Noise issues in vibration sensing and isolation for Advanced
Virgo Eric Hennes
GWADW, Hawaii 2012, May 15
Advanced Virgo
Nikhef 2011-2014 “vibration” activities on AdV
144 m
3 km
EIB
angular alignment port
Laser
detection and alignment ports
3 km
z
dihedron
IMC
2GWADW May 15, 2012
• External Injection Bench Seismic Attenuation (SAS) • Isolation of 5 new suspended benches for linear alignment
• High sensitivity diagnostic accelerometer• Input Mode Cleaner:
- new End mirror + suspension
- new Dihedron + support
External Injection Bench (EIB) Seismic Attenuation (SAS)
Motivation (since 2009): • beam jitter noise is spoiling sensitivity • Ascribed to mechanical modes of
- bench frame (15-20, 30-60 Hz) - optical mounts (200-300 Hz)
• excited by both seism and “a bit of” sound• to be minimized for Virgo+ and AdV
Projection of vertical beam jitter on hrec Projection of EIB displacement on hrec
hrec (VSR2)
EIB displ projection on hrec
Coherent part
Virgo+
AdV
hrec (VSR2)
Jitter HOR projection on hrec
Coherent part
Virgo+
Jitter HOR projection on hrec
3
Old bench support
GWADW May 15, 2012
Fiori et al.Fiori et al.
EIB-SAS Requirements
4
• Up to 40 dB attenuation > 10 Hz• Long term stability (RMS 20 m, 15 rad, max 1 oC variations• DC-control
GWADW May 15, 2012
Re
sid
ual
dip
lace
men
t [m
/√H
z]
EIB-SAS Solution option history
Chosen solution:Highly-compliant suspension
5GWADW May 15, 2012
Stiffer legs
Stiffer support
Active isolation (TMC-Stacis)
GWADW May 15, 2012 6
6-DOF passive attenuation, anti-spring technology• 3 GAS filters, 3 Inverted Pendulum legs
Active positioning and low-f resonance damping• 6 x LVDTs, 6 x geophones, 6 x voice coil actuator
• 800 kHz 18 bit ADCs
• 7 stepper motors
Optical bench
Built and tested at Nikhef, 2011
EIB-SAS: adopted from HAMSAS project (Caltech) & AEI-Hannover (2010)
EIB-SAS complication: horizontal GAS modes < 20 Hz
GWADW May 15, 2012
7/20
Yaw + z (12 Hz)
• Modes due to lateral GAS spring compliance• Can be damped by extending control bandwidth
x + (bench) roll (18 Hz)
z + yaw +(bench) pitch (16 Hz)
7
Damped
Ty
[rad
/√H
z]
7
xy
z
EIB-SAS complication: spring box modes 30–60 Hz
GWADW May 15, 2012 8
Springbox vertical bouncing mode ~ 50Hz
• Modes due to upper IP flex and spring box internal compliances• Can partially be damped with eddy current damped resonators
50 Hz
coupled tilt modes (~ 30-40 Hz)
10 cm
EIB-SAS: First major Advanced Virgo installation (Dec 2011)
GWADW May 15, 2012 9
EIB in acoustic enclosure
EIB-SAS after installation
EIB-SAS commisioning: beam jitter
GWADW May 15, 2012 10
The good news:Beam jitter significantly reduced < 100 Hz
But:Optics mounts keep resonating (200-300 Hz)
Fiori et al.
And …
EIB-SAS commisioning: bench motion and acoustic noise
GWADW May 15, 2012 11
Conclusions after many tests
•1 – 100 Hz: bench moved by acoustic noise from central building Airco
•40 - 50 Hz: second noise source is present
•Limited shielding by acoustic enclosure
•Optical mounts are moved by sound
Bench displacement
Sound level in acoustic enclosure
Fiori et al.
EIB-SAS: conclusions
GWADW May 15, 2012 12
• Commissioning EIB-SAS revealed prominent roll of acoustic noise
• Laser lab walls are cleanroom walls → do not shield acoustic noise
• EIB-SAS is doing only most of the job, but ….
• It is usefeul for Advanced Virgo ..
- to replace these walls by concrete walls
- to move Airco’s outside main building
MultiSAS: Isolating suspended benches
for linear alignment
GWADW May 15, 2012 13
5 similar systems consisting of: • MultiSAS (Nikhef)• Optical bench (DET/LAPP)• UHV vacuum mini-tower (LAPP)
3 km
3 km
z
MultiSAS: multi-stage 6DOF bench isolation & control
GWADW May 15, 2012 14
Schedule•Design approved•Prototype assembled•In-air testing started with dummy bench•Integration test at Nikhef/LAPP from Sep 2012•AdV production and first chamber installation 2013
Characteristics two-stage vertical isolation (GAS) 3-stage horizontal isolation (IP, P, P) 320 kg single-wire suspended bench sensing/control technology like EIB-SAS
Requirements @ 10 Hz (VIR-0128A-12)
h 2 E-12 m/√Hz
3 E-15 rad/√Hz
hRMS 2 E-5mm
RMS 3E-9 rad
tough!
MultiSAS: special GAS mechanics
GWADW May 15, 2012 15
Bottom GAS filter GAS blades below Filter body
Top GAS filter Down-tunable to
100 mHz thanks to combination of
over-compression and hysteresis
SiC magic wandsTransfer function down to ~ 10-
5 at 30 Hz thanks to excellent percussion effect compensation
TF MultiSAS intermediate filter with Magic Wands
1.E-05
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
1.E+01
1.E+02
0.1 1 10 100f (Hz)
yL/y
0
Transfer bottom filter with Magic wand
MultiSAS: IP equal bottom & top flexures (to be tested)
GWADW May 15, 2012 16
• Both flexures in compression• Allows for significant raise oftop plate bouncingand rigid leg modes
thin + thick flex equal flexes
material alu steel alu steelwall thickness 1 2 1 2thickness 7.4 6.7length 40top plate y 64 80 83 133rigid leg x/z 145 100 270 295
MultiSAS IP modes (FEM)
mmlower flex
8.760
modes Hz
thin (2mm) thick
leg
upper flex
MultiSAS Modal analysis: modes < 250 mHz
GWADW May 15, 2012 17
Top stage yaw (230 mHz)
Common yaw(10 mHz)
Common lateral(80 mHz)
…. and 3 more
MultiSAS: Modal analysis: modes < 2 Hz
GWADW May 15, 2012 18
Pendulum mode (1.7 Hz)
GAS mode (0.5 Hz)
Filter tilt (0.43 Hz)- coupling to horizontal
…. and 4 more
MultiSAS: Modal analysis > 60 Hz
GWADW May 15, 2012 19
Keystone rotation (250 Hz)
Keystone coupled mode (70 Hz)
…. Etc etc
Blade mode (340 Hz)
Multi-SAS horizontal residual displacement @ Virgo (FEM)
1.E-24
1.E-22
1.E-20
1.E-18
1.E-16
1.E-14
1.E-12
1.E-10
1.E-08
1.E-06
1.E-04
1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03f (hz)
from horizontal ground motionfrom vertical ground motion (if 0.1% coupling)total
MultiSAS: residual horizontal displacement
GWADW May 15, 2012 20
keysto
ne & to
p stage
reso
nances
< 2 10-12 m/√Hz required
Margin 105 !
10-4
10-6
10-8
10-10
10-12
10-14
10-16
10-18
10-20
10-22
10-24
Res
idua
l dip
lace
men
t [m
/√H
z]
Frequency [Hz]
Calculated from FEM transfer functions (open loop)
Multi-SAS bench residual tilt @ Virgo (FEM)
1.E-24
1.E-22
1.E-20
1.E-18
1.E-16
1.E-14
1.E-12
1.E-10
1.E-08
1.E-06
1.E-04
0.01 0.1 1 10 100 1000
MultiSAS: residual tilt (FEM)
GWADW May 15, 2012 21
< 3 10-15 rad/√Hz required
Margin 3*103
!
10-4
10-6
10-8
10-10
10-12
10-14
10-16
10-18
10-20
10-22
10-24
Res
idua
l tilt
ang
le [r
ad/√
Hz]
Frequency [Hz]
MultiSAS: current state of prototype
GWADW May 15, 2012 22
• Cabling, Electronics and DAQ to be installed• Mechanics to be fine-tuned• Performance to be measured• Dummy bench in production• Looking forward to MiniTower for tests in vacuo• Development of diagnostic interferometric
accelerometer
GWADW May 15, 2012 23
High-sensitivity diagnostic accelerometer
Mechanics: monolithic single-axis Watt-linkage oscillator (Bertolini, 2005*)Projection of thermal noise xth reduced by choosing– low resonance frequency (~ 0.5 Hz)- large quality factor (> 100)- large mass (~ 1 kg)
Readout: based on balanced Michelson interferometer (Gray et al, 1998**)- LED source allows for low Relative Intensity noise (RIN)- Sensing of both reflected and transmitted beam >40 dB RIN
cancelling- fringe locking by feed back loop to piezo actuator or voice coil
On behalf of David Rabeling
**Optics & Quantum Electronics 31, 1999, 571-582*Nuclear Instruments and Methods in Physics Research A 556, 2006, 616–623
x
z
LED
1
2
PT
PR
Inverted pendulum
pendulumM P
M IPL
to PZT or Voice coil
GWADW May 15, 2012 24
Accelerometer mechanics
Transfer function:
0:resonance angular frequencyQ : Quality factor >>1A : percussion factor <<1K : total elastic stiffnessM : total mass (M P+MIP) kB : Boltzmann’s constant : loss angle in flexes
Projected thermal noise PSD:
Construction• Wire-cut monolithic aluminum• Symmetric pendulum + inverted
pendulum• All flexes (4x5x0.1 mm) in tension• Adjustable mass distribution (tuning)• Optional: voice coil actuator
k
MMgLQ IPP
)(1
1
For structural damping:
L=7 cm
Voice coil actuator
frame displacement
Mass-to-frame displacement
25
Interferometric read-outPhotodiode incident power
z : mass-to-frame displacementPinc: incident LED power: carrier wavelengthC : complex temporal coherence
LED
1
2
PT
PRPinc
zczC
PP
zczC
PP
inc
R
inc
T
4cos))/(Re(1
8
4cos))/(Re(1
4
z
Coherence length cRelated to source bandwidth
GWADW May 15, 2012 5 -4 -3 -2 -1 0 1 2 3 4 5
Relative PD intensity pattern (T and R)
c/2
zm
Readout current sensitivityLock on half fringe:
DCTincOut
IP
dz
dI,
4
To PZT or Voice coil
IOut
GWADW May 15, 2012 26GWADW May 15, 2012 26
Optical and electronic readout current noiseShot noise (SN) current
)(22
RTSNPPqi
Intensity noise
ADC noise 2
ADCi
Electronic noise
2
OpAmpi
2
,termalRi
< -40 dB of shot noise (?)Dark current noise
BW
Ii dark
dark
2
2
4log/)log(o
os
f
fm
n : number of bitsfs : sampling frequencyfos: over-sampling frequencyrf : transimpedance resistanceVop, Iop: OpAmp noise levelscv, cv,cv,: corner frequenciesVpp,: ADC peak-peak voltage
26
1
fsrf
GWADW May 15, 2012 27
Projections of noise on frame displacement
3 fm /rt(Hz)
MultiSAS Requirement @ 10 Hz
measurable bench motion
Femtometer noise
For accelerometer with f0 = 0.5Hz, Q =100
Bench residual
Input mode cleaner (IMC) modifications: Dihedron support
28
• modes 55, 78, 82 Hz disappeared
• new support: 240 Hz (Hertz contact)
SSFS corr signal
Old leg support
New stiff support
replaced Oct 2011
GWADW May 15, 2012
Genin et al.
144 m
EIB
ITF
IMC modifications; 2. New Dihedron
GWADW May 15, 2012 29
• Shaped and polished by Optronica (Den Helder, NL)
• Mirrors connected by Van der Waals bonds (no adhesive)
• Angle (89o 58’ 58’’) machined to 0.4’’ accuracy (max 2” allowed)
• To be coated (by IPN-Lyon)
• To be installed in 2013
IMC modifications: New End Mirror suspension
30
• mirror twice as thick• including front (and back?) baffle • adapting Marionette and gear box• new installation procedure /tools• largest challenge: connect baffles stiffly to reference mass
GWADW May 15, 2012
144 m
EIB ITF
GWADW May 15, 2012 31
Summary
• EIB meets requirements, provided acoustic sources be removed or better shielded.
• MultiSAS prototype is working so far; performance to be fully tested
• New IMC dihedron mirror waiting for coating, than ready to be installed on its new support
IMC end mirror suspension: a challenge regarding stiff baffle mounting
• Diagnostic accelerometer R&D looks useful tool for bench seismic attenuation performance testing
The tracking interferometer can measure RMS test bench motions up to 2.8 m. Larger excursions can be compensated using the voice coil actuator.
Projections of noise and bench residual on sensor displacement z
System design parameters
Component list