Summary of Magnet Discussions

BNL -FNAL - LBNL - SLAC Summary of Magnet Discussions P. Wanderer LARP CM9 October 19, 2007 1 We magnet folk would like to thank our SLAC hosts, particularly Tom and Naomi. We would also like to thank our colleagues from CERN for their comments and (bold) proposal. 2 Key questions & answers • Where are we? We discussed data from several 1m quad and other recent Nb3Sn dipoles • Where do we want to go? a) 200 T/m, 4 m quad by 12/31/09 b) more understanding of materials & magnets c) Carefully evaluate proposed “Phase I” goal • When do we get there? a) Three “long, strong” quads by 12/31/09 b) Initial evaluation Dec. 1; report/revu June08 3 Where are we? We can make Nb3Sn magnets that •reach ~ 90% of conductor limit (4.5K) •with predictable harmonics Examples: •1 m cos2θ models (several) •1 m and 2 m cosθ mirror dipoles • 4 m racetrack coils • But … we don’t understand everything yet 4 • Mirror = one coil + iron in place of 2nd coil Materials – the “fundamental building blocks” of our magnets 5 RRP Strand Development with OST (Barzi) Increase Subelement Spacing Increase Subelement Number 54/61 restack Jcmax~3000 A/mm2 RRP0 108/127 restack Jcmax~2400 A/mm2 60/61 restack with spaced SE’s Jcmax3000 A/mm2 RRP1 114/127 restack with spaced SE’s Jcmax~3000 A/mm2 6 TQ Data: BNL-FNAL-LBNL 1.9 K calc Ic calc +/- 0.1% strain 7 Strand Production 54/61 114/127 Strand Inventory kg kg MAGNET Req. kg of 54/61 kg Oct-05 33 SR01 7 26 Nov-05 70 96 Dec-05 96 Jan-06 96 Feb-06 96 Mar-06 90 TQC02 40 146 Apr-06 LRS01-C01 27 119 May-06 90 TQC02-R 35 174 Jun-06 TQS02 35 139 Jul-06 139 Aug-06 139 Sep-06 LRS01-C02, SQ 36 103 Oct-06 103 Nov-06 103 Dec-06 30 LQM01 37 96 Jan-07 90 186 Feb-07 30 216 Mar-07 216 Apr-07 50 LRS01-C03 30 236 May-07 236 Jun-07 SQ03 236 Jul-07 236 Aug-07 PCX01 0 236 Sep-07 150 386 Oct-07 LQ01, LQM02 200 186 60kg R&D Stock, only used for practice coils. PC: Practice coil for LQ coil winding Oct-07 Nov-07 Dec-07 Jan-08 Feb-08 Mar-08 Apr-08 May-08 Jun-08 Jul-08 Aug-08 Sep-08 Oct-08 Nov-08 Dec-08 Jan-09 Feb-09 Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 54/61 114/127 Strand Inventory of Inventory of kg kg MAGNET Req. kg 54/61 kg 114/127 kg 186 0 186 0 80 186 80 186 80 100 LQ02 165 186 15 186 15 186 15 186 15 90 186 105 186 105 110 HQ01 165 186 50 186 50 186 50 200 FNAL 85 186 165 186 165 CDP 180 6 165 6 165 6 165 6 165 6 165 6 165 6 165 6 165 6 165 6 165 8 RRP-8648, 0.7 mm RRR ~ 310 2.09 K 1600 PS. LIMIT 1400 1200 1000 Ic (A) 800 600 400 200 0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 9 Ic 2.09 K FIT Iq V-H No Quench H (T) TQ - 1 m quadrupoles – quench performance at 4.5 K, 1.9 K TQS – shell support TQC – collar support 10 TQS01 (technology quad – shell support structure) 11 Summary (Caspi) 4 TQS tests: Maximum gradients at 4.4K - 178-222 T/m Plateau reached after 4-20 quenches with 80-90% of short-sample Maximum gradients at 1.9-3.2K - 192-225 T/m Plateau reached after 20 quenches with 80-84% of short-sample Bronze/Titanium islands under axial tension/compression TQS01 quenches in inner layer straight section (near island gaps) TQS02a dominated by outer layer quenches At 12 kA layer 2 (outer) field is 1.2 T less than layer 1 (inner) 12 12 Summary – issues (Caspi) Unexpected outer layer quenches and plateau. PW: discussion  outer layer quenches very likely originate in turn with maximum preload (> 150 MPa?) A monotonic increase in training current between 4.4K and 1.9K (no jump) No gain in current returning from 1.9K to 4.4K Epoxy “bubbles” at 1.9 K on the inner layer free surface High MIITS (7+) degrade the plateau and impacts coil stress But heating can improve performance (reduces strain ?) 13 13 TQS01 & TQS02 Plateau Quenches Lietzke & Lizarazo Three TQS-magnets performed below “short-sample” projections: TQS01a (coils 5, 6, 7, 8) TQS01b (coils 7, 8, 14, 15) TQS02a (coils 20, 21, 22, 23)  coils into “TQE02” Quench onset and propagation examined: Short-sample error? Mechanical motion? Flux-jumps? Local conductor degradation? – PW: most likely cause … but source degradation unknown 14 TQE02 (likely to be renamed TQC02) uses coils from TQS02 (Bossert/Ambrosio) TQE02: Quench History 14000 12000 10000 8000 6000 Training 4.5K 4000 2000 0 0 5 10 15 20 25 30 35 40 45 50 lead ramp rate Training 1.9K 15 Magnet data  cable testing as a function of the usual variables (temperature, background field) + strain is important  discuss collaboration with CERN 16 HQ – Explore larger aperture, higher gradient 17 HQ Magnet (Caspi slide; + Felice) 130 mm bore quadrupole with a gradient of ~ 200 T/m ( 2 layers). Option to inserts TQ coils ( 2+2=4 layers) ,90 mm bore, Gradient ~ 300 T/m •Cross-section • • • • • • 2 layers Clear bore diameter – 133.6 mm (use with TQ coils) Outer coil diameter – 198.3 mm Strand diameter 0.8-1.0 mm (so far: 0.7 mm) Cable width 15 mm (28-36 strands) - key-stone 1 wedge per layer •Gradient ( 205 T/m) in 130 mm aperture •1 m long Nb3Sn coils •Azimuthal Stress ~ 150 MPa  Shell support structure specifically designed for HQ (Felice) 18 18 Where do we want to go? Long range: develop Nb3Sn magnet as candidate for Phase II upgrade of LHC IR (2016) FY08: Evaluate opportunity to contribute Nb3Sn quads to Phase I upgrade (2012) [HQ?] FY09: Demonstrate good quench performance of 4 m quads, G >200 T/m As resources permit: Larger aperture/higher gradient (HQ) Improve Nb3Sn strand (60/61114/127) 19 LQ: 4 m, G > 200 T/m, 90 mm aperture, Nb3Sn quadrupoles 20 LQ features (Ambrosio) • Change as few things in TQ (1m) series as possible – Conductor – Coil • Major issues due to extra length: – Stored energy/quench protection – Bow in reacted coil caused by coil (Fermilab 4m dipole) • Major choice: – Support structures: shell, collar 21 LRS01 – long racetrack, shell 4m (4.5 K) . 10000 Iss = 10600A (4.5K) Quench Current (A) 8000 6000 I (plateau) ~ 91% Iss 4000 2000 50 A/s 20 A/s 20 A/s 10 A/s 20 A/s 30 20 A/s 0 0 5 10 15 50 A/s 100 A/s 25 Quench Number 20 22 5 A/s LQ with collar support structure 23 LQ Status (Nobrega + Schmalzle) Reaction tooling is being fabricated and is a single cavity, laminated fixture that can be used for both reaction and epoxy impregnation. Practice 24 Assembly – LQ based on TQC (Bossert) Coil Midplane “Gap” is equal to coil oversize plus coil midplane shim determined from FEA and previous experience with short models. 4. Assembly of coils with midplane shims. 5. Collaring in 5-6 passes to achieve target preload. 25 LQ with shell support structure 26 2D magnet design from TQS02 to LQS01 (Ferracin) TQS Structure designed for a 4-layer coil Initial design included ss pads Key location determined by bladder width LQS Stainless steel pads (not brittle at 4.2 K) Original TQ magnetic design TQS02 Interference keys moved towards mid-plane Optimized coil stress (big benefit) 8 bladders (small price) LQS01 4 auxiliary bladders for yoke pre-assembly Aluminum shell from 22 mm to 20 mm thick Axial rods closer to coil Less deflection of end plate 27 Assembly process • Separate assembly of 4 shell-yoke sub. assemblies – Similar to TQS – Yoke laminations compressed with 1 m long tie rods – Shell pre-tensioned with yoke bladders (0.8 m) and gap keys • Assembly of 4 segments – Segments connected with pins – Yoke laminations compressed with 3.3 m long tie rods • Insertion of coil-pad sub-assembly • Bladders (1.6 m) to align and pre-load 28 Alignment features • Holes for pins shell - yoke • Slots for keys yoke – yoke • Hollow pins between yoke laminations • Slots for keys pad - yoke • Interference keys • Masters to facilitate alignment during assembly • Alignment options – No alignment (TQS case) – Shell-yoke • Necessary for assembly – Yoke-yoke and pad yoke • To be discussed 29 How do we get there? “Projectize” 4 m quad R&D (LQ) • Review R&D plan, two candidate support structures • “Prioritize” LQ •Track progress closely • Learn from 1 m R&D (e.g., lots of coils) • First test ~ 1 year from now • Continuing resolution is a handicap 30 • review: last week of November LQ Schedule & Budget (Ambrosio) | 2008 | 2009 | Plan: LQ01: shell; LQ02: collar; LQ03: TBD FY08 budget (w/o contingency): $3.1M 31 Opportunity/challenge Make 4/8/16 Nb3Sn quads for replacement of IR quads for Phase I upgrade (2012) Quads for Phase I  new funds (inside/outside LARP?) LARP, CERN discussions underway Evaluate cost, staff, schedule June 08: IRUWG report, DOE revue 32