PIT conductor development and production for NED Outline

PIT conductor development and production for NED Outline: 1. Introduction 2. PIT conductors at SMI before NED 3. R&D phase: a. Step 1 b. Step 2: strands B 207, B 215, characterization results c. Heat treatment optimization studies on B 215 4. Final strand production and characterization 5. a. Strands already delivered: amounts and characterization 6. b. Delivery planning 4 2008 conductor review, T. Boutboul 5. November Conclusions and NED perspectives 1

Introduction • Following call for tender in Summer 2004, Shape. Metal Innovation (SMI, Enschede, The Netherlands) awarded a NED contract (300 k€) for Powder-In-Tube (PIT) conductor. • PIT technique developed by ECN (since 1975) and purchased by SMI in 1992. • PIT technology acquired by EAS (Germany, now Bruker EAS) in December 2006 • PIT: Nb. Sn 2 powder (possibly enriched with pure Sn) compacted and inserted into a Nb-X/Cu tube. After tube drawing to smaller diameter, stacked into Cu tube with a Cu core for final cold drawing. November 4 2008 NED conductor review, T. Boutboul 2

PIT conductor situation at SMI before NED Before NED, SMI “state of the art” strand: B 179 • Jc ~ 2280 A/mm 2 @ 12 T. • 192 filaments, deff ~ 54 µm but for strand diameter of 1 mm (~70 µm at 1. 25 mm vs < 50 µm). • For D=1. 25 mm, needed more filaments. • Strand very sensitive to deformation. For NED (Step 1), SMI worked on improving Jc (powder, barrier, HT) November 4 2008 NED conductor review, T. Boutboul 3

R&D phase: Step 1 During Step 1, SMI developed two strands (1 mm in diameter instead of 1. 25 mm, ~ 50 µm filaments): B 201 and B 205 (~ 320 m for both). B 201: 192 filaments, to improve Jc SMI added free Sn to Nb. Sn 2 powder but, during HT, Sn leaks!! B 201, SEM B 201, Optical B 201, no heat treatment B 205: less Sn-enriched powder used than B 201 but Sn leaks as well… November 4 2008 NED conductor review, T. Boutboul 4

R&D phase: Step 1 (end) • Upon CERN request, B 201 strand design incorporated more copper between the filaments (tubes) than B 179 for an improved mechanical behavior following cabling deformation: Flat rolling deformation B 179 25 % def. B 201 • Two billets with Ta barrier tried for full reaction but very serious workability troubles… • HT: 84 h (instead of 64 h) @ 675 o. C on B 179: Jc ~ 2410 A/mm 2, 12 T November 4 2008 NED conductor review, T. Boutboul 5

R&D phase (Step 2): B 207 • For the Step 2, SMI produced a strand, B 207 (~ 300 m long), including 288 filaments, nominal NED diameter (1. 25 mm) but Cu/non-Cu = 0. 96 (instead of 1. 25) due to bad die selection at drawing beginning. • Even more copper inserted between filaments than B 201 strand to further improve mechanical behavior. Powder composition as B 179. • However: Jc = 2077 A/mm 2 (12 T), 1118 A/mm 2 (15 T), i. e. ~ 13 % lower than B 179: disappointing. RRR = 250 (virgin), 40 -80 (def. ). Low Jc due to a mistake during powder preparation. November 4 2008 NED conductor review, T. Boutboul 6

R&D phase (Step 2): B 215 • B 215 strand fabricated (summer 2006) according to B 207 design. • 950 m produced in a lone length. • 288 filaments (~ 50 µm) fulfilling NED specification. • Adequate Cu/non-Cu (~ 1. 22) and Strand diameter (1. 257 mm). B 215 cross-section (no heat treatment) November 4 2008 NED conductor review, T. Boutboul 7

B 215 strand characterization (standard HT) • Critical current measurements done at CERN, INFN/Milan and Twente University with the heat treatment recommended by SMI (84 h @ 675 o. C). • Max. measured Jc ~ 2500 A/mm 2 in non-copper part, Ic=1397 A (Twente), 12 T and 4. 2 K. • Combination of high Ic (12 T) and fine filament size (~ 50 µm): world record!! • Ic=756 A @ 15 T, 4. 2 K (only 8 % below spec. ), Jc=1350 A/mm 2 (Geneva). November 4 2008 NED conductor review, T. Boutboul 8

B 215 strand characterization (standard HT) • Acceptable: RRR ~ 70 -80 (virgin samples). • Magnetization measurements (Frascati) on VSM instrument: few flux jumps! November 4 2008 NED conductor review, T. Boutboul 9

HT optimization studies on B 215: motivation • • In PIT reacted strands: A 15 phase composed of coarse grains (1 -2 µm, ~ 30 %) and fine grains (~ 200 nm). Jc contribution of coarse grains negligible. 0. 20 µm 0. 21 µm The idea: try to improve Ic and RRR by reaction temperature decrease in order to (tentatively): 1. To reduce fine Nb 3 Sn grain size 0. 23 µm 0. 22 µm B 215 sample, 84 h @ 675 o. C 2. To reduce coarse grain area (possible? ) November 4 2008 NED conductor review, T. Boutboul 10

HT optimization studies on B 215: method • Due to limited amounts of available B 215 strand to limited reaction oven resources , a systematic optimization study with broad scanning of treatment temperature and duration not possible. • Instead, a kind of “trial and error” method used: a specific HT schedule is selected, then Ic, RRR measured, cross-section examined at microscope to evaluate the extent of the reacted area. According to these results, the next schedule is chosen… • Various temperatures and durations already tried with the same temperature ramp of 50 o. C/hour: 660 o. C: 84 h 650 o. C: 84 h, 120 h 625 o. C: 200 h, 260 h, 320 h and 400 h November 4 2008 NED conductor review, T. Boutboul 11

Prominent results 120 h @ 650 o. C • Ic measurements performed at CERN and Twente. Consistent Ic data measured with maximal value of 1410 A (12 T, 4. 2 K), corresponding to Jc ~ 2520 A/mm 2, ~ 1 % larger than 84 h @ 675 o. C. • Although slight increase not significant, measurements more reproducible (Ic scatter of less than 3 % versus ~ 6 % for standard HT). • Measurements at Geneva: 1 % as well at 15 T and 4. 2 K, Bc 2 K~25. 6 T. • From RRR point of view, 120 h @ 650 o. C appears to be more favorable than 84 h @ 675 o. C. Indeed, RRR values, generally larger than 100 (max. 143) for 120 h @ 650 o. C as compared to values not exceeding 80 (standard HT schedule). November 4 2008 NED conductor review, T. Boutboul 12

320 h @ 625 o. C After treatments at 650 o. C, decided to try 625 o. C. As preliminary test, B 215 sample treated 61 h @ 620 o. C to check reaction feasibility at such low temperature. • Good news: thin A 15 layer already reacted (dark grey). • Bad news: large area of Nb 6 Sn 5 (white layer close to core) also present, to be converted to Nb 3 Sn coarse grains (at least 25 % of reacted Nb 3 Sn phase). Courtesy of G. Arnau and C. Scheuerlein November 4 2008 NED conductor review, T. Boutboul 13

320 h @ 625 o. C (continuation) • After two trials at 625 o. C (200 and 260 hours), 2 samples were treated 320 h @ 625 o. C at CERN, 2 additional samples reacted at Twente. • Ic data consistently measured at CERN and Nijmegen in the range: = 1494 -1539 A @ 12 T, 4. 22 K, corresponding to Jc ~ 2700 A/mm 2, + 10 % as compared to standard HT. • 15 T, 4. 222 K: Ic > 818 A (NED spec. ), Jc ~ 1500 A/mm 2 • New record!! B 215 strand completely fulfilling NED specification. • RRR data impressive as well since RRR ~ 220 for virgin strands!! Better for stability. • Last treatment: 400 h @ 625 o. C but no Ic gain and lower RRR (~ 120) November 4 2008 NED conductor review, T. Boutboul 14 Ic

320 h @ 625 o. C (continuation) Bc 2 K = 26. 3 T (vs 25. 6 T) Stringent NED specification fulfilled!! November 4 2008 NED conductor review, T. Boutboul 15

Specific heat measurements on B 215 (University of Geneva) Courtesy of C. Senatore and R. Flukiger (Geneva Univ. ) November 4 2008 NED conductor review, T. Boutboul 16

SEM examinations on B 215 polished and fractured samples Un-reacted barrier ratio Sn content (average) Coarse grain ratio Mean fine grain size 675 o. C/84 h 625 o. C/320 h 23 % of total filament 25 % of total filament 24. 2 at. % Sn 24. 7 at. % Sn ~ 30 % of A 15 phase ~ 180 nm ~ 160 nm Sample with optimized HT: less reacted Nb 3 Sn but with higher quality: higher Sn content, Bc 2 K, Tc and smaller grain size (160 nm vs 180 nm) November 4 2008 NED conductor review, T. Boutboul 17

Final strand production and characterization (1) • Following successful R&D, EAS/SMI was awarded the green light to launch the final strand production in fall 2007. • Final strand manufacturing (12. 7 km of strand overall) is currently underway and the full strand delivery to CERN has taken a few months delay (was expected by summer 2008). • A first 1000 m long strand, B 228, produced partly at Hanau (EAS) and partly at Enschede (SMI) delivered to CERN in April 2008 into 2 strand lengths (400 m + 600 m). • A second strand, B 230 (1650 m), was completely produced at Hanau, except final strand drawing to nominal diameter. This strand was delivered to CERN in July 2008 into 4 strand lengths. Since July 2008, equipment transfer from SMI to EAS completed. November 4 2008 NED conductor review, T. Boutboul 18

Final strand characterization B 228 strand: • Strand diameter: 1. 254 mm, Cu/non-Cu = 1. 19 • When reacted 120 h @ 650 o. C, Jc ~ 2450 -2510 A/mm 2, RRR > 140 • Very similar to B 215 strand: EAS/SMI able to reproduce nice perf. !! B 230 strand: • Strand diameter: 1. 253 mm, Cu/non-Cu = 1. 26 • According to EAS measurements: Jc = 2513 A/mm 2, 4. 2 K and 12 T (650 o. C/120 h). However, at CERN, Jc ~ 2400 -2430 A/mm 2 (- 3 -4 %). RRR ~ 50 -90, lower than B 228 (Sn leaks? ) • Indication of smooth technology transfer from SMI to EAS, despite 2 -3 % Jc discrepancy between B 228 and B 230. Radial expansion of ~ 2 % for both B 228, B 230 (very typical for PIT). November 4 2008 NED conductor review, T. Boutboul 19

Final strand production: delivery planning • 3. 8 km of strand just arrived to CERN (October 31 2008)! It was delivered into two lengths (1. 4 km + 2. 4 km). • 6. 3 km in two strand batches should arrive to CERN till end of 2008 (December 2008). November 4 2008 NED conductor review, T. Boutboul 20

Conclusions and perspectives • During R&D phase for NED, SMI developed a strand with ~ 50 µm filament size and Jc ~ 2500 A/mm 2 at 12 T and 4. 2 K with HT schedule recommended by firm (84 h @ 675 o. C). • For 320 h @ 625 o. C, increase of 10 % as compared to standard HT observed in critical current (Ic ~ 1500 A @ 12 T and 4. 2 K, Jc ~ 2700 A/mm 2, Ic ~ 840 @ 15 T and 4. 2 K, Jc ~ 1500 A/mm 2), new record!! RRR high (~ 220) as well for this treatment. • This strand fully within NED specifications, due to optimized HT. • EAS/SMI currently producing final strand according to R&D strand design. 6. 4 km already delivered and part of it found with current abilities almost similar to R&D. Remaining 6. 3 km to be delivered by end of 2008. November 4 2008 NED conductor review, T. Boutboul 21