Long Quadrupole Giorgio Ambrosio

1
Long Quadrupole Giorgio Ambrosio
DOE review of LARP program Jun. 19-20, 2008
LBNL

• OUTLINE
• The road to the LQ
• LQ main features and plans
• FY08 development and issues
• Schedule and budget

2
Long Quadrupole

• Main Features
• Aperture 90 mm
• magnet length 3.6 m
• Goal
• Gradient 200 T/m
• Timeframe
• Performance and reproducibility by the end of
  2009
• Testing 3 LQs

3
The road to the LQ
4
Plan for Length Scale-Up
LENGTH SCALE-UP CHALLENGE No Nb3Sn accelerator
magnet longer than 1m has ever been built LARP
plans to have a successful 4m long quadrupole by
end of 09
FNAL Long Mirrors
LARP TQs
5
Long Coils RD - Results

• 2nd Long Racetrack (4m coils) 96 of SSL
• So we can successfully make long Nb3Sn coils!
• But the LR had flat coils without ceramic binder,
  and coils were not heat treated in a closed
  cavity under pressure

• 1st Long Mirror (2m coil) SSL
• Accelerator quality coil using PIT conductor
• Cos-theta coil w wedges, end spacers

98 SSL taking into account simulated temperature
increase

• 2nd Long Mirror (4m coil) 87 SSL
• Accel. quality coil using RRP 114/128
• Performance improved by heating the outer layer ?
  instability

Front view of mirror magnet
6
LR result segmented shell
7
Original Present plans
2N 1R
No 3rd generation
In FY08
In FY07-08, No ceramic binder

• We had delays, some parts were skipped, but we
  want to keep the 2009 LQ goal
• ? it can be done, but not without pain!

8
Mitigation factors

• Significant effort for fabrication and inspection
  of practice coils
• Started before definition of latest details
  (holes for coil lifting, pockets for trace
  wiring)
• Saved time with some extra costs
• LQ coil reaction and impregnation at two labs
  (BNL, FNAL)
• This is going to save time during LQ coils
  production
• Providing risk mitigation (equipment failure)
• Cost of fixtures and tooling at two labs
• Continuous coil fabrication until successful
  inspection
• Travelers continuously updated to be able to
  start production right away

9
LQ main features
LQ Design Report available on LARP web site
at https//plone4.fnal.gov/P1/USLARP/MagnetRD/lon
gquad/designreport/
10
Conductor

• Cable
• LQ cable design TQ cable design
• 27 strands, 0.7 mm diameter,
• 10 mm wide cable, 1 keystone angle
• Strand
• OST-RRP 54/61 for LQ01
• Strand used in TQS02 and TQC02 coils and LR
• Good performance at 4.5 K
• Performance at 1.9 K under investigation
• Higher number of subelements (108/127) may be
  used in following LQs
• See schedule for options

11
Magnetic Design
Parameter Unit Collars Shell
N of layers - 2 136 29.33 2 136 29.33
N of turns - 2 136 29.33 2 136 29.33
Coil area (Cu nonCu) cm2 2 136 29.33 2 136 29.33
4.2 K temperature Jc 2400 A/mm2 4.2 K temperature Jc 2400 A/mm2 4.2 K temperature Jc 2400 A/mm2 4.2 K temperature Jc 2400 A/mm2
Quench gradient T/m 221 233
Quench current kA 13.3 13.4
Peak field in the body at quench T 11.5 11.9
Peak field in the end at quench T 11.9 11.4
Inductance at quench mH/m 4.6 4.9
Stored energy at quench kJ/m 406 439
1.9 K temperature Jc 2400 A/mm2 1.9 K temperature Jc 2400 A/mm2 1.9 K temperature Jc 2400 A/mm2 1.9 K temperature Jc 2400 A/mm2
Quench gradient T/m 238 251
Quench current kA 14.4 14.5
Peak field in the body at quench T 12.4 12.9
Peak field in the end at quench T 12.9 12.4
Stored energy at quench kJ/m 472 512
2800 A/mm2 2800 A/mm2
231 244
14 13.9
12 12.5
12.5

443 477
2800 A/mm2 2800 A/mm2
249 263
15.1 15.1
12.9 13.5
13.5
516 557

• Coil layout TQs
• 2 layers (NO splice)
• iron closer to coils in shell structure
• Peak field in ends with collar structure

12
Mechanical Design - I

• LQ Magnet Structure Review
• Nov 28-29, 2007 at BNL
• LQC Long TQC - LQS based on TQS

13
LQMS Review plan

• Plan
• Procure in FY08 both shell and collar long
  structures
• Provides options back up
• 1st LQ with shell-based structure
• Best performance, shorter assembly, easier to
  replace coils
• 2nd LQ with collar-based structure
• Reusing LQ01 coils (done w TQs)
• 3rd LQ with structure depending on previous
  results

• Scale up issues
• TQC models didnt exceed 200 T/m
• Collaring long coils
• TQS structure needs modifications for long
  magnets
• Segmented shell
• Improvements introduced based on TQS test results

TQS LQS
14
Advantages of Proposed Plan

• Larger probability of success within FY09 by
  developing both structures
• Providing back-up (LQC01 back-up for LQS01) and
  options (several options after LQS01 tests)
• We are building a large and unique set of
  expertise and experimental data for the design of
  the structure for the LHC phase-II upgrade
• All 3 labs are strongly involved with this plan
• ? the best intellectual contribution from all
  experts
• ? very high level of internal scrutiny

15
LQMS Review close-out
16
LQMS Review close-out

• However, from our perspective
• Concerns about schedule, changes from TQS to LQS,
  stresses, coil alignment, end-load, TQ
  performances
• Addressed
• Shear stress addressed in dedicated technical
  note
• Available at https//dms.uslarp.org/MagnetRD/long
  quad/shear_note_V4.doc
• Schedule concerns addressed by 2 lab for RI, LQ
  priority, contingency
• LQS with iron pads at TQS test of 1m LQS
  structure at LN, and of whole LQS structure at
  300 K with dummy coils
• TQS02 tested (2 times) replacing limiting coils
• Replies/responses to all recommendations are in
  the back-up slides

17
Quench Protection

• Goal
• MIITs lt 7.5 ?? Temp 380 K (adiabatic
  approx)
• Quench protection param. (4.5 K) conservative
  hypothesis
• Dump resistance 60 mW (extract 1/3 of the
  energy Vleads 800 V)
• 100 heater coverage (? heaters also on the
  inner layer)
• Detection time 5 ms based on TQs with I gt 80
  ssl
• Heater delay time 12 ms based on TQs with I gt
  80 ssl

18
LQ plans and work in progress
19
Fabrication and test plans

• Coils are being fabricated at FNAL and BNL
• 2 practice coils at FNAL, 1 at BNL
• 4 coils by the end of FY08
• 3 at FNAL, 1 at BNL
• Start spare coils in Q4
• 6 coils in FY09 (more using contingency)
• 2 at FNAL, 4 at BNL
• Shell structure design, procurement test at
  LBNL
• Ready by the end of FY08
• Test of 1m model at LN whole structure at 300K
  with dummy coils
• LQS01 mechanical assembly and pre-load at LBNL
• LQS01 electrical assembly and prep for test at
  FNAL
• Collar structure procurement at FNAL
• Some parts held in contingency until Q4.
• All LQ models to be tested at FNAL
• LQ01 test in Feb 2009

20
LQS design

• 20 mm shell
• 4-split iron yoke
• Iron masters with 2 bladders and 2 interference
  keys
• Iron pads with holes for coil end support and tie
  rods
• Stainless steel sheet between coil and pad
  laminations

• Shell 4 segments, 0.8 m long
• Yoke 50 mm laminations with 3.3 m long tie rods
• Pads 50 mm laminations with 3.3 m long tie rods
• Masters 2 segments, 1.6 m long
• Stainless steel axial rods w 24.5 mm diameter

21
Assembly - I

• Coil-pad sub-assembly
• Pads bolted around the coil
• Bolts disappear under compression
• Yoke-shell sub-assembly
• Gap keys keep yoke stacks apart and pre-tension
  the shell

22
Assembly - II

• Assembly of 4 single shell-yoke sub-assemblies
• Connection of shell-yoke sub-assemblies with tie
  rods
• Insertion of coil-pad sub-assembly

23
LQS status

• All parts in house
• Instrumentation of 0.8m section in progress
• 0.8m section will be fully assembled with dummy
  coil and tested at LN

Coils FNAL, BNL Structure LBNL Test FNAL
24
Long Nb3Sn coil challenges

• Conductor
• Need km-size strand piece length, and long
  cabling runs (250 m for 4m long quad coils)
• Insulation
• Need technique for long coils
• Reaction
• Need long oven
• The displacements due to differential expansions
  scale with length
• Total friction force scales with length
• Impregnation
• Impregnation time increases with length
• Handling
• LARP set criteria for Max strain -0.15lt e lt0.05

LQ coil
25
LQ Coil Fabrication

• Coil design
• LQ coils TQ coils w minor modifications

• ReactImpr fixture change
• From 2-in-1 used for TQ coils to single coil
  fixtures for LQ
• More symmetric coils ()
• New parts, new procedures (-)
• Symmetric plates (as in LR)

26
LQ coil fabrication issues

• Practice coil issues
• Coil bowing after reaction (PC 2)
• ? Pre-heat treatment of all fixture parts
• ? Symmetric fixture (add top plate)
• ? Reduce friction (mica)
• Damaged lead (PC2)
• Due to coil bowing because of winding tension
• ? Keep coil always under load
• ? Introduce gaps only for HT?
• Damaged leads (PC 3)
• Due to shims overlapping
• ? Continuous shims
• ? Connect saddle to pole tip
• Inner pole shorter after HT (PC 3)
• Still under investigation
• ? Introduce gaps only for HT?

solved
solved
good plan
27
Coil Instrumentation

• Will use Kapton Traces as in TQs
• Voltage taps 13 IL 7 OL
• Protection heaters on both layers
• Two traces (1.7 m each) per layer
• Large ss strip with narrow heating areas
• Successfully tested on Long Racetrack
• Bubbles may reduce heater efficiency
• Option test at 4.5, 2.5 K and 1.9K (at the end)
• Strain gauges 4 IL 1 OL
• Wires on trace for outer layer strain gauge
• Gauges on the inner layer will be instrumented
  with wires

28
Magnet tests

• Preparation for test
• Adaptive QP threshold
• Symmetric grounding
• Handling and support of LQS at VMTF
• Handling fixture modified for pivoting
• Test all LQ magnets
• Test at 4.5K and 2.5K (1.9K at the end)
• Magnetic measurement, ramp rate dependence, RRR
• Magnetic measurement only ½ length
• Thermal cycles to check training memory

Pivoting LRS01 in prep. for test at BNL
29
QA Plan

• QA for LQ coil production
• Travelers
• Each lab is responsible for its own travelers,
• Travelers will be distribute to the 3 labs
• Fabrication steps (HT cycle, impregnation,
  insulation) and measurement plan are the same
• Discrepancy Reporting
• Assures efficient reporting and recording of all
  discrepancies
• All LQ task leaders will receive all DRs
• New feature for LARP
• Documents on LARP web site (plone)
• Easily available to the whole collaboration
• Address https//plone4.fnal.gov/P1/USLARP/MagnetR
  D/longquad/

30
Projectized task

• LARP is a collaboration with a program
• LQ project-like features
• Plan Task sheets with milestones, budget for
  each milestones, and commitment (technical, not
  financial) by task leaders and supporting labs to
  do the job
• Budget LQ had priority in the FY08 plans, and in
  the use of magnet contingency
• No core-programs support
• QA developed and implementing LQ QA plan

31
FY08 Budget

• FY08 LQ budget (k) after all contingency was
  allocated

• Budget at FY08 start 3.4M
• After mid-year contingency allocation 4.0M
• After June contingency allocation 4.2M
• including 413k for LRS02

32
Budget and spending profile

• LQ budget and expenses (w commitments) in k

We have similar plots for each task, For FNAL
also with labor and MS breakdown
33
Schedule

• Options
• - LQC01 with S01 coils, LQS02 with new coils
  (FY09 budget 2.3M)
• - coils with new conductor for LQ03 (FY09
  contingency)

34
Conclusions

• The Long Racetrack has successfully opened the
  way for the use of coils and structures for long
  Nb3Sn accelerator magnets
• The Long Quadrupole FY08 work is completing the
  development of coils and structures for this goal
  adding accelerator quality features
• The Long Quadrupole FY09 plan aims at testing 3
  LQs by the end of CY2009 exploring the use of
  shell and collar structures, providing
• larger probability of success
• unique data and expertise for LHC-IR upgrades
• ? and for any future use of Nb3Sn magnets for HEP
  appl.

35
Extra
36
LQMSR concerns I

• The structure plan and decisions are perhaps a
  bit late for the stated goal of a demonstration
  in 2009- and you have to do better with respect
  to schedule that the program has done previously.
• Yes, there have been delays in the whole program.
  Therefore the LQ plan has features to reduce
  their impact, and the impact of other possible
  causes of delays coils will be reacted and
  impregnated at two labs (BNL and FNAL), two long
  structures are under procurement, we are planning
  to reuse some LQ coils in different structures
• The 4m shell design and process have numerous
  changes over what was demonstrated in the TQ
  program, raising the potential for unknown risks,
  and has challenging assembly operations. This
  should be carefully reconsidered.
• The changes should lead to performance
  improvement and help during the assembly of the
  long structure. Two tests of the LQS structure
  are planned with dummy coils a test at LN of a
  1m model, and a test at room temperature of the
  whole structure with Fuji films. Having both
  structures available provides further options and
  backup.

37
LQMSR concerns II

• The impregnated coil development to date has
  eliminated the positive coil-collar registration,
  an essential feature of most successful collared
  accelerator magnet campaigns, and this appears to
  complicate the absolute alignment scale up to
  longer lengths, and potentially raise the
  conductor strain in ways that affect the ultimate
  gradient performance.
• Other parts of the program (such as the HQ) are
  looking at coil alignment. This is not among the
  first LQ goals, although it can be tested by
  introducing keys in the collar structure
  (possibly after successful test of the LQC
  without keys, and of a TQC with keys). Alignment
  features were removed from the collar structure
  in order to allow more uniform stress
  distribution among the coils.

38
LQMSR concerns III

• The collared design performance results appear
  marginal both mechanically and in magnetic
  gradient. Have you really been careful enough
  with the coil strain state management conditions
  including low tech handling and tooling issues?
• See note by R. Bossert (TQC task leader).
• The shell design places emphasis on axial
  restraint (all load), while the collared design
  does not (30 of axial load seen at the ends).
  The test results even qualitatively do not
  resolve this issue, raising doubt about its
  importance, and emphasis, and thus how this
  engineering issue really scales scales to 4 m,
  and why there are so many changes in LQ01.
• Because of the importance of selecting the better
  of the two azimuthal coil support options
  (collar/shell), we have not changed the axial
  support levels associated with the collar/shell
  designs. The LQ structures aim at reproducing the
  pre-loading conditions of the best TQ models for
  each structure. Most changes of the LQS
  structures aims at easy assembly (masters), same
  load (SS rods) and lower pre-stress in the outer
  layer.

39
LQMSR concerns IV

• The anomalous results at 4.5K and 1.9K beg
  interpretation- at a minimum, this should include
  strand strain degradation studies coupled with
  analysis conductor sub-models that characterize
  peak strains in cable strands, and possibly
  retesting TQS02 with new coils to replace those
  suspected of being inferior.
• TQS02 has been tested with new coils showing some
  improvement at 4.5K, and no improvement at 1.9K.
  The understanding of these quench performance is
  among the main goals of the LARP magnet RD
  (strand and cable tests, TQ models with new
  conductor, HQ).
• We are concerned that all this may beat on the
  already narrow performance margins observed in
  the just completed 1m tests
• We agree. For this reason the results of upcoming
  conductor and magnet tests will be taken into
  account by the LQ RD (possible use of different
  conductor, and modifications coil fabrication,
  magnet assembly and pre-loading procedures).

40
Impact of skipped/delayed parts

• 3rd TQ generation LQMS (LQ short models)
• No test of LQS cross-section in TQ magnets
• No test of TQC with alignment features
• No test of 1m single-coil reaction fixture
• Test of 127 sb conductor planned in FY09
• 2nd LR using coils reacted in closed cavity using
  ceramic binder
• No test of reaction of long coils in a closed
  cavity using the ceramic binder (? accelerator
  quality coils)
• The LM compensated only partially because used
  Al-bronze poles w gaps
• LQ practice coils
• No feedback about fixtures and procedures before
  2008