With contributions of: TE/MPE: Arjan Verweij TE/MSC-CI: N. Bourcey TE/MSC-TF: M. Bajko, G. Deferne, G. Dib, M. Charrondiere TE/MSC-SCD: L. Bottura, D. Richter, G. Peiro, C. Scheuerlein, S. Heck TE/MSC-LMF: P. Fessia, K. Chaouki, R. Principe, S. Triquet EN/MME: T. Regnalia, P. Perret TE/EPC: G. Hudson, M. Cerqueira EN/ICE: A. Rijllart, D. Kudryavtsev TE/CRG: V. Benda and many more... Thermal runaways in LHC main circuit interconnections: Experiments Gerard Willering 1 Technology Department Gerard Willering – Splice review – 18 October 2010 - CERN
27
Embed
Thermal runaways in LHC main circuit interconnections: Experiments Gerard Willering
Technology Department. Thermal runaways in LHC main circuit interconnections: Experiments Gerard Willering. With contributions of: TE/MPE: Arjan Verweij TE/MSC-CI: N. Bourcey TE/MSC-TF: M. Bajko, G. Deferne, G. Dib, M. Charrondiere - PowerPoint PPT Presentation
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
With contributions of:TE/MPE: Arjan VerweijTE/MSC-CI: N. BourceyTE/MSC-TF: M. Bajko, G. Deferne, G. Dib, M. CharrondiereTE/MSC-SCD: L. Bottura, D. Richter, G. Peiro, C. Scheuerlein, S. HeckTE/MSC-LMF: P. Fessia, K. Chaouki, R. Principe, S. TriquetEN/MME: T. Regnalia, P. PerretTE/EPC: G. Hudson, M. CerqueiraEN/ICE: A. Rijllart, D. KudryavtsevTE/CRG: V. Bendaand many more...
Thermal runaways in LHC main circuit interconnections: Experiments
Experiments to the effects of defected and consolidated LHC main bus splices are conducted in three phases
1. Thermal runaways in interconnections with defectsAugust 2009 – Februari 20105 quadrupole busbar samples in test station FRESCAGoal: Validation of model → safe operating current before consolidation
2. Proof of principle of the consolidation with shuntsMarch 2010 – June 20104 quadrupole busbar samples in test station FRESCAGoal: Validation of model → Proof of principle of the consolidation proposal
3. Final validation of the consolidation with shunts in a realistic test setupMarch 2010 – October 20102 dipole busbar samples inbetween two special SSS magnetsGoal: Validation of model and final validation of the shunts
- In the FRESCA teststation the sample length is limited to 1.7 m, which gives 0.8 meter of busbar on each side of the interconnection. -24 Voltage taps- 10 Thermocouples- 5 heaters- The ends of the busbars are thermalized (a lot of copper in direct contact with helium).- Measurements are performed with constant current.
- Due to limitations of the test station (Helium volume, length of sample, vincinity of the current leads) the quadrupole interconnections are chosen to test.
Fingerprint of a local thermal runaway:- Relatively low busbar temperature.- Accelerated voltage increase in the non-stabilized cable.Main characteristic: Thermal runaway time trun
- Except for sample 3B, all samples would melt within 1 second with a current of 12 kA.- The MIITs (kA^2/s) for an exponentially decaying current with timeconstant τ is reached by a constant current in t = 0.5*τ. - For the quadrupole circuit with τ = 20 s, we can correlate the safe currents for the sample conditions with a cross-section at trun = 10 s.
- Although there is a correlation, safe currents can not be drawn from the measurements.
The current at trun versus the additional resistance R add shows a good correlation.The allowed power at 10 K is between 16 and 27 W.
Since we varied the applied field on the sample, the effective Radd varied giving us a wider range in measurements. Therefore more than 5 points (number of samples) are shown.
To perform multiple thermal runaway measurements, the current is cut-off when the maximum temperature reaches in between 100 and 300 K. Out of 175 run-aways we did, we choose the smallest defect of 20 mm at 9 kA to demonstrate that the incident can be reproduced. In fact, each of the 175 measurements would lead to a melt-down.
With an increased protection cut-off voltage the thermal runaway was conducted until the cable melted over the full width over a length of 1.5 to 3 mm.- The temperature was at least 1360 K to melt the copper in the cable.- Remarkably, at the moment of melt-down, the thermocouple in the busbar 15 mm from the hotspot only measured 50 K.
Experiments to the effects of defected and consolidated LHC main bus splices are conducted in three phases
1. Thermal runaways in interconnections with defectsAugust 2009 – Februari 20105 quadrupole busbar samples in test station FRESCAGoal: Validation of model → safe operating current before consolidation
2. Proof of principle of consolidation with shuntsMarch 2010 – June 20104 quadrupole busbar samples in test station FRESCAGoal: Validation of model → Proof of principle of the consolidation proposal
3. Final validation of the consolidation with shunts in a realistic test setupMarch 2010 – October 20102 dipole busbar samples inbetween two special SSS magnetsGoal: Validation of model → safe operating current before consolidation
-Runaway time for the shunted samples much higher than for non-shunted samples.
- All the shunted samples can carry 13 kA for more than 24 seconds.
- The same data, but the MIITs are calculated (kA2*s) - The shunted samples with 1.5 and 3 mm thick shunts can handle the MIITs of 15.5 kA with τ = 20 s.- These samples do not have the worst case parameters and not the worst case conditions. Therefore no direct conclusions for LHC conditions.
Experiments to the effects of defected and consolidated LHC main bus splices are conducted in three phases
1. Thermal runaways in interconnections with defectsAugust 2009 – Februari 20105 quadrupole busbar samples in test station FRESCAGoal: Validation of model → safe operating current before consolidation
2. Proof of principle of consolidation with shuntsMarch 2010 – June 20104 quadrupole busbar samples in test station FRESCAGoal: Validation of model → Proof of principle of the consolidation proposal
3. Final validation of the consolidation with shunts in a realistic test setupMarch 2010 – October 20102 dipole busbar samples inbetween two special SSS magnetsGoal: Validation of model → safe operating current before consolidation
Goal: Test in realistic conditions of a worst case scenario, with a non-soldered shunt length of 8 mm and low RRR values.- 2 Special SSS spare magnets are connected to the testbench in SM18.- In total 35 meter of RQ busbar and 35 meter of RB busbar. - Two instrumented RB (M3) interconnections.- No magnets in the test-circuit
High precision measurements on resistance are important for the validation of shunt and model.- In the test the U-profile/wedge have a low RRR- In the tests the shunts have a much lower RRR than foreseen for the LHC conditions since they are not annealed
Test cycle: 14 kA, τ = 100 sTest cycle: 14 kA, τ = 140 sTest cycle: 14 kA for 22 s, then τ = 140 s. Still no signs of thermal runaway in the most critical shunt!!
Therefore we went to constant currents of 13 and 14 kA (power supply limit).
- No significant heating of the interconnection in 180 s.- No significant heating in the busbar Q9-1 in 180 s.- Normal zone does not enter the Q8 busbars.
- Very stable conditions at 13 kA in busbar and interconnection!!!
Q9-busbar have an RRR ≈ 250 → R9.4meter ≈ 2.3 µΩ at 10 KQ8-busbars have an RRR ≈ 300 → R16.5meter ≈ 3.5 µΩ at 10 K
- Small temperature increase in the interconnection in 85 s.- The full 35 meter of busbar between the quench-stoppers become normal- Accelerated heating effect in busbars Q8 and Q9-2.
- Limitation factor is not the shunted interconnection, but the busbar.
Q9-busbar have an RRR ≈ 250 → R9.4meter ≈ 2.3 µΩ at 10 KQ8-busbars have an RRR ≈ 300 → R16.5meter ≈ 3.5 µΩ at 10 K
- In the straight section the busbars are encapsulated in a G10 casing and close to each other. - In the region closer to the interconnection superfluid helium is available for cooling.
MIITs >> 30000 kA2s at 13 kA (no sign of thermal runaway) MIITs > 18000 kA2s at 14 kA (Start of thermal runaway in busbars)(LHC 13 kA, 100 s – MIITs = 8500 kA2sLHC 11.8 kA, 100 s – MIITs = 6800 kA2s)
Thermal runaways in interconnections with defects- Clear proof of the damage a defect can have with the melted sample.- Measurements provided largely sufficient experimental data for model validation (by A. Verweij).- Conclusions on safe current/energy cannot be drawn directly from this measurements, since test conditions are different from machine conditions.
Proof of principle of consolidation with shunts- Clear improvement of the thermo-electric stability by applying shunts on the samples with defects.- Boundary conditions of the test-station prohibit direct conclusions on the stability of the consolidated interconnection, but indicates that the principle good.- Sufficient experimental data for model validation (by A. Verweij).
Final validation of the consolidation with shunts in a realistic test setup- A consolidated interconnection with a copper shunt having a cross-section of 45 mm^2, a double defect in the interconnection, a non-soldered lenght of 8 mm and an RRR of 160 is more stable than the busbar itself in the straight section. - In the condition a quench starts in the interconnection itself a continuous current of 13 kA does not show any sign of a thermal runaway in the first 180 seconds.- At a continuous current of 14 kA provokes an excellerated temperature increase in the encapsulated part of the busbars, with a temperature of about 40 K after 85 s.- In terms of thermo-electrical stability the shunt is overdesigned.