logo area 11T Hybrid Assembly 1. Disassembly and Post Mortem Analysis 2. Revised Project Plan 78 th Meeting of the HL-LHC Technical Coordination Committee T.A. Bampton, M. Bernardini, S.E. Bustamante, A.P. Foussat, O. Housiaux, F. Lackner, M. Michels, J. Petrik, H. Prin, D. Pulikowski, J.L. Rudeiros Fernandez, F. Savary, D. Schoerling, E. Tsolakis, G. Willering CERN – Room 30/7-010 – 2019-07-04 – https://indico.cern.ch/event/829607/
65
Embed
11T Hybrid Assembly 1. Disassembly and Post Mortem Analysis … · 2019. 8. 7. · Part 1: Disassembly and Post Mortem Analysis Quality control plan Inspection of the cold mass assembly
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
logo
area
11T Hybrid Assembly
1. Disassembly and Post Mortem Analysis
2. Revised Project Plan
78th Meeting of the HL-LHC Technical Coordination Committee
T.A. Bampton, M. Bernardini, S.E. Bustamante, A.P. Foussat, O. Housiaux, F. Lackner, M. Michels, J. Petrik, H. Prin, D. Pulikowski, J.L. RudeirosFernandez, F. Savary, D. Schoerling, E. Tsolakis, G. Willering
Test results for GE-02 (hybrid) [5.10-3 – 800 mbar]
QH Left: Paschen minima at 2 mbar, indicating a defect size of 0.4 mm. Multiple Paschen regimes were not observed CR-02 (not cold-tested), which may indicate progressive degradation of the insulation system in the case of GE-02
QH right: do not show a full Paschen curve. It seems that the minima is shifted towards higher pressure, 90 mbar, indicating a defect size of 80 𝜇m (that is radial, through insulation thickness)
F. Savary @ 78th Meeting of the HL-LHC TCC 42Courtesy A. Foussat & J. Petrik
logo
area
Conclusion of Paschen tests
For CR-02:
A breakdown voltage minima was found at 575 V @ 0.4 mbar
for a QH-Right, and 675 V @ 0.5 mbar for the QH-Left
Considering the minimum of the product P・d, as invariant in
air at 8 mbar ・ mm, then the minimum defect involved in the
breakdown in CR02 would be around 14 mm
For GE-02:
The breakdown voltage minima was found at 680 V for both
QHs, indicating the same gas conditions as for CR02, i.e.
presence of air, not helium!
F. Savary @ 78th Meeting of the HL-LHC TCC 43Courtesy A. Foussat & J. Petrik
logo
area
Content
Introduction – corrigendum
Part 1: Disassembly and Post Mortem Analysis Quality control plan
Inspection of the cold mass assembly
Inspection of the collared coils / coils (visual, and metrology)
Paschen tests
Thermo-mechanical FEA model
Part 2: Revised Project Plan WUCD process
Revised plan & schedule
F. Savary @ 78th Meeting of the HL-LHC TCC 44
logo
area
1. 11T FE thermo-mechanical model
Objective: Investigate the impact of large delta temperature during the magnet cool down
and warm up, on coil stress
F. Savary @ 78th Meeting of the HL-LHC TCC 45
CoilsShell
Top
Bottom
Top splice
Coil head
Temperature measurements during testing
MBHSP109
Sta
tion
Lon
g
SP109 – 3rd Cool down
(i.e. 2nd thermal cycle)
Vertical ΔT
Shell
Radial
ΔT
Courtesy J.L. Rudeiros Fernandez
logo
area
1. 11T FE thermo-mechanical model
F. Savary @ 78th Meeting of the HL-LHC TCC 46
The FE models
Sta
tion
Lon
g
2in11in1
Geometries
Temperature is
imposed, for each
step, based on
experimental
measurements, in
the coil and the
shell. Thermal Model (Temperature Field)
ANSYS Steady-State Thermal
Mechanical Model
(Stress – Strain) ANSYS
Static Structural
Body Temperature
Courtesy J.L. Rudeiros Fernandez
logo
area
1. 11T FE thermo-mechanical model
F. Savary @ 78th Meeting of the HL-LHC TCC 47
Preliminary results
Sta
tion
Lon
gMeasurement
0
50
100
150
200
250
300
350
-1.5
-1.4
-1.3
-1.2
-1.1
-1
-0.9
-0.8
-0.7
3 8
Te
mp
era
ture
(K
)
Colla
r n
ose
str
ain
(S
ca
led
)
Cool Down - Simulation Steps
Collar nose strain
ΔT (Tcoil-Tshell)
Shell Temperature
Simulation
SP109 – 3rd Cool down (i.e. 2nd thermal cycle) 1. In terms of the global mechanical
response, the relative increase of the measured strain in the collar nose, can be replicated by the thermo-mechanical model
2. The model doesn’t account for dynamic and fracture behaviour (e.g. high thermal strain rates, crazing, cracking and debonding), potentially present within the coil due to thermal shocks or high rates of temperature variation, during the cooling down or warming up of the magnet
Considerations
The scaling was considered because of (1) uncertainty in material parameters (like thermal contraction and
variation of stiffness vs. temperature, and (2) unexplained stress variation , as measured for SP109 between
end of collaring and beginning of cool down) Courtesy J.L. Rudeiros Fernandez
logo
area
Content
Introduction – corrigendum
Part 1: Disassembly and Post Mortem Analysis Quality control plan
Inspection of the cold mass assembly
Inspection of the collared coils / coils (visual, and metrology)
Paschen tests
Thermo-mechanical FEA model
Part 2: Revised Project Plan WUCD process
Revised plan & schedule
F. Savary @ 78th Meeting of the HL-LHC TCC 48
logo
area
Connection sideCFB
300 K (for CD)
or 80 K (for
WU) helium
gas inlet
through N-line
(TT 161)
300 K or 80 K helium gas injected
in cold mass on connection sideHelium gas
exits the
cold mass
by a M-line
(TT 150)
The helium gas flows
through holes in the cold
mass (but not in the
beam tube, nor in the
heat exchanger tube)
The head on the connection side has
seen the largest thermal gradient
This is also the part that limited the
quench current in the coil after the first
thermal cycle
Standard Warm Up and Cool Down process in SM18
Horizontal test benches
F. Savary @ 78th Meeting of the HL-LHC TCC 49Courtesy G. Willering
logo
area
Large thermal
gradient during
first warm up
Controlled thermal
gradient during warm up
Nb-Ti magnet warm up cycle using mainly
300 K He gas, also used for the 11T
prototype, and 11T hybrid magnet
New: manually controlled warm up cycle
using a blend of 300 K and 80 K gas
This delta T control has been implemented in
automated process by TE-CRG and is ready for the
cool down of the 1st series magnet, LMBHB002 (cool
down expected next week), see next slide
Note that the TT821 probe in the middle of the magnet
shows that the temperature gradient is entirely on the first
half of the magnet, and possibly an even a shorter part of it
The local gradient may be much higher than 150 K / 5.5 m
Cool down and warm up process improvement
F. Savary @ 78th Meeting of the HL-LHC TCC 50Courtesy G. Willering
logo
area
CERNOX
Temperature
probe on cold
mass
Maximum delta T regulated between 3 probes: TT150 – gas inlet temperature (in CFB)
TT161 – gas outlet temperature (in CFB)
New CERNOX placed on the cold mass on the CFB side
Control For the first cool down/warm up a delta T of 30 K will be requested
Technical difficulty / risk Mixing of 80 K + 300 K gas happens just before the magnet
Restarting the cool down or warm up phase can only be done with
80 K or 300 K gas, so if the cool down/warm up is interrupted
(accidentally) in the middle of the process (for example at 200 K), it
leaves little options than temporary exceeding the requested delta T,
even at minimum gas flow which is rather high (~ 30 g/s)
Details of the process explained in EDMS 2136536
TT161TT150
F. Savary @ 78th Meeting of the HL-LHC TCC 51
New cool down & warm up process
Courtesy G. Willering
Warm thanks to TE-CRG (N. Guillotin, J-P. Lamboy et al.) for the implementation
Determine irreversibility limits of conductors and fatigue behavior under thermal (≈10 WUCD) and electromagnetic (≈1000 EM cycles) conditions
Carry out comprehensive characterizations of the behavior under transverse stress of final strands and cables, including:
Assessment of maximum allowable stress at warm on cables (at FRESCA)
Assessment of reversible and irreversible degradation vs. stress at cold on strand (University of Geneva setup) and cables (at FRESCA and Twente University)
Assessment of thermal cycling on cables (TBD)
Assessment of stress cycling at cold on cables (at Twente University)
Assessment to be accompanied by micrographic examination to determine the onset and extent of filament cracking on the final strands (at FSU/ASC and CERN/EN/MME)
F. Savary @ 78th Meeting of the HL-LHC TCC
A. Devred
62
logo
area
Action item 6 – Numerical modeling & analysis
Establish coherence and reference numerical models for the single, and two-in-one structure (for MQXFB also) Material properties (mechanical, physical)
Level of detail (coil bloc, cable, …), meshing
Boudary conditions
Mechanical and thermal stresses
Format for the outputs (strain/stress, average vs local stress, @ pole/mid plane, radial direction, assembly to operation conditions, …)
Make sure the link with the mechanical measurements will be possible and relevant
Develop an understanding and analyze thermal stresses that arise in magnetcoils during various phases of testing and operation (WUCD, …) Develop an instrumentation plan such that comparison and validation will be possible
Assess in-coil performance of conductors, i.e. understand the degradationsobserved after Thermal cycles
High MIITs quenches resulting in high hot spot temperature
F. Savary @ 78th Meeting of the HL-LHC TCC
A. Devred
63
logo
area
MBH-hybrid second, and third cooldown
Performance
- After thermal cycle, the magnet
performance was limited by a
single point in the magnet
- Strong ramp rate dependency
- All observations, and magnet
behavior, point to local damage
to the Nb3Sn cable
- At 4.5 K, the quench current is
still @ 12.2 kA
Retraining memory
- Quench 22 is seen as a
training quench, in the other
head (NCS) of coil C03
- All the other quenches were in
the weak spot. Otherwise, up to
12.5 kA the mechanical training
memory seems to be good
Localisation of degradation
- In the head of the upper coil
(C03) on the connection side.
In segments 2 to 3 of the
quench antennas
- From quench 10 to 30, all but
one started in this location
F. Savary @ 78th Meeting of the HL-LHC TCC 64
logo
area
LMBHP001 – Hybrid – first cool down
Significantly better training in first cool
down than all earlier model magnets
Higher coil limit at 4.5 K, 300 A higher
than for the best model magnet
“Clean” training curve
No detraining
Single quench to nominal current
Even at 4.5 K reaching almost ultimate current level,
showing good temperature margin for nominal operation
94% of the conductor short sample was reached @ 4.5 K
Significantly better than the prototype, which had a