Quench Protection for the MICE Cooling Channel Coupling Magnet

Abstract

This paper describes the passive quench protection system selected for the muon ionization cooling experiment (MICE) cooling channel coupling magnet. The MICE coupling magnet will employ two methods of quench protection simultaneously. The most important method of quench protection in the coupling magnet is the subdivision of the coil. Cold diodes and resistors are put across the subdivisions to reduce both the voltage to ground and the hot-spot temperature. The second method of quench protection is quench-back from the mandrel, which speeds up the spread of the normal region within the coils. Combining quench back with coil subdivision will reduce the hot spot temperature further. This paper explores the effect on the quench process of the number of coil sub-divisions, the quench propagation velocity within the magnet, and the shunt resistance.

Full text

Page 1

eScholarship provides open access, scholarly publishing
services to the University of California and delivers a dynamic
research platform to scholars worldwide.
Lawrence Berkeley National Laboratory
Peer Reviewed
Title:
Quench Protection for the MICE Cooling Channel Coupling Magnet
Author:
Guo, Xing Long
Publication Date:
02-26-2009
Publication Info:
Lawrence Berkeley National Laboratory
Permalink:
http://escholarship.org/uc/item/9920d08g

Page 2

IEEE Transactions on Applied Superconductivity 19, No. 3 MICE Note 2351

Abstract—This paper describes the passive quench protection
system selected for the muon ionization cooling experiment
(MICE) cooling channel coupling magnet. The MICE coupling
magnet will employ two methods of quench protection
simultaneously. The most important method of quench
protection in the coupling magnet is the subdivision of the coil.
Cold diodes and resistors are put across the subdivisions to
reduce both the voltage to ground and the hot-spot temperature.
The second method of quench protection is quench-back from the
mandrel, which speeds up the spread of the normal region within
the coils. Combining quench back with coil subdivision will
reduce the hot spot temperature further. This paper explores the
effect on the quench process of the number of coil sub-divisions,
the quench propagation velocity within the magnet, and the
shunt resistance.
Index Terms—Passive quench protection, quench back,
superconducting magnets, subdivision protection.
I. INTRODUCTION
HE muon ionization cooling experiment (MICE) will be a
demonstration of ionization cooling in a short section of a
muon cooling channel. The MICE cooling channel contains
three absorber focus coil (AFC) modules and two RF coupling
coil (RFCC) modules that reaccelerate the muons back to their
original momentum [1]. The RFCC module consists of a 1.9 m
long vacuum vessel that contains four 201.25 MHz RF
cavities that are bounded by thin beryllium windows. The
coupling magnet located outside of the RF cavity vacuum
vessel is a superconducting solenoid that produces enough
magnetic field (up to 2.2 T on axis) to guide the muons and
keep them within the iris of the thin RF-cavity windows [2].
The MICE coupling solenoid uses multifilament Nb-Ti in a
copper matrix. The magnet is cooled with a pair of pulse tube
coolers, hence the magnet current is low. During a quench,
the magnet stored energy and the high conductor current
density causes a large temperature rise where the quench
originated. In addition, high voltages between the coil and
Manuscript received 17 August 2008. This work was supported by the fund
of cryogenics and superconductivity engineering technology innovation
platform, the second phase of “985 Project” of Harbin Institute of
Technology. This work was also supported by the Office of Science of the
US Department of Energy under DOE contract DE-AC-02-05CH11231.
X. L. Guo, F. Y. Xu, L. Wang, H. Pan, H. Wu, X. K. Liu and L. X. Jia are
with the Institute of Cryogenics and Superconductive Technology, HIT,
Harbin 150001, China. M. A. Green is with Lawrence Berkeley National
Laboratory, Berkeley CA 94720, USA (e-mail: magreen@lbl.gov). K. Amm
is with the General Electric Research Center, Niskayuna, NY 12309, USA
ground will develop. The magnets in MICE will be passively
quench-protected through coil subdivision and quench-back.
This paper describes the magnet quench protection system.
The effects of coil sub-division, quench propagation velocity,
and the shunt resistance across each subdivision on the magnet
hot-spot temperature and the voltage to ground are studied.
II. PASSIVE QUENCH PROTECTION SYSTEM DESIGN
A. Coupling Coil Design
The coupling magnet will work in two modes due to the
polarity change of two focusing coils in the AFC module. One
is gradient mode (flip mode), and the other is solenoid mode
(non-flip mode). Table I shows the coupling coil design
parameters. The worst-case is to operate the MICE in the flip
mode at a muon average momentum of 240 MeV/c. The beam
beta at the center of the absorbers is designed to be 420 mm.
The coupling coils will be made from a commercial copper
matrix Nb-Ti conductor originally used for MRI magnets. The
conductor Ic is >760 A at 5 T and 4.2 K. Using this conductor,
the magnet margin is expected to be >0.8 K when the
induction at high field point is 7.4 T, the current is 210.1 A,
and the cold mass temperature is 4.2 K [2].
Fig.1 shows the coupling magnet basic structure. The MICE
coupling magnet consists of a single 285 mm long coil. At
room temperature, the coil inner radius is 750 mm and its
thickness is 102.5 mm. The coil is wound on a forged 6061-T6
aluminum mandrel. The over all design dimensions for the
coil cold mass are: The inner radius is 736 mm; the thickness
is 143.5 mm; and the length is 329 mm. The G-10 insulation
thickness between the coil and the bobbin, the end flange and
the banding are 1.0 mm, 3.0 mm and 1.0 mm respectively. The
coil is indirectly cooled by liquid helium flowing in extruded
cooling tubes attached to the outer surface of the coil cold
mass case [3].
B. Protection Circuit Design
Fig. 2 shows the proposed coupling magnet circuit. A 300 A
10 V power supply is used charge the magnet. The magnet
will be discharged through a water-cooled varistor circuit. The
current into the magnet will be carried by a single pair of
copper and HTS leads. The coil will have eight subdivisions
with a pair of back-to-back R620 diodes at ~5 K and a resistor
across each subdivision. The coil rapid discharge system will
consist of 25 diodes mounted on a water-cooled plate. The
mandrel acts as a shorted secondary circuit inductively
coupled with each of the coil subdivisions [2].
Quench Protection for the MICE Cooling
Channel Coupling Magnet
X. L. Guo, F. Y. Xu, L. Wang, M. A. Green Member IEEE, H. Pan, H. Wu, X. K. Liu,
L. X. Jia, K. Amm Member IEEE
T

Page 3

IEEE Transactions on Applied Superconductivity 19, No. 3 MICE Note 2352
Magnet subdivision is a passive quench protection method
long used in MRI magnets. Back-to-back cold diodes allow
the magnet to be safely quenched at either magnet polarity.
The high diode forward voltage (at least 4 V at 5 K) across the
diodes prevents current from bypassing the magnet coil during
a magnet charge or discharge at its design charging and
discharging voltage. Magnet coil subdivision reduces both the
voltage to ground and the hot spot temperature [4], [5].
The magnet aluminum mandrel is inductively coupled to all
of the coil subdivisions. The mandrel will act as a shorted
secondary circuit that absorbs energy from the magnet during
a quench. The current in the mandrel will heat the mandrel,
which will eventually heat up the adjacent coil subdivisions
and induce new normal zone. This process is called quench
back. Quench back will speed up the quench process, and thus
reduce the hot spot temperature [6], [7].
III. THE QUENCH MODEL DESCRIPTION
A semi-empirical quench model considering both the
subdivision and quench back has been developed. A quench is
initiated in the mid-plane and starts to expand in the three
directions with velocities vφ (longitudinal propagation along
the coil conductor), vr (radial propagation) and vz (axial
propagation). The normal zone shape is assumed to be an
ellipsoid [8]. In the model, the average constant quench
propagation velocities in three directions are used during
quench process. The calculation proceeds in time steps. At
each step another layer is added to the surface of the normal
zone like the skin of an onion. After a time step the current
will have decayed. The calculation continues until each coil
subdivision current is less than the 1 percent of its initial
current. Helium cooling is ignored in this simulation.
Each coil section is assumed to be adiabatic, for each time
step. The temperatures of each successive quenching volume
are determined by the joule heating within that volume. The
hot-spot temperatures are the temperatures of the start point in
each subdivision. The power supply is disconnected once the
quench starts. The cold diodes are assumed to be a short
circuit. The rapid discharging diode stacks are assumed to be
closed. The current in all conducting loops including the
mandrel are calculated using the inductance matrix, the shunt
resistance and the temperature-dependent normal zone
resistance [9]. The voltages-to-ground are estimated by using
the resistive voltage drop across the normal zone in each
magnet coil subdivision.
The quench-back time is the sum of two time periods. The
first time period is the time it takes for the aluminum mandrel
heat up to the superconductor critical temperature (~9 K).
This time is associated with the shift of the current from the
superconducting coil to the conductive mandrel. A high
resistivity mandrel will take longer to heat to 9 K than a low
resistivity mandrel, hence a stainless steel mandrel is not used.
The second time period in the quench back process is the time
needed for heat to flow from the mandrel to the
superconductor. This time is associated with the heating of the
superconductor and the heating of the insulation between the
superconductor and mandrel. This time is of the order of 0.08
to 0.10 s when the insulation thickness between the coil
conductor and the mandrel is ~1 mm. [7]. After quench back
occurs, the magnet normal region is the normal region due to
quench propagation plus the normal region induced due to
quench back from the mandrel.
He Tubes
Coil
Bobbin
End Plate
Banding
Cover Plate
Insulation
Fig.1 Cross subdivision of the coupling magnet cryostat
Rapid Discharge Switch
Power Supply
±10V, 300A
Coupling Magnet
Rapid Discharge Diode
Cold Resistor
Mandrel
Cold Diode
Copper Lead HTS Lead
Fig.2 Protection circuit of coupling magnet
TABLE I. COUPLING MAGNET SPECIFICATION
ParameterFlipNon-flip
Coil Length (mm) 285
Coil Inner Radius (mm)750
Coil Thickness (mm) 102.5
Number of Layers 96
No. Turns per Layer166
Magnet Self Inductance (H)592.5
Magnet J (A mm--2)* 114.6108.1
Magnet Current (A)*210.1198.2
Magnet Stored Energy (MJ)*13.111.6
Peak Induction in Coil (T)*~7.40~7.12
Coil Temperature Margin (K)*~0.8~1.0
*Worst case design based on p = 240 MeV/c and β = 420 mm

Page 4

IEEE Transactions on Applied Superconductivity 19, No. 3 MICE Note 2353
IV. A PARAMETER STUDY OF THE QUENCH PROCESS
A. The Effect of Magnet Coil Subdivision
The effects of subdivision number of 2, 4, 6 and 8 were
studied. All simulation cases use the quench propagation
velocities shown in Table II and the shunt resistance across
each subdivision is zero. Fig. 3 shows the peak hot-spot
temperature as a function of the number of sub-divisions. The
coil hot-spot temperature is 135 K for eight subdivisions and
149 K for two subdivisions. The peak hot-spot temperature
decreases as the number of subdivisions increases. Because of
inductive coupling between subdivisions, the temperature of
the coil will become more uniform with more subdivisions.
Fig. 4 shows the voltage to ground with different subdivision
numbers. The voltage to ground is ~2.7 kV for 8 sub-divisions
and ~10.6 kV for 2 sub-divisions. The maximum voltage to
ground decreases as the subdivision number increases.
Without subdivision the maximum coil voltage is ~22.5 kV.
The voltage-to-ground is slightly lower without quench-back.
Fig. 3. Peak Hot-spot Temperature versus Time and Sub-division Number
Fig. 4. Peak Voltage to Ground versus Time and Sub-division Number
B. The Effect of Quench Propagation Velocity
The effect of quench propagation velocity within the
coupling magnet on the magnet quench process was studied.
Table II shows two different quench velocities cases studied.
The Case 1 quench velocities are based on Nb-Ti quench
propagation measurements made at LBNL in the 1970’s.
According to this correlation, the propagation velocity is
dependent only on the current density in the conductor cross-
section and the conductor magnetic induction. Both quench
velocities are for an average magnet field of 2.5 T and a
matrix current density of 1.4x108 A m-2 [4]. Case 1 velocities
correspond to using Nb-Ti in the coil. Case 2 velocities are
the Case 1 velocities divided by three, which corresponds to
using Nb3Sn in the magnet. Both cases were simulated for an
operating current of 210 A, an operating temperature of 4.2 K
and a resistance of 0 ohms across each subdivision.
TABLE II CASES FOR QUENCH PROPAGATION VELOCITIES
Vφ (m/s)
vr (m/s)vz (m/s)
Case 1 3.4770.0650.085
Case 2 1.159 0.0220.028
Fig. 5 shows the peak hot-spot temperature for the two
cases. In Case 1 the hot-spot temperature is ~135 K; in Case 2
the hot-spot temperature is ~216 K. High quench propagation
velocities make the temperature within the coil more even.
Without quench-back the hot-spot temperature in the magnets
would be much higher. Fig. 6 shows the voltage to ground for
the quench propagation velocity cases. In Case 1 the peak
voltage-to-ground is ~2.7 kV; in Case 2 the peak voltage-to-
ground is about ~2.6 kV. In this model the voltage to ground
is the voltage drop across the resistance of the normal zone. A
fast propagation velocity induces a fast growth of normal zone
resistance. With a rapid growth of normal zone resistance, the
magnet current rapidly decreases. The voltage to ground is a
function of both the resistance and dI/dt.
Fig. 5. Peak Hot Spot Temperature versus Time and Quench Velocities
Fig. 6. Peak Voltage to Ground versus Time and Quench Velocities

Page 5

IEEE Transactions on Applied Superconductivity 19, No. 3 MICE Note 2354
C. The Effect of Shunt Resistance across a Sub-division
The effect of a resistance of 0 and 5 ohms (at 4 K) across
each subdivision was studied. In both cases the voltage drops
across diodes were ignored. The resistance change with
temperature was also ignored. In both cases the simulations
were done using the Case 1 quench propagation velocities.
Fig. 7 shows the effect of the shunt resistance on the hot
spot temperature. In the 0-ohm case, the hot spot temperature
is ~135 K. In the 5-ohm case, the hot spot temperature is
~100 K. The hot-spot temperature decreases with an increase
in the shunt resistance. The reason for this is that the shunt
resistor absorbs stored magnetic energy from the magnet.
This energy doesn’t end up in the coil winding. Fig. 8 shows
the effect of the shunt resistance on the voltage-to-ground. In
the 0-ohm case, the voltage to ground is 2.7 kV. In the 5-ohm
case, the voltage-to ground is 1.35 kV. The quench velocities
in the two cases are same. Quench back increases the voltage-
to-ground slightly for both cases.
Fig. 7. Hot-spot Temperature versus Time for Two Shunt Resistances
Fig. 8. Peak Voltage to Ground versus Time For Two Shunt Resistances
A large shunt resistance results in a faster current decay and
a slower increase of the coil normal zone resistance. All these
effects result in lower peak voltage-to-ground. When the
shunt resistance per coil sub-division is 5 ohms, ~5.5 MJ of
the coil energy ends up in the resistors. The resistors must
have a mass of ~1.5 kg per ohm in order to absorb magnetic
energy from the coil without going above 350 K. If the shunt
resistance is too high, the voltage to ground and the hot spot
temperature increase, because less coil energy is removed.
V. CONCLUSION
A passive quench protection system for coupling magnet
based on coil subdivision and quench back has been designed.
A special semi-empirical model was developed to analyze the
quench process and study the effect of coil sub-division, of
quench propagation velocities, and of the shunt resistance
across each subdivision.
More magnet sub-divisions result in lower hot-spot
temperatures and lower voltages-to-ground. Quench-back
from the mandrel reduces the hot-spot temperature and
increases the voltage to ground a few percent.
Faster quench propagation velocities will result in lower
hot-spot temperatures, but the voltages-to-ground may
increase. Quench-back will have a large effect on the hot-spot
temperature when the quench propagation velocities are low.
The cases illustrated were for Nb-Ti and Nb3Sn. Hot-spot
temperatures would be much higher if the coil superconductor
was an HTS conductor, because HTS conductors have much
lower quench propagation velocities.
A larger shunt resistance across each subdivision will result
in lower hot-spot temperatures and lower peak voltages-to-
ground provided significant magnetic energy is removed from
the system by the shunt resistors. Since the shunt resistors are
cold, they must have enough mass to keep their temperature
from going above room temperature when the quench is over.
In the coupling magnet, a 50 m-ohm stainless steel will be
used per magnet sub-division. A larger resistor would be
desirable, but the space available for the eight shunt-resistors
is limited.
The MICE coupling magnet can have a passive quench
protection system that is based on sub-dividing the coil and
putting a cold diode and resistor across the subdivisions. Even
when the resistance across the sub-division is low, eight sub-
divisions of the coupling magnet coil are adequate for
effective quench protection of the magnet.
REFERENCES
[1]
G. Gregoire, G. Ryckewaert, L. Chevalier, et al, “MICE and
International Muon Ionization Cooling Experiment Technical Reference
Document”, [on line] http://www.mice.iit.edu, 2001.
L. Wang, M. A. Green, F. Y. Xu, et al, “The Engineering Design of the
1.5m Diameter Solenoid for the MICE RFCC Modules”, IEEE
Transactions on Applied Superconductivity 18, No. 2., p 937, 2008.
L. Wang, H. Wu, L. K. Li, M. A. Green, et al, “The Helium Cooling
System and Cold Mass Support System for the MICE Coupling
Solenoid”, IEEE Transactions on Applied Superconductivity 18, No. 2, p
941, 2008.
M. A. Green and H. Witte, “Quench Protection and Power Supply
Requirements for the MICE Focusing and Coupling Magnets,” MICE
Note 114, http://www.mice.iit.edu, LBNL-57580, May 2005.
M. A. Green, L. Wang, X. L. Guo, et al “Quench Protection for the
MICE Cooling Channel Coupling Magnet”. LBNL-63698, MICE Note
193, http://www.mice.iit.edu, 31 January 2008.
M. A. Green, “Quench Back in Thin Superconducting Solenoid
Magnets”. Cryogenics 24, p 3-10, 1984.
M. A. Green, “The Role of Quench Back in Quench Protection of a Thin
Superconducting Solenoid”, Cryogenics 24, p 659-668, 1984.
M. N. Wilson, Superconducting Magnets, Oxford University Press,
Oxford, 1983. p 209-219.
C. H. Joshi, Y. Iwasa, “Prediction of current decay and terminal voltages
in adiabatic superconducting magnets”. Cryogenics 29, p 157-167, 1988.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]

Page 6

IEEE Transactions on Applied Superconductivity 19, No. 3 MICE Note 2355
DISCLAIMER
This document was prepared as an account of work sponsored by the United States Government.
While this document is believed to contain correct information, neither the United States
Government nor any agency thereof, nor The Regents of the University of California, nor any of
their employees, makes any warranty, express or implied, or assumes any legal responsibility for the
accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed,
or represents that its use would not infringe privately owned rights. Reference herein to any specific
commercial product, process, or service by its trade name, trademark, manufacturer, or otherwise,
does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United
States Government or any agency thereof, or The Regents of the University of California. The views
and opinions of authors expressed herein do not necessarily state or reflect those of the United States
Government or any agency thereof, or The Regents of the University of California.