
and cable Ic are low (see Fig. 14). Using spot heaters followed
by power supply current extraction after a safe set time interval
(See Fig. 6) we could fine tune the FPGA program to detect
Fig. 14. The CERN test station “SM18”, project students, & SM18 test team.
Key: 1) Variable temperature gas supply cryostat. 2) Magnet test cryostat. 3)
Variable temperature Helium gas supply line. 4) Connection to the 20 kA power
supply through water 15 kA cooled leads. 5) Instrumentation shielded cables
between magnet and protection system. 6) Instrumentation electronics racks. 7)
15 kA current leads. 8) Fisher instrumentation connector box.
quenches and map short sample down to approximately 60 K
the limit of the current lead capacity. The goal of this test was to
commissioning of the advanced quench detection system, to be
sure we can rely on it for Feather-M2 protection. Temperature
stability at the higher temperatures was very good, less than 1 K.
Temperature difference between the satellite vessel and the test
cryostat with the HTS Feather-M0.4 coil was approximately 5 K.
After a thermal cycle to room temperature, the 20 kA power supply was connected. Working at higher temperatures than normal
care was taken with the 12 kA LTS current leads. Some adjustment was needed to balance the circuit as the control system
gave an unusual current oscillation at the point of switching
on, because of the very small inductance and almost negligible
resistance, as the power supply was turned on the current sat at
20 A then jumped to 600 A, followed by what seemed to be a
controlled ramp down during several minutes to 80 A. After this
the current controlled within milliamps. At 1700 A we triggered
the energy extraction with an 80 mOhm resistor. After an initial
400 A over 1 ms increase in current, the current was extracted in
a total of 4 ms which is important for future protection planning.
During joint resistance measurements which follow an up and
down current staircase powering profile. At 2600 A and 60 K,
we saw the voltage across joint 2 slowly rising (Black curve
in the plot of Fig. 15). The current was manually driven down.
Running the current back up to 2600 A taking careful note of
voltage, the voltage increase was reproduced. The Helium mass
flow rate was then increased from 1.5 g/s to 3 g/s and we saw a
small voltage reduction while maintaining the same temperature
and current. Further tests with reduced temperature 56 K saw
the joint-2 voltage start to runaway at 3000 A (see Fig. 15). Coil
was powered to 3200 A for short periods. The joint-2 voltage
was limiting our ability to quench the magnet in the coil. After
adding additional cooling to all joints, the magnet was powered
to quench over a range of gradually reducing temperatures. The
joint temperatures were always approximately 20 K lower than
the magnet. In Fig. 16 we see a selection of joint trips and coil
KIRBY et al.: FIRST COLD POWERING TEST OF REBCO ROEBEL WOUND COIL FOR THE EUCARD2 FUTURE MAGNET DEVELOPMENT
4003307
liquid Helium is promising but still needs to be proven. The
first high current HTS coil, Feather-M0.4 containing Bruker
Tape and KIT Roebel cable has been powered above 12.9 kA
in 25 K gas and quenched over 100 times without degrading
the coil. This is a significant milestone. The next Feather-M0
coils, exploring different features have started to be produced.
The first full-scale 5 T Feather-M2.1 Aligned Block design and
its alternative, CEA Saclay Paris, Cosine Theta coil are both expected to be cold tested in spring 2017. The technical challenges
that we are still facing are: detecting quenches with the magnet
submerged fully in liquid helium, the high current HTS joint
design and operation, operating in high external magnetic field
and finally controlling magnetic field-quality using the 5 mm
wide tapes.
REFERENCES
Fig. 15. Test day 3, Feather0-4 test plot showing increase in joint-2 resistance during powering. Temperature stability, joint resistance as temperature is
decreased and current increased.
Fig. 16. Feather M0.4 cold test quenches v temperature plot. We see a selection
of coil quenches and joint voltage trips.
quenches. A maximum current of just over 12.9 kA was reached
in 25 K helium gas, when a joint voltage tripped the protection.
This exceeded our target current value 10 kA. At 13 K the cryostat started to fill with liquid helium. The test was then stopped,
due to the uncertainty in being able to operate safely in liquid
helium. Further testing is planned.
VI. DISCUSSION AND CONCLUSION
Developing high current cables, coils, joints and quench protection for the HTS high current magnets is a highly demanding
challenge and has advanced greatly with the EuCARD2 HTS
magnet program. The recent success in being able to simulate
current distribution and quench in the multi-strand Roebel cable is fundamental to all tape based, good field quality magnet
developments. The potential to use innovative quench detection
systems, such as pick-up coil arrays, or temperature sensors in
[1] F. Bordry, S. Baird, K. Foraz, A. L. Perrot, R. Saban, and J.P Tock,
“The first long shutdown (LS1) for the LHC,” in Proc. Int. Particle Accelerator Conf., 2013, pp. 44–48. [Online]. Available: http://accelconf.
web.cern.ch/AccelConf/IPAC2013/papers/mozb202.pdf
[2] L. Rossi, “LHC upgrade plans: Options and strategy,” in Proc.
Int. Particle Accelerator Conf., San Sebastian, Spain, 2011, pp.
908–912. [Online]. Available: http://accelconf.web.cern.ch/AccelConf/
IPAC2011/papers/tuya02.pdf
[3] L. Bottura, G. de Rijk, L. Rossi, and E. Todesco, “Advanced accelerator
magnets for upgrading the LHC,” IEEE Trans. Appl. Supercond., vol. 22,
no. 3, Jun. 2012, Art. no. 4002008.
[4] L. Rossi et al., “The EuCARD-2 future magnets European collaboration for accelerator-quality HTS magnets,” IEEE Trans. Appl. Supercond.
vol. 25, no. 3, Jun. 2015, Art. no. 4001007.
[5] G. A. Kirby et al., “Accelerator quality HTS dipole magnet demonstrator designs for the EuCARD-2, 5 tesla 40 mm clear aperture magnet,” IEEE Trans. Appl. Supercond., vol. 25, no. 3, Jun. 2015, Art.
no. 4000805.
[6] G. Kirby et al., “Status of the demonstrator magnets for the EuCARD-2
future magnets project,” IEEE Trans. Appl. Supercond., vol. 26, no. 3,
Apr. 2016, Art. no. 4003307.
[7] J. Van Nugteren et al., “Measurement and numerical evaluation of ACloss in a ReBCO roebel cable at 4.5 K,” IEEE Trans. Appl. Supercond.,
vol. 26, no. 3, Apr. 2016, Art. no. 8201407.
[8] J. Van Nugteren, “High temperature superconductor accelerator magnets,”
Ph.D. dissertation, University of Twente, Enschede, The Netherlands,
2016.
[9] J. Van Nugteren et al., “Numerical analysis of magnetisation in coated
conductor cables and coils,” in Proc. Appl. Supercond. Conf., 2006, Paper
1SPOr1A-06.
[10] J. Murtomäki et al., “Mechanical effects of the non-uniform current
distribution on HTS coils for accelerators wound with ReBCO Roebel
cable,” in Proc. Appl. Supercond. Conf., 2006, Paper 3LPo2F-03.
[11] P. Gao et al., “Effect of tape layout and impregnation method on the
transverse pressure dependence of the critical current in REBCO Roebel
cables,” in Proc. Appl. Supercond. Conf., 2006, Paper 2LPo2D-06.
[12] 3T,
Additive
manufacturer’s
website.
[online].
Available:
http://www.3trpd.co.uk/.
[13] W Goldacker et al., “Roebel cables from REBCO coated conductors: A
one-century-old concept for the superconductivity of the future,” Supercond. Sci. Technol., vol. 27, 2014, Art. no. 093001.
[14] C. Lorin et al., “Development of a Robel-cable-based Cosϑ dipole: design
and wind ability of magnet ends,” presented at Eur. Conf. Appl. Supercond.
(EUCAS), 2015 Lyon, France.
[15] cRIO-9068 FPGA system NI 9068. [Online]. Available: http://sine.ni.com/
nips/cds/view/p/lang/en/nid/211620
[16] Micron Optics sm130. [Online]. Available: http://www.micronoptics.com/
product/dynamic-optical-sensing-interrogator-sm130/
[17] NI 9205. [Online]. Available: http://sine.ni.com/nips/cds/view/p/lang/en/
nid/208800
[18] NI 9401. [Online]. Available: http://sine.ni.com/nips/cds/view/p/lang/en/
nid/208809