Transcript Kirby_4Nov08_Ned - Indico

High Field Magnet Fresca 2
Glyn A. Kirby
&
Bernardo Bordini, Franck Borgnolutti, Jens Erik Bruer, Ezio Todesco,
Gijs De Rijk, Stephan Russenschuck.
Introduction
Existing strand designs, PIT and OST’s RRP are being used in the conceptual designs
for two types of magnet, classical Cos and a hybrid block design. Which eventually will
lead to building a high field 100mm clear aperture accelerator quality model magnet.
This work is still ongoing however we present some initial ideas and observations
related to the strand design and operating conditions for the strand.
G.A.Kirby 4th Nov.08
Talk outline
Comparing the available materials.
Magnet quench protection.
Mechanical protection of the Nb3Sn in the strand some ideas .
Compare to magnet designs, Cos  & Hybrid-block
Closing comments.
G.A.Kirby 4th Nov.08
Comparing the available materials
Ic for one strand at 4.3K
OST 0.8 Jc 3000 A/mm^2 @12 T & 4.5K
PIT 1.25 Jc 2500 A/mm^2 @ 12T & 4.5K
3000
Ic Current [A]
2500
2000
1500
OST 0.8Dia
1000
PIT NED 1.25Dia
MY PIT 0.8Dia
500
NED 0.8 Current
0
0
1
2
3
4
5
6
7
8
9
Field [T]
G.A.Kirby 4th Nov.08
10
11
12
13
14
15
16
17
Quench simulations
Accelerator magnets are designed to absorb their energy by spreading heat through the coils.
However this heating must be limited so not to damage the cables performance. Nb3Sn has a high
temperature margin ~15K in low field volumes at the operating point, this makes quenching the
cables very difficult.
FRESCA 16T 100 App.
No Dump resistor
100% of coils quenched
500
Current (A)
400
Cable 2 temp
300
Cable 1 temp
Hot spot
200
Voltage (V)
100
0
-100
-200
0
0.1
0.2
0.3
time [s]
G.A.Kirby 4th Nov.08
0.4
0.5
0.6
0.7
16800
15800
14800
13800
12800
11800
10800
9800
8800
7800
6800
5800
4800
3800
2800
1800
800
-200
-0.1
200
150
Current (A)
100
Cable 2 temp
Cable 1 temp
50
Hot spot
Voltage (V)
0
-50
-100
-150
0
0.1
0.2
0.3
time [s]
0.4
0.5
0.6
0.7
Hot spot [K] , Voltage
[V]
600
Current [A]
16800
15800
14800
13800
12800
11800
10800
9800
8800
7800
6800
5800
4800
3800
2800
1800
800
-200
-0.1
Hot spot [K] ,
Voltage [V]
Current [A]
FRESCA 16T 100 App.
No Dump resistor
15% of coils quenched
Strand design in relation to
quench performance
-
If we can quench all the coil volumes the existing strand copper volume is sufficient.
However the recent magnet tests have only manages to quench the high field volumes, about
15% of the coils. We need to develop a working quench heater, or put more copper in the strand.
We could also use Nb-Ti in the low field volumes where we have a working quench heater design.
-
If we use a energy extraction dump resistor the existing strand is ok. But this would imply
that every magnet would have its own set of current leads and dump resistor. For development
magnet and special one off magnets we can use dumps. However this is not practical for long
strings of magnets. (we could develop a cold dump?) For a dump resistor can be used.
G.A.Kirby 4th Nov.08
Mechanical protection of the
Nb3Sn in the strand
At 150 MPa the strand starts to degrade.
What can be done? :
– Don’t let the strand see that stress!
– Make larger coils/magnets with low stresses.
– Protect the filaments with a tough collar and or, more copper
around the filaments (this may be needed for quench).
– Change the strand geometry?
– Put Nb-Ti in low field volumes.
– Arrange to put the low fields in high stress areas. Then
degradation may not be seen in the magnet performance?
High stress
Low field
Use Nb-Ti
G.A.Kirby 4th Nov.08
Cu
Tough collar
Stresses in Conductors
Try to support the turns by bonding and adding new features.
Put Nb-Ti in low field volumes < 8 Tesla
Classical magnet with surfaces that are free to slide above any
friction forces
Magnetic forces
Coil bonded to fix its external
surface
Collaring forces
Stress
Stress uniform
Magnet coil pushing against its support
Compression
stresses
Tensile
stresses
G.A.Kirby 4th Nov.08
Stress redistributes as magnet is
energized however the final maximum
stress much the same as a totally
free coil.
A bonded coil could have lower stress
compared to an identical classic
design.
Avoiding the high stresses with a
Hybrid design
74
conductors
138
Conductors
A hybrid design that uses Nb-Ti material in low field volumes also has the
advantages that it is not limited by the same stress limits as the Nb3Sn. This
also helps with the quench protection. There is also a cost saving using Nb-Ti
instead of Nb3Sn.
The block design lends itself better to a hybrid design than the Cos .
G.A.Kirby 4th Nov.08
Hybrid Nb3Sn and Nb-Ti
Nb3Sn
Nb3Sn
Nb3Sn
Nb-Ti
Nb-Ti
Nb-Ti
Nb3Sn Nb-Ti
Although the Nb-Ti is not
damaged at 150MPa we still need
to take care with the insulation.
G.A.Kirby 4th Nov.08
Low field, high current
instabilities
RRP 0.8 mm - 4.3 K - RRR 270
RRP 0.8 mm - 1.9 K - RRR 270
Ic Jc=3030 A/mm^2 @ 12 T; Bc2=24.92 T
Ic V-I 5 A/sec
Iq V-I 20 A/sec
Iq V-H 0.3 T/min
Iq V-I 5 A/sec
2000
Ic (Jc=3917 A/mm^2 @ 12 T; Bc2=27.65 T)
Ic V-I 5 A/sec
2000
1800
Iq V-I 20 A/sec
Iq V-I 5 A/sec
Current, I (A) .
1600
Current, I (A)
No quench: Measurement
System Limit
1400
1200
1000
1800
Iq V-I 2 A/sec
1600
Poly. (Ic (Jc=3917 A/mm^2 @ 12 T;
Bc2=27.65 T) )
1400
1200
1000
800
800
600
600
400
400
y = -0.5901x3 + 31.065x2 - 649.92x + 5369.4
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16
Applied Magnetic Field, Ba (T)
0
1
2
3
4
5
6
This should not be a problem for high field (15 to 17 T)
magnets as the current in the strand is under 400A
G.A.Kirby 4th Nov.08
7
8
9
10 11 12 13 14 15 16
Applied Magnetic Field, Ba (T)
Data from Bernardo Bordini
 Magnet design using HD2 cable 51 stand
Cos
Sliding coil
mid-plane stress
300
0A
5000A
1000A
250
15000A
17000A
stress (MPa)
200
20000A
21700A
150
100
13 T
50
0
0
5
10
15
20
25
30
35
40
45
coil width (mm)
150 MPa @ 13 T
G.A.Kirby 4th Nov.08
Field in coil @ 13 T
Magnet Designs
OST RRP 51 - 0.8mm dia. Strands.
16.40 Tesla Central field at 99% short sample
18.05 Tesla maximum in the coil all @1.9K
51 strand OST cable Ic
90000
4.3K
80000
1.9K
70000
16.3 T ss
Ic in cable
60000
50000
40000
LL
14.7 T central
18.05 T ss
4.5 K
16.4T central
1.9 K
30000
20000
10000
0
10.0
11.0
12.0
13.0
14.0
15.0
16.0
17.0
18.0
19.0
Field [T]
Sliding coil @ 15.5T
93%ss @ 1.9K
G.A.Kirby 4th Nov.08
Bonded coil @ 15.5T
93%ss @ 1.9K
Stress comparison at 15.5T
PIT 40 strands 1.25mm diameter block design
V
HD2 cable RRP 51 strand 0.8 diameter block design
PIT 40 strand
25mm wide cable
15.5T
G.A.Kirby 4th Nov.08
HD2 51 strand
20.8mm wide cable
15.5T
Iron Yoke diameter is large at
High fields
13 T
750mm
diameter
G.A.Kirby 4th Nov.08
16.5T
Stray fields <5 Gauss
@ surface of Iron
1600mm diameter
Stray fields
250 Gauss @
5m
Closing Remarks
•
•
•
•
•
Lots of development work is needed!
Quench heaters that work in Nb3Sn at 1.9K low field volumes!
Some problems can be avoided by design!
– Use Nb-Ti in low fields, then we have a quench heater that work.
– Use Nb-Ti in low fields, high stress not a problem for Nb-Ti.
– Change the strand design so that it can withstand the high stress. Internal
protection for the filaments,
What temperature to use 4.5K or 1.9K?(at 1.9K many magnets show damage!)
Small Strands in Wide Cables using block design hybrid magnets.
–
•
Smaller strands are more stable!
For the “Fresca” replacement high field, large aperture magnet planed at CERN an
operating current of under 17KA.
G.A.Kirby 4th Nov.08