rana adhikari caltech the next gravity wave interferometers
Post on 22-Dec-2015
217 views
TRANSCRIPT
Rana AdhikariCaltech
The Next Gravity Wave
Interferometers
2
Gravitational Waves = “Ripples in space-time” Two transverse polarizations - quadrupolar: + and x
Gravitational Waves?
Example:
Ring of test masses
responding to wave
propagating along z
Amplitude parameterized by dimensionless strain h: L ~ h(t) x L
Need to measure strain of ~ 10-21-10-22
We want a very large ‘L’
3
Compact binary inspirals: “chirp” NS-NS waveforms are well described. 1.4 Msolar NS/NS
inspiral is a standard candle.standard candle. BH-BH waveforms are rapidly improving
Supernovae / Mergers: “burst” Short signals. Waveforms not well known. Search in coincidence between two or more interferometers and
possibly with electromagnetic and/or neutrinos signals
Spinning NS: “continuous” search for signals from observed pulsars all-sky search computing challenging
Cosmic Background: “stochastic” Metric fluctuations amplified by inflation, phase transitions in
early universe, topological defects Unresolved foreground sources
GW Sources in LIGO Band 50-1000 Hz
4
LIGO Observatories
Livingston, LA (L1 4km)
Hanford, WA (H1 4km, H2 2km) - Interferometers are aligned to be as close to parallel to each other as possible
- Observing signals in coincidence increases the detection confidence
- Determine source location on the sky, propagation speed and polarization of the gravity wave
LIGO GEO VirgoTAMA
AIGO (proposed)
5
Michelson Interferometer
Anti-Symmetric (Dark) Port
Ly
LxReflected Port
PAS PBS x sin2()
dP/d PBS x
sin()cos()
= 2 (Ly - Lx) /
ddh L
Laser
Noise PAS
6
End Test Mass
50/50 Beam Splitter
Photo detector
Laser
MichelsonInterferometer
MichelsonInterferometer 4 km Fabry-Perot
arm cavitywith Fabry-Perot Arm Cavities
Input Test Mass
6 W
Power Recycled
Power RecyclingMirror
300 W 20 kW
Interferometer Optical Layout
Signal Phase shift between the arms due to GW
7
Science Requirements Doc:
The LIGO-I Sensitivity Goal
Seismic:Natural and anthropogenicground motions, filtered byactive/passive isolation systems.Depends strongly on in-vac seismic isolation.
Thermal:Brownian noise in the mirrors and in the mirrors’ steel suspension wires.Depends mostly on internal rubbing in the suspension wires.
Shot Noise:Photon counting statistics -- > 10 kW in the cavities ~ 200 mW detected power
- Goes down with increased laser power and better fringe contrast
8
5 years of debuggin’ in Louisiana…
9
10
What Is Inside1.2 m diameter - 3mm stainless 50 km of weld 10-9 torr vacuum and no leaks!
Seismic isolationStack of masses and springs
Coils and magnets to control the mirror
Fused silica mirror25 cm diameter10 kg mass
11
Dark Port Optical Table
2 mm diameterInGaAs photodiode
12
Time Line
NowInauguration
1999 2000 2001 2002 20033 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
First Lock Full Lock all IFO
10-17 10-18 10-20 10-21
2004 20051 2 3 4 1 2 3 4 1 2 3 4
2006
First Science Data
S1 S4Science
S2 RunsS3 S5
10-224K strain noise at 150 Hz [Hz-1/2]
2006
HEPI at LLO
13
Caltech LIGO Laboratory MIT
LIGO Hanford Observatory LIGO Livingston Observatory
University of Adelaide ACIGA
Australian National University ACIGA
Balearic Islands University
Caltech LIGO
Caltech Experimental Gravitation CEGG
Caltech Theory CART
University of Cardiff GEO
Carleton College
Cornell University
Embry-Riddle Aeronautical University
University of Florida-Gainesville
Glasgow University GEO
NASA-Goddard Spaceflight Center
Hobart – Williams University
India-IUCAA
IAP Nizhny Novgorod
IUCCA India
Iowa State University
Loyola New OrleansLouisiana State University
Louisiana Tech University
MIT LIGO
Max Planck (Honnover) GEO
Max Planck (Potsdam) GEO
University of Michigan
Moscow State University
NAOJ - TAMA
Northwestern University
University of Oregon
Pennsylvania State University
Southeastern Louisiana University
Southern University
Stanford University
Syracuse University
University of Texas-Brownsville
Washington State University-Pullman
University of Western Australia ACIGA
University of Wisconsin-Milwaukee
LIGO Scientific Collaboration~40 institutions, ~550 scientists
14
LIGO Science Run The fifth science run started in November 2005 S5 goal is to collect one year of triple coincidence data at the
design sensitivity Optimistic event rates: NS/NS ~3/year, BH/NS ~30/year Nakar,
Gal-Yam, Fox, astro-ph/0511254 Plan to reach the Crab pulsar spin down limit Expect to beat the Big-Bang Nucleosynthesis limit on
gravitational wave density in the LIGO band GEO interferometer joined the S5 run in January 2006. Virgo interferometer plans to join S5 later this year.
15
NS-NS Inspiral Range Improvement
Time progression since the start of S5
HistogramDesign Goal
Commissioningbreaks
Stuck ITMY opticat LLO
16
S5 Duty Factor
S5 Goal is 85% for single
interferometer and 70% for
triple coincidence
One week running average
Commissioningbreaks
Stuck ITMY opticat LLO
17
H1H1 H2H2 L1L1
UptimeUptime 72% 79% 60%
Wind, Storms, Earthquakes 4.5% 9%
Nearby Logging, Construction, Trains - - 10%
Maintenance, Commissioning, Calibration
10% 9%
Hardware and Software Failures 3.5% 7%
Lock Acquisition, Other 10% 5%
S5 Duty Factor
H1&H2&L1 = 45% H1||H2||L1||G1 close to 100%H1&H2&L1 = 45% H1||H2||L1||G1 close to 100%
18
Triple Coincidence Accumulation
100%
~ 45%
~ 61%
Expect to collect one year of triple coincidence data by summer-fall 2007
19
Sometimes You Get Lucky Large mirror (ITMY) was wedged into the earth quake stops Vented the vacuum and released it. Adjusted EQ stop. Noise improved!? 12->14 Mpc
Earth quake stop
before
after
ChargeDissipation
on the optic?
20
Advanced LIGO LIGO mission: detect
gravitational waves and initiate GW astronomy
Next detector Should have assured detectability
of known sources Should be at the limits of
reasonable extrapolations of detector physics and technologies
Must be a realizable, practical, reliable instrument
Daily gravitational wave detections R&D is mature, prototypes exist Installation start in 2011
Advanced LIGO
21
101
102
103
10-24
10-23
10-22
Frequency (Hz)
Str
ain
No
ise,
h(f
) /H
z1/2
10 Hz 100 Hz 1 kHz
10-22
10-23
10-24
10-21
Anatomy of the projected Adv LIGO detector performance
Newtonian background,estimate for LIGO sites
Seismic ‘cutoff’ at 10 Hz
Suspension thermal noise
Test mass thermal noise
Unified quantum noise dominates at most frequencies for fullpower, broadband tuning
Initial LIGO
Advanced LIGONS-NS Tuning
22
End Test Mass
50/50 beam splitter
GW signal
Optical Configuration
Laser
MichelsonInterferometer
MichelsonInterferometer 4 km Fabry-Perot
arm cavitywith Fabry-Perot Arm Cavities
Input Test Mass
125 W
Power Recycled
Power Recyclingmirror
2 kW 500 kW
Signal Recycling mirror
Dual recycled
23
Detuned Signal Recycling
frequency offset from carrier [Hz]
•Responses of GW USB and GW LSB are different due to the detuning of the signal recycling cavity.
•IFO Differential Arm mode is detuned from resonance at operating point
00
SRC DARM0
Car
rier
fre
quen
cy
-10000 -5000 0 5000 10000
50
100
200
500
1000
Sid
eban
d am
plit
ude
[a.u
.]
FWHM
USBLSB
fsig
IFO DARM/CARM
from R. Ward
24
Opto-mechanical Spring
Optical Spring stiffness ~ 107 N/m
Angular spring resonance ~ 2 Hz
• ½ MW in the arms ->
• ‘Optical Bar’ detector
• ~75 Hz unstable opto-mechanical resonance
• High Bandwidth servos
Measured Transfer Functions from the 40m prototype
BMW Z4 ~ 104 N/m
Radiation pressure: F = 2 P / cDetuned Cavity -> dF/dx
25
The next several years
Between now and AdvLIGO, there is some time to improve…
1) ~Few years of hardware improvements + 1 ½ year of observations.1) Factor of ~2.5 in noise, factor of ~10 in event rate.2) 3-6 interferometers running in coincidence !
S5 S6
4Q‘05
4Q‘06
4Q‘07
4Q‘08
4Q‘10
4Q‘09
Adv
LIGO~2 years
Other interferometers in operation (GEO, Virgo)NOW 4 yrs
26
NS/NS Binary
Most of the sensitivity comes from a band around 50 Hz
50 Hz
Area proportional to SNR
27
30/30 M BH/BH
Most of the sensitivity comes from a band around 30 Hz
Area proportional to SNR
30 Hz
28
Astrophysical Motivation
How does the number of surveyed galaxies increase as the sensitivity is improved?
From astro-ph/0402091, Nutzman et al.
Power law: 2.7For NS-NS binaries
Prop. to inspiral range
Factor of 2.5 reduction in strain noise,
factor of 10 increase in # of sources
S4
29
Baseline Goals
1. Dark Port Filter Cavity (Caltech/MIT)1. Reduce amount of junk light, reduce shot noise
2. Reducing detected light power allows higher laser power
3. Upgrade the detection system to the Advanced LIGO style.
2. Higher power laser (Hannover)1. Our (10 W) laser company was bought out by JDS Uniphase.
2. Collaborators at AEI/LZH are offering us 35 W lasers (for free!)
3. High Power Input Optics (UF, Gainesville)
4. Miscellaneous …
30
Seismic:No modification in seismic isolation systems
Thermal:Good wires, good mirrors, and control of “technical” noises
Shot Noise:- New in-vac filter cavity-- 4-5x more laser power-- Advanced readout technique
31
Increased Power + Enhanced ReadoutLower Thermal
Noise Estimate
32
The Plan
Improvements on the 4km IFOs starting in Sep07Do Louisiana first (pathfinder). Start Hanford in Jan08.Some modest Suspension electronics fixesThen some more science running.
Not enough time/manpower to do all 3 IFOs.
A factor of 2.5 on H1/L1 is better than a factor of 2 on all three.
We don’t gain more AdvLIGO knowledge by doing 3 IFOs.
After the pumpdown, H2 can join Virgo in a science run.
33
Conclusions
1 more year of R&D to support the upgrade Starting installation in Sep ’07 Next Science Run (w/ improved sensitivity) starts
in Sep ’09.
Reduces much technical risk for Advanced LIGO Its time to make the first gravitational wave
detection.
34
Short, Hard GRBs Several Short Hard Gamma-Ray Bursts
since last year Detected by Swift, HETE-2, then Hubble,
Chandra, and BATSE SHGRBs (< 2 s) different from ‘Long Soft
GRBs’ (supernova explosions) Candidates for progenitors of SHGRBs:
double neutron star (NS/NS) or neutron star-black hole (NS/BH 1.4/10) coalescences
Optimistic rates for Initial LIGO (S5) could be as high as a few/year
Nakar, Gal-Yam, Fox, astro-ph/0511254
Double Neutron Star MergerDana Berry / NASA
35
Suspension wire re-working Change the clamp to reduce excess noise (no evidence so far) Change the wire to reduce the intrinsic noise Needs some serious coil driver redesigns capitalize on lower noise. Need to know more about the excess noise first.
Squeezed Light Implement on one IFO instead of the laser upgrade; more
speculative, but doesn’t require new IO equipment. An opportunity to commission another AdvLIGO system
Signal Recycling No real sensitivity improvement; lots of work.
Double Suspension Not directly applicable to AdvLIGO. Substantial reworking req. Not clear if we can get the technical noises out of the way.
Beyond ‘Fixes’
36
Cross-correlation estimator
Theoretical variance
Optimal Filter
Strain Power:
Choose N such that: HY
)100/()( HzfHfH )()(
)(),(1),(
21
point~
fPfP
fHft
NftQ
,,2,1
)(2
point
^
),(A
At
At
c
txfi
FFeft
Detection Strategy, point source
Point Spread Function
37
S4 Upper Limit map , H(f)=const
148%90 10)1.685.0( HzH