CERN, Geneva, Switzerland
EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH
European Laboratory for Particle Physics
Long Term Stability of the LHC Superconducting Cryodipoles
after Outdoor Storage
F. Seyvet, G. Arnau Izquierdo, A. Bertarelli, O. Denis, P. El-Kallassi, E.D. Fernandez Cano,
P. Fessia, S.D. Ilie, J.B. Jeanneret, D. Letant, A. Poncet, P. Pugnat, F. Savary, S. Sgobba,
A. Siemko, E. Todesco, D. Tommasini, R. Veness, B. Vullierme, E. Wildner
The main superconducting dipoles for the LHC are being stored outdoors for periods from a few weeks
to several years after conditioning with dry nitrogen gas. Such a storage before installation in the 27 km
circumference tunnel may affect not only the mechanical and cryogenic functionality of the cryodipoles
but also their quench and field performance. A dedicated task force was established to study all aspects
of long term behaviour of the stored cryodipoles, with particular emphasis on electrical and vacuum
integrity, quench training behaviour, magnetic field quality, performance of the thermal insulation,
mechanical stability of magnet shape and of the interface between cold mass and cryostat, degradation of
materials and welds. In particular, one specifically selected cryodipole stored outdoors for more than one
year, was re-tested at cold. In addition, various tests have been carried out on the cryodipole assembly
and on the most critical subcomponents to study aspects such as the hygrothermal behaviour of the
supporting system and the possible oxidation of the Multi Layer Insulation reflective films. This paper
summarizes the main investigations carried out and their results.
Presented at the 19th International Conference on Magnet Technology (MT19)
18-23 September 2005, Genova, Italy
Geneva,
Large Hadron Collider Project
CERN
CH - 1211 Geneva 23
Switzerland
LHC Project Report 895
Abstract
19 May 2006
MOA06PO02
1
Abstract—The main superconducting dipoles for the LHC are
being stored outdoors for periods from a few weeks to several
years after conditioning with dry nitrogen gas. Such a storage
before installation in the 27 km circumference tunnel may affect
not only the mechanical and cryogenic functionality of the
cryodipoles but also their quench and field performance. A
dedicated task force was established to study all aspects of long
term behaviour of the stored cryodipoles, with particular
emphasis on electrical and vacuum integrity, quench training
behaviour, magnetic field quality, performance of the thermal
insulation, mechanical stability of magnet shape and of the
interface between cold mass and cryostat, degradation
of materials and welds. In particular, one specifically selected
cryodipole stored outdoors for more than one year, was re-tested
at cold. In addition, various tests have been carried out on the
cryodipole assembly and on the most critical subcomponents to
study aspects such as the hygrothermal behaviour of the
supporting system and the possible oxidation of the Multi Layer
Insulation reflective films. This paper summarizes the main
investigations carried out and their results.
Index Terms— Cryogenics, Superconducting accelerator
magnets, Materials science and technology, Electrical
Engineering.
I. LHC CRYODIPOLE GLOBAL DESCRIPTION
O achieve a final proton-proton collision energy of
14 TeV in the 27-km circumference tunnel, the
LHC will
be composed of 1232 horizontally curved 15 m long dipole
magnets [1], generating a nominal magnetic field of 8.33 T, for
a current of 11.85 kA flowing through the NbTi
superconductors of the coil winding. The dipole cold masses
are cooled by superfluid helium down to 1.9 K. A two-stage
temperature thermal shielding system aims at intercepting the
largest fraction of applied heat loads at higher temperature; it is
composed of a first major heat intercept at 50 K to 75 K
around the cold mass, and of a second lower temperature heat
intercept at 4.6 K to 20 K designed to protect the magnet from
both beam induced heat loads (by cooling the beam screen),
and conduction heat transfer from the vacuum vessel through
the magnet supporting system. Multi Layer Insulation (MLI) is
wrapped around both the cold mass and the first major heat
intercept to shield the superconducting magnet from heat
radiated by the vacuum vessel at room temperature. An
Authors are within CERN, Geneva, Switzerland
Manuscript received September 20, 2005.
Instrumentation Feedthrough System (IFS) was developed for
the routing of the instrumentation wires from the 1.9 K
environment to the outside. The cold mass is supported within
its vacuum vessel by three Glass Fibre Reinforced Epoxy
(GFRE) support posts of low thermal conductivity and high
stiffness; the two extremity supports are free to slide on a
PTFE® coated surface to cope with differential thermal
contractions of the cold mass and the vacuum vessel. A cross
section of the dipole superconducting magnet in its cryostat is
presented in Fig.1.
Fig. 1. Cross-section of the LHC main superconducting dipole in its cryostat.
No cryodipole components were explicitly designed to meet
external storage environmental effects. Delays in the
installation of the machine entailed the decision to store cold
masses and assembled cryomagnets outdoors. A dedicated task
force was established to study the Dipole Long Term Stability
(DLTS) and determine the effects of long term outdoor storage
on magnet electrical insulation and continuity, field harmonics,
quench training memory, geometry stability, degradation of
materials and welds. A risk assessment was finally completed.
II. S
TORAGE TIME AND CONDITIONS
The dipole cold masses are delivered at CERN from three
manufacturers and are first stored outdoors. The magnet is then
assembled into its cryostat and the performances of each
individual dipole are assessed under cryogenic operating
conditions. After the cold tests, a second outdoor storage
period occurs. The cleaning of the beam tubes and subsequent
insertion of the beam screens are performed before lowering
down the cryomagnet into the 27-km circumference tunnel.
Sometimes, a third outdoor storage period occurs between
Long term stability of the LHC superconducting
cryodipoles after outdoor storage
F. Seyvet, G. Arnau Izquierdo, A. Bertarelli, O. Denis, P. El-Kallassi, E. D. Fernandez Cano, P. Fessia,
S. D. Ilie, J. B. Jeanneret, D. Letant, A. Poncet, P. Pugnat, F. Savary, S. Sgobba, A. Siemko,
E. Todesco, D. Tommasini, R. Veness, B. Vullierme and E. Wildner
T
MOA06PO02
2
these last two operations. An illustration of the storage
conditions of the cold masses before assembly into their
cryostat is presented in Fig. 2. The conditioning of the cold
masses and cryodipoles includes protection of electrical
cabling, installation of leak tight covers on both magnet and
vacuum vessel extremities and pressurization of the cold mass
with 1.2 bar of gaseous nitrogen.
Fig. 2. External storage of LHC main dipole cold masses (winter 2005).
The distribution of the storage time for both the cold masses
and the already assembled cryodipoles, at the start of the DLTS
investigation early 2005, is summarized in Fig. 3. 86 cold
masses or assembled cryodipoles were stored outdoors for
more than one year after their delivery at CERN.
Fig. 3. Outdoor storage time for both the LHC dipole cold masses and the
assembled cryomagnets at the start of the DLTS work.
III. ELECTRICAL INSULATION AND CONTINUITY
Two samples of 33 units were built to compare the electrical
insulation of dipole cold masses stored outdoors for long
periods, between 5 and 14 months, and for short ones, below
10 days.
Fig. 4. Leakage current measured for dipole magnets that were stored outdoors
for long and short time periods.
The leakage currents flowing in-between main insulated
magnet components were measured at 1.9 K with maximum
applied voltages of 2.7 kV and 3.1 kV as a function of the
electrical circuits. Results presented in Fig. 4 show that the
electrical insulation of the magnets is stable regardless of the
magnet storage time. Furthermore, a complete electrical check
is performed on each cryomagnet before their installation in the
tunnel in order to ensure the electrical system integrity. The
check consists of a continuity and an insulation tests. Up to
now, more than 80 cryodipoles stored outdoors for periods up
to one year have already passed this check successfully.
Finally, the IFS was designed to sustain external storage
conditions, and specific conditioning procedures are applied
that ensure the IFS functionality in the long-term.
IV. F
IELD HARMONICS
The extended duration of outdoor storage was questioned to
eventually enhance the creep in the coil structure [2]. Induced
relaxation of the coil pre-stress would perturb the coil
geometry, resulting in non-negligible variations in the magnet
field quality.
A mechanical model of the dipole cross-section, previously
developed and validated with experimental data [3], was used
to evaluate the influence of coil pre-stress on field quality. At
the nominal pre-stress of 70 MPa, a loss of 10 MPa would give
+0.5 units of b3, +0.12 units of b5, -0.015 units of b7. The
effect is thus modest but not negligible.
Measurements of the field quality at CERN show that the
offsets between warm and cold field measurements in b3 are
stable within ±0.5 units, regardless of the time of storage of the
magnets. Pre-stress in the coils is thus stable in time at least
within ±10 MPa; results are shown in Fig. 5. The results are
very similar for b5 and b7 with stability within ±0.2 and ±0.1
units respectively.
Fig. 5. Stability of the warm and cold measurement correlation with time for
the b3 harmonics of the LHC arc superconducting dipoles.
V. QUENCH TRAINING
Similarly, the relaxation of the coil pre-stress in the cold
mass could cause quench training degradation. In order to
investigate this, a magnet displaying a relatively slow training
and an initial quench below the nominal value was chosen. It
was then cold tested for a second time after one year of outdoor
storage. Quench training results are presented in Fig. 6. During
its first cold test, this magnet initially quenched at 8.23 T,
below the nominal value of 8.33 T, and seven quenches were
required to reach a maximum field of 8.95 T. After the
standard thermal cycle, the magnet had “memorized” the
MOA06PO02
3
quench training and its first quench occurred at 8.76 T.
Fig. 6. Quench training of a dipole magnet after one year of outdoor storage.
During the second test, after one year of outdoor storage, the
first quench of this dipole occurred at 8.67 T, a value largely
above the first training quench but slightly lower than the first
one after the standard thermal cycle. No important degradation
of the quench training behaviour was observed for this magnet
after one year of outdoor storage. For a firm statement
concerning the whole population of LHC superconducting
magnets, a statistical study is required.
VI. G
EOMETRY AND ALIGNMENT
The geometry stability of the LHC cryodipoles has been
assessed by comparing their geometry between two stages;
before outdoor storage once they are assembled and cold
tested; and after storage when the beam screens are inserted
into the beam tubes. Measurements of 329 cryodipoles, that
were stored outdoors for periods varying between one month
and two years, were used to do this analysis. The position
change of the cold mass extremities within the cryostat has
been studied, as it is representative of both the corrector
magnet alignment and the sagitta change of the dipole. Results
are presented in Fig. 7. Statistically, more than 95% of the
cryomagnets remained stable within +/-0.5 mm both
horizontally and vertically. 8 magnets that show geometry
change outside this range are under study. The mean and the
standard deviation of the movements of the ends of the dipoles
have been calculated using windows sliding over storage time.
No indication of long term trends has been detected.
Fig. 7. Stability of the cold mass extremity positions before and after long term
storage (connection side).
VII. DEGRADATION OF MATERIALS
A series of cold masses and assembled cryodipoles stored
outdoors for periods of the order of one year has been
inspected for potential degradation [4], with the aim to identify
the components susceptible to degrade and assess the nature
and importance of the degradations as well as their effects on
thermal, structural, and vacuum performance of the magnets.
During the inspections, focus was put on stainless steel cold
mass extremity welds, cold foot pad and bellow welds, copper
heat exchanger tube and stainless steel to copper joints, copper
to copper welded joints, the general aspect of the cryostat
(made out of C steel), GFRE support posts and sliding surfaces
of the centering pieces, helium transport lines including their
aluminium to aluminium welds and bimetallic junctions, as
well as the MLI.
The excessive penetration of a weld between the beam tube
ultra high vacuum and the helium pressurized cold mass was
specifically monitored. The risk of loss of He leak tightness of
this weld is critical for machine operation and for beam-gas
background in the LHC experiments. The lack of back gas
shielding during this welding operation associated to a full
penetration resulted in a corroded aspect of the weld root
visible after some storage time. Visual inspections of every
LHC cryomagnet were thus carried out and further X-rays and
repair will be performed, when necessary, to ensure that only
magnets adequate for LHC operation will be installed in the
machine. It should be noted that this issue is coming from a
cold mass manufacturing non conformity, and is thus only
indirectly linked to the DLTS work.
A. Cryodipole components, cold mass welds and bellows
Further results from the inspections are summarized in the
following. A few decilitres of liquid water were found in a cold
mass heat exchanger line because of poor storage conditioning.
Since one of the three producers used cast 304 H instead of
cast 304 L for the cold feet pad, particular attention was paid to
the possible sensitization of the Heat Affected Zones (HAZ) of
the welds of this component. No corrosion of this weld was
observed after outdoor storage for 9 months. On the other
hand, a « rusty aspect », was locally observed in the magnet
extremity welds, mainly in the HAZ. The rust extent was found
to depend on duration of external storage and the cold mass
assembler.
Ageing tests of welds after removing rust stains were further
carried out. Furthermore, representative weld samples that were
recovered from the three cold mass producers were also
subjected to tests representative of the tunnel conditions, and to
complementary destructive tests after a significant storage time.
Finally, a storage inspection and re-conditioning campaign was
launched on all cryodipoles. These efforts led to the
conclusions that first, some dipole welds are indeed affected by
rust but they are not critical and they can be repaired, and
second, the functionalities of the LHC cryodipole components,
welds and bellows remain stable in the long-term.
MOA06PO02
4
B. Vacuum
Condensation and re-evaporation cycles in the vacuum
enclosures are possible during outdoor storage; the oxidation of
the vacuum system was therefore investigated. Furthermore,
the leak tightness of the beam screen cooling capillary was
checked because of possible combined effects, during storage,
of halogen residues in the system inducing corrosion, residual
stresses due to welding, and temperature and humidity cycles.
Magnets stored outside for a period of one year that had
already been equipped with beam screens and cooling circuits
were inspected, and 3 of them were leak tested. The vacuum
enclosure was found leak tight, regardless of the outdoor
storage time.
C. Thermal insulation
The air moisture in the vacuum vessel during outdoor
storage might oxidize the 40 nm aluminium coating deposited
on each side of the MLI layers. Degradation of the shielding
properties of the MLI protecting the superconducting magnets
from radiant heat loads was thus analysed. The thermal
performance of a production sample that was stored in an
unprotected environment for a period of more than two and a
half years has been analyzed and tested [5]. The thickness of
the aluminium coat on each MLI polyester layer was
determined by measuring the electrical resistivity of the sheet,
and the thermal performance of two superimposed 15 layers
blankets was measured with a heatmeter. Results show that the
thickness of the aluminium layer was found unchanged except
for the two external layers of the units, where it was found to
be reduced to a minimum of 20 nm. Similarly, the overall MLI
thermal performances have been degraded by 17% with respect
to the test performed on a new reference sample.
A degradation of the thermal performance of a MLI sample
that was poorly stored for a long time has been observed, but
the degraded MLI was still found to meet the initial
specifications for the operation of the LHC machine [6].
D. Interface between the cold mass and the cryostat
During outdoor storage, the following environmental
conditions could be suspected to affect long-term behaviour of
the supporting system of the cold mass inside the cryostat:
moisture absorption, the night and day as well as the winter
and summer thermal excursions, the change of phase of the
moisture absorbed, creep phenomena under hygrothermal
effects, dirt and dust.
The main risks have been identified as, first, a degradation
of the friction coefficient of the PTFE® coated surface
allowing sliding of the extremity support posts on the cryostat;
and, second, a decrease of the stiffness of the GFRE support
posts eventually degrading the alignment stability of the LHC
magnets in the machine.
A specific hygrothermal ageing treatment of both GFRE
supports posts and low-friction centering pieces has been
defined, based on a series of humidity absorption and de-
sorption cycles on samples of the GFRE composite material. In
an environmental chamber, four supports and four centering
pieces were pre-conditioned during 3 days at 40°C and 95% of
humidity, and they were then subjected to 91 cycles of 8 hours
each; each cycle consisting of 6 hours at 40°C and 95%
humidity, and then 2 hours at -20°C.
The GFRE support posts were mechanically tested before
and after hygrothermal treatment. The rigidity of the units in
cantilever bending degraded by 6% after the environmental
conditioning, whereas their compression rigidity remained
unaffected [7]. This was shown not to affect the alignment
stability of the magnets in the LHC machine.
The static friction force between the centering piece and the
support post was found to increase by 45% after the
hygrothermal treatment [7]. This degradation was attributed to
the rust and wear-induced dust lying on the pad. It will
generate increased loads in the supporting system during
magnet cool down, but these loads remain five times below the
admissible ones. This is thus acceptable for the machine
operation.
VIII. C
ONCLUSION AND RISK ASSESSMENT
The long-term behaviour of the LHC main superconducting
dipoles stored outdoors has been investigated, with a focus on
electrical integrity, magnetic field quality, quench training,
geometry, degradation of materials and welds, performance of
the thermal insulation and interface between the cold mass and
the cryostat. The analysis has shown that dipole cryomagnets
remain functionally unaffected or weakly affected by long-term
outdoor storage, even if such a constraint was not considered
during the design phase. Thanks to the inspection work done, a
manufacturing non conformity in a cold mass critical weld was
monitored and treated, and the quality control of the storage
conditions was improved.
A
CKNOWLEDGMENT
This paper synthesizes the work done by many colleagues
who have contributed to the DLTS Task Force activities.
Specific contributions of the CERN CryoLab, D. Bozzini,
R. Principe and V. Parma are acknowledged.
R
EFERENCES
[1] O. Brüning et al. “LHC Design Report” CERN 2004-003, Vol. I The
LHC Main Ring, 2004.
[2] W. E. Cooper et al "Long term stability of Fermilab energy saver
magnets", IEEE Trans Nucl Science 30, 1983.
[3] P. Ferracin et al. “Azimuthal coil size and field quality in the main CERN
Large Hadron Collider dipoles”, Phys. Rev. ST Acc. Beams 5, 062401,
2002.
[4] S. Sgobba et al. “Dipole Long Term Stability, Sub-Unit 2”, Presentation
to CERN Main Arc Ring Committee (MARIC), March 9
th
, 2005.
[5] G. Ratcliffe et al. “Etude d’un matelas de super-isolation Jehier stocké
dans de mauvaises conditions”, CERN Cryolab Tehnical Note 06-05,
May 2
nd
, 2005.
[6] L. Tavian et al. “Heat Load Working Group (HLGP)” http://lhc-mgt-
hlwg.web.cern.ch/ CERN, 2000.
[7] P. El-Kallassi et al. “Mechanical tests of the LHC cryodipole support
posts and sliding pads, before and after hygro-thermal treatment” CERN,
DLTS Test Report 633462, April 20
th
, 2005.