Transcript 161109_MQXFS3_Protectionx

MQXFS3 Test Results:
Quench Protection
Susana Izquierdo Bermudez, Hugo Bajas, et. al.
logo
area
Protection studies goals and constrains
 Follow the test plan objectives and
guidelines defined for MQXFS1b.
 A maximum QI of 28 MIITs was set
as the limit for in MIITs for the first
thermal cycle.
 Due to the limitation in magnet
performance, protection tests were
performed only up to a level of
current of 0.8*Inom (13.18 kA).
 It was not possible to perform CLIQ
studies (problems on electrical
integrity of one CLIQ lead).
 The primary goals of the tests were:
 Quench integral studies
 Protection heater delay
logo
area
Current [kA]
1.65
3.30
4.94
8.24
13.18
16.48
17.76
I/Inom
0.1
0.2
0.3
0.5
0.8
1.0
1.08
2
MQXFS3 heater design
 Each of the four coils is
equipped with 6 Protection
Heaters per coil on their inner
and outer surfaces
 The heater design is the
“baseline” design for the
prototype magnets.
logo
area
3
MQXFS3 Heater Circuits and Wiring

SM18 constrains:
 The maximum number of wires that can
be routed outside the Cryostat are 32 
HFi-HF*i & LFi-LF*i (i=1,4) are connected
in series inside the cryostat.
 8 quench heater power supplies
available

*
*
*
*
*
*
*
*
*
*
*
*
Remark: The baseline for MQXFA considers 12
quench heater circuits per magnet, 24 heater strips
per magnet, 48 heater wires per magnet
logo
area

HF and LF strips are connected in series to
reduce the number of outer layer heater circuits
from 8 to 4
The inner layer strips in the same coil are
connected in series.
Circuit #
Strips
1
LF1-LF1*-HF1-HF1*
2
LF2-LF2*-HF2-HF2*
3
LF3-LF3*-HF3-HF3*
4
LF4-LF4*-HF4-HF4*
5
IN1-IN2
6
IN3-IN4
7
IN1*-IN2*
8
IN3*-IN4*
4
Heater Powering
 Resistor added in series with the heaters to have a
peak current and time constant on the heaters as
close as possible to the full size magnet.
 Constrain: The maximum heater current in SM18 is
limited to 150 A
 To be upgraded for future tests
Voltage
Powering parameters Capacitance
inner layer heater Peak current
circuits
RC
Peak power density
Energy density in the heater stations
MQXFS3 MQXFA
(V)
900
900
(mF)
7.05
7.05
(A)
150
134
(ms)
42
47
2
(W/cm )
122
98
(J/cm2)
2.59
2.32
Voltage
Powering parameters Capacitance
outer layer heater Peak current
circuits
RC
Peak power density
Energy density in the heater stations
MQXFS3 MQXFA
900
900
(V)
7.05
7.05
(mF)
150
198
(A)
42
32
(ms)
2
123
213
(W/cm )
(J/cm2)
2.60
3.42
logo
area
5
Heater powering
 According to the model, the selected powering scheme is
conservative when comparing to the actual situation for the full
length magnet.
logo
area
6
Definition of protection parameters
 During training:
 Rdump = 30 mΩ
 All inner layer and outer layer heaters fired
 After the second training quench, the inner layer heater in coil 105 (right
hand) failed;
 After the fourth training quench, the inner layer heater in coil 106 (left hand)
failed;
 Quench integral studies
 Dump delayed by 990 ms.
 Two cases studies:
 Inner layer and outer layer heaters fired
 Remark: it corresponds to a “failure case” as two inner layer heater strips were
not working.
 Only outer layer heaters fired
 Heater delay studies
 Only specific heater circuits are fired.
 Once the quench is detected by the detection system, the rest of
the heaters and the dump are fired.
logo
area
7
Quench integral studies
 No dump, no real quench
(quench induced due to the firing
of the heaters)
 Two cases are studied:
 All heaters are fired
 Only outer layer heaters are fired
 Remark: This plot represents the
quench integral after start of the
quench (excluding the time
needed for the heaters to
quench the magnet)
QI from quench start (MA2s)
30.00
25.00
20.00
15.00
10.00
5.00
0.00
0
5000
10000
15000
Imagnet (A)
20000
IL + OL
only OL
Adiabatic, OL+IL quenched at t=0
Adiabatic, only OL quenched at t=0
logo
area
8
Quench integral studies
25.00
QI (MA2s)
20.00
15.00
10.00
5.00
0.00
0
Exclude heater delay
Include heater delay
logo
area
2000
4000
6000
I (A)
IL + OL, from quench start
IL+OL, from QH on
8000
10000
12000
14000
only OL, from quench start
only OL, from QH on
9
Outer layer quench heater delays
 At high field, heater
delays are slightly
faster than predicted
by the model.
 At low field, large
variation on the
measurements (large
incertitude at the level
of the measurements
for low magnet current)
logo
area
Quench heaters powering:
Pd = 123 W/cm2
RC = 42 ms
10
Inner layer quench heater delays
MQXFS3 - Inner layer
160
140
QH delay (ms)
 As in the outer layer
heaters case, the
incertitude in the
measurements
increases for low
magnet current level.
 Delays a bit longer
than expected, but not
so far.
120
100
80
60
40
20
0
0
Quench heaters powering:
Pd = 123 W/cm2
RC = 42 ms
logo
area
5000
Computed
IL, coil 106
10000
I (A)
Il, coil 106
IL, coil 07
15000
IL, coil 106
11
MQXFS1 vs. MQXFS3 (IL)

Even if the heater powering
conditions are rather different, inner
heater delays in MQXFS1 and
MQXFS3 are similar.

MQXFS1:



MQXFS3:


logo
area
Pd = 202 W/cm2
RC = 18 ms
Pd = 123 W/cm2
RC = 42 ms
12
MQXFS1 vs. MQXFS3 (OL-HF)


At low field, heater delays are very
different, but the incertitude in the
measurements is also large.
At high field, heater delays are
similar.
logo
area

MQXFS1:
 I > 8 kA: Pd = 148 W/cm2 , RC = 21 ms
 I < 8 kA: Pd = 172 W/cm2 , RC = 21 ms

MQXFS3
 Pd = 123 W/cm2 , RC = 42 ms
13
Summary
 Quench heater powering scheme was adapted to have
conditions as close as possible to the full length magnet
for the given constrains in SM18.
 Outer layer heaters perform as expected.
 Two inner layer heater strips were lost. Investigations
needed to understand the source of the problems.
 Quench integral studies indicate that the heaters
effectively quench the magnet.
logo
area
14
To be done in MQXFS3b
 Protection heater delay studies:
 Outer layer, low field region
 Delays at nominal and ultimate current (if possible)
 Quench integral studies
 OL + CLIQ
 Quench integral at higher currents (without the
constrain on MIITs<28 MA2s)
 Heater performance study:
 Minimum heater power density to quench
 Heater delay study as a function of the heater power
density (at a magnet current as high as possible)
logo
area
15
Additional slides
logo
area
Quench integral studies
 For Imagnet < 0.3 Inom, there is still current in the magnet
at t = 990 ms so the cannot be directly used for the QI
studies as there is an effect from the dump.
I/Inom Imagnet
%
0.1
0.2
0.3
0.4
0.5
0.7
0.8
A
1647
3294
4941
6588
8235
11529
13176
QI from
QI from QI from
QI from
quench
QH trig QH on quench start
detected
MA2s
MA2s
MA2s
MA2s
3.14
9.16
14.72
18.15
20.48
23.80
25.04
logo
area
3.13
9.11
14.64
17.98
20.21
23.27
24.35
2.89
8.45
13.08
15.93
17.77
20.61
21.65
1.36
7.04
12.11
15.24
17.23
20.21
21.65
QI from
I/Inom Imagnet QH trig
%
A
MA2s
16.47
0.3
4941
20.59
0.4
6588
23.14
0.5
8235
QI from QI from
QH on quench start
MA2s
MA2s
16.37
14.79
20.46
18.42
22.87
21.07
QI from
quench
detected
MA2s
13.81
17.55
19.82
17
Nominal Quench Heater Circuits and Wiring
[Emmanuele Ravaioli]
logo
area
 Each QH power supply
is connected to two
strips in series.
 Heater wiring scheme
has been defined to
optimize the quench
voltage distribution for
CLIQ + heaters com
 12 heater circuits per
magnet.
 24 heater strips per
magnet.
 48 heater wires per
magnet
HiLumi Collaboration Meeting, Paris 2016
18
LF3
HF3
LF2
HF2
HF4
HF1
IN3
LF4
IN4
IN1
LF1
LF1
IN1
IN4
LF4
IN2
HF1
IN3
HF4
HF2
logo
area
IN2
LF2
LF3
HF3
19