Transcript cavalier_3jul06

The Quest for Gravitational Waves
Cavalier Fabien
LAL Orsay
NIKHEF
July, 3rd 2006
Virgo and the quest for Gravitational Waves
I.
The interferometric detection of gravitational waves
1.
2.
3.
Gravitational Waves: nature and effects
Sources
Principle of interferometric detection and improvements
II. The Virgo Challenge
1.
2.
3.
4.
5.
The infrastructures
The seismic noise and the super-attenuator
The thermal noise
The control system
The Virgo sensitivity
III. Experimental Results
1.
2.
3.
Control of optical cavities
Sensitivities
Data analysis
IV. Interferometric search of GW in the world
V.
The future of Gravitational Waves
Gravitational Waves
• A little bit of General Relativity
• In Special Relativity, the space-time interval is given by:
ds2 = dx2 + dy2 + dz2 – c2 dt2 = hmn dxm dxn where hmn is the Minkowski metric tensor
• In General Relativity, we have:
ds2 = gmn dxm dxn
with gmn metric tensor which follows Einstein’s equation
• Weak Field Approximation
gmn = hmn + hmn with || hmn || << 1 and
hmn can follow a propagation equation
2•h
mn
= - 16 p G Tmn
c4
where Tmn is related to the source
Gravitational Waves
• Properties
•
•
•
•
Helicity 2
Celerity c
Dimensionless amplitude h
Quadrupolar emission  Can be generated only by motions without axial symmetry
• Effect of free particles
L+dL
• h ~ dL/L
• Differential effect
L
L
L-dL
An Hertz Experiment ?
source
distance
h
P (W)
Steel Bar, 500 T,  = 2 m
L = 20 m, 5 turn/s
1m
2x10-34
10-29
H Bomb 1 megaton
Asymmetry 10%
10 km
2x10-39
10-11
Supernova 10 M asymmetry 3%
10 Mpc
10-21
1044
Coalescence of 2 black holes 1 M
10 Mpc
10-20
1050
Einstein Quadrupole Formula:
P  G5
5c
mn
 Q

Q
mn
G/5c5 ~10-53 W-1
Quadrupole Moment
G/c5 very small, c5/G will be better
© J. Weber (1974)
• GW Amplitude
P~
e2
c5 Rs2
v6
G R2 c6
e source asymmetry
Rs Schwarzschild radius of the source
R source radius
v source typical speed
 Cataclysmic Astrophysical Phenomena needed
for production of detectable GW
An Indirect Proof: PSR 1913+16
(Hulse & Taylor, Nobel’93)
PSR 1913+16 : binary pulsar (couple of 2 neutron stars)
 tests of gravitation in strong field and dynamic regime
Loss of energy due to GW emission: orbital period decreases
Gravitational Waves exist
The Sources
• Coalescence of binary systems
• Neutron Star-Neutron Star
• Neutron Star-Black Hole
• Black Hole-Black Hole
• Precise theoretical prediction of the waveform before merging phase
• Huge incertitude on annual rate
• Duration from few seconds to few minutes (for Virgo)
The Sources
• Supernovae
• Signal poorly predicted
• Rate: 1/30 year per galaxy
• Duration : few milliseconds
• Black Hole formation
• formation poorly predicted
• Good predictions for Ringdown phase
• Rate: ?
• Duration : few milliseconds
•Pulsars :
• Periodic signal
• If they have a quadrupolar moment
• Stochastic Background
• Incoherent sum of individual sources
• Cosmological Background
(like 2.7 K CMB for photons)
Historical View
1960 First detector (Weber)
1963 Idea of ITF detector (Gersenshtein&Pustovoit, Weber)
1969 First false alarm (Weber)
197X Golden Age for Weber-like detectors
1972 Feasibility of ITF detector (Weiss) and first prototype (Forward)
1974 PSR1913+16 (Hulse&Taylor)
Late 70s Bars cooled at 4 K, ITF prototypes (Glasgow, Garching, Caltech)
1980 First activities in France
1986 Birth of VIRGO collaboration (France+Italy)
1989 proposal VIRGO, proposal LIGO (USA)
1992 VIRGO FCD French Approval. LIGO approved
1993 VIRGO approved in Italy
1996 Start Construction VIRGO et LIGO
2001-2002 VIRGO CITF. LIGO : engineering runs
Fin 2005 LIGO reaches its nominal sensitivity
200X VIRGO at its nominal sensitivity
The Interferometric Detection
End Mirror M22
h 1 1
L P
Table Top experiment:
hMin  10-17 Hz-1/2
Fabry-Perot 2
Min
Virgo :
hMin  10-23 Hz-1/2
Input Mirror M21
Recycling Mirror Mrc
Fabry-Perot 1
Laser
Beam-Splitter
Mirror Mbs
Input Mirror M11
Photodiode
End Mirror M12
The “Historical” Laboratories
• LAL Orsay:
• Vacuum
• Laser Control
• Global Control
• Simulation
• LAPP Annecy:
• Detection
• Standard Electronic Components
• Tower
• Data Acquisition
• Simulation
• Nice Observatory:
• Laser
• Input Optics
• IPN Lyon : Mirror Coating
• ESPCI Paris : Mirror Metrology
• INFN Pisa:
• Super-attenuator
• Vacuum
• Infrastructure
• INFN Florence : Super-attenuator
• INFN Naples :
• Acquisition
• Environmental Monitoring
• INFN Perugia : Suspension wires
• INFN Frascati : Alignment
• Univ. Rome :
• Local Controls
• Marionette
Vacuum Chamber
• Pressure Fluctuations:
• P < 10-7 mbar (H2)
• P < 10-14 for hydrocarbons
• Tube:
• Diameter 1,2 m
• 6 km long
• V  7000 m3
• Diffused Light
• light traps
• deflectors
Vacuum Chamber
Entry West
Power Recycling
Laser Lab
Detection Lab
Beam Splitter
Entry North
Seismic Noise
• Measurement:
h seismic ( n )  10-10 n-2
Hz-1/2
• Isolation Principle:
• chain of pendulums with internal dissipation
• each pendulum behaves as a low pass filter:
H( n ) = ( n0 / n )2
for n > n0
The Super-Attenuator
Performances
• mirror motion with few microns amplitude
• mirror speed about few microns per second
The thermal Noise
• Each suspension wire and each mirror behaves as
an oscillator excited by thermal agitation
• Characterized by w0 and Q quality factor
• Q Measurements:
• silica : 106
• steel wire : 104 – 105
• Limiting factor between 3 and 500 Hz
• Mirror weight: 30 kg (noise  when M )
• Test of new materials (sapphire, silicon)
• Monolithic suspensions
The mirrors
• Reflectivity defined better than 0,01 %
• Reflectivity of end mirrors > 0.9998
• Losses (absorption, diffusion) about few ppm
• High Radius of curvature (3400 m) and defined with 3 % precision
• Surface defined with l/40 precision over 30 cm of diameter
• Coating realized by SMA at IPN Lyon
• Metrology made at ESPCI
Solution : silica mirrors (SiO2)
 = 35 cm and h = 10 or 20 cm
Position Control
• Fabry-Perot resonant: dL < 5 10-10 m
• Recycling Cavity resonant: dlR < 2.5 10-10 m
• Dark Fringe (coupling with laser power noise): dlDF < 10-10 m
• Alignment End Mirrors: 3 10-9 rad
• Alignment Entry Mirrors: 2 10-8 rad
• Alignment Recycling Mirror: 10-7 rad
• Fully Digital System running at 10 kHz for Locking and 500 Hz for Alignment
The errors signals
Pound-Drever technique
for Fabry-Perot cavity
• phase modulation of laser frequency
• side-bands anti-resonant
• use reflected beam
Generalization for Virgo
Use all signals coming out of the ITF
The Virgo Sensitivity
Virgo and the CITF
The CITF
(Central area InTerFerometer)
Central Part (no kilometric arm) used from June 2001 to July 2002.
Tests and validation :
 super attenuators
 electronic and software
 data acquisition
 output mode cleaner
 injection optics
Main Output: Learn how to control a suspended interferometer with digital systems
CITF Engineering runs : results
Alignment Noise
Frequency Noise
The Virgo Commissioning
• Started in September 2003 after the upgrade to full Virgo
• Strategy:
• North arm
• Lock acquisition
• Frequency stabilization
• Auto Alignment
• Hierarchical control (top stage, marionette, reference mass)
• West arm (same activities)
• Recombined ITF (no recycling mirror) (same activities)
• full ITF (same activities)
North Arm
• Locking at the first trial
Transmitted power
• first lock ~ 1 hour
• frequency noise and
alignment noise
Frequency noise
reduction
North Arm with Automatic Alignment
Linear alignment OFF
Linear alignment ON
Hierarchical control: 3 points
Hierarchical control: top stage
Slow corrections (f < 70 mHz)
Done during the CITF commissioning
Force applied to mirror
No feedback to top stage
3.5 mN
with feedback to top stage
Fast corrections (f > 70 mHz)
Recombined Interferometer
B8_demod
“3 steps” strategy
west arm
north arm
B5
B2
B1
B7_demod
Michelson
locked
Recombined Interferometer
North locked
West locked
DC
Error
signal
Correction
Recycled Interferometer
B8_ACp
B8_ACp
B8_ACp
B2_3f_ACp
 PR
(PRCL)
B2_3f_ACp
B2_3f_ACp 
 PR
PR (PRCL)
(PRCL)
B1p_DC

BS
(MICH)
B1p_DC

BS
(MICH)
B1p_DC  BS (MICH)
B7_ACp
 FP
Nord
B5_ACp
(CARM)
B5_ACp 
 SSFS
SSFS (CARM)
B8_ACp
 FP
Ouest
B8_ACp
B8_ACp 
 NE-WE
NE-WE (DARM)
(DARM)
• Misalignment of Recycling Mirror
• Lock ITF half distance from Dark
Fringe
• Start Second Stage of Frequency
Stabilization
• Alignment of Recycling Mirror
• Decrease Dark Fringe offset
• Switch to error signal
10
10 μrad
μrad
B7_ACp
LASER
LASER
Offset sur B1p_DC
Offset sur B1p_DC
1
1
0.8
1
0.8
0.6
0.6
0.8
B2_3f_ACp B1p_DC
B2_3f_ACp B1p_DC
B5_ACp
0.4
0.6
0.2
0.4
0
0.2 0
0
10
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Recycled Interferometer
Power in recycling cavities
Power on Side-Bands
Automatic procedure lasting few minutes
Power in the arms
The Various Sensitivities
Noise Sources
C7 Limitations
• Several problems forced us to run with reduced input laser power
• Mechanical problems with Recycling mirror
 new injection bench for the laser source:
• Faraday isolator to avoid problems with the light reflected by the ITF towards the laser
• Better mechanical properties
• Better optics
 New Recycling mirror :
• better mechanical properties
• increased reflectivity to gain on power recycling

After the modifications
• Restart in December 2005
• First stable lock since June (10 hours)
• Impinging power on Beam-Splitter increased by a factor 10 (280W)
• 10 alignment loops closed (7 with low bandwidths)
• Noise hunting restarted
• Science Run at the end of the year
Data Analysis: Coalescing Binaries
Horizon with detection threshold at SNR=8 supposing an optimal orientation
Data Analysis: Bursts
Main activity :
Definition of vetoes on auxiliary channels
 Feedback to commissioning
The other ITF detectors for GW
LIGO
GEO
VIRGO
TAMA
AIGO
3 kilometric antennas :
• VIRGO (3 km)
• LIGO (2 antennas, 4 km)
 Coincidences and position reconstruction
GW Astronomy needs at least 3 detectors
Why a network of detectors ?
• Mandatory for Stochastic
• Reduce false alarm rate (bursts and coalescences)
• Increase detection probability (bursts and coalescences)
Efficiency
HL
HV
LV
HL  HVLV
HLV
41 %
22 %
22 %
60 %
19 %
• Reconstruct the source position (precision about one degree)
• Reconstruct the gravitational waveform
Antenna Pattern
(Sensitivity as a function of the source position)
Hanford
Livingston
Virgo
LIGO Sensitivities
LIGO Detection Range and Duty Cycle (February 2006)
NS-NS Detection Range
Duty Cycle
L1
~12 Mpc
L1
55.1%
H1
~14.5 Mpc
H1
63.9%
H2
72.5%
Any two
66.7%
Triple
38.4%
H2
~7 Mpc
The Future of Gravitational Waves
Increase the sensitivity by a factor 10
 Gain a factor 10 on detection distance
 Gain a factor 1000 on the volume of possible sources
 Start GW astronomy
The « second generation » detectors are mandatory (~2010):
• studies started in Virgo … but not yet a complete design for
« advanced Virgo »
• « white paper » in preparation
• Advanced LIGO on the track
Virgo + (~2008)
• Monolithic Suspensions (fused silica)
• More powerful laser (50 W)
• Thermal Compensation
• Upgrade of the control system
-19
10
(a) Virgo +
(b) Virgo + (old mirror th. noise model)
(c) Nominal Virgo
(d) Pendulum Thermal Noise
(e) Mirror Thermal Noise
(f) Optical Readout Noise
(d)
-20
h(f) [1/sqrt(Hz)]
10
-21
10
(c)
(f)
-22
10
(b)
(e)
(a)
-23
10
1
10
100
Frequency [Hz]
1000
10000
The Sensitivity for 2nd generation
-20
10
AdvVirgo
Virgo
Virgo semi-advanced
Virgo semi-advanced "new mirror model"
Advanced LIGO
Advanced Virgo without thermo-refractive
-21
h(f) [1/sqrt(Hz)]
10
-22
10
-23
10
-24
10
10
100
1000
Frequency [Hz]
10000
Conclusions
• Significant improvements in 2005
• After several difficult months in 2006, we reach C7 level with full power
• A factor 20 to gain at high frequency to reach nominal sensitivity
• Data Analysis really started, mainly focused on detector behavior
• Science Run at the end of the year
• LIGO at its nominal sensitivity and will run to get one integrated year of data
• Joint analysis in preparation :
• working group LIGO-Virgo
• MoU soon signed for data exchange
• Virgo + foreseen for 2008
• 2nd generation under definition and foreseen after 2010
• R&D 3rd generation starting
GW: a never ending story
The future of gravitational astronomy looks bright.
1972
That the quest ultimately will succeed seems almost assured.
The only question is when, and with how much further effort.
1983
Kip S. Thorne
[I]nterferometers should detect the first
waves in 2001 or several years thereafter (…)
1995
Km-scale laser interferometers are now coming on-line, and it
seems very likely that they will detect mergers of compact
binaries within the next 7 years, and possibly much sooner.
2002