SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 1/31
React and Wind Magnet TechnologyProgram Plans and Special Conductor Requirements
Ramesh GuptaSuperconducting Magnet DivisionBrookhaven National Laboratory
Upton, NY 11973 USA
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 2/31
Outline of the Presentation
• Motivation behind “React and Wind” Approach
• Special Requirements for “React & Wind”
Conductor/Cable Issues - dialogue with industry
• Program Experience (BNL and Fermilab)
• Program Plans (BNL and Fermilab)
• Update on React & Wind HTS
• Summary
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 3/31
Motivations for React & Wind Approach
• React & Wind approach eliminates the need to deal with the differential thermalexpansions between various materials of coil modules during high temperaturereaction process. The issues become more critical as magnets get longer.
Wind & React technology will require a number of long furnaces; React &Wind does not.
In Wind & React approach, the integrated build-up of differential thermalexpansion and associated build-up of stress/strain on brittle Nb3Sn duringreaction process is proportional to the length of magnet. This could have asignificant impact on magnet manufacturing and on magnet performance.
• React & Wind approach allows one to use a variety of insulation and othermaterials in coil modules as the coil and associated structure are not subjected tothe high reaction temperature.
• React & Wind approach appears more adaptable for building long magnets byextending present NbTi manufacturing techniques and tooling. One must look intogeneral differences between long and short magnets. However, unlike in Wind &React technology, no new complications/issues are expected.
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 4/31
Challenges with React & Wind Approach
• The conventional pre-reacted Nb3Sn Rutherford cable is brittle and isprone to significant degradation or even damage during winding andother operations.
• Bend radius degradation is an important issue and plays a major rolein developing conductor designs, magnet designs and magnet tooling.
Flexible cable approach is an example of working onconductor/cable design and common coil on magnet design.
• The magnet design and manufacturing process must be developedand proven by a successful test to demonstrate that the react and windtechnology can be used in building high field Nb3Sn acceleratormagnets.
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 5/31
Conductor R&D for React & Wind Approach
• Bend strain issue is much more critical for React & Wind designs.Nb3Sn superconductor made with different manufacturingtechnologies may have quiet different bend strain properties.
Study differences between Modified Jelly Role, Internal Tin,Powder in Tube.
R&D for increasing bend strain tolerance in each (new design?).
• Reaction process is important. Sintering between wires within thecable must be avoided.
Need more R&D on the treatment of cable before hightemperature reaction and on the design of reaction spool, etc.
• Are there alternatives to Rutherford cable that may be moresuitable for carrying high currents?
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 6/31
Axial Strain Studies
20
30
40
50
60
70
80
90
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Nb3Sn wire; Oxford Ins truments
Ic (A)
Ic -Unloaded (A)
Crit
ical
Cur
rent
(A)
Applied Strain (%)
B = 16 .5 TT = 4 K
εm
0 .29 %ε
irr 0.73 %
AA
B
C
D
E
F
G
H
I
J
A'B'
C' D'E'
F'
G'
H'
I'
J '
K'
--
000822/M010322
High niobium density (50% High niobium density (50% ncanca) Nb) Nb33Sn (Oxford)Sn (Oxford)
J. E
kin,
N. C
hegg
or, e
t al,
NIS
T, L
TSW
‘01
~0.3 % axial strain seemsto be acceptable.Perhaps ~0.5% may betolerable, if “high strain”and “high field” are not atthe same location (as isthe case in the mostdesigns of acceleratormagnets).
A direct bend strain study is, however, more relevant for accelerator magnets.
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 7/31
Bending Strain in 6x1 Cable(Cable made with 0.33mm ITER Strand)
y = 8864.2e-0.0214x
y = 2038.9e-0.0213x
100
1000
30 40 50 60 70 80 90 100 110 120
Field, kG
I, A
100
1000
10000
Jc, A
/mm
2
After-insulationBefore-InsulationJc-H curveExpon. (Jc-H curve)Expon. (After-insulation)
N-value @ 8T ~ 30 for 25 mm bend radius.
When bent around a 16 mm radius mandrelJc does not change but the n-value dropsto ~ 25, indicating a degradation.
Cable made with Mobil 1coated strand. After HeatTreatment, cable insulatedwith 25 mm Kapton film.
0
20
40
60
80
100
120
140
160
0 100 200 300 400 500 600
I, A
V, µ
ΩµΩ µΩµΩWrapped around a 16 mmradius mandrel
25 mm bend radius
Arup Ghosh, BNL, LTSW’01
Bending strain degradation:None in Ic till 8T.
Observed degradation in n-value: Small for 0.66% (R=25 mm) Noticeable for ~1% (R=16 mm)
•Performance of high Jc wire?
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 8/31
Bend Strain Studies at Fermilab
Fermilab has made a number of studies on bend strain tolerance onwire and some on cable. Most of them have been reported earlier.
10/28/02 G. Ambrosio - Conduc tor R &D at Ferm ilab for Nb3Sn R eact-and-Wind 6
Bending degradation of wires
Reaction sample holder
Measurement sample holders
o HT on reactionsample holder(diam = φφφφ1)
o Measurement onITER sample holder(diam = φφφφ2)
o Bending is given byφφφφ1 < φφφφ2
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 9/31
Results from Fermilab
10/28/02 G. Ambrosio - Conduc tor R &D at Ferm ilab for Nb3Sn R eact-and-Wind 8
Bending degradation of High-Jc wires
0.7
0.75
0.8
0.85
0.9
0.95
1
8 9 10 11 12 13 14 15Magnetic Field [T]
I c(be
nt)/I
c(unb
ent)
OST 0.24%
OST 0.47%
0.8
0.85
0.9
0.95
1
12 13 14 15 16Magnetic Field [T]
I c(be
nt)/I
c(unb
ent) IGC 0.23%
IGC 0.44%
Courtesy of E. Barz iThe critical current degradationdue to bending of 0.7 mm wiresis 5-7 % @ 12T, 0.24% εεεεmaxfor IGC (IT) and OST (MJR)
• diameter: 0.7 mm• Jc = 1904 A/m m2 @ 4.2K 12T• C opper: 47 %• twist pitch: 13 mm• sube lem ents: 54• “ thic k” Nb barrier
• diameter: 0.7 mm• Jc = 1676 A/m m2 @ 4.2K 12T• C opper: 38 %• twist pitch: 13 mm
Larger bending degradation in high Jc wires as compared to low Jc wires •Degradation depends on the wire manufacturing process
Courtesy: E. Barzi
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 10/31
Results from Fermilab on Cable
Useful studies; we need more such studies for various bendingparameters for various cables/wire/heat treatment, etc.
Courtesy: P. Bauer
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
8 9 10 11 12
µ0H (T)cr
itica
l cur
rent
deg
rada
tion
bent wo/cstrand - cablingstrand - cabling and 0.08% bendingstrand - cabling and 0.25% bendingstrand - cabling and 0.53% bending
0.3 mm stran d cable
Bending degradation of cablesBending degradation of cables- results -- results - †
bold lines - m easure ments,dashed lines: calculations using Ekin’s model.
†P. Bauer et al. “Fabrication and Testing of Rutherford-type Cables for React and Wind AcceleratorMagnets” IEEE Trans. On Applied Superconductivity, vol. 11, no.1, 2457, March 2001.
Critical currents of 0.7 mm ITER strand cables w/wout core at 20 MPa transverse pressure w/wout bending strain - relative to virgin strand
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
8 9 10 11 12µ0H (T)
Ic re
lativ
e to
ext
rapo
late
d vi
rgin
cab
le
straight wc run 1 sample 1a&bbent wc run 1 sample 2a&b
bent wc run 1 sample 3astraight wc run 2 sample 1a&b
bent wc run 2 sample 2a&bstraight woc run 3 sample 1a&bbent woc run 3 sample 2a
bent wout c run 3 sample 2b
Courtesy of P. B aue r
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 11/31
Reaction Spool and Tooling at Fermilab
10/28/02 G. Ambrosio - Conduc tor R &D at Ferm ilab for Nb3Sn R eact-and-Wind 4
Assembly procedure - 1
Synthetic oil is used in order toprevent sintering between thetwo layers of wires in the cable,
o Some synthetic oil is used duringcabling,
o More synthetic oil is added beforeheat treatment
Cable is reacted inside a retorto Single layer spool,o A gap is left between the core of
the spool and the first turn
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 12/3110/28/02 G. Ambrosio - Conduc tor R &D at Ferm ilab for Nb3Sn R eact-and-Wind 25
Quench history
Computed Ic frommeasurement of thecritical temperatureof the cables facingthe spot heaters~ 97%
Iq = 71% @ 20A/s 4.5KIq = 78% @ 75A/s 5.1KLocal degradation
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 13/3110/28/02 G. Ambrosio - Conduc tor R &D at Ferm ilab for Nb3Sn R eact-and-Wind 29
Temperature marginin the innermost turn
Current kept at constant valueo I = 7, 10 , 11.5 kA
Spot heater was connected toa DC power supply
o Heater current was raised veryslowly
Temperature was recorded byadjacent sensor
ANSYS model was used tocompute the temperature inthe hot spot
Generation temperature wascomputed taking into accountthe bending degradation
o Several models used (range shownby error bar)
4
5
6
7
8
9
10
11
12
13
14
6 7 8 9 10 11 12
Current (kA)
Crit
ical
tem
pera
ture
(K)
measured "critical" temperature bottom coil (K)
measured "critical" temperature, after correction
average model calculation
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 14/3110/28/02 G. Ambrosio - Conduc tor R &D at Ferm ilab for Nb3Sn R eact-and-Wind 30
Possible cause of local degradation
Inspection of cable leftoversfour bumps in a 5” long region repeated every 44” the print is less and less sharp as the regions move away
from the beginning of the cable it is the print of the copper shims used to have a smooth
transition from the first to the second turn bottom coil leftover: 10 bumps (the cable was much longer) top coil leftover: 6 bumps (the cable was six turn long) some bumps should be in the winding of the top coil
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 15/3110/28/02 G. Ambrosio - Conduc tor R &D at Ferm ilab for Nb3Sn R eact-and-Wind 34
Strain in the Racetrack coilStrain in the OST cable used in the 2nd Racetrack
innermost coil turn outermost coil turnWire OST OST OST OSTstrand diameter d mm 0.7 0.7 0.7 0.7outer filam. diam / strand diam. 0.88 0.88 0.88 0.88outer filament diameter φ mm 0.616 0.616 0.616 0.616starting radius (in the spool) mm 253.5 253.5 180 180final radius (in the magnet) infinite 90 infinite 132.3Max strain (strand diameter) ε1 % 0.121 0.221 0.171 0.062Max strain (sintered strands) ε2 % 0.260 0.472 0.366 0.132Position 1 2 3 4
−+=
−=
122
121
112
112
RRd
RR
φε
φε
12
34
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 16/31
10/29/02 G. Ambrosio - FNAL Single Layer C om mon Coil 19
Technological model winding
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 17/3110/29/02 G. Ambrosio - FNAL Single Layer Common C oil 22
Status and plans
The technological model has been completedo Some small modifications end parts
We are practicing with:o The inner spliceo The instrumentation
The tensioners have been modified
The cable for both coils has been reacted
The winding of the first model will begin soon
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 18/31
React & Wind Program at BNL and at FNAL
•React & Wind programs at BNL and Fermilab have some similarities butmostly they are complementary.
•This is an ideal case as we do not want to repeat every thing but it ishealthy (perhaps necessary) to have more than one group working on a newand technologically challenging R&D. Some time little details and biasesmake or break a critical item and hence may cast an incorrect impression onthe viability of the entire technology. In this connection, a recent visit byGiorgio Ambrosio of Fermilab to BNL was quiet useful.
•Fermilab (in collaboration with LBNL) has worked more on the conductorand cable issue and BNL on 10-turn coil rapid turn around program.
•Fermilab uses 0.7 mm wire in their common coil magnet design with aminimum bend radius of 90 mm. BNL uses 0.8 mm wire in our common coildesign with minimum bend radius of 70 mm (more aggressive approach).
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 19/31
Reaction Process at BNL
BNL has four reaction spools. The bending radii of smallspool (on left) happens to be twice the minimum bendradius of our common coil design.Below (right) is a new oil impregnation setup (made afterGiorgio’s suggestions) to vacuum impregnate the cablebefore reaction to minimize the chances of sintering.
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 20/31
Construction & Test Results
7000
7200
7400
7600
7800
8000
8200
8400
0 1 2 3 4 5 6 7 8Quench Number
Cur
rent
(Am
ps)
@ 4.507 K@ 4.574 K
@ 4.34 K
Temperature excursion to establish that quenches are
conductor limited
BNL has made nineteen 10-turn coils using “React & Wind” Technology with Nb3Sn & HTScable. Nine tests have been carried out so far.In two tests we had problem, the same problem.Thank God it was a serious problem -– only 10-20% of short sample – as serious problems aregenerally easier to locate and fix. We blame it on the cable getting highly damaged from a wiremesh during reaction. FNAL points to similar excuse in one case.
The last test withITER cable (oneafter shown or right),the magnet went tocable short sampleon first quench itself.
Lessons learned: Treat Nb3Sn with respect.
A Perfect Test Result of a “React & Wind” Test Magnet:
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 21/31
Next in 10-turn Coil Program
• Make two coils with high performance vacuum impregnated cable.
• React cable on a spool that has about twice the radius of that in themagnet coil. This lowers the effective bend strain by a factor of two.
• Eliminate the machine insulation step and be extra watchful in rest of theconstruction to avoid any potential of damage/degradation.
• Make one coil with almost no instrumentation (to minimize the potential ofdamage) and other with as much instrumentation as possible (to helplocate the problem spot). We have been putting one voltage tap each turn.
• We hope that the above coils produce good results=> retain 12T design.• A still poor performance would indicate that though we did not exceedthe bend strain tolerance in ITER conductor, we did in the highperformance Nb3Sn. In that case, build two 10-turn coils with a lower bendstrain design. A good performance in the last two coils point to a requiredmodification in the 12 T design with a lower bend strain.
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 22/31
Experience withRapid Turn Around Program
We have been arguing for rapid turn around program for about ~5 years.
In last 2-3 years, we have made 20 test coils with brittle pre-reactedmaterial (15 with Nb3Sn cable and 5 with HTS cable) and tested 10+ testmagnets in a cost effective magnet program.
The program has been successful in producing what it was expected to:Aggressive magnet R&D - generating both good and bad results(learning experiences).
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 23/31
BNL 12 T Nb3Sn Common CoilBackground Field Dipole
Nb3Snconductorfor both
inner andouter layersis provided
by OST
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 24/31
Insert Coil and Sample Test Scenarios
An interesting feature of the design, which will make it a truly facility magnet, isthe ability to test short sample and HTS insert coils without disassembling it.
HTS insert coil test configuration Short sample test configuration
HTS Coil SS Fixture
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 25/31
Magnets with Flexible Wire
• The Lorentz forces are containedin the individual blocks and do notpile up on the midplane as inconventional cos Θ magnets
Erich Willen
Recently flexible pre-reacted Nb3Sn wire has become available. BNLis trying to use that in magnets in magnets that require small bendradii in the ends (example LHC IR upgrade and muon collider)
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 26/31
Progress in BSCCO2212 at Showa(Investment in HTS and Current Capacity)N
ot for Distribution
Not
for D
istr
ibut
ion
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 27/31
Progress in BSCCO2212 at Showa(Progress in Ic and It’s Uniformity)
Not for D
istributionNot
for D
istr
ibut
ion
Tested at BNL also(Cable 5)
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 28/31
BNL Measurements of Various Cables from Showa(Note: Continuous progress in cable performance)
Ic as a function of T in various Bi2212 cable from Showa
0
50
100
150
200
250
300
350
400
55 60 65 70 75 80T (K)
Ic, A
mp
(for 1
µµ µµV/c
m)
Cable 5 (1 mm)Cable 4 (1 mm)Cable 3 scaled to 1mmCable 3 (0.8 mm)Cable 1 scaled to 1mmCable 1(0.8 mm)
(without self field correction)
Extrapolated 4 K performance of 20 strand cable (#5) (wire dia = 1 mm) :~5 kA at high fields and ~9 kA at zero field
Test Dates:Cable 1: 06/00Cable 3: 01/01Cable 4: 07/01Cable 5: 11/02
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 29/31
Performance of HTS Coil in theBackground Field of Nb3SN Coils
Short sampledefinition for HTS
Measured electrical Resistance of HTS coilin the background field provided by variousNb3Sn coils in the magnet DCC008R
Field in various coilsNb3Sn HTS Nb3Sn
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 30/31
Performance of HTS Coil in theBackground Field of Nb3SN Coils
Performance of HTS cable in coil (before and after winding)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
0 0.5 1 1.5 2 2.5 3 3.5 4H(T)
I(kA
)
Ic,kA Iss(kA)I(Nb3Sn)=9kA I(Nb3Sn)=6kAI(Nb3Sn)=0kA I(Nb3Sn)=3kA
INb3Sn=0kA INb3Sn=0kA INb3Sn=0kA INb3Sn=0kA
DCC008R
Ic Before Winding
Ic After Winding
HTS coil was subjected to various background field by changing current in“React & Wind” Nb3Sn coils (HTS coil in the middle and Nb3Sn on either side)
SuperconductingMagnet Division
R. Gupta, BNL, React & Wind Talk, Napa Valley, CA Nov. 11, 2002 Slide No. 31/31
Summary•React and Wind approach has a potential to offer asignificant advantage in developing Nb3Sn magnettechnology that is scalable for large scale production oflong magnets.
•The performance of Nb3Sn React & Wind magnets withITER has been impressive in low to medium field magnets.
•We need to do more conductor and magnet R&D todemonstrate similar success with high performance Nb3Snin high field React and Wind magnets.
•React & Wind HTS conductor and magnet programcontinue to show progress.