Transcript transparencies - Rencontres du Vietnam

Status and perspectives of the
gravitational wave detector Virgo
Raffaele Flaminio
European Gravitational Observatory and CNRS/LAPP
Summary
I. A bit of gravitational wave physics
II. The Virgo design
III. Virgo status and plans
IV. Virgo+, Advanced Virgo and beyond
Conclusions
Hanoi, August 8th, 2006
1
I. A bit of gravitational wave physics
Hanoi, August 8th, 2006
2
Gravitational waves
• Waves of the space-time metric
 effect? distances variation
GW
L-DL
L+DL
• Transverse
 effect perpendicular to the wave direction of propagation
DL  1 hL
2
• Quadrupolar
h = GW amplitude
 opposite effect along x and y
• Tide-like
 effect larger on longer distances
• Produced by time varying quadrupole moment
2G d 2Q 1
h 4
c dt 2 d
d = source distance
Q = quadrupole moment
• Small coupling factor (“gravity is weak”)
 GW generation on earth not possible
 Astrophysical sources
Hanoi, August 8th, 2006
3
Gravitational waves are there
• Binary pulsar 1913+16 (Hulse and Taylor)
- binary system formed by two neutron stars (one pulsar)
- orbital period (~ 8 h) is decreasing
- due to energy loss via GW emission
- excellent agreement with general relativity
• Six binaries of this kind known in the galaxy
- stars will coalescence in a few 100 Myr
- frequency of emitted GW will sweep across the
detectors bandwidth
h
“Chirp” waveform
Hanoi, August 8th, 2006
4
Coalescing binaries
• Binaries formed by compact stars (NS/NS, NS/BH, BH/BH)
- Inspiral phase accurately predictable with GR
- Merger phase unknown
- Ringdown predictable
• Strong scientific potentials:
- Standard candles
distance of the source can be found out of the
waveform
- Test of GR
accurate measurements of inspiral waveform can
test gravity in the strong field regime
- Nuclear physics
before coalescence waveform sensitive to the equation of state
• NS/NS Event rate: 2/yr – 3/day in a range of 300 Mpc (for advanced detectors)
• More events expected for BH/BH
Hanoi, August 8th, 2006
5
More GW sources
• Supernovae
- star core collapse (non-spherical collapse)
- impulsive event
- waveform and amplitude difficult to predict
- rates: tens/year in the VIRGO cluster
• Rotating neutron stars
- GW emitted if non perfectly spherical star
- periodical signals, amplitudes unknown, upper limits
from pulsars slow down
- > 103 pulsars known today, ~109 neutron stars in the galaxy
• Relic stochastic background
- imprinting of the early expansion of the universe
- stochastic signals (two correlated detectors needed)
- signal too weak if standard inflation, signals larger from some string models
• The unknown
Hanoi, August 8th, 2006
6
Scientific motivations
• First direct detection of gravitational waves
• Study of the gravitational force
- GW can be generated by pure space-time (black holes)
- GW can reveal the dynamic of strongly curved space-time
• New window to observe the universe
- GW are produced by coherent relativistic motion of large masses
- GW travel through opaque matter
- GW dominate the dynamics of interesting astrophysical systems
Hanoi, August 8th, 2006
7
II. The Virgo design
Hanoi, August 8th, 2006
8
GW detection with laser
interferometers
• GW are quadrupolar (spin 2 wave)
• Michelson interferometer ideal tool
• All mirrors suspended through pendulums
= ‘free falling masses’
Suspended
mirror
DL  
Suspended
mirror
hL
2
DL  
• h = 10-22 , L = 3 km  DL  3 10-19 m
• GW detection
= measure tiny displacements
= measure tiny phase shift
Hanoi, August 8th, 2006
Beam splitter
LASER
Light
Detection
9
hL
2
Limits: sensing noise
• GW  phase shift D  4 hL

• Minimum measurable phase shift D  1 / N
photons
• To get h ~ 10-22 need L = 100 km and P = 500 W
Fabry-Perot
• Increase effective length
use Fabry-Perot in the arms
L
L
• Increase photons stored in interferometer
use light recycling
Hanoi, August 8th, 2006
P’
P
Recycling
mirror
10
Virgo optical lay-out
• Need for powerful and stable lasers
- 20 W Nd:YVO4 laser used for Virgo
- 10-8 relative power stability
- mHz frequency stability
• Need for LARGE high quality mirrors
- 35 cm diameter, 20 kg mirrors used
Hanoi, August 8th, 2006
11
Large mirror technology facility
• Development of a coating facility in Lyon (LMA)
• Development of metrology facilities to characterize
large mirrors (up to 400 mm2)
- flatness: few nm
- roughness: < 0.1 nm
- scattering losses: few ppm
- absorption losses: few ppm
Hanoi, August 8th, 2006
12
Limits: displacement noises
• Seismic noise
Displacement
m/Hz 10-4 10-5 10-6 10-7 10-8 10–9 10– 10
–
10 11
10–
12
10– 13
10– 14
10–
15
10– 16
10-1
1
• Thermal noise
10
102
Frequency (Hz)
~1010
Photons shot noise
- Need for good seismic isolation
17
- Determines lower frequency cut-off
August 8th, 2006
–
10Hanoi,
“seismic
wall”
18
10–
- Use high mechanical quality suspension
- Use high mechanical quality mirrors
13
- … or cool down
Seismic isolation
• Multi stage pendulum
- pre-isolator: inverted pendulum
- cantilever springs for vertical isolation
- cascade of six stages
- height ~ 10 m
Super-attenuator
• Inverted pendulum
• Seismic filter
Hanoi, August 8th, 2006
• Expected attenuation > 1014 @ 10 Hz
14
Mirror suspensions
• Mirrors suspended to marionette by four wires 200 mm diameter
• Marionette
- allow mirror angular position control
by means of coils attached to last seismic filter
• Recoil mass
- support coils for mirror position control
Marionette
Mirror
Hanoi, August 8th, 2006
Recoil mass
15
A large vacuum system
• Protect against acoustic noise
and air motions
- 3 km long tubes, 1.2 m diameter
- one 10 m high vacuum chamber per suspension
- 10-9 mbar
Hanoi, August 8th, 2006
16
h (Hz-1/2)
Design sensitivity
-18
10
Pulsars
hmax – 1 yr integration
-19
10
-20
10
Credit: P.Rapagnani
LIGO
Virgo
BH-BH Merger
Oscillations
@ 100 Mpc
GEO
-21
10
Core Collapse
@ 10 Mpc
QNM from BH Collisions,
100 - 10 Msun, 150 Mpc
QNM from BH Collisions,
1000 - 100 Msun, z=1
Resonant
antennas
BH-BH Inspiral, 100 Mpc
-22
NS-NS Merger
Oscillations
@ 100 Mpc
10
BH-BH Inspiral,
z = 0.4
NS, e=10-6 , 10 kpc
-23
10
NS-NS Inspiral, 300 Mpc
-24
10 Hanoi, August 8th, 2006
1
10
100
1000
17
Hz
4
10
III. Virgo construction,
status and plans
Hanoi, August 8th, 2006
18
What, where and who
• What: gravitational wave detector based on a laser interferometer with 3km long arms
• Where: located near Pisa in the Tuscany countryside (Italy)
• Who: built by a French-Italian collaboration supported by INFN in Italy and CNRS
in France. Virgo collaboration composed by 11 laboratories:
LAPP Annecy, INFN Firenze, INFN Frascati, IPN Lyon,
INFN Napoli, Observatoire de Nice, LAL Orsay, ESPCI Paris,
INFN Perugia, INFN Pisa, INFN Roma
• A new actor since 2002: European Gravitational wave Observatory (EGO)
consortium set-up by INFN and CNRS at the Virgo site to:
- support the commissioning of Virgo, its operation, maintenance and upgrades,
- create and run a computing center for the analysis of data on site,
- promote R&D useful for the detection of gravitational waves
• Two new laboratories joining the Virgo Coll. in 2006: Nikhef and INFN Tor Vergata
Hanoi, August 8th, 2006
19
The construction of Virgo
• Main steps
1989
First proposal
1992
Final conceptual design
1994
CNRS-INFN approval and agreement
1996 - 1998 Construction of central area
1999 - 2002 Installation and commissioning of the central interferometer
Construction of arms and terminal buildings
2001 - 2003 Vacuum tubes installation
2002 - 2003 Installation of large mirrors and of terminal suspensions
2003 - today Detector commissioning
Hanoi, August 8th, 2006
20
What is “Commissioning” ?
1) Set-up and tune all the interferometer control systems
- suspensions and mirror position controls
- laser stabilization controls
- interferometer alignment control
- interferometer length control
- etc. etc.
> 150 control systems
2) Reach the design sensitivity
- ITF noise identification (due to controls and environment)
- noise reduction down to “fundamentals noises”
• Each control is a potential noise sources: 1 and 2 correlated
- more controlsmore stability BUT more controlmore noise
- less noisenew noise sources visiblechange controlless noise ….
• A long process. Six years needed at LIGO.
Hanoi, August 8th, 2006
21
Commissioning main steps
• Steps of increasing complexity
• Single arms
- Started in September 2003
- Completed in spring 2004
• Recycled interferometer
- Since summer 2004
- Problems with diffused light
detected in recycled mode
- Injected laser power reduced
by 10 in 2004
Hanoi, August 8th, 2006
• Recombined interferometer
- Completed in December 2004
- All control systems tested in
this configuration
• Regular commissioning runs to:
1) Collect real data for data analysis preparation
2) Test interferometer performances on longer
time-scale (noise stationarity, duty-cycle)
22
Measured sensitivity
(7 W)
(7 W)
(7 W)
(7 W)
(0.7 W)
(0.7 W)
(0.7 W)
C7 NS/NS maximum distance ~ 1.5 Mpc
Design NS/NS maximum distance ~ 30 Mpc
Hanoi, August 8th, 2006
23
Status today
• Commissioning interrupted in
September 2005 to solve the
diffused light problem:
- new injection bench
- plus other changes ….
7 hours
• Interferometer fully controlled again
in June 2006
- 10 input power (7 W)
- 10 power on the beam splitter (290 W)
• Now: looking again into interferometer
sensitivity
Hanoi, August 8th, 2006
24
Next steps
• Complete commissioning of recycled interferometer
- set-up and tuning of the last missing control systems (mainly automatic alignment)
- then focus on noise reduction (mainly control noise and technical noise)
Goal: reach design sensitivity above few hundred Hz by the fall (10 better than C7)
• Soft transition to long data taking periods
- start with weekly science run (over the weekends) next September
- then weekends and nights ….
Goal: start Science Run by the end of the year
• Collect six months of integrated data in 2007
Hanoi, August 8th, 2006
25
Data Analysis preparation …
• Large amount of data produced (~8 MB/s 24h/2h)
- need to be ready with data analysis pipelines
• Preparation of data analysis hardware infrastructure
- 120 TB disk space on site; six months look back
- cluster of ~200 PC on site for on-line analysis
- use of computing centers for off-line analysis
(CCIN2P3-Lyon, CNAF-Bologna)
- use of GRID tools for more demanding analysis (pulsars)
• Preparation of data analysis pipelines
- calibration and h-recostruction (see S. Karkar talk)
- development of epoch vetoes
- development of search algorithms
specific to each type of source
- development of a-posteriori vetoes
using output of search algorithms
using monitoring channels (~1500 channels available)
Hanoi, August 8th, 2006
26
…. with real data
• Test of search algorithms
- hardware injection of fake events
bursts & coalescing binaries
- calibration  h-reconstruction  event searches
run on-line during C6-C7
- parameters estimation of detected signal
(SNR, time of arrival, etc.)
- comparison with injected signals
• Development of vetoes
- searches very sensitive to noise stationarity
- epoch vetoes based on ITF state
- vetoes based on monitoring channels
acoustics, seismic, electromagnetic, photodiodes signals,
control systems signals, laser monitoring, etc.
- a posteriori vetoes
c2 , search filter output etc.
- link with noise characterization (and then noise hunting)
• Search of rare events (~1-2/yr)
- need for data exchange with other detectors
Hanoi, August 8th, 2006
27
International collaborations
• Need for data exchange with the other detectors
- Common data format (“frame format”) adopted since several years
• Virgo-ROG-AURIGA
- Exchange of real data + software injection of GW events
- Ongoing joint analysis:
bursts from galactic center or stochastic background
• Virgo-LSC
- Joint data analysis with simulated data since 2 years
burst, coalescing binaries, now starting with SB
- ‘Physics gain’ in adding Virgo to the LIGO network
reconstruction of the source location
signal parameters estimation
detection efficiency enhanced by 50% (burst) or 30% (binaries)
- Now starting with real data
• MOU to be signed between the LIGO Scientific Collaboration and Virgo:
- Full exchange of data, joint analysis
- Coordinate science runs, commissioning and shutdowns
- Joint publications
Hanoi, August 8th, 2006
28
IV. Virgo+, Advanced Virgo
and beyond
Hanoi, August 8th, 2006
29
Toward GW astronomy
• Present detectors will test upper limits
• Even in the optimistic case rate
too low to start GW astronomy
• Need to improve the
sensitivity
• Increase the sensitivity
by 10  increase the
probed volume by 1000
LIGO - Virgo
• Plans to improve the
present detectors
LIGO+ - Virgo+
Credit: R.Powell
AdvLIGO - AdvVirgo
Hanoi, August 8th, 2006
30
Improving the sensitivity
-18
10
• Thermal noise: main limitation
(a) Virgo Nominal sensitivity
(b) Seismic noise
(c) Pendulum thermal noise
(d) Mirror thermal noise
(e) Shot Noise
-19
10
- decrease friction
- better wires
- better clamping
2. Mirror thermal noise
h(f) [1/sqrt(Hz)]
1. Pendulum thermal noise
- better mirror substrates
- decrease coating mechanical losses
-20
10
(b)
-21
10
(d)
-22
10
(a)
• Shot noise
- higher laser power
- decrease optical losses
- change optical configuration
(e)
(c)
-23
10
1
10
100
1000
10000
Frequency [Hz]
• And keep technical control noise under control
Hanoi, August 8th, 2006
31
The Virgo+ upgrade
• Virgo has already an “advanced” seismic isolation system
• Possibility to improve the sensitivity with a set medium scale incremental improvements
• Planned upgrades
- Fused silica suspensions
- Larger laser power: 20W  50W
 mirror shape thermal compensation
- Electronics and control system
h(f) [1/sqrt(Hz)]
Compatibility with the main Virgo subsystems
Same optical lay-out
Limited shutdown
Limited commissioning
• Coalescing binaries ‘horizon’
-18
10
50W/2 + new losses model
50W/2 + current mirrors
Nominal Virgo
50W/2 + new losses model+FS suspensions
Virgo+ with Newtonian Noise
Virgo
Virgo+
-19
10
-20
10
-21
10
-22
10
-23
maximum distance for optimally oriented source
10
1
10
100
1000
10000
Frequency [Hz]
Virgo
Virgo+
NS-NS
30 Mpc
114 Mpc
BH-BH
145 Mpc
584 Mpc
Hanoi, August 8th, 2006
Coalescing binaries rates increased by ~ 30-100
32
Virgo+ status and plan
• Status of the project:
- approved by CNRS and INFN
- new mirror substrates purchased
- infrastructure for fused silica fiber production and
mirror assembly ready
- laser amplifier to be delivered next fall
- new electronics being developed
• Plan
Hanoi, August 8th, 2006
- test prototype of fused silica suspension next year
- start upgrade after extended Virgo data taking
33
- shut down for upgrade ~ 2008
Toward Advanced Virgo
• Sensitivity goal: 10 better than Virgo (similar to Advanced LIGO)
• Four working groups set-up (with participation of GEO scientists)
• Expected area of improvement:
- High power lasers and injection optics
- Laser beam geometry
- Interferometer optical configurations
- Mirror coatings and shape
- Monolithic suspensions
- Larger mirrors
- Signal recycling
• A possible sensitivity
• Envisaged timeline
- by 2009: R&D and final design
- 2009-10: engineering
> 2010: construction
• EGO is now launching an R&D program to support the preparation of Advanced Virgo
Hanoi, August 8th, 2006
34
A joint LIGO-Virgo plan
• The upgrades of LIGO and Virgo calls for a joint plan to
minimize network down time.
• Discussion with LIGO started last spring
• First very preliminary joint plan sketched
Hanoi, August 8th, 2006
35
Beyond Advanced Virgo
• Advanced interferometer will reach some of the present infrastructures limitation (seismic)
• Need for a new generation interferometer based on an underground facility
- reduce “Newtonian” noise
- construction of longer arms easier
• Third generation ITF keywords:
- squeezed light
- low losses mirrors/suspensions
- massive mirrors
- cryogenics
- ….
• Sensitivity goal:
10 better than Adv ITF
Enter the 10-25 scale
• A will to start a design study in Europe
1) start work within ILIAS-GWA (EU supported network dedicated to GW)
2) prepare
a design
proposal for FP7
Hanoi,
August study
8th, 2006
36
Conclusions
Status of Virgo
- Sensitivity at low laser power a factor of 20 from the design
- Interferometer now running at high laser power
- Entering the last phase of commissioning: noise hunting
- Preparing data analysis and the agreement with LIGO
Next steps
- Soft transition from commissioning to data taking starting next fall
- Long data taking in 2007
- The LIGO-Virgo network starting a new exploration: a first detection is possible
Toward GW astronomy: sensitivity improvements
1. Virgo+
- project defined and started
- plans for installation after a long Virgo data taking (2008)
2. Advanced Virgo
- R&D in progress
- conceptual design by the end of next year, construction > 2010
Hanoi, August 8th, 2006
37
C7 sensitivity
Hanoi, August 8th, 2006
38
Design sensitivity
Hanoi, August 8th, 2006
First attempt to extend detection band down to 10 39
Hz !
Data Analysis preparation
• Large amount of data produced (~8 MB/s 24h/2h)
- need to be ready with data analysis pipelines
• Preparation of data analysis hardware infrastructure
- 120 TB disk space on site; six months look back
- cluster of ~200 PC on site for on-line analysis
- use of computing centers for off-line analysis
(CCIN2P3-Lyon, CNAF-Bologna)
- use of GRID tools for more demanding analysis (pulsars)
• Preparation of data analysis algorithms
- development of search algorithms
- test with real data and hardware injections
- study accuracy of parameters estimate
• Search of rare events
- development of vetoes using auxiliary monitoring channels
(~1000 channels available)
- detector characterization and noise understanding
- development of vetoes using algorithms output
- Need for data exchange with other detectors
Hanoi, August 8th, 2006
40