Transcript PH-2L (2 layers) - Indico

Test Program and Results
Guram Chlachidze for
FNAL-CERN Collaboration
September 26-27, 2012
Outline
• Test program
• Quench Performance
• Quench Protection
• Magnetic Measurements
• Summary
MBHSP01 Test at VMTF
• MBHSP01 - the first 11T demonstrator dipole magnet was tested at
Fermilab’s Vertical Magnet Test Facility (VMTF) in June-July 2012
• VMTF was designed for testing magnets up
to 4-m length and 0.6-m diameter
at temperatures between 1.8 K and 4.6 K
– 30 kA power system
– 30 kA/1 kV DC rated dump resistor
(15 mΩ – 120 mΩ)
– Two 15 kA/1 kV DC rated solid-state dump
switches
• 60-mΩ dump resistor was used for the stored
energy extraction
Guram Chlachidze, FNAL-CERN Collaboration
2
MBHSP01 instrumentation
• Voltage tap system covers the inner and outer coil layers, pole turn, multiturn and splice sections. There are 10 voltage taps on the inner layer and 9
voltage taps on the outer layer
Guram Chlachidze, FNAL-CERN Collaboration
3
Magnet instrumentation (cont’d)
• Each coil was equipped 2 pairs of protection heaters. Each pair covers 28%
of the total coil surface
• 1-layer or 2-layer 5-mil (125 µm) Kapton insulation was placed between the
heaters and the outer-layer coil block
PH-1L (1 layer)
PH-2L (2 layers)
PH-1L (1 layer)
PH-2L (2 layers)
Guram Chlachidze, FNAL-CERN Collaboration
4
Magnet instrumentation (cont’d)
• 64 strain gauges (SG) were installed on shell, coils and bullets for
monitoring mechanical strain and coil stresses during the magnet
construction and testing
• 4 resistive temperature sensors (RTD) were mounted at top, middle and
bottom of the magnet outer skin
• No quench antenna was available for this 60-mm aperture magnet
Guram Chlachidze, FNAL-CERN Collaboration
5
MBHSP01 Test at VMTF
• Cold test program included quench training, ramp rate dependence study,
protection heater study and field quality measurements both at 4.5 K and
1.9 K
• Temperature dependence study, RRR and splice resistance measurements
also are part of our standard test program. AC loss measurement was not
included in the test program
• The original test program was modified due to limited magnet performance
–
–
–
–
Quench training at various ramp rates
Additional tests to investigate magnet stability
Field quality measurements at a maximum current of 6500 A
Heater study at a maximum of ~ 65% of SSL
• Long test with several interruptions
Guram Chlachidze, FNAL-CERN Collaboration
6
Quench Performance
Unscheduled thermal cycle at the very end of the test
TC-1
4.5 K
TC-2
1.9 K
Guram Chlachidze, FNAL-CERN Collaboration
4.5 K 1.9 K 2.6-4.5 K
7
SG data: cool-down
Pole gauges in coils 2 (top)
& 3 (bottom)
Guram Chlachidze, FNAL-CERN Collaboration
8
SG data: cool-down
Strain gauges on shell (skin)
and bullets
LE and RE Skin
LE Bullets
RE Bullets
Guram Chlachidze, FNAL-CERN Collaboration
9
Ramp rate dependence
10.4 T or 78% of SSL
SSL:
4.6 K
1.9 K
13 kA
15 kA
No quench when ramping down from 8 kA at a ramp
rate of 120 A/s
Possible conductor damage in the mid-plane block
may be reason for a limited magnet performance
Guram Chlachidze, FNAL-CERN Collaboration
10
Quench locations
2b2_b1 x 10
2b2_b1 splice segment picks up a
signal from the quenching 2b3_b2
Segment (mid-plane block)
- checked for a PH induced quench
Heater induced quench
2b2_b1 x 10
Upper limit of b2_b1 splice
resistance estimated as <2 nΩ
Guram Chlachidze, FNAL-CERN Collaboration
11
Quench Locations
Coil-2/3 IL/OL MP: A2-A3, B2-B3
Coil-2/3 IL:
A4-A5, A5-A6
Coil-2/3 OL:
B3-B4, B5-B6
• Only few training quenches
in pole turns
Guram Chlachidze, FNAL-CERN Collaboration
12
Holding quenches
Ramp to a pre-set current at 20 A/s and hold this current until quench
All holding quenches initiated in the mid-plane block of coil 2 OL (B2-B3)
– Zero holding time is a regular quench at 4.5 K or 1.9
Reproducibility test, tests at different ramp rates were not done
Guram Chlachidze, FNAL-CERN Collaboration
13
Test with DC powered heaters
• Instability test - reducing Jc in the mid-plane area
• Small DC current through the protection heaters
– PH cover also multi-turn pole block
• No improvement in quench performance when DC current in PH is ON
Guram Chlachidze, FNAL-CERN Collaboration
Name
14
Temperature Dependence
50 A/s ramp rate in most quenches. All quenches at intermediate
temperatures were initiated in the mid-plane block of coil 2 OL (B2-B3)
Magnet showed temperature dependence of quench current, but exhibited
degradation and instability
SSL
Guram Chlachidze, FNAL-CERN Collaboration
15
RRR measurement
Average RRR ~ 100 (lowest 80, highest 118)
Same segments in different coils have similar RRR
Coils with RRP 108/127 strand
TQ coil 34:
185
HQ coil 14:
80
LQ coil 14
(RRP 114/127):
180
Guram Chlachidze, FNAL-CERN Collaboration
16
More tests: voltage spikes,
quench propagation speed
Voltage Spike Detection System captures half-coil signals at a sampling
rate of 100 kHz
Quench propagation speed was estimated
in ramp # 5 (Iq 9.4 kA) : ~27 m/s
Guram Chlachidze, FNAL-CERN Collaboration
17
Summary on Quench Performance
•
Magnet showed limited quench performance and reached only 10.4 T or 78%
of SSL at 1.9 K. Magnet did not reach quench plateau
•
Most quenches at low ramp rates, as well as holding quenches and quenches
at intermediate temperatures were initiated in the mid-plane block of the
outer-layer coil
•
Only few training quenches occurred in the high field area at the very
beginning of test at 4.5 K and 1.9 K
•
Plan to have a quench antenna for a better quench localization
•
Quench location, ramp rate and temperature dependences and additional
tests indicate magnet degradation and related instability. Possible conductor
damage in the mid-plane area (presentation by F. Nobrega) could be a reason
for this degradation
Guram Chlachidze, FNAL-CERN Collaboration
Name
18
Protection Heater Study
• Heat transfer from the heater to the outer coil layer and then from the
outer-layer to the inner-layer coil helps to spread and absorb the magnet
stored energy
• Temperature profile in the magnet after 48 ms (left) and 96 ms (right) from
the PH induced quench at a nominal current of 11.8 kA
Experimentally verified for a
PH induced quench at 8 kA
with dump delayed for 120 ms
After 65 ms quench starts in
the outer coil layer and after
150 ms - in the inner coil layer
- More tests expected with next 11T
prototypes
Guram Chlachidze, FNAL-CERN Collaboration
19
Protection Heater Study (cont’d)
• PH-1L and PH-2L are heaters with one and two layer Kapton insulation
respectively
PH-1L
• PH peak power 25 W/cm2, HFU voltage decay
PH-2L
PH-1L
time 25 ms
PH-2L
• PH-2L delay is large at low currents
Guram Chlachidze, FNAL-CERN Collaboration
20
Protection Heater Study (cont’d)
Peak power in heaters only 25 W/cm2
Compare to 50 W/cm2 in LQ magnets
Guram Chlachidze, FNAL-CERN Collaboration
21
Summary on PH study
• Heat transfer from the PH to the outer layer coil and then to the inner
layer coil was experimentally observed. This effect helps to spread
and absorb magnet stored energy
• More tests will be done with next 11T prototypes
• Protection heaters with different insulation were tested both at 4.5 K
and 1.9 K
• Protection heater with 2 layers of 5-mil (125 µm) Kapton insulation
found less efficient than PH with 1 layer of Kapton insulation
• Heaters with 1 layer of Kapton insulation will be used in next magnet
for protection and heater study
Guram Chlachidze, FNAL-CERN Collaboration
22
Magnetic measurements
• Warm and cold magnetic measurements were performed at VMTF with
the magnet in a vertical position
• Fast rotating coil magnetic measurement system based on a digital signal
processor (DSP) was used for the measurements
• 250 mm long and 25 mm diameter tangential probe, as well as printedcircuit board (PCB) based 26 mm long and 130-mm long probes were used
for measurements
– 25-cm tangential probe was used only at room temperature and for few “cold”
measurements at 4.5 K . This probe was rejected after signals were found noisy
Guram Chlachidze, FNAL-CERN Collaboration
23
Magnetic measurements (cont’d)
• All harmonics are presented at a reference radius of 17 mm
• Data measured with the tangential probe and 130-mm PCB probe are
consistent
• Sign inconsistency was found for a2 and other even-ordered harmonics
– 180 degree phase shift between the probes
– Absolute values are consistent
– Currently under investigation
sign mismatch corrected
Guram Chlachidze, FNAL-CERN Collaboration
24
Impact of ramp rate
Guram Chlachidze, FNAL-CERN Collaboration
25
Impact of temperature
• Only tangential probe data available for comparison
• We see about 3% decrease in strand magnetization at 1.9 K
- 7% increase expected from the simulation (see next presentation by Mikko
Karppinen)
• Need to verify with PCB probes during the next test
Guram Chlachidze, FNAL-CERN Collaboration
26
Impact of reset current
20 A/s loops
Guram Chlachidze, FNAL-CERN Collaboration
27
Accelerator cycle measurements
•
•
•
•
•
•
Only tangential probe data available
Maximum current of cycle 6500 A
Reset current 100 A, ramp rate 10 A/s
Injection plateau at 760 A, dwell time 900 s
No snap-back or decay was observed
Injection plateau close to the b3 minimum
Guram Chlachidze, FNAL-CERN Collaboration
28
Stair-step measurements
• 130-mm long PCB probe data
• Stair steps up and down from 1000 A to 6500 A
• Dwell time at flattop 120 s
Guram Chlachidze, FNAL-CERN Collaboration
29
Summary
• The first 11T demonstrator magnet was assembled and tested at Fermilab
• Magnet showed limited quench performance and reached only 10.4 T or
78 % of SSL at 1.9 K
• Most quenches at low ramp rates, all holding quenches and quenches at
intermediate temperatures initiated in the mid-plane block of the outer
coil layer
• Quench location, ramp rate and temperature dependence studies indicate
magnet degradation and related instability. Possible conductor damage in
the mid-plane area during fabrication could cause the observed
degradation
• Protection heaters (PH) with different insulation thickness were tested
Guram Chlachidze, FNAL-CERN Collaboration
30
Summary (cont’d)
•
PH with 2 layers of 5-mil (125 µm) Kapton insulation found not efficient,
therefore PH with 1 layer of Kapton insulation will be used in next magnet for
protection and heater study
• Heat transfer from the PH to the outer layer coil and then to the inner layer coil
was experimentally observed. This effect helps to spread and absorb magnet
stored energy
– More tests to be done
•
Magnetic measurements were performed with tangential probe and PCB based
probes
– Need to investigate source of noise in signals from the tangential probe
– Need to understand source of 180 degree phase shift in data from the
tangential and PCB probes
•
Measurements at two different facilities (FNAL and CERN) would be very useful
•
First demonstrator test experience will be used in test preparation and test of
next 11T magnets
Guram Chlachidze, FNAL-CERN Collaboration
31
Backup Slides
Guram Chlachidze, FNAL-CERN Collaboration
Name
32
SG data: cool-down in coil 2
Coil 2
Guram Chlachidze, FNAL-CERN Collaboration
33
SG data: cool-down in coil 3
Coil 3
Guram Chlachidze, FNAL-CERN Collaboration
34
Excitation SG data: Pole gauges
Guram Chlachidze, FNAL-CERN Collaboration
35
Excitation SG data: Coil gauges
Guram Chlachidze, FNAL-CERN Collaboration
36
Excitation SG data: Bullets
Guram Chlachidze, FNAL-CERN Collaboration
37