HIGH-C
M.Kor
Abstract At each in
low-resistancsuperconductcurrent path surrounding [1]. About extensive caand map thsuperconductno splices wNon-invasivemade at 300sectors. Morsuspect intercwere repaireimperfect ststatistical anremaining wmaximum opextensive int
Types of spThe main
two quadrupsplices in tosplices (Figu
Magnet spdiodes and incurrent decayare not protewith the timwhich is mucircuit – betwat 7 TeV). Tabout 300 pΩresistance of 17 μΩ for the
Individual methods. Unmeasured is taps. A typicor 3 splices, bus containssuperconductthe splices bBusbar excesbusbar minsuperconduct
CURREN
ratzinos, F. R.ScR.H.
nterconnectionce solder joting cables inthrough the copper stabil10,000 such mpaign has b
he resistancesting cable spliere found wite measuremen0 K in 5 of thre precise locaconnects that ed. Howevertabilizer jointnalysis is us
worst-case resiperating energtervention [2].
SP
plices circuits of th
pole (RQ) perotal. Out of
ure 1) and the rplices are pron case of a quys in less than
ected by diodeme constant ouch larger (anween 10 s for The nominal iΩ at 1.9 K, wf 16cm of coppe RB and RQ
splices are nder non-inva busbar seg
cal busbar segwhereas a typ
s 8 splices. Nting temperat
but also the ress resistance nus its noting temperatu
NT BUS STHE OPE
Bertinelli, Zchmidt, A.S.Flora, H.Pf
n between LHoint must bn order to prosuperconductizer in case tjoints exist
been undertaks of these jices were meath a resistancents of the stahe 8 sectors, al measuremewere opened
r, it is likelts still exist sed to placeistances. Thisgy of the LHC
PLICES
he LHC – oner sector – conthese 10,170
rest are magnotected by thuench, heatern a second. Ines and have toof the energynd depends oRQ at 3.5 TeVinterconnect s
whereas at rooper stabilizer ibuses respectmeasurable oasive conditi
gment betweengment for the pical busbar sNon-invasive tures measureesistance of th
(REXC,BB) is tominal resiures.
PLICE RERATIN
Z.CharifoulSiemko, A.feffer, J.Str
HC main magne made betovide a contintor and also tthe cable quen
in the LHCken to charactoints. All of
asured at 1.9 Ke larger than 3abilizer joints
and at 80 Kents were madup, and poor jly that addit
in the LHe bounds on sets limits oC, prior to a
e dipole (RBntain about 240 are intercoet splices. he magnet byrs are fired annterconnect spo withstand cuy extraction con the energyV and 100 s fosplice resistanom temperaturis about 10 μΩtively. only with invions what can adjacent voRB bus contaegment for th
methods at e the resistanhe busbar segmthe resistancestance at
RESISTANG ENER
lline, K.DahVerweij, CE
rait, Fermila
nets a tween nuous to the nches
C. An terize f the K and 3 nΩ. were
K in 3 de on joints tional C. A
n the on the
more
) and 4,000 nnect
ypass nd the plices urrent circuit y and or RB nce is re the Ω and
vasive an be oltage ains 2 he RQ
non-ce of ment.
e of a non-
Figupoor
MeaM
the analyaccidresisprobnommeasvarie
Capropcapathe Resi(16 detec1 mKusing
Elmeaswhic10 m
Ussplicthe m
Thoperseriethe segmRega
NCES ANGY OF T
hlerup-PeterERN, Genevab, Batavia I
ure 1: (a) An inr copper stabil
asurements Motivation for m
19 Septembysis of a circudent yields thestance at 1.9blem. This is
minal splice ressure the resistety of ways: alorimetry [4
perties of heliacity of the syresistive heatstive power amagnets, 350cted using thK/hour/W. REg this methodlectrical: Wsurements usich covers 6 ou
magnet splicessing both metces were foundmagnets were he enhancedrational in 20es of accurate above metho
ment REXC,SC warding magne
ND IMPLTHE LHC
rsen, R.Denva, SwitzerIL 60510 U
nterconnect splizer continuit
at cold measuring splber 2008 acuit test performe value of 2349 K (REXC,SC)
orders of msistance and htance of as ma
4]: taking bium, together ystem at 1.9 Kt released by
as low as 1 W 0 t of cold mhis method.
EXC,SC as low .
We have aling the quencut of 7 magne (RQ). thods, two sigd, one 100 nΩreplaced.
d QPS (nQP009/2010. Opvoltmeters, itds, especiallywhere the accet splice res
LICATIONC
nz, C. Scheuland.
U.S.A.
plice and (b) aty (2D projecti
lice resistanceccident. The med a few da4±15 nΩ [3] f) in the vic
magnitude highas prompted aany splices as
benefit of thwith the ver
K, it is possiby a high-resisin one cryoge
mass, 250 m The temper
as 40 nΩ can
lso performeh protection s
et splices (RB)
gnificant REXCΩ and the othe
PS) system perateed esset has a better ay when meascuracy is bettesistances no
NS FOR
uerlein,
an example of ion).
es comes fromcalorimetric
ays before thefor the excesscinity of thegher than thea campaign tos possible in a
he superfluidry small heatle to measurestance splice.enic subsectorlong) can be
rature rise isn be detected
ed (ad-hoc)system (QPS)) and 6 out of
C,SC in magneter 50 nΩ, and
[5] becameentially as aaccuracy thansuring busbarer than 1 nΩ.value above
f
m c e s e e o a
d t e . r e s d
) ) f
t d
e a n r . e
Proceedings of IPAC’10, Kyoto, Japan MOPEB044
07 Accelerator Technology
T10 Superconducting Magnets 373
25 nΩ has resistances, anywhere inseen is 2.99±couple of REX
REXC,SC ofoperation; hoor a problemresult in a hig
Measuremetemperatur
A copper superconduct(Figure 1) lim[6]. Such a higher than nresistance is 300 K for thethe RB busbgood knowlecopper stabilon the limit 80 K is very the resistanceresistance ofto measure. change in theof an increasby increasing50 μΩ, wherRRRCS incre
Table 1: Fa
Busbar segment
Typre(μ
RB@ 300K 20
RQ@ 300K 11
RB@ 80K 2
RQ@ 80K 15
A substa
resistances w2009 [1]. measurementhand-held de
The overaon the sectorRQ at roorespectively)4-5.5 μΩ for
been found, no REXC,SC
n the machine±0.02 nΩ (RQXC,SC above 2
f a few nΩ poowever it migm with the sgh REXC,BB.
ents at non-res – non-in
stabilizer witing cable bamits the maxim
bad joint mnominal resist
challenging, e RB bus beinar segment reedge of the telizer (RRRCS)of accuracy, difficult. Tab
e measured atf a segment anColumn 4 she segment by se in the crossg the width areas the last casing from 10
actors affectin
pical esis. μΩ)
Typical defect (μΩ)
000 50 (2.5%)
000 50 (0.5%)
270 7 (2.5%)
500 7 (0.5%)
antial effort was undertaken
Measurements) were take
evice in the tunall analysis acr) of 9-17 μΩ om temperatu) whereas at 8r the RB and
whereas forabove 3 nΩ
e. The highesQ circuit, secto
nΩ. oses no problght suggest a soldering pro
-supercondunvasive ith no continadly solderedmum safe eneanifests itselfance at warmwith only th
ng relatively eaesistances at 8emperature an. Measuring twhereas mea
ble 1 shows tht warm, togethnd of a defectows the effec1 K. Column
s section of thand the heightcolumn show00 to 150.
ng segment res
Temp. change +1 K
Xc
7 (0.4%)
40 (0.4%)
8 (2.9%)
44 (2.9%)
to map all n during the pnts (referreden manually nnel in all sec
chieved an acfor the RB anure (about 80 K the accud 6-12 μΩ for
r busbar segΩ has been st splice resisor 23), with o
lems under nostructural pro
ocedure that c
uctive
nuity coupled d to the stabergy of the maf as a splice . Measuring s
he measuremeasy. For meas80 K we neednd the RRR othe RQ at 300asuring the Rhe factors affeher with the tyt that we are tct of a temper
5 shows the ehe copper stabt of the copps the effect o
sistance at wa
Xsection change
RRcha
12 (0.6%) (0.3
85 (0.8%)
4(0.3
2 (0.6%) (2.6
11 (0.8%)
4(2.6
busbar segperiod April tod to as Busing an acc
ctors. curacy (depen
nd 25-36 μΩ fo0.7% and
uracy was betr the RQ (2%
gment seen
stance only a
ormal oblem could
to a bilizer chine with
uch a ent at suring d very of the 0 K is RQ at ecting ypical trying rature effect
biliser er by
of the
arm
RRCS ange
7 3%)
40 3%)
7 6%)
40 6%)
gment o July Biddle curate
nding or the 0.3%
tween % and
0.7%300 80 Kand tsectoin th(whecolumsectocolum10-1
T
CircuSecto
A12 A34 A45 A56 A67
MeaIn
indivinvatheseafterBiddwas contihigh
Thsplic
Figuresis42 μ
WorSin
exist
% respectivelyK but was m
K. Five sectothe worst spliors were meashe RB of the ere accuracy mn shows theor at the 90%mn 4, as the m7 μΩ.
Table 2: recap
uit/ or
Temperatspread (K
RB 1.1 RB 1.9 RB 0.9 RB 0.4 RB 0.6
asurements addition to tvidual interconsively (referre were made fr an indicatiodle measureme70μΩ, whereinuous RB b
hest R16 exceshe gamma-rayce interconnec
ure 2: The intestance. The siμΩ (the other s
rst remaininnce reliable ints only for par
y). This was smarginal for tors were measices were opensured at 80K. T
sectors measis good) af
e estimated hig% confidence measurement
of RB Biddle
ture K)
REXC,BB spread (μΩ)
13 10 17 9 14
at warm – ithe Biddle mnnect splices w
red to as R16following an on of a high ents. The wor
e the average busbar is aboss resistance (Ry picture of
ct can be seen
erconnect witide seen has side measured
ng splice nformation abort of the mach
ufficient for tthe RQ and isured at roomned up and reTable 2 showssured at roomfter the repaghest excess rlevel. This isaccuracy is o
measurement
Highest remaining REXC,BB( μΩ
37 35 53 20 31
invasive measurements,
were opened a6 measuremenopening of an
excess resisst R16 resistaresistance of out 10 μΩ. TREXC,R16) is 60one side of in Figure 2.
th the worst man absolute
d at 28 μΩ).
out the resistahine, a statisti
the RB bus atinsufficient at
m temperatureepaired. Threes the situation
m temperatureairs. The lastresistance in as higher than
of the order of
ts at 300 K
Ω)
REXC,BB 90%CL (μΩ)
51 47 78 34 48
a number ofand measurednts). Most ofn interconnectstance by thence measureda 16 cm-long
Therefore the0 μΩ. that specific
measured R16resistance of
ance of splicesical method is
t t e e n e t a n f
f d f t e d g e
c
6 f
s s
MOPEB044 Proceedings of IPAC’10, Kyoto, Japan
374
07 Accelerator Technology
T10 Superconducting Magnets
needed to estLHC, Rmax. the RB meameasured RE
(sector A45)repairs 53±1is 60±1 μΩmeasurementexponential fnumber of saThis fit predwhich these 0.8 joints abois REXC,R16 >
The next limits on whsample contain the machindication ofsufficient resthey provideand for decidsectors for thmeasurementthe high-currsampling framaximum ex90% confidsystematic efof the form ogiving an ovfrom this anderived, howworst identifsuperconductbusbar segmwas used forthe LHC was
CONSTRThe safe e
worst splice concentratedof the energycalculations can be toler=100) give [τEE=15 s):
• RB: • RQ:
For 3.5 TeV • RB: • RQ:
If the RRRthe above va
However, the 3.5 TeV lsuch large d
timate the woThe only rel
asurements atEXC,BB before r), whereas the5 μΩ (again s
Ω. The cumults can be seenfunction, is camples abovedicts that thedata are a sa
ove 60 μΩ. T82 (106) μΩ. question is hhat remains iains about 2.3hine, they wef the non-invasolution in teed a meaningfding which inthe RB bus. ts represents rent joints in action leads xcess resistandence level: ffects were coof the distribuverall uncertanalysis of thewever, is still fied failure moting cable re
ment). Thereforr subsequent ds Rmax ≈ 90 μΩ
RAINTS Oenergy limit oin the machin
d on one side oy extraction (τof what is th
rated as a fun[6]: For 5.0 Te
43 μΩ 41 μΩ operation (RB76 μΩ 80 μΩ
RCS is 200 inslues increase bif the defect ilimits becomedefects canno
orst possible spiable Biddle
t 300 K (5 serepairs for thee worst remasector A45). Tlative distribun in Figure 3.onstrained to
e 20 μΩ, and e underlying ample, would,The limit for 0
how to use thin the machin3% of the totaere selectivelyasive measure
erms of REXC,Rful basis for lterconnects toIn this case tabout (1/3) xthe machine.to the follow
nce for all mRmax,R16 =
onsidered bothution and the ainty to the ce order of ±1larger than thode (it would emained unsre, a more rea
discussions onΩ.
ON THE LHof the LHC isne and dependof a splice, onτEE ) and on thhe worst singnction of eneeV operation
B τEE=50 s, RQ
stead of the (by 10 μΩ for is split on bote 120 μΩ and t be concentr
plice present ameasurement
ectors). The we RB is 74±1
aining REXC,BB
The worst REution of the The red curvgive the obsefits the data distribution,
, on average, .1 (0.01) bad j
hese data to ne. Althoughal number of jy chosen afteements. TheseR16 (~15 μΩ)locating bad j
o open in 5 outhe sample ofx (5/8) =~ 20 This value o
wing value omain circuits a
98 μΩ. Vah in the estimsampling frac
confidence bo10 μΩ. The he resistance omean that 10coldered fromalistic value w
n the safe ener
HC ENERGs a function os on if the defn the time conhe RRRCS. Detgle-side defectergy (using R
(RB τEE=75 s
Q τEE=10 s):
(conservative)both RB and th sides of a s140 μΩ. In re
rated only on
at the ts are worst
15 μΩ after
EXC,R16 R16
ve, an erved well. from have
joints
place h the joints er an e had , that joints
ut of 8 f R16 0% of of the of the at the arious
mation ction, ounds value of the cm of
m the which rgy of
GY of the fect is nstant tailed t that
RRRCS s, RQ
) 100, RQ. splice eality n one
side comp
Figusplic
Thinterhas bof thwhicof mtargewerestatisin thaccuthe Lposs
[1]
[2]
[3]
[4]
[5]
[6]
of the splicpatible with 3
ure 3: The cumce excess resis
he splice rerconnect splicbeen seen. Th
he high currench the LHC cameasuring splieted repairs e performed istical methodhe sectors thauracy. This defLHC can opeibly, repairs) c
M. Koratzinoof main circwarm”, ATS rF. Bertinelli eSuperconductproceedings.M. Koratzinomain circuitLHC using caL. Tavian etdetection of Sectors”, PAC2009-006. F. Formenti eSystems for thMachine at CFirst OperatioA. Verweij, splices for 7Chamonix 20
ce, therefore .5TeV operati
mulative distristance data.
CONCLUsistance has es and no exc
he quality of tnt circuits of than safely operices in a nonand accurate n part of the
d to estimate tat were not mfines the maxierate before can take place
REFEREos et al., “Meacuit busbar sreport in prepaet al., “Towardting Splices fo
os and J. Strasplices with
alorimetry”, At al., “Helium
Abnormal RC ’09, Vancou
t al., “Upgradhe Supercond
CERN: From Conal Experienc“Minimum re7TeV operati
010 workshop,
the 90 μΩ Rion.
ibution of all a
SIONS been measu
cess resistancethe copper stahe LHC limit
rate. An extendn-invasive wa
resistance mmachine. We
the worst remmeasured withimum safe enmore measur
e.
NCES asurements of segments at aration. ds a Consolid
for 7 TeV Ope
ait, “A methoabnormal resi
ATS report in pm II CalorimResistive Zonuver, Canada,
de of the Quenducting CircuitConcept and ce”, these proequirements fion”, proceed, CERN-ATS-
Rmax limit is
available R16
ured for alle above 3 nΩabilizer jointsthe energy at
ded campaignay as well asmeasurementse have used amaining spliceh the requiredergy at whichrements (and,
the resistancethe LHC at
dation of LHCeration”, these
od to identifyistance at thepreparation. metry for thenes in LHC CERN-ATS-
nch Protectionts of the LHCDesign to the
oceedings. for the 13kAdings of the-2010-026.
s
6
l Ω s t n s s a e d h ,
e t
C e
y e
e C -
n C e
A e
Proceedings of IPAC’10, Kyoto, Japan MOPEB044
07 Accelerator Technology
T10 Superconducting Magnets 375