Controlled ID 1652363
1
Reference Design of the Mu2e Detector Solenoid
S. Feher, N. Andreev. J. Brandt, S. Cheban, R. Coleman, N. Dhanaraj, I. Fang, M. Lamm, V. Lombardo, M. Lopes,
J. Miller, R. Ostojic, D. Orris, T. Page, T. Peterson, Z. Tang, R. Wands
Abstract—The Mu2e experiment at Fermilab has been approved
by the Department of Energy to proceed developing the
preliminary design. Integral to the success of Mu2e is the
superconducting solenoid system. One of the three major
solenoids is the Detector Solenoid that houses the stopping target
and the detectors. The goal of the Detector Solenoid team is to
produce detailed design specifications that are sufficient for
vendors to produce the final design drawings, tooling and
fabrication procedures and proceed to production. In this paper
we summarize the Reference Design of the Detector Solenoid.
Index Terms—Detector, magnet, solenoid, superconducting.
I. INTRODUCTION
HE Muon-to-Electron conversion experiment (Mu2e),
under development at Fermilab [1], will search for
evidence of new particles (beyond the very successful
Standard Model) by detecting electrons originating from
neutrino-less conversion of the muon under the nuclear force.
This simple goal, however, can only be achieved with a
complicated experimental setup that includes an 8 kW 8 GeV
proton beam interacting with a tungsten target to generate
muons, and a complex solenoid magnetic field that filters and
guides the muons into the detector region where the
conversion electrons are identified and their momentum is
measured.
The Mu2e magnet system (see Fig. 1) [2] consists of three
large superconducting solenoids: Production Solenoid,
Transport Solenoid and Detector Solenoid. The last in the
chain of magnets, the Detector Solenoid (DS) [3], houses the
muon stopping target and the detectors. The role of the DS is
to i) maximize muon yield by efficiently focusing muons to
the stopping target, ii) efficiently focus conversion electrons
from the stopping target
towards the detectors and iii) provide
a solenoid field as part of the tracker spectrometer. These
goals can be achieved by a solenoid field that has 4.2 m long
graded (2 T – 1 T) and a 5.2 m long uniform (1T) sections.
In this paper we summarize the Reference Design (RD that
contains all design specifications of the Detector Solenoid.
Manuscript received July 16, 2013. This work was supported in part by the
Fermi Research Alliance under DOE Contract DE-AC02-07CH11359.
S. Feher is with Fermilab, Batavia, IL 60510. Phone: +1-630-840-2240,
fax : +1-630-840-2323, e-mail: fehers@fnal.gov.
R. Ostojic is with CERN, Geneva, Switzerland.
N. Andreev, J. Brandt, S. Cheban, R. Coleman, N. Dhanaraj, I. Fang, M.
Lamm, V. Lombardo, M. Lopes, D. Orris, T. Page, R. T. Peterson, Z. Tang
and R. Wands are with Fermilab, Batavia, IL 60510, USA.
J. Miller is with Boston University, Boston, MA 02215, USA.
II. DETECTOR SOLENOID OVERVIEW
The 10.75 m long and 1.9 m inner bore diameter Mu2e
Detector Solenoid utilizes two and one layer coils that are
wound from two different types of aluminum stabilized NbTi
superconductor cable. At the magnet operational current of
6114 A (at 5 K coil temperature) the coil peak field on the
conductor in the graded section (2 T – 1 T linearly decreasing
field in the bore) is 2.2 T while in the spectrometer section
(1 T in the bore) is 1.3 T. The DS cold mass and cryostat
weighs about 9 t and 30 t respectively. The stored energy is
30 MJ.
III. M
AGNETIC DESIGN
The magnetic design specifies the solenoid field map for
the Mu2e experiment, the coil layout (coil dimensions and
spacing) and the operational current. The coil structure of the
Mu2e solenoid system is shown in Fig. 2. DS overlaps the
three last coils of the Transport Solenoid consequently the
field map takes into account the field originated from the
Transport Solenoid as well.
Fig. 2. Mu2e Solenoid System Coil layout; Production Solenoid (PS),
Transport Solenoid (TS) and Detector Solenoid (DS).
The magnetic field at two radial positions (in the center of
the solenoid and R = 0.4 m) is shown in Fig. 3. In the graded
section (S1) at the entrance to the DS the field is 2 T then it
gradually drops to 1 T at the beginning of the spectrometer
T
Fig. 1. Layout of the Mu2e Solenoid system.
FERMILAB-PUB-13-399-TD
Controlled ID 1652363
2
section (S3). In the spectrometer section the field is 1 T and it
has a slight negative gradient. In Fig. 4 the field tolerance
values obtained by extensive studies that are based on coil
manufacturing tolerances are shown. The detailed studies of
the tolerances and their importance to the Mu2e experiment
are presented in [4].
Fig. 3. – Magnetic field at different radii. S1 is the gradient region, S2 is
transition region, S3 is the spectrometer region.
Fig. 4. – Magnetic field difference between the nominal field and the field
when conductors have a 50 micron variation in thickness.
Fig. 5. Detail of the DS coils and the three TS coils overlapped by DS.
The coil dimensions and spacing in the Mu2e coordinate
system are presented in Table I. The coils labeling is presented
in Fig. 5. The field specifications tolerances could
accommodate the following manufacturing tolerances: i) ± 50
µm cable thickness, ii) ± 5 mm radial and ± 1 mm longitudinal
positioning of the coils with respect to each other, iii) coils
must be aligned within 2 mrad with respect to the magnet
axes. These tolerance values assume that the coil positions are
adjusted by choosing the correct spacer length after the coil
lengths are precisely measured.
TABLE I
S
UMMARY OF DETECTOR SOLENOID COIL PARAMETERS
Coil
#
Coil IR
[m]
Coil
OR [m]
Coil
Length
[m]
Z at the
Center
[m]
layers
Total
turns
1
1.0500
1.0915
0.4198
3.7489'
2'
146'
2
1.0500
1.0915
0.4198
4.1739'
2'
146'
3
1.0500
1.0915
0.4198
4.5989'
2'
146'
4
1.0500
1.0915
0.4198
5.2519'
2'
146'
5
1.0500
1.0915
0.3623
5.8801'
2'
126'
6
1.0500
1.0915
0.3623
6.5701'
2'
126'
7
1.0500
1.0915
0.3623
7.3971'
2'
126'
8
1.0500
1.0705
1.8300
8.8178'
1'
244'
9
1.0500
1.0705
1.8300
10.6528'
1'
244'
10
1.0500
1.0705
1.8300
12.4883'
1'
244'
11
1.0500
1.0915
0.3623
13.6425'
2'
126'
IV. C
ABLE DESIGN
The DS utilizes two different types of aluminum stabilized
Rutherford cables (DS1 – used for the double layer coils and
DS2 – used for the single layer coils). Based on detailed
specifications
1
the prototype versions of these two types of
cables have been ordered from two different vendors. Table II
and Table III summarizes the main parameters of the strand
and cables respectively.
To avoid unnecessary risk procuring the strand and cable
we have chosen specifications that are similar to specs of
Detector Solenoids that have already been successfully used
by other experiments. The specifications for the conductor
includes a rigorous Quality Assurance program require
qualifications tests and material certifications during all phases
of the fabrications process. The fine details of the specs also
reflect the input from vendors during the procurement process
of the prototype cables.
V. C
OIL DESIGN
As shown in Fig. 5 the DS is made of 11 coils. Seven two
layer coils are in the gradient section (S1); three (long) one
layer and one (short) two layer coils at the downstream end of
the spectrometer section (S3). From the manufacturing point
of view there are only three different types of coils.
The DS employs a composite cable insulation made of
polyimide and pre-preg glass tapes. This type of insulation,
originally developed for the TRISTAN/TOPAZ solenoid [5],
was also used in the ATLAS Central Solenoid [6]. The cable
is insulated with two layers of composite tape consisting of 25
1
The specifications documents are available upon request.
Controlled ID 1652363
3
µm Kapton tape sandwiched between 25 µm of a semi-dry
(BT) epoxy on one side and 75 µm of pre-preg E-glass on the
other side. The total thickness of this composite tape is 125
µm.
The cable is butt-wrapped with 2 layers of the composite
tape. Therefore the cable insulation thickness is 250 µm.
TABLE II
S
UMMARY OF DETECTOR SOLENOID STRAND PARAMETERS
Parameters
Unit
Value
Tolerance
Outer Diameter
mm
1.303
± 0.005
Cu/Sc ratio
1 : 1
± 0.1
Filament OD
µm
40
-
Ic (4.22 K, 5 T)
A
1850
Min RRR
80
-
Twist direction
Left
Twist pitch
mm
30
± 4
TABLE III
S
UMMARY OF DETECTOR SOLENOID CABLE PARAMETERS
Rutherfod
Cable
Parameters
Unit
DS1
DS2
Toler.
No strands
12
8
-
width
mm
7.88
5.25
± 0.01
thickness at 5 kPsi
mm
2.34
2.34
± 0.01
Transp. angle
degree
15
15
± 0.5
Lay direction
Right
Right
-
Minimum Ic
(at 4.22 K, 5T)
kA
20.9
13.9
-
Min RRR
≥ 60
≥ 60
Cable residual twist
degree
< 45
< 45
Min bending radius
mm
20
20
Aluminum
cladded
cable
Minimum Ic
(at 4.22 K, 5T)
kA
> 18.8
> 12.5
width
mm
20.1
20.1
± 0.1
thickness
mm
5.27
7.03
± 0.03
Strand Copper RRR
> 80
> 80
Aluminum RRR
> 800
> 800
Yield stress at 293 K
MPa
> 30
> 30
Yield stress at 4.22 K
MPa
> 40
> 40
Shear stress Al-strand
MPa
> 20
> 20
The coils are wound onto a collapsible mandrel. The
recommended tension and the compaction criteria during the
winding process are similar to what was for the CDF central
solenoid [7]. The layers for the two layer coils are wound
continuously and sheets of dry E-glass insulation (0.5 mm
thick) are introduced between the coil layers. There is also
extra (2 mm thick) ground insulation between the coil and the
support structure that consists of dry E-glass and 2x25 µm
layers of Kapton. The extra thickness of E-glass between the
coil and support structure allows for the machining of the
outer coil surface after the impregnation to obtain the 0.05 mm
tolerance value for the outer radius of the coil that is required
for controlled shrink-fitting the surfaces of the support shells
and the coils. The support shells are made of 5083-0
aluminum and they are 1 cm (for single layer coil) and 2 cm
(for double layer coil) thick. Before the slightly oversized
shell is milled to precise ID (0.05 mm tolerance) all the
necessary welding activities need to be completed.
The coils will be cured and vacuum-impregnated. These
two processes can be performed simultaneously or in series.
The mandrel will remain in the coil during these
manufacturing processes. The mandrel will be removed after
the shrink fit is completed. The maximum interference value
between the coil and the support cylinder needs to be within
0.5 mm with 0.05 mm tolerance values in order to allow
adequate preload on the coil. Before the coil is inserted the
cylinder will be heated up to 130 C and a thin layer of epoxy
will be put on the cylinder inner surface.
VI. C
OLD MASS DESIGN
After each coil module is built, their length will be
measured precisely. Based on the coil dimensions the axial
spacers’ length will be determined and the spacers will be
fabricated within 1 mm axial tolerance values. Radial
tolerance values of the spacers are not critical; however, the
bolt hole patterns of the flanges need to be matched with that
of the coil support cylinders. To avoid losing pre-compression
the coils are bolted together with high strength Aluminum
bolts. The detailed calculation of sizing the screws are based
on DS loads and screw properties.
Fig. 6. DS coil and axial spacer assembly with cooling tubes and cryostat
supports.
The coils are electrically connected in series. The cable
joints are welded together on the wide sides of the cable.
Exact length still needs to be specified. The internal bus
utilizes DS1 type of conductor. The Helium cooling tubes are
made from 2 inch OD Aluminum tubing. The cooling tube
arrangement (Fig. 6) was optimized for the force flow-cooling
scheme. The cooling tubes are attached to the support cylinder
through flexible high heat conductivity aluminum strips.
VII. C
RYOSTAT DESIGN
The DS cryostat (Fig. 7) is 10.75 m long. The cryostat
vessel is made of two concentric SS cylinders that have 2 cm
wall thicknesses. The design satisfies ASME code Section
VIII, Div. 1 rules for cylindrical shells under external pressure
of 1 atmosphere at room temperature. The inner shell will be
supporting the detectors sitting on a rail system. Due to the
load of 10 metric tons (detectors, shielding and other
equipment) the maximum vertical deformation of the inner
shell is 1.0 mm.
The DS cold mass support system (see Fig. 6) uses 16
tangentially-arranged metallic rods, eight on each ends, to
Controlled ID 1652363
4
support against dead weight and radial de-centering forces.
Eight metallic rods at the downstream end only provide
support against the axial magnetic forces.
The radial and the axial support rods are made from Inconel
718, and all supports are connected to their 5083-0 Al bracket
at the cold mass by one inch diameter Inconel 718 pins. Both
support systems use spherical rod ends to account for motion
during cool-down, including the axial thermal contraction of
36 mm at the upstream end.
The cryostat sits on two saddles, positioned very close to the
ends of the vessel, right there where the warm end of the axial
supports are attached to the cryostat through towers (Fig. 7).
This way the cryostat outer shell is directly transmitting the
forces (100 tons axial) to the saddle support producing no
stresses on the cryostat.
VIII. I
NSTRUMENTATION
The DS instrumentation serves for: 1) quench protection
and monitoring (QPM); 2) cryogenic monitoring and controls
(CMC); and 3) mechanical characterization (MC). The
primary (P) sensors for quench detection are voltage taps.
Holes will be drilled and tapped in the aluminium conductor
so voltage taps can be attached using screws. Two voltage taps
will be installed across each splice for redundancy (R).
Temperature sensors will also be installed on each support
cylinder for a secondary quench trigger.
Instrumentation will also be implemented for cryogenic
monitoring and controls for cool down, warm up, operational
steady state, and changes in response to a heat load.
Temperature sensors will be installed for monitoring the
temperatures of the 80 K shield, the cold mass, and the support
posts. In addition, temperature sensors and heaters will be
installed along the bore for monitoring and controlling its
temperature.
The mechanical state of the solenoid will be monitored
using strain gauges and position sensors. Several strain gauges
will be installed on the upstream, downstream, and
longitudinal supports. Table IV summarizes the different type
and number of sensors envisioned to be used in the DS.
TABLE IV
DS Instrumentation List
Sensor Type
Description
Function
Coil Voltage Taps
12P/12R
Primary QPM
Cold Mass RTDs
11P/11R Cernox
TM
Secondary QPM
Lead Voltage Taps
4P/4R
Primary SC Lead QPM
Lead LTS Wire
NbTi SC Wires,
2P/2R
Secondary SC Lead QPM
80K Shield RTDs
10 Platinum
CMC
Cold Mass RTDs
10 Cernox
TM
CMC
Bore RTDs
12 Platinum
CMC (Bore Temp. Control)
Bore Heaters
1P/1R
CMC (Bore Temp. Control)
Phase Separator
Liquid Level
He LL Probe,
1P/1R
CMC (Helium Flow
Control)
Phase Separator
RTD Liquid Level
5 Cernox
TM
CMC (Helium Flow
Control)
Support Post
RTDs
20 Cernox
TM
CMC
Support Post
RTDs
20 Platinum
CMC
Support Post
Strain Gauges
32 Half-Bridge
MC (Coil Support Stress)
Cold Mass
Position Sensors
6P/6R
MC (Coil Displacement)
IX. INTERFACES
The DS has many interfaces with other parts of the Mu2e
Experiment: cryostat inner shell, flanges between DS and TS
and flanges between DS and vacuum pump spool piece, port
for the transfer line and the magnet support. The design of
these interfaces are in the final design phase.
X. C
ONCLUSIONS
The Mu2e experiment requires 2 T – 1 T strong, 10.75 m
long, large (1.9 m) OD bore magnets to house the stopping
target and the detectors. Most of the design considerations
have been completed and the reference design is close to
completion. A summary of the specifications was presented
for the field, conductor, coil, cold mass, cryostat and
instrumentation design.
R
EFERENCES
[1] Mu2e Collaboration, "Mu2e Conceptual Design Report",
arXiv:1211.7019, http://arxiv.org/abs/1211.7019.
[2] M.J. Lamm et al., “Solenoid Magnet System for the Fermilab Mu2e
Experiment”,IEEE Transaction on Applied Superconductivity, vol. 22,
Issue 3, pp. 4100604, June 2013.
[3] R. Ostojic et al., “Challenges in the Design of the Detector Solenoid for
the Mu2e Experiment”,IEEE Transaction on Applied Superconductivity,
vol. 23, Issue 3, pp. 4500404, June 2013.
[4] M. Lopes et al. "Tolerance studies of the Mu2e solenoid system" - IEEE
Trans. Appl. Supercond. 23 submitted for publication.
[5] A. Yamamoto et al. " A Thin Superconducting Solenoid Wound With
Internal Winding Method for Colliding Beam Experiments “,J. Phys.
Colloques, vol. 45, pp. C1-337-C1-340, 1984.
[6] Atlas Central Solenoid, ATLAS TDR-9; CERN/LHCC 97-21, April
1997.
[7] H. Minemura et al., “Construction and testing of a 3 m diameter x 5 m
superconducting solenoid for the Fermilab collider detector facility
(CDF) ”, Nucl. Instrum. Methods A, vol. 238, pp. 18-34, July 1985.
Fig. 7. Cross section of DS magnet-in cryostat. Coils are colored to
distinguish them from axial spacers.
