Gravitational Waves
Physics, Technology, Astronomy and Opportunities
Unnikrishnan. C. S.
Tata Institute of Fundamental Research
Mumbai
[email protected]
www.tifr.res.in/~filab
IIT-K, 2014 Gravitational Waves
The Structure of Physics: Matter, Fields, Space and Time
a = F / m and Principle of relativity
Laws of Motion
Physical Understanding
Laws of Fundamental Forces
F = −GMm / r 2
Conservation Laws (Constraints)
Conservation of momentum. Energy etc…
Four Fundamental Interactions and several fundamental particles
Gravity, Electromagnetism, Weak interactions and sub-nuclear
strong interactions.
Electromagnetism
Weak Interaction Gravity
Strong Interaction
IIT-K, 2014 Gravitational Waves
Magnetic Field
Electric Field
Charges and Currents are fundamental, fields are ‘theoretical’.
Current Magnetic field
Electromagnetic waves need electric AND magnetic field for generation
and propagation.
∂B ∂E
− = ∇ × E , µ0ε 0 = ∇ × B ...Maxwell
∂t ∂t
Unobservables in Physics
Fields, Wavefunction, Space and Time…
IIT-K, 2014 Gravitational Waves
GRAVITY, ITS FIELDS and THEIR WAVES
IIT-K, 2014 Gravitational Waves
F = −GMm / r 2
ag =
−GMm / r 2m =
−GM / r 2
a = F /m
Independent of the mass of the falling body
F = kQq / r 2
kQ q
= =
aem kQq /r m 2
2
a = F /m r m
Gravity seems to be a special interaction
Inertia turns out to be identical to the gravitational charge –
Equivalence Principle (physics of gravity identical to physics in
accelerated frames)
IIT-K, 2014 Gravitational Waves
Universality of Free-Fall
δa
δ a < 10−12 m / s 2 → < 10−13
a
IIT-K, 2014 Gravitational Waves
So, gravitational field ‘g’ and
acceleration ’-a=g’ seem equivalent
This is called the Equivalence
Principle
This is the same as saying that in
g free-fall, there is no gravitational field
But it does not mean that in free-fall there is no gravity!
IIT-K, 2014 Gravitational Waves
So, tidal deviations cannot be eliminated by free-fall
(description in the General Theory of Relativity)
IIT-K, 2014 Gravitational Waves
Torsion balance: Harmonic potential
Ultra-sensitive
IIT-K, 2014 Gravitational Waves
0.3 mm
0.3 mg
g ≈ 5 × 10−8 cm / s 2
1 mm
Flux − density ∝ 1/ R 2 0.01 gm
(For 3D space)
What if space is higher dimensional at some tiny scales (micrometers or less) ?
Inverse-square law for ‘g’ will change!
IIT-K, 2014 Gravitational Waves
TIFR Gravitation Laboratory
Gauribidanur, Karnataka
IIT-K, 2014 Gravitational Waves
‘g’ of Sun ~0.6 cm/s2
δ a (earth − moon ) < 10−14 cm / s 2
IIT-K, 2014 Gravitational Waves
Dark matter
IIT-K, 2014 Gravitational Waves
3% of total estimated matter: So 97% is invisible, and NOT
made of ANY known particle
So, we are not sure of gravity’s behaviour at very small
scales and at very large scales.
IIT-K, 2014 Gravitational Waves
Magnetic Field
Electric Field
Current Magnetic field
Electromagnetic waves need electric AND magnetic field for generation
and propagation.
∂B ∂E
− = ∇ × E , µ0ε 0 = ∇ × B ...Maxwell
∂t ∂t
IIT-K, 2014 Gravitational Waves
What is an electromagnetic wave?
IIT-K, 2014 Gravitational Waves
IIT-K, 2014 Gravitational Waves
Gamma rays and galaxies
IIT-K, 2014 Gravitational Waves
Spectral view – multi-wavelength
Radio
IR
H-α
UV
X-ray
IIT-K, 2014 Gravitational Waves
Spectral view – multi-wavelength
Crab Nebula
IIT-K, 2014 Gravitational Waves
Multi-wavelength galaxy
IIT-K, 2014 Gravitational Waves
Gravity and Electromagnetism
Both have ‘electric’ and ‘magnetic’ parts Charges and currents.
Mass is the ‘charge’ of gravity and Spin is its ‘gravito-magnetic moment’
One important different between the two is that while electric and magnetic
fields have no electric charge, gravitational field has gravitational charge!
With m=E/c2, all forms of energy is equivalent to mass, and hence generate
gravity. Therefore, all fields including the gravitational field, which carry
energy, also generate gravitational fields. This is one reason why the theory
of gravity (The General Theory of Relativity) is complicated to work with.
IIT-K, 2014 Gravitational Waves
Gravity and electromagnetism
Charges (static): Coulomb force – electric fields Electromagnetic
Currents (motion): Ampere’s force – magnetic fields Waves
What about relativistic gravity?
We know static gravitational charge (mass/energy) generates g-field.
Does moving and rotating masses generate a gravito-magnetic field?!
If so, then there is a possibility of gravitational waves…
IIT-K, 2014 Gravitational Waves
Gravito-magnetism
A natural consequence of relativistic gravity, and yet, was not
detected experimentally till recently.
GJ GI Ω GM
µ0 M Bg = = → Ω
B= 2 3
cr 2 3
cr 2
cr
r3
IIT-K, 2014 Gravitational Waves
The real gravitational field near the earth
GM
ω≈ 2
Ω ≈ 5 × 10 −14
rad / s
cr
Gravity Probe – B (Stanford U.)
IIT-K, 2014 Gravitational Waves
Physics of Gravitational Waves
What is an electromagnetic wave?
So, can we expect gravitational waves from oscillating (accelerated) masses?
IIT-K, 2014 Gravitational Waves
Radiation
Et vt at aR
= ≈ = 2 and Er = q / R 2
Et Er ct c c
Er q aR sin θ qrsin θ
=Et =
ct R2 c2 Rc 2
ct
d sin θ
vt
qr =d → Et =
Rc 2
Since area=R2, number of flux lines/area 1/R2
So, radial Electric field (flux/area) ~ 1/R2
What about the Transverse Electric field (radiation)?
Since, circumference of a great circle on the
sphere increases only as R, transverse
radiation field decrease as 1/R.
IIT-K, 2014 Gravitational Waves
0 Phase diff.
GM r sin θ
=Egt 2
×k ⋅r
Rc
GM r sin θ ω r cos θ
= 2
×
Rc c
GMr 2ω 3
= sin θ cos θ
0 3
cR
IIT-K, 2014 Gravitational Waves
The force-field of gravitational waves
+
IIT-K, 2014 Gravitational Waves
The relation between spin of the field and polarization of the force field
IIT-K, 2014 Gravitational Waves
What is the physical effect of a passing gravitational wave?
Quadrupole Radiation formula
∆L G Mr 2ω 2 G Mv 2
Strain h = ≈ 4 ≈ 2
IIT-K, 2014
L c R c R c2
Gravitational Waves
Are we confident that Gravitational Waves exist, apart from the
belief in the correctness of the theory?
Jl. Franklin Inst. 1937
IIT-K, 2014 Gravitational Waves
Binary Pulsar 1913+16 (Hulse-Taylor)
dEG 32G 2 6 4
=
EG ≈ 5 M ωr
dt c ω 2
≈ GM / a 3
→ν ≈ 1
2π Gρ
tc ≈ 3 × 108 yrs
IIT-K, 2014 Gravitational Waves
Orbital decay and speeding up of the binary pulsar:
dEG 32G 2 6 4
=
EG ≈ 5 M ωr
dt c
IIT-K, 2014 Gravitational Waves
Signal Strength at Earth for neutron star spiral in milky way:
Distance: 10 kpc ~ 1020 meters
G Mv 2
Strain h ≈ 4 ≈ 10−64 Mv 2
c R
Gm
v≈ = 0.1 c (3 × 107 m/s) for neutron stars at r ≈ 100 km
2r
∆l
With M~10 kg, v~3x10 m/s, Strain =
30 7 h ≈ 10−19
l
If the event happens in another galaxy, 100 Mpc ( 1024 m away),
Strain h ≈ 10−23
This small strain requires the measurement of <10-20 meters in a
detector of size 1 km! (almost million times smaller than the atomic
nucleus).
Is it a mad venture trying to make a ‘detector’?
IIT-K, 2014 Gravitational Waves
h ≈ 10−17
IIT-K, 2014 Gravitational Waves
A modern cryogenic resonant mechanical detector
IIT-K, 2014 Gravitational Waves
When these waves reach earth, what can they do to free masses?
Michelson Interferometer
G Mv 2 6.7 ⋅ 10−11 × 3 ⋅ 1030 −19
Strain h ≈ 2 → 0.052
< 10 m / m!
( 3 × 10 ) × 10
2 2
c R c 8 20
Much less than the size of the nucleus.
This is the primary device for gravitational wave detection
IIT-K, 2014 Gravitational Waves
Signal Thermal noise,
1/R, G/c2,
Random source…
<< Seismic Noise,
Quantum Noise
Tidal Noise,
All instrument Noise,
Any Noise one can think of…
Signal Lower Limits of all
these noises allowed
1/R, G/c2,
Random source…
> or ~ by Physics and
today’s technology…
IIT-K, 2014 Gravitational Waves
I/
I Signal in the interferometer
∆l
The general problem of ‘fringe splitting’
(centroid, locking…)
Mean number of photons @ intensity I= N, N = P / hν
N ∆φ
Change in the number of photons, ∆N ≈
π
Noise = N ≈ ∆N min
With 1 W of optical power, N=1019 / s,
π∆N min π
∆φmin ≈ = ≈ 10−9
N N
IIT-K, 2014 Gravitational Waves
There is another equivalent way to talk about photon shot noise that
explicitly brings out the basic feature of quantum mechanics involved.
The energy-time uncertainty relation ∆E ∆t ≥
t ω (∆N ) ∆φ / ω ≈
∆E ∆=
1 1
→ ∆N ∆φ ≥ 1 → ∆φ ≈ =
∆N N
IIT-K, 2014 Gravitational Waves
Detection of gravitational waves requires the measurement of
movements 10-17 to 10-20 meters in a detector of size 1 km.
With 1 W of optical power, N=1019 / s, and 10 ms (100 Hz),
π
∆φmin ≈ ≈ 10−8 → ∆L = ∆φmin × λ ≈ 10−14 m
N
STABLE
Laser
1) Increase Laser Power 10 kW → ∆L ≈ 10−16 m
2) Increase Length up to 4 km: Not much gain, though very important (1/R).
3) Fold optical path
n 10 → 10−17 m
=
Reaching there, but not comfortable yet!
IIT-K, 2014 Gravitational Waves
Optimal length of the Interferometer arm:
λg c /ν g 3 ×108 m /100 Hz
Lopt ≈ ≈ ≈ =
750 km !
4 4 4
This is the optimal distance the light should travel for maximal signal. In
other words light should be in the interferometer for an optimal duration of
about 750 km/c seconds or a quarter of the GW period of 10 ms or so.
This is achieved by multiple bounces with average time equal to about
quarter of the GW period such that 4 km x nB = 750 km. So, the number
of bounces is about 200.
IIT-K, 2014 Gravitational Waves
Detection of gravitational waves requires the measurement of
movements 10-17 to 10-19 meters in a detector of size several km.
Improvements:
1) Folding Fabry-Perot Cavity
Finesse ~ n : 300 +
10 kW , 4 km → 10−16 m → 3 ×10−19 m with F ≈ 300
Intra-cavity power > 1 MW !
Radiation Pressure Noise and Thermal Lensing are problems
IIT-K, 2014 Gravitational Waves
Large Interferometer VIRGO at Pisa, Italy ( 3 km)
IIT-K, 2014 Gravitational Waves
LIGO-HO
IIT-K, 2014 Gravitational Waves
∆L G Mr 2ω 2
Strain=
h ≈ 4 ω 2 ≈ GM / a 3 → ν ≈ 1
Gρ
L c R 2π
IIT-K, 2014 Gravitational Waves
But, every bit counts… because waves strength is 1/R
If sensitivity is increased by factor X, then the distance reach increases by X,
and the number of astrophysical sources increases as X3!
So, a factor of 10 in sensitivity means a factor of 1000 in number of possible
detections.
IIT-K, 2014 Gravitational Waves
Mode cleaner cavity
FP1
Laser FP2
Power recycling
Signal recycling
Scheme of the Interferometric detector
3m scale prototype being built in TIFR – can measure <10-17 m
IIT-K, 2014 Gravitational Waves
Schematic of Advanced LIGO detectors
Large scale Ultra high Vacuum to be fabricated in India
10 mega -litres at nano-torr
Noise from Light:
λπ
∆l = λ∆φmin =
N
1
Photon Shot Noise: hsn = α N = Pi / ( hc / λ )
Pi
hPi
Radiation Pressure Noise Frad = N ( h / λ ) =
cλ
Movement noise due to this force: hrp ∝ Frad / mω 2 =
β Pi
Standard Quantum Limit
1
htotal = h + h → hmin = hSQL
2 2
=
sn rp
π Lf m
IIT-K, 2014 Gravitational Waves
How can we hope to measure 10-19 m when the ground vibrations
are like 1 micrometer?!
Immunity to vibrations needed by a factor of 1014!
Solution:
1 1/f2/stage
R
fres
Frequency
Ground vibrations: 10-6 m at 1 Hz, 10-9 m at 30 Hz
3 stages of springs and pendulum with each resonance at around 1 Hz
Response down by a factor (103)3 = 109 at 30 Hz!
Possible to isolate from vibrations at the 10-20 m level at 100 Hz with 3-4 stages.
IIT-K, 2014 Gravitational Waves
IIT-K, 2014 Gravitational Waves
IIT-K, 2014 Gravitational Waves
Absorption/Scattering ppm level
GEO600
IIT-K, 2014 Gravitational Waves
1
htotal = hsn2 + hrp2 → hmin = hSQL =
π Lf m
10-17 Seismic
10-18
10-19
h 10-20
10-21 RP
10-22
Shot noise
10-23
1 10 100 1000
Frequency (Hz)
IIT-K, 2014 Gravitational Waves
A sample of Vacuum chambers (LIGO): < 10-8 mbar, all 4x2 km
2m
IIT-K, 2014 Gravitational Waves
Quantum Noise
∆x∆p ≥
Light: Energy Ε ≈ E 2 + B2 → p2 + q2 Harmonic Oscillator
Main features: Zero point energy, equally quantized energy levels interpreted as
number states of photons
p Coherent state of light - laser
IIT-K, 2014 Gravitational Waves
∆A
p Coherent state of light - laser
∆θ
Amplitude and Phase quadratures
Same Heisnberg uncertainty
x
∆A ∆θ
Amplitude Noise: Phase Noise:
N N
Therefore, if we can somehow SQUEEZE the uncertainty is amplitude or phase,
we can improve measurements by the squeezing factor.
IIT-K, 2014 Gravitational Waves
∆A
p
∆θ ∆A
p
x
∆θ
∆x∆p ≥
IIT-K, 2014 Gravitational Waves
∆A
p
∆θ
If the injected ‘vacuum’ is squeezed by a factor 10, then the quantum noise is
reduced by the same factor, and this is equivalent to increasing the laser
power by a factor 100, because noise reduces only as square root of power!
IIT-K, 2014 Gravitational Waves
Roman Schnabel
American Jl. Physics 2013
IIT-K, 2014 Gravitational Waves
Why do we need more detectors?
GW detectors are not telescopes. They cannot pin-point a source.
c∆t =d
Timing (and only timing ) can fix a direction Need at least 3
IIT-K, 2014 Gravitational Waves
The LIGO-India Concept
IIT-K, 2014 Gravitational Waves
Good angular resolution Base-lines to Japan comparable
IIT-K, 2014 Gravitational Waves
The LIGO-India Idea and Opportunity
The LIGO-India proposal envisages the hardware meant for one of the
LIGO detectors (out of 3) to be given to India and Indian scientists and
engineers will build and operate the detector at a suitable site in India.
Thus the LIGO-India detector will be the third vertex of the LIGO
network, working like a large gravitational wave telescope.
The Dept. of Atomic Energy, Dept. of Science and Technology and the
Planning Commission have approved the Rs.1300 Crore (15 years)
proposal. A cabinet approval is required and awaited. Site selection is
in progress.
www.gw-indigo.org
IIT-K, 2014 Gravitational Waves
IIT-K, 2014 Gravitational Waves
Science Gain from Strategic Geographical Relocation
Source localization error Also Polarization
S. Fairhurst
arXiv:1205.6611v2 B. S. Sathyaprakash et al., LIGO document T1200219-v1
Original Plan
2 +1 LIGO USA+ Virgo
LIGO-India plan LIGO-Aus plan
1+1 LIGO
IIT-K, 2014 USA+ Virgo+ LIGO-India 1+1 LIGO USA+ Virgo+Gravitational
LIGO-Aus Waves
IIT-K, 2014 Gravitational Waves
IIT-K, 2014 Gravitational Waves
4 km
IIT-K, 2014 Gravitational Waves
Gravitational Waves: The Indian Initiative
LIGO-India Project
There are great possibilities and bright future for
gravitational wave-based astronomy… if we manage
to detect gravitational waves with these detectors
The IndiGO Consortium
www.gw-indigo.org
IIT-K, 2014 Gravitational Waves
