FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe,...
-
Upload
humberto-fayer -
Category
Documents
-
view
213 -
download
1
Transcript of FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe,...
![Page 1: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/1.jpg)
FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION
K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens*
The research leading to these results has received funding from the European Commission under the FP7 Research Infrastructures project EuCARD
![Page 2: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/2.jpg)
2
Outline
S. Janssens, P. Fernandez, A&T Sector Seminar, Geneva, 24 November 2011
How to perform active control on the preisolator to damp the 1 Hz resonance
Issues with using the preisolator as it is Further research proposal Question about accelerometer feedback
![Page 3: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/3.jpg)
3
Preisolator
m=50ka=480e6ξ=0.01 Mp =40000fp =1 Hz ξp=0.01
What is needed to damp/influence peak of pre-isolator mode?
Perfect feedback (no filtering and sensors)
![Page 4: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/4.jpg)
4
Active Isolation Strategies
Add virtual mass
Feedback control principle
![Page 5: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/5.jpg)
5
Active Isolation Strategies
Sky-hook damper
(D.C. Karnopp, 1969)
Feedback control principle
![Page 6: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/6.jpg)
6
Active Isolation Strategies
Feedback control principle
S. Janssens, P. Fernandez, A&T Sector Seminar, Geneva, 24 November 2011
Virtual stiffness added
![Page 7: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/7.jpg)
7
Active Isolation Strategies
Closed Twx
10-10
10-5
100
Mag
nitu
de (
abs)
10-1
100
101
102
103
104
-360
-270
-180
-90
0
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
Control on
Control off
10-2
100
102
104
Mag
nitu
de (
abs)
10-1
100
101
102
103
104
-180
-135
-90
-45
0
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
Preisolator mode
Top mass mode
Reduces transmissibilityDoesn’t damp first modeAlways stableAuthority at 1 Hz needed with factor 100
Qd0 stabilization system as it isPosition feedback
![Page 8: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/8.jpg)
8
Active Isolation Strategies
10-10
10-5
100
Mag
nitu
de (
abs)
10-1
100
101
102
103
104
-360
-270
-180
-90
0
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
control off
control on
Preisolator mode
Top mass mode
Reduction of transmissibilityNo damping of the modeTop mass mode doesn’t move (as no added stiffness due to ff)Authority at 1 Hz needed with factor 100
QD0 stabilization as it isFeedforward of
![Page 9: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/9.jpg)
9
Active Isolation Strategies
10-10
10-5
100
105
Mag
nitu
de (
abs)
10-1
100
101
102
103
-360
-270
-180
-90
0
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
control off
control on
10-4
10-2
100
102
Mag
nitu
de (
abs)
10-1
100
101
102
103
104
-90
-45
0
45
90P
hase
(de
g)
Bode Diagram
Frequency (Hz)
Preisolator mode
Top mass mode
Closed TwxOnly effect on second modeAlways stable(similar results with IFF+noise issue)
Damping with velocity feedback
![Page 10: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/10.jpg)
10
Active Isolation Strategies
10-20
10-10
100
1010
Mag
nitu
de (
abs)
10-2
10-1
100
101
102
103
104
105
106
-360
-270
-180
-90
0
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
Control off
Control on
Preisolator mode
Top mass mode
Closed TwxDoesn’t work as it just injects noise due to phase difference
Velocity feedforward
![Page 11: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/11.jpg)
11
Active Isolation Strategies
10-10
10-5
100
105
From: In(3)
To:
Out
(2)
10-1
100
101
102
103
104
105
106
-90
-45
0
45
90T
o: O
ut(2
)
Bode Diagram
Frequency (Hz)
Mag
nitu
de (
abs)
; P
hase
(de
g)
Proposed velocity feedback of
Damps first mode critically
![Page 12: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/12.jpg)
12
Active Isolation Strategies
Damps the peak and reduces transmissibility
Combination of and feedback
![Page 13: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/13.jpg)
13
Issues for the preisolator
3D coherence of ground motion
C. Collette, ILC-CLIC LET Beam Dynamics Workshop (23-25 June 2009)
Measurements performed in LHC tunnel by K. Artoos and M. Guinchard.
Ground motion > 12 m @ 1 Hz is not coherent
=> Active damping and transmissibility reduction necessary at 1 Hz
![Page 14: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/14.jpg)
14
Issues for the preisolator
+pressure on vertical plane on surface of magnets which has a moment arm!! +other noise sources coming from the detector=> Increasing with position feedback necessary!
![Page 15: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/15.jpg)
15
Issues for the preisolator
15
J. Pfingstner
Preliminary max roll simulations vs luminosity loss=> Max 1 μrad!
![Page 16: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/16.jpg)
16
Further simulations with PID
More complex model has been made
With PID feedback for =>improved position, damping and compliance
With proposed CERN MBQ stabilization (as example)
![Page 17: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/17.jpg)
Further simulations with PID
Possible transfer function between ground and QD0 + no drift at low freqPossibility to change Preisolator position!!!No noises or filters in it!!!!
![Page 18: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/18.jpg)
Further Research proposal (tender/inhouse)
More complex model=>Investigate modes with changing magnet positions
![Page 19: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/19.jpg)
Further Research proposal (tender/inhouse)
Research goals1. Make a dynamic model of the system presented. For this model:
• give the transfer functions between all degrees of freedom and the ground• make a table of all the modal frequencies and their decomposition in eigenvectors• graphical representation of the evolution of the modal frequencies and their decomposition
for a changing γ from 0 up to 15 degrees which changes the position of mg,mQF1, mst, mstb, mQD0.• What is the effect of increasing the 1st mode to 20 Hz?
2. Propose the best active damping (velocity feedback, Integrated Force Feedback,…) system which:• damps the 1st mode of Mpi critically• Reduces the 1st mode from 20 Hz to 1 Hz through active control• uses existing technologies compliant with the environmental parameters,• does decrease the drop off above 2ωpi in the transfer function between wpi and xpi,
for the ground vibrations specified, due to noise or any other limitations (actuator or sensors).• Specify the number actuators/sensors (The 4 specified are a suggestion).• Is it better to use a global controller or have each leg have its own SISO controller
and decouple them with joints?• Simulate the performance of the proposed isolation system in an environment with
ground vibrations and applying actual sensor/actuator, sensitivity, noise and resolution.
~2 months
~5 months
![Page 20: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/20.jpg)
Further issues
The alignment stage (would fall off now) Effect of the 2nd and 3rd mode on
luminosity Effect of phase difference between two
preisolator blocks Further issues?
![Page 21: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/21.jpg)
Question accelerometer feedback
101
102
10-3
10-2
10-1
100
101
102
Frequency [Hz]
T_
{en
de
vco
/ge
op
ho
ne
int acc vs T6int acc vs L4C
Measured sensitivity endevco M86
Sensitivity curve of accelerometer has resonance at ~200-300 Hz
->issue with stability when resonance of system and accelerometer meet
![Page 22: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/22.jpg)
Question accelerometer feedback
10-4
10-2
100
102
104
Mag
nitu
de (
abs)
101
102
103
-360
-180
0
180
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
Openloop acc pos fbOpenloop acc vel fbOpenloop acc acc fb
system modemode accOpenloop:Stable if for Mag>1Phase -180<phi<180
System mode high:Acc fb okVel fb unstable Could be made stablePos fb unstable Could be made stable
Openloop cut
![Page 23: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/23.jpg)
Question accelerometer feedback
10-4
10-2
100
102
104
Mag
nitu
de (
abs)
101
102
103
-360
-180
0
180
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
Openloop acc pos fbOpenloop acc vel fbOpenloop acc acc fb
accmodesystem mode
Pole cancellation -> very good knowledge of system necessary, difficult for complex system -> acc mode needs to be far away from system modes
System mode low:Acc fb okVel fb unstablePos fb unstable Only pole cancellation
![Page 24: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/24.jpg)
Conclusion
Active control of the preisolator block is needed:-to perform damping of the resonance-to provide positioning/alignment capabilities/synchronize with other preisolator-to improve compliance of the system
Proposed solution for a PID on the preisolator
Further study required with full model=> Tender/inhouse study?
Stef’s learning moment about accelerometer feedback
![Page 25: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/25.jpg)
Spares
![Page 26: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/26.jpg)
Pre-isolator feedbackacc fb
10-10
10-5
100
105
From: In(2)
To:
Out
(1)
10-1
100
101
102
103
-360
-270
-180
-90
0
To:
Out
(1)
Bode Diagram
Frequency (Hz)
Mag
nitu
de (
abs)
; P
hase
(de
g)
Control off
Control on (acc fb)
Closed Twx
10-10
10-5
100
105
Mag
nitu
de (
abs)
10-1
100
101
102
103
104
0
45
90
135
180
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
Preisolator mode
Top mass mode
Only effect on second modeAlways stable
![Page 27: FINAL FOCUS: COMBINATION OF PRE ISOLATOR AND ACTIVE STABILISATION K. Artoos, C. Collette, R. Leuxe, C.Eymin, P. Fernandez, S. Janssens * The research leading.](https://reader036.fdocuments.in/reader036/viewer/2022070306/551738de55034603568b6073/html5/thumbnails/27.jpg)
10-20
10-10
100
1010
Mag
nitu
de (
abs)
10-2
10-1
100
101
102
103
104
105
106
-360
-270
-180
-90
0
Pha
se (
deg)
Bode Diagram
Frequency (Hz)
controll off
control on
Pre-isolator feedbackacc ff
Closed Twx
Preisolator mode
Top mass mode
Doesn’t work!