hq01e3 test summary december 2012 m. marchevsky, lbnl
TRANSCRIPT
HQ01e3 test summaryDecember 2012
M. Marchevsky, LBNL
Original test plan
• Verify training and magnet status after CERN test - The primary objective for HQ01e3
- Correct operation of the newly installed SCRs, safety equipment (new valves), detection and acquisition systems (CryoDAC, fast event DAQ, acoustic piezo-sensors) following the upgrades will be verified.
• Validate and extend the magnetic measurements The plan is to use updated injection levels, test new MM probe and study connection between the flux jumps and field quality. AC loss studies with varying ramp rates will be conducted. • Heater efficiency and MIITs studies We have several proposals that address: - heater efficiency through measuring quench delay and MIITs evolution as
function of magnet current- resistance development in the magnet during fast rampdown and using
provoked quenches - spontaneous quenches with increasing detection/protection delays,
gradually increasing the MIITs - Study of self-protection capabilities of the magnet; also important for
establishing operational margins of the MTF.
- SCR operational issue has been discovered
- Indirect assessment of magnet state has been done through flux jump analysis
- FastDAC, acoustic data have been collected
- Magnetic measurements have been conducted as planned, but only up to 9 kA current.
- AC loss data analysis are underway
- No dedicated MIITS studies were conducted, but the 2 high MIITS provoked extractions and one ramp-rate quench were conducted with effectively absent dump resistor.
Heater configuration
PHA01 PHA02 PHB01 PHB02
Coil 9 1000 / 1000/100
1000 /1000/1000
851 /660/686
922(**)/1000/890
Coil 8 1000 /960/706
1000 /1000/1000
1000 /1000/960
567 / 590/590
Coil 7 797 (*) / 750/895
1000/1000/1000
1000 / 1000/1000
1000 /450/366
Coil 5 1000 / 1000/1000
1000 / 1000/1000
1000 /550/718
921 / 550/0
PHs to Coil Hi-Pot: LBL data (HQ01e) / CERN data (HQ01e2)/ HQ01e3
F5: C9A01+C8A02+C5A01+C7A01 -> powered to 190 VF6: C9A02+R+C7A02+C5A02 -> powered to 270 VF7: R+C8B01+C5B01+C9B02 -> powered to 150 V
Heater circuits formed:
->Heater degradation since the CERN test:May be related to the - high MIITS quench- testing at 1.9 KHowever, some degradation was seen already in HQ01
01: 500 A Vdump = 51 V02: 995 A Vdump = 101 V04: 1000 A Vdump = 94 V05: 2000 A Vdump = 181 V06: 3000 A Vdump = 274 V08: 4000 A Vdump = 378 V09: 5000 A Vdump = 406 V10: 5700 A Vdump = 407 V11: 5717 A Vdump = 408 V12: 175 A Vdump = 24V13: 300 A Vdump = 32 V15: 498 A Vdump = 50 V16: 1052 A Vdump = 98 V17: 1552 A Vdump = 137 V18: 2065 A Vdump = 172 V19: 3141 A Vdump = 256 V20: 3083 A Vdump = 274 V21: 3567 A Vdump = 326 V22: 3064 A Vdump = 275 V23: 5598 A Vdump = 409 V24: 5086 A Vdump = 163 V25: 3743 A Vdump = 119 V26: 1202 A Vdump = 39 V27: 3344 A Vdump = 74 V28: 2995 A Vdump = 96 V29: 998 A Vdump = 34 V
24: 5086 A Vdump = 163 V25: 3743 A Vdump = 119 V26: 1202 A Vdump = 39 V27: 3344 A Vdump = 74 V28: 2995 A Vdump = 96 V29: 998 A Vdump = 34 V
MM02: 1629 A Vdump = 54 VMM07: 6642 A Vdump = 218 VMM08: 5583 A Vdump = 178 V
RR01 (100 A/s): 7144 A Vdump = 234 VRR02 (100 A/s): 7154 A Vdump = 234 VRR04 (75 A/s): 10871 A Vdump = 406 V
Extractions / quenches summary
Rdump 120 mW
Rdump 30 mW
Rdump 30 mW
Extraction system has malfunctioned at Vdump>(406-409) V
System tests
Magnetic measurements
Ramp-rate quenches
SysTest11: extraction at 5.6 kA
~400 A
Only ~0.5% of magnet energywas extracted
Due to SCR malfunction, the extraction was interrupted at +38 ms and magnet re-connected back to the power supply
SysTest23: extraction at 5.7 kA
0 200 400 600 800 1000 1200 1400 16003
4
5
6
7
8
9
10
He
level (in
)Pcr
yo (
psi
)
t (s)
Lost ~ 40L of LHe over ~130 s
~15.7 MIITS
Flux jump statistics
Each flux-jump event was assigned to one particular coil. For every event the selected coil was the one with the largest absolute value of integrated flux imbalance. The histogram shows the number of events assigned to each coil.
Unusually large number of FJ events in Coil 7; was not seen in the HQ01d,e tests when quench antennas we re used to monitor flux jumps – possible degradation in C7
Each flux-jump event was assigned to one particular coil. For every event the selected coil was the one with the largest absolute value of integrated flux imbalance. The histogram shows the sum of the integrated flux imbalance of all the events that were assigned to each coil.
54/61 strands
108/127strands
MM ramp up to 9 kA and back down
Ramp-rate quenches
HQ01e35
8
8
5,
8
8
85
55
5
5,8
Ramp-rate quench current values and also quenching coils are is consistent with the earlier HQ tests
RR04: 10872 A quench at 75 A/s
Quench starts in the outer layer multi-turn of C5
0 200 400 600 8000.0
2.0x107
4.0x107
6.0x107
8.0x107
1.0x108
I2 (
A2 )
Time (ms)
12.3 MIITS
0 100 200 300 4000.00
0.01
0.02
0.03
0.04
R (
Ohm
)
Time (ms)
0 100 200 300 4000.00
0.01
0.02
0.03
0.04
0.05
0.06
Rm
ag (
Oh
m)
Time (ms)
H. Bajas et. al., MT22 presentation
Magnet resistance with no dump
Effective Rdump is very low, can be neglected
𝑅𝑚𝑎𝑔 (𝑡)=− 𝐿(𝐼𝑚𝑎𝑔 (𝑡 ))𝑑 𝐼𝑚𝑎𝑔(𝑡)/𝑑𝑡𝐼𝑚𝑎𝑔(𝑡)
Magnet resistance reached maximum of ~52 m .W - another reference datapoint adding to the CERN data of Rmag~120 mW at 15.9 kA / 18.3 MIITS
Independent coil resistance measurements
0 100 200 300 4000
2000
4000
6000
8000
10000
12000 Sound triggered Imbalance triggered Sound and imbalance
I (A
)
t (s)
0 100 200 300 4000
100
200
300
400
500
Sound triggered Imbalance triggered Sound and imbalance
Um
ax, (
mV
)
t (s)
Current and sound at the triggered events
0.5 1.0 1.5-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
0.112 ms
Sensor 1 Sensor 2
Usn
d (V
)
t (ms)
-4
-2
0
2
4
6
8
10
Imbalance
Uim
b (V)
Sound signals example at Imag=10036 A
Acoustic emission studies (RR04)
0 10 20 30 40 50 60 70 800.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Sensor 1 (bottom) Sensor 2 (top)
mV
rms2 /
Hz
f (kHz)
Sensor 1 is installed at the bottom load plate; sensor 2 is installed at the top plate
• HQ magnet produces increased acoustic emissions (seemingly unrelated to FJ) and high-frequency (>50 kHz) vibration “bursts” when energized above 9kA. The latter are occasionally correlated with the short imbalance spikes and most likely caused by stick-slip motion of the conductor
• Inductive pickups sensors can be developed and used in conjunction with acoustic devices to improve selectivity for the specific mechanical and electrical events
Also presented at WAMSDO 2013:http://indico.cern.ch/contributionDisplay.py?contribId=34&confId=199910