Quench Performance of Fermilab’s Nb 3 SN Magnets

1. Quench Performance of Fermilab’s Nb3SN Magnets Sandor Feher for the HFM group

2. Outline • Introduction • Magnet Test Goals • Magnet instrumentation • Test procedure • Results • HFDA02-04 – quench location near splice • HFDM01 – key experiment, splice test, instabilities in cable (1st, stimulated cable test with SR and short samples) • HFDM02 – pre-preg (effect of binder – current sharing) • SR01, SR02 – PIT is stable and MJR (even high RRR) is unstable • HFDM03 (PIT) – stable, SSL • HFDA05 (PIT) – stable, SSL • HFDM04-05 (RRP) - instabilities • HFDA06 (PIT) – stable, SSL, reproducible quench performance (1st) • Conclusions • Documentation: test reports, TD notes, papers HFDA/HFDM Quench Performance

3. Introduction • Fermilab has a world class test facility • Hardware & Software • Staff • Tested more Nb3Sn magnets than anyone in the world • 5 cos(Θ) dipoles, 5 mirror dipoles, 3 Race track dipoles, 1 double aperture common coil dipole, 2 small race track coils • Developed a detailed and thorough approach to testing Nb3Sn magnets • Test procedures • Instrumentation • Provides timely feedback to the design team • Can explore and probe issues that are hard to analyze and calculate • LARP provides an opportunity to collaborate even more closely with LBNL and BNL HFDA/HFDM Quench Performance

4. Magnet Test Goals • To be able to characterize and/or measure: • Quench current limit of the magnet • Conductor critical (quench) current limit: • Quench current plateau at 4.5K • Quench location (typically the high field region) • Temperature and ramp rate dependence (in detail) • After lower temperature (2.2K) quenches (high Lorentz forces) • return to 4.5K • determine if quench current remains the same • Quench velocity – high value; monotone change with time → uniformity • Quench performance limitation – instability (mechanical, conductor) etc. • Quench protection studies • Magnetic Field Quality HFDA/HFDM Quench Performance

5. Instrumentation • Voltage taps • Localize quenches • Spot heaters • Initiate and measure quench velocities • Temperature sensors • Measure Ic by warming up the coil using spot heaters in DC mode • Strain gauges • Used in assembly phase • Quench Antenna • Using pick up coils – only used few times when we had a bore in the magnet HFDA/HFDM Quench Performance

6. Test Procedures General plan for all the magnets – easy to compare them • Test Cycle I • Quench training – 4.5K, 20A/s • Quench locations (V-taps, Quench Antenna) • Voltage spikes • Current Ramp Rate Dependence • Magnetic Measurements • Temperature Dependence • Training the magnet first at 2.2K (20A/s) • Quenching at different temperatures • Quench heater • Strip heater – quench protection • Spot heater – quench velocities, quench integral, DC heating • AC losses • Splice Resistance • RRR • Test Cycle II • Quench Training HFDA/HFDM Quench Performance

7. The Magnets Cos(Θ) Dipoles Mirror Dipoles Cable/splice Small Race track Reached critical current limit HFDA/HFDM Quench Performance

8. Results Two types of magnets: • Quench current was only 50-60% of expected short sample limit (Bmax~6-7 T) • Magnets reached their critical current limit ( ~ 10T) HFDA/HFDM Quench Performance

9. Results Cos(Θ) Dipoles Mirror Dipoles Cable/splice Small Race track This group of magnets had poor quench performance. We made lot of investigations to understand the cause. HFDA/HFDM Quench Performance

10. Reasoning • Premature quenching due to: • Mechanical instability • Conductor movement – inadequate mechanical support; large spontaneous energy release • Epoxy cracking • Splices are not appropriate or other conductor damage • Conductor instability • Strand - Flux jump (and related – cooling, RRR, Cu/Sc etc.) • Cable - BICC • Tests performed to identify or to narrow down the cause: • Quench locations and velocities • Ramp rate dependence studies • Temperature dependence studies • Temperature margin measurements • Voltage spike and flux change studies HFDA/HFDM Quench Performance

11. HFDA02-HFDA04 Low quench current; Quench Location for HFDA02-04 magnets were at the splice region which is also a low field region; for HFDA04 splice was improved 20A/s quench locations are at the splice and cable junction Mirror magnet quench location as an example HFDA/HFDM Quench Performance

12. Splice resistances Rsplice= 0.60 ± 0.25 nΩ HFDA/HFDM Quench Performance

13. HFDM01 • Mirror magnet configuration was chosen because of its fast turn-around time so that we can determine the root-cause for the poor quench performance observed in the previous dipole magnets, HFDA-02 through 04 • Half coil from HFDA-03 dipole magnet was picked to fabricate the first mirror magnet with several additions • Two new sets of splice joints and various instrumentation were added to investigate the following issues • Current Sharing: Both lead end and return end splice joints would improve current sharing in the coil • Splice Joints: The new splice joints at lead end could be used to bypass the old splice joints to check if the poor quench performance observed in HFDA-03 was due to the result of conductor degradation during splicing operation • Conductor Damage: Spot heaters and temperature sensors were installed to measure the local behavior of the conductor HFDA/HFDM Quench Performance

14. HFDM01 Quench History • Short sample limit is 25 kA for the Mirror Magnet • At TC I the quench current is even lower than HFDA03 but at TC II it is higher. However, it is well below expectations • Different ramp profiles were used to test BICC: • Ramp with constant ramp rate to quench • Ramp to current plateau hold steady current for some time (max 3 hours) then ramp to quench • Do few current cycles before quench • No effect of these ramp profiles on quench current HFDA/HFDM Quench Performance

15. HFDM01 Quench propagation studies using spot heaters Slope of the voltage trace and time of flight velocity measurements indicate => Ic uniformity along the cable HFDA/HFDM Quench Performance

16. HFDM01-A (Splice/Cable) • The goal was to test the splice joints in a configuration similar to that in a magnet • The return end shunts were removed from HFDM01 coil and new NbTi lead cables were spliced • The cable next to the return end splice joints was cut to separate the mid-plane cable from the rest of the turns • This configuration would allow us to test the four splices in series or the two splices on the inner or outer layer separately HFDA/HFDM Quench Performance

17. HFDM01-A (Splice/Cable) Quench Locations TCI => outer cable quenches; not very well known the location due to long segment and low quench velocity TCII => some of them are multiple quenches within 25 msec time window Cold quenches: less than 4.3K LHe bath temperature HFDA/HFDM Quench Performance

18. HFDM01-A (Splice/Cable) Temperature dependence and Ic measurements • Direct Ic measurement using spot heaters and temperature sensors next to them showed that the cable has plenty of quench current margin • Quench current improved by lowering the LHe bath temperature • Quenches at low (<4.3K) temperatures are located at the junctions of the cable and splices HFDA/HFDM Quench Performance

19. Temperature dependence • Little gain of quench current at lower LHe bath temperatures • No direct evidence of mechanical limitation HFDM02 special ceramic pre-preg was introduced to reduce interstrand resistance and to increase strand stability • No significant qeunch current improvement HFDA/HFDM Quench Performance

20. Voltage Spike Studies • During magnet ramps we also observed voltage spikes • Two half coil signal are bucked • Spikes occur on every ramp • Spikes occur in both half coils at roughly the same time, but with somewhat different amplitude and detailed time structure • Spike frequency and amplitude increase with ramp rate • Spikes are associated with quench initiation in all plots; in a few cases there is on the order of 20 milliseconds between the spike and quench start; in one case, a spike occurs after the quench has begun propagating. • The magnet current always drops (sometimes dramatically, up to 20A in ≤ 1-2msec) at the time of a spike; a current increase has not been observed. HFDA/HFDM Quench Performance

21. Quench performance for MJR magnets • Although the High Field dipoles (02, 03, 04) quench behavior were quite below their expectations, we found that the quenches were at the splice and cable junctions; Mirror magnet and splice/cable tests revealed that splices are OK => low splice resistances and quench location for the splice /cable tests are not at splices • Quenches tend to occur in a relatively low field region • Only the mirror magnet with shunts showed strong ramp rate dependence which might mean that BICC is present, however, this effect was not a limiting factor at low ramp rate quenches HFDA/HFDM Quench Performance

22. Quench Performance of MJR Magnets • No evidence for local cable degradation since splice/cable quenches vary along the cable. On the other hand, between the “old” and “new” splices the cable exhibited Ic degradation. • Direct Ic measurement using spot heaters and temperature sensors next to them showed that the cable has significant of quench current margin • Lots of voltage spikes Conductor instability HFDA/HFDM Quench Performance

23. Results Cos(Θ) Dipoles Mirror Dipoles Cable/splice Small Race track To investigate conductor instability  Flux jump modeling, strand instability studies, cable tests  Built two race track (SR01 &SR02) structures (LBNL collaboration) HFDA/HFDM Quench Performance

24. Small Racetrack (SR) • LBNL design • Simple to fabricate • Quick cable test HFDA/HFDM Quench Performance

25. SR01 and SR02 • SR01 stable PIT conductor  reached its critical current limit • SR02 relatively high RRR still not stable MJR RRR = 129 RRR=125 HFDA/HFDM Quench Performance

26. Results Cos(Θ) Dipoles Mirror Dipoles Cable/splice Small Race track From stable PIT conductor we built first a mirror magnet, than after the successful test we built a dipole using the mirror half coil plus a newly wound coil. Later to check reproducibility we built another dipole using completely new coils made from PIT HFDA/HFDM Quench Performance

27. HFDA05 and HFDA06 Test Good conductor EXCELLENT MAGNETS • One of the HFDA05 coils were previously tested in HFDM03 → robust coil structure no retraining of the coil • Reached critical current limit • In the Bore 10.2 T • In the conductor 10.6T • Two identical magnets with very similar excellent quench performance → first achieved using Nb3Sn technology HFDA/HFDM Quench Performance

28. HFDA05 and HFDA06 Test • Good agreement with HFDA05 • Within 2% quench current limit HFDA/HFDM Quench Performance

29. HFDA05 and HFDA06 Test Strong Ramp rate dependence at higher ramp rate Steeper fall for HFDA06 At 5A/s HFDA6 Quench current is higher by 2% HFDA/HFDM Quench Performance

30. Results Cos(Θ) Dipoles Mirror Dipoles Cable/splice Small Race track Meantime we studied RRP cable →smaller Deff and higher RRR. HFDM04 cable made by LBNL → problem with cabling expected strand damage. HFDM05 cable partly made by LBNL and final key-stoning at FNAL HFDA/HFDM Quench Performance

31. HFDM04 & HFDM05 Test • In order to reach higher fields two high Jc RRP coils were tested in two magnetic mirror configurations. • Both magnets exhibited poor quench performance due to conductor instabilities. No quench current improvement observed • Low quench current values • Erratic quench behavior • Intensive voltage spikes • Multiple quench locations →Conductor instabilities HFDA/HFDM Quench Performance

32. Multiple Quench Locations Observed in the past → Quench acceleration Evidence of many quench locations simultaneously HFDA/HFDM Quench Performance

33. Fermilab contribution to LARP • FNAL unique 1.8K - 4.5K test capability • A standard test plan, developed under the Fermilab base program, helped us to identify complicated problems experimentally LARP accepted the test plan developed at Fermilab • Test Integration Group activity Joint effort among the three labs to share the resources and knowledge for all test activities • LBNL built Small Quadrupole (SQ01) which was tested at FNAL • It was a great success - demonstrated a close collaborative effort between LBNL and FNAL • MJR coil exhibited the same conductor instability at 2.2K • SQ02 will be tested at FNAL • TQSxx (built by LBNL) series are planned to be also tested at FNAL HFDA/HFDM Quench Performance

34. Conclusions • Excellent test facility, test procedures and personnel • Magnet tests demonstrated the importance of the conductor stability in building high field accelerator magnets • once the conductor is stable, the magnet technology developed at Fermilab reliably builds 10 T NB3SN accelerator magnets • Fermilab’s test capabilities will be a key component of the LARP program HFDA/HFDM Quench Performance

35. Documentation • TD notes • Test reports • Test facility description • Special experiments • Papers • ASC • Magnet Technology • PAC • EPAC • CEC/ICMC HFDA/HFDM Quench Performance