Transcript slides - GSI IndiCo

Quench protection and wire properties_Introduction
- Strands with a Cu matrix
- Resistivity of bulk CuMn
- Examples of cables for SIS100 dipoles (1 and 2 layers, Cu and CuMn)
- Increase of the linear resistance and hotspot temperature with using CuMn
- Possible strand specification for the SIS300 dipole
- Cable linear resistance as a function of =(Cu+CuMn)/NbTi, CuMn/Cu, ds, RRRCu, CuMn
- Minimum values of Ic and JcNbTi (influence on temperature margin)
- Possible templates for strand specification
- Proposed measurement during strand series production
- Quench calculations for the SIS100 single layer dipole (Cu and CuMn inter-filamentary
matrix)
E. Floch.GSI.
Intas_wire_19Fe09
Strands with a Cu matrix
- For the first 100 LHC dipoles: 130 < RRRCu < 280 with an average of 210 (more than 200
measurements)
- LHC specification RRRCu > 70 (probably copied from SSC or RHIC wire specifications)
- Proposed FAIR wire specification: 100 < RRRCu < 280
- The highest hot spot temperature (Tm) is reached for:
RRRCu_min when using dump resistors (SIS100 dipoles, quadrupoles, SIS300 quadrupoles)
RRRCu_max when using dump resistors + quench heaters and cold or warm bypasses (SIS300 dipoles)
- Computations done for the SIS100 2 layer dipole:
ds = strand diameter (mm)
0.508
0.492
0.492
0.492
 = Cu/NbTi
1.5
1.5
1.3
1.3
RRRCu
280
280
280
100
Tm_5.41E6A2.s (K)
196
255
297
450
Tm_7.32E6A2.s (K)
450
620
745
>900
- RRRCu, ds and  very strongly influence Tm,
- the variations of these 3 parameters must be specified and measured on each spool piece during series production
- the SIS100 quench protection scheme with resistors must be defined for the minimum values of RRR Cu, ds and 
E. Floch.GSI.
Intas_wire_19Fe09
Resistivity of bulk CuMn0.5%wt
Resistivities of CuMn0.5%wt found in the literature
CuMn0.5%wt_bulk (.m) measured by
cold
warm
RRR
[email protected]
2.135 10-8 at 4 K
3.60 10-8 at 295 K
1.69
Bochvar after cold drawing
1.62 10-8 at 4 K
3.39 10-8 at 295 K
2.09
IEEE Trans. On Mag., Vol24, N°2,
pp1145-1148, March 1988.
1.5 10-8 at cold
IEEE Trans. on Applied Sup. Vol3, N°1,
p 859, 1993
2.5 10-8 at 12 K
Given by Luvata for the SIS300 dipole strand
2.135 0-8 at 10 K
3.6 0-8 at 295 K
1.69
1.6 10-8  1.9 10-8
3.3 10-8   3.6 10-8
proposal for the FAIR magnets
- The resistivity will depend on the % of Mn which could change during series production
- Proposal : measurement of CuMn0.5%wt_bulk at 4, 77 and 300 K on each billet during series production
 = (Cu+CuMn)/NbTi is no more sufficient to have the different material cross-sections
proposal : give  and CuMn/(NbTi+Cu+CuMn) or CuMn/(Cu+CuMn) or CuMn/Cu
E. Floch.GSI.
Intas_wire_19Fe09
SIS100 cables
(single or 2 layer dipoles, Cu and CuMn inter-filamentary matrix)
C2LD
C2LD
CSLD8b
CSLD8b
2
2
1
1
6500
6500
12746
12746
inner diameter CuNi tube
(mm)
4
4
4.7
4.7
n° of strands
31
31
23
23
strand diameter (mm)
0.5
0.5
0.8
0.8
 = (Cu+CuMn) / NbTi
1.38
1.38
1.5
1.5
Inter-filamentary matrix
Cu
CuMn
Cu
CuMn
18144
18144
18144
18144
df: filament diameter (micron)
2.41
2.41
3.76
3.76
s/df
0.1
0.1
0.1
0.1
0.241
0.241
0.376
0.376
CuMn_IF_12K (ohm.m)
1.89E-08
1.89E08
1.79E08
1.79E08
CuMn_IF_295K (ohm.m)
3.66E-08
3.66E08
3.56E08
3.56E08
0
0.85
0.00
1.55
3.53
2.67
6.94
5.39
0
14.04
0.0
13.4
curved dipole type
n° of layers
I (A)
n° of filaments
s: inter-filamentary spacing
(micron)
ACuMn (mm2)
ACu (mm2)
ACuMn/A_all_strands (%)
E. Floch.GSI.
Intas_wire_19Fe09
SIS100 cable linear resistance and Miits curves
(ohm/m)
8.00E-03
7.00E-03
rl_C2LD_Cu (ohm/m)
rl_C2LD_CuMn (ohm/m)
rl_CSLD-8b_Cu (ohm/m)
6.00E-03
CL2D: 2 layer dipole
rl_CSLD-8b_CuMn (ohm/m)
CS LD-8b: single layer dipole
5.00E-03
4.00E-03
3.00E-03
2.00E-03
1.00E-03
T (K)
0.00E+00
0
50
100
150
200
250
300
350
400
(K)
500
For the 2 layer dipole, Tm increases from 300
Tmax_C2LD-a-Cu (K)
Tmax_C2LD-a-CuMn (K)
450
400
to 500 K for Miits = 6 MA2.s when the inter-
350
filamentary matrix uses CuMn
300
250
200
150
100
50
Miits (1E6A2.s)
0
0
1
2
3
4
5
6
7
8
E. Floch.GSI.
Intas_wire_19Fe09
Current dumping for SIS100 dipoles
- The current is dumped with a time constant = Lstring / Rd (Rd: total dump resistance)
- The maximum coil to ground voltage (Vcgm) occurs when one dump resistor fails to open
Curved dipole version
2 layer
I (A)
Inter-filamentary matrix
6500
Cu
1 layer
6500
CuMn
12746
Cu
12746
CuMn
 = (Cu+CuMn) / NbTi
1.38
1.38
1.5
1.5
strand diameter (mm)
0.5
0.5
0.8
0.8
5.54
4.46
17.61
14.44
 (s)
0.174
0.139
0.144
0.125
Vcgm (V)
1643
2060
1026
1185
Mitts',tf,  = Miits350K (1E6A2.s)
- Computations were done with average and ds_average (should be redone with min and ds_min)
- Using CuMn forces to dump the current faster which leads to higher maximum coil to ground voltages (V cgm)
- Single layer dipole is much easier to protect than the 2 layer dipole
E. Floch.GSI.
Intas_wire_19Fe09
SIS300 cable with CuMn inter-filamentary matrix
specified in
Sep_2007
FDR / Nov_2008
FDR / Nov_2008.
used in quench calculations
1st generation
2nd generation
0.825
0.825 ± 0.003
0.825 ± 0.003
(Cu+CuMn)/NbTi
1.6
1.56+/-0.1
1.56+/-0.1
filament diam. (micron)
2.5
3.5
2.5
%_NbTi
38.5
39
39
%_CuMn_0.5%wt
17.6
17.1
24.7
%_Cu
43.9
43.9
36.3
36
36
36
Transposition pitch (mm)
100  5
100  5
100  5
stainless steel core 316L
13mm*25 micron
13mm*25 micron
13mm*25 micron
>70
>70
>70
.
wire diam (mm)
n° of strands
RRR_Cu
For the specification of 2007:
RRR_Cu
RRR_cable
100
200
300
87
172
257
Cable specification established
by INFN for
the 4.5 T SIS300 curved dipole
- For a Cu matrix, the RRRCu measurement is sufficient
- When using CuMn, the wire producer has to give rl _strand at 10 and 300 K (in ohm/m)
E. Floch.GSI.
Intas_wire_19Fe09
Linear resistance of a cable with CuMn
- The billet is designed to achieve the average value: av = (Cu+CuMn)/NbTi and the corresponding proportions
pCub = ACu/A, pCuMnb = ACuMn/A, pNbTib = ANbTi/A with A=ACu+ACuMn+ANbTi (b stands for billet)
-  can be measured on strands but pCu and pCuMn probably not
- To compute the real values of pCu and pCuMn in the strands, we are obliged to assume : pCuMn/pCuMn = pCuMnb/pCub.
With this assumption, we have:
pCu 

1
p
  1 1  CuMnb
pCub

and
pCuMn 

1
  1 1  pCub
pCuMnb

The linear resistance of a SIS300 cable with CuMn inter-filamentary matrix is given by:




1
 1 
1
1

r( RRR Cu , B, T ) 



2
 
d




p
p
 s
  Cu ( RRR Cu , B, T )  1  CuMnb   CuMnIF (T )  1  Cub  
4
pCub 
pCuMNb  



1
E. Floch.GSI.
Intas_wire_19Fe09
Variations of the Linear resistance of a cable with CuMn
The cable linear resistance depends on PCuMnb/PCub, , ds, RRRCu and CuMnIF (IF for inter-filamentary matrix):
r(T ) min 
r(T ) max 
1
d s2max

4




 1 
1
1

 max



 max




p
p
  Cu ( RRR Cu  280, T )  1  CuMnb   CuMnIF min (T )  1  Cub  
pCub 
pCuMNb  



1
d s2min

4




 min  1 
1
1





 min




p
p
  Cu ( RRR Cu  100, T )  1  CuMnb   CuMnIF max (T )  1  Cub  
pCub 
pCuMNb  



1
1
- To make Quench calculations, we must know in which interval rl varies
- This interval must be defined in advance: that means the interval is in the cable specification:
CuMnbulk = bulk resistivity of CuMn0.5%wt
(.m)
1.6 10-8  1.9 10-8 at 4 K
3.3 10-8   3.6 10-8 at 300 K
r= cable linear resistance (in /m)
r(10K)min   r(10K)max
r(300K)min   r(300K)max
r( SIS300 dipole)
0.00025  r(10K)  0.00075 /m
0.060  r(300K)  0.065 /m
- The measurement on each spool piece must give rl(10K) and rl(300K) and not RRRstrand
E. Floch.GSI.
Intas_wire_19Fe09
Minimum critical current density
- The knowledge of Ic and Jc strand is of importance to analyze the training behavior
and compute the temperature margin (which will be used to analyze beam induced quenches)
- The strand specification gives : IcNbTi(Bref) > IcNbTi(Bref)min. If there is no current sharing between strands,
the corresponding minimum critical density is:
J c (4.2 K , Bref ) min  I c (4.2 K , Bref ) min 
- For SIS300 dipoles and quadrupoles : JcNbTi(5T)min  2700 A/mm2,
 1
4
 max
2
1
  d s min
- For the SIS100 dipole: Bref = 2 or 5 T?, if 2 T, what value for JcNbTi(2T)?
Influence of Jc on Tcs (SIS100 straight dipole, load lines computed by C. Muehle in 2007)
(these load lines not the same than those measured on the BNG prototype)
Bc(4.7K,I) computed considering
Jc_NbTi(4.22K,B) of EAS wire
produced in 2004
actual strands
produced in 2006
JcNbTi(4.22K,5T) (A/mm2)
3000
2662
JcNbTi(4.22K,2T) (A/mm2))
5490
4324
Iml_4.7K_computed= max on load line (A)
10040
8950
71
80
Ic(2.26T, 4.7K) (A)
11737
9878
Tcs(7166A, 2.26T)
6.11
5.70
Tma=Tcs(7166A, 2.26T)-4.7 (K)
1.41
1.00
I0/Iml_4.7K_computed (%)
The influence of Jc on Tcs-THe is strong. The variations of  and ds on Tcs-THe must also be considered
E. Floch.GSI.
Intas_wire_19Fe09
Proposal for cable specification templates
The values inside the templates are only indicative. Real values will be defined by magnet designers
For a strand having a Cu inter-filamentary matrix
specification
dsaverage= strand diameter
0.5 mm
ds
 2.5 m
average = Cu/NbTi
1.5

 0.08 or  0.1
RRRCu
100   280
Bref (T)
2 T for SIS100 and 5 T for SIS300
Ic(4.2K,Bref) = critical
current of the strand
Warning:
the specification is on Ic not
on JcNbTi
 Ic(4.2K,Bref)min
with
I c (4.2 K , Bref ) min  J cNbTi (4.2 K , Bref ) min 
  d s2min
4

1
 max  1
JcNbTi(4.2K,2T)min = 4700 A/mm2
or
JcNbTi(4.2K,5T)min =2700 A/mm2
E. Floch.GSI.
Intas_wire_19Fe09
Proposal for cable specification templates
For the SIS300 dipole cable with a CuMn inter-filamentary matrix
The values inside the templates are only indicative. Real values will be defined by magnet designers
proposed specification
dsaverage= average
strand diameter
0.825 mm
ds
 2.5 m
average =
(Cu+CuMn) /NbTi
1.6

 0.1
Ic(4.2K,5T) =
critical current of the
strand
 Ic(4.2K, 5T)min= 547 A with
I c (4.2 K , Bref ) min  J cNbTi (4.2 K , Bref ) min 
  d s2min
4

1
 max  1
JcNbTi(4.2K,5T)min =2781 A/mm2 (quite an ambitious value)
CuMnbulk =
bulk resistivity
of CuMn0.5%wt
1.6 10-8  1.9 10-8 .m at 4 K
3.3 10-8   3.6 10-8 .m at 300 K
r= cable linear
resistance
0.00025  r(10K)  0.00075 /m
0.060  r(300K)  0.065 /m
(Ic_strand(4.2K,5T)=547 A was chosen by INFN in Autumn 2007)
E. Floch.GSI.
Intas_wire_19Fe09
Aim of quench protection, example of SIS100 dipole
- Like LHC magnets, SIS100 and SIS300 dipoles and quadrupoles will have their coil to ground voltage (Vcg)
tested at 3 kV at warm and at cold.
- The EU law for AC applications states: Vtest = 2 * Vmax + 500, Vmax = Vcgm should be kept below 1250 V
- Aim of the protection system for SIS100 and SIS300 dipoles and quadrupoles:
Vcgm < 1250 V
hotspot temperature Tm < 350 K when protection scheme works fine
Tm < 350 K for a defined failure of the protection scheme
- The 108 SIS100 dipoles (in series) are protected with 6 dump resistors
- Because quench threshold in bus bars (Vthb) = 2 * quench threshold in magnet (Vthm),
the hotspot temperature in bus bar (Tmb) > hotspot temperature in magnet (Tmm )
- The dump resistance (Rd) in chosen so that Tmb = 350 K when 6 six dump resistor are activated,
- If one dump resistor opens of a delay of tf:
tf < tfm for which Tmb = 450 K
Vcgm = 7/36*Rd*I < 1250 V
E. Floch.GSI.
Intas_wire_19Fe09
Strand example for SIS100 single layer dipole
Strand proposed by H. Mueller in Feb 2009
n° strands
23
ds = strand diameter (mm)
0.8
ds
=(Cu+CuMn) / NbTi

proposed by
 0.005
1.4
0.1
HM
EF
4600
5030
Ic_min(max=1.5,,ds_min=0.795 mm,4.2K,2T)
(A)
741
810
Ic_min(al_max,ds_min,THe,Bmax) (A)
735
803
5
5
Tcs-THe_max=1.5,ds_min=0.795 mm (K)
0.83
1.05
Tcs-THe_ min=1.3,ds_max=0.805 mm (K)
1.11
1.31
Jc_NbTi_min(4,2K, 2T) (A/mm2)
THe (K)
(in 2007, the single layer dipole version CSLD-8b considered = 1.5)
E. Floch.GSI.
Intas_wire_19Fe09
Quench study for SIS100 single layer dipole (CSLD-8b)
I (A)
12746
L (mH)
0.553
300
(Cu+CuMn)/NbTi
1.3
Inter-filamentary matrix
Cu
8.17
8.17
11.42
11.42
A_NbTi (mm2)
4.96
4.96
500
A_(Cu+CuMn) (mm2)
6.45
6.45
n° of filaments
18144
18144
3.89
3.89
0.1
0.1
0.389
0.389
ro_cuMn_bulk_max_4K (ohm/m)
1.60E-08
1.60E-08
Tmax2_RRR=80, beta=1.38 (K)
ro_cuMn_bulk_max_300K (ohm/m)
3.60E-08
3.60E-08
Tmax1_RRR=80, beta=1.38 (K)
ro_cuMn_IFM_4K (ohm.m)
1.77E-08
1.77E-08
ro_cuMn_IFM_300K (ohm.m)
3.77E-08
3.77E-08
ACuMn (mm2)
0.00
1.66
ACu (mm2)
6.45
4.79
strand diameter (mm)
CuNi
CuMn
(K)
Vthb (quench threshold in bus bars) (mV)
0.795
A_all_strands_min (mm2)
400
df: filament diameter (micron)
Only the 31 strands
warm up
s/df
The whole cable
300
xf = 0.47
s: interfilamentary spacing (micron)
1-xf=0.53
200
Tmax_measured RRR=80 (K)
100
Tm_computed_RRR=80,B=2T (K)
Miits (1E6 A2.s)
0
0
1
2
3
4
5
6
7
8
(3 experimental points measured in Dubna on Nuclotron dipole)
9
E. Floch.GSI.
Intas_wire_19Fe09
Quench study for SIS100 single layer dipole (CSLD-8b)
I (A)
12746
Interfilamentary matrix
Cu
xf
12746
CuMn
0.47
0.47
15.60
12.35
Vpf (m/s)
18.7
18.7
trt (time to reach Vth = 300 mV) (ms)
14.1
11.2
trt+tv+to=tRd (time at I = cte) (ms)
25.1
22.2
0.142
0.108
0.4212
0.555
17.10
13.69
90
94
1044
1376
Mitts350K (1E6 A2.s)
350K=2*( Mitts350K / I2-tRd) (s)
Rd = Lstring /350K (ohm)
Mitts450K (1E6 A2.s)
tfm (ms)
Vcgm (V)
IFM
xf
Vcgm (V)
Cu
CuMn
0.3
0.4
0.47
0.6
0.3
0.4
0.47
0.6
1199
1103
1044
952
1591
1457
1375
1244
- Using CuMn: the increases of Vcgm is between 300 and 400 V and Vcgm
- With CuMn : 1244 < Vcgm < 1591 V which is above our target of 1250 but could still be acceptable
- The best would be to have 2 coils for the SIS100 single layer prototype (one with Cu and the other with CuMn)
so that we can make a choice based on experimental data
E. Floch.GSI.
Intas_wire_19Fe09
Conclusions
- To perform appropriate quench calculations, the intervals in which the cable characteristics vary must be defined
- For a Cu inter-filamentary matrix, specifying 100 < RRRCu < 280 is sufficient
- For a CuMn inter-filamentary matrix, RRRstrand ≠ RRRCu. The specification will be on the linear resistance:
rl(10K)min < rl(10K) < rl(10K)max and rl(300K)min < rl(300K) < rl(300K)max
- Ic_strand(Bref)min should be defined using Jc(Bref)min, min and dsmin.
- Bref = 5 T for SIS300 dipoles, Bref and corresponding Jc should be defined for SIS100 dipoles (2 or 5 T?)
- For the strand series production:
- one complete set of measurement : , ds, RRRCu or (rl(10K) and rl(300K)),
Ic(4.2K,Bref) or Ic(4.2K) at 0.5, 1,2 ,3, 5 and 6 T
- one complete set of measurement every spool piece
- every billet, measurement of CuMn at 4, 77 and 300 K
- Measurements after the cable production are still to be defined
(Detailed explanations on strand specifications are given in the internal note "MT Internal Note: MT-INT-ErF-2008-007")
- Quench calculations on single layer dipole (CSLD-8):
with Cu matrix: the max coil to ground voltage 950 < Vcgm < 1200 V below the upper limit of 1250 V
with CuMn inter-filamentary matrix:1250 < Vcgm < 1600 V above the upper limit of 1250 V
- Best procedure to decide Cu or CuMn is to test 2 different coils inside the single layer dipole prototype
E. Floch.GSI.
Intas_wire_19Fe09