SOVIET PHYSICS JETP VOLUME 1^6, N U M B E J l ^ FEBRUARY, 1963
ON THE DETECTION OF LOW FREQUENCY GRAVIT^iTlONAL WAVES
M. E. QERTSENSHTEIN and y . L PUSTOVOIT /
Submitted to JETP editor March 3, 1962
κ
J. Exptl. Theoreti Phys; (U.S.S.R.) 43, 605-607 <August, 1962)
It iS'shown that the sensitivity o£ the electromechanical experiments for detecting gravita-
tional, waves by means of piezocrystals is .ten orders of magnitude ^ o r s e tha^ that^estipiated
by Weber. ^^^ In the low frequency range'it should be possible to detept gravitational wa,yes
by the. shift of the bands in an optical interferonieter. The sensitivity'of this method "isanr
vestigated, t
T h e problem^ of ^he discovery ofxgravitational sensitivity of defection r e p o r t e d j n the^work of
waves has recently beguh to be discussed in the Weber ^^^ is too high by 10 orders of magnitude.
literature, where discussion is given of the Linear processes, which dominate in weak fields;
electromechanical'experiments. However, the determine the limiting sensitivity of the experi-
nonrelativistic bodies at the disposal 0f-the ex'- ments for,the discovery pf weak gravitationa^l ,
perimenter interact very weakly with the gravi- waves. If the equations of an arbitrary lineai;^^sys-
tational wave* Let us consider the equations of tem a r e reversed in time, then a relation results
motion of a particle in the nonrelativistic approxi- connecting the energy losses by radiation and the
mation in the presence of an external electromag- effective diameter σ in reception:
netic fiejd, f®^
ο-σ,ση, il-Qo/Q/?. ' (3)
:1
rnc
ds -l· Ttku'W where σο is the effective cross section of an ideal
(1) lossless antenna, G is" the gain^ resulting from the
A.plane gravitational wave (goo = - Eoa ^ directivity, η is the efficiency of the antenna in
does not chgtnge the proper time of the nonrelativr the transmission mode, Qo is the quality factor of
istic (u*^ =>0) body, Γ ^ · = 0 ; an unoharged-non- the real antenna, and ,Qp.is the quality factor a s -
relativistic particle doe^ not sense the wave. If sociated with radiation, Tlie cross section is
the field F^^ is produced by specified nonrela- σο = λ^ with accuracy to within a factor of order
tivistic charges and curj^ents which do not change unity. A relation of the type (3), known in "the the-
under the action of the .passing gravitational wave, ory of antennas, ^^^ can be obtained from the f o r -
then the field, likewise does not<;hange as mulas of the paper of Weber, ^^^ and also follows
seen directly frpm the equations o^motioni^^^ from the principle of detailed balancing {see
Sec. 117 ). Experiihent shows that tlie results'oi
f^'^],df'/dx' = (4it/c) i\ Weber do riot satisfy the relation (3), and the di-
since. V - g = 1 in the.gravitational wave. The vergekce amounts to 10^®. Thus, for example,
presence of a gravitational waye: leads to tbe^'ap- according to Weber, for a" wave with, λ = 100 cm,
pearance of an additional force on-the .right hand the power fed to the crjrstal in the radiation mode
side pf (1), which vanishes ^n the .non- is equal to 10^ W, the power radiated^'is li)'^^ e r g /
relativistic approximation (k = OJ. Detection of' sec, whence η = 10'^®, the radiation^is quadrupole
gravitational waves by a nonrelativistic body (-as and G = 15, In thfe i^eieption mode f^r an ideal
well as by a piezoelectric one) is but weakly e f - antenna, σο = 3 χ 10^ cm^; for a real antenna Eq.
fective. (3) gives b· = 4 X 10"^^ cm^. At threshold, the
Starting out from the'general relations for lin- value oi etiergy flux according to'Webei is Ρ
ear processes^^ we shall shqw below that the high = 10"^erg/sec-cm^, the electromagnetic energy
received is equal to 4 x 10"^^ W = — 334 dB/W*
which is lower by^ 11-0'dfe thab the threshold power
^ t is well known that a rotating rod or k binary^ star raciiates
for an ideal receiVer with a noise temperature of
in quadrupole fashion, the frequency of the radiation-being
twice as great as the frequency of motion existing in.the" s y s -
3°K and b^dwidth of 1 cps. The tin^e requ^ired
tem. Such processes in which'double frequencies appear are for detection of a 'signal which is lower than
not linear and are not considered here. threshold by 110 dB'exceeds 10^ years.
433
434 Μ. Ε . G E R T S E N S H T E I N a n d V. Ι ; PUSTOVOIT
Calculation of the radiated power was carried ^22 = ""^33 = h is different from zero, and there-
out by Weber according to the ^'quadrupole" f o r - fore the flux of gravitational energy is equal to
mula, ^^^ the validity of which there is no reason
Ρ- / 16jik .:: co^c/i^ / 2x ; κ - Hnk / c^, (8)
to doubt. Therefore, it follows from Eq. (3) that
the absorbed power was incorrectly'computed by where κ is the Einstein gravitational constant.
him. The reason for the mistake lies in the fact Making use of (7), we have
that in a piezocrystal, the piezoelectric s t r e s s e s
are compensated by mechanisms which were not - ti-lj^ Y P —2— (9)
ί, - V C - ~ i{cpsy erg/cm^ · sec.
taken into account by Weber. Piezoelectric
s t r e s s e s do not satisfy the virial theorem, The minimum measured ΔΖ with ordinary light
Since the reception of gravitational waves is a sources amounts to 10"^ A, which is equal to
relativistic effect, one should expect that the use cm, for a time constant of the apparatus τ
of an ultrarelativistic body—light—can lead to a -- 1 sec. It can be expected that application of
more effective indication of the field of the gravi- strong sources and amplifiers of monochromatic
tational wave. The optics of rays in a gravita- directed light radiation (lasers'^^) will make it
tional field is'determined by the eikonal equation ^^^ possible to decrease this factor by at least three
orders of magnitude.
0ik ^ ^ i-^V (4) Assuming the interferometer arm to be f 10^
^ dx' dx^ Vlv' /
cm, we have for the minimum observable change
where ψ is the eikonal. This is equivalent to a AI/Iq ^ 10-14-10-^ τ - 1 sec.
medium with index of refraction Thus the interferometer makes it possible, at
least In principle, to detect the very weak gravita-
η = \ + ^h^fifi^n^ (5)
tional waves. For fg =10"^ cps and Ρ = 1 erg/cm^
where n*^ is a unit vector along the propagation of -sec we get A l / l « 8 χ lO'^*^, which is approxi-
the ray. For propagation of the ray along and p e r - mately 10^—10^^ times better than the possibilities
pendicular to the gravitational field, we have of electromechanical experiments.^^^^
Further gain in the sensitivity can be obtained
= «J. = 1 cos 2φ 2φ, by increase in the time of observation and by
cos φ " n.^. (6) known methods for the separation of a weak sig-
nal from the noise background. Evidently, times
In apparatus such as the Michelson interferom- of observation τ 10"^—10® sec are realistic; in
eter, the relative difference of the optical lengths this case Pmin erg/cm^-sec. Bernstein,
of the light rays traveling along and perpendicular detected a monochromatic'sinusoidal signal
to the gravitational wave will be against the noise. In the case of a useful signal
with a complex spectrum, these estimates must
M/L - (7) be changed somewhat but we shall not dwell on
where is the unperturbed length of the inter- this subject since these changes are the same for
ferometer arm. We note that Eq. (7) for the both the interferometer and the electromechanical
Michelson interferometer can be obtained directly. experiments. Technical experiments with an
In a gravitational field, the optical length of the interferometer on the discovery of gravitational
interferometer arm is changed and the relative waves of low frequencies existing in nature are
difference Sec. 84) is equal to very complicated. It is necessary to have stable
» h u apparatus, and the air must be pumped out along
^ (7a) all the optical paths. Inasmuch as neither the fre-
0 0 quency nor the polarization nor the direction of
In the derivation of Eq. (7), it was assumed that propagation of the wave are known, it is necessary
the period of the gravitational wave is much greater to have several interferometers and to study the
than the time of flight of the ray in the interferom- correlation among them.
eter. In conclusion, the authors express their deep
Thus the gravitational wave produces a p e r i - gratitude to Professor V. L. Ginzburg for valu-
odic displacement of the interference bands. We able observations and discussions of the given
Y»<a c2iia νΓ»ν»
express (7) in t e r m s of the radiation energy flux
P, first rotating the 0x2 and 0x3 axes so as to J. Weber, Phvs. Rev. 117. 305 η 960^.
eliminate the component hoo. In this svstem onlv
ON T H E D E T E C T I O N OF LOW F R E Q U E N C Y GRAVITATIONAL WAVES 435
^B. B. Braginskii and G. L Rukman, J E T P 41, ®L L. Bernshtein, DAN SSSR 94, 665' (1954),
304 (1961), Soviet Phys. J E T P 14, 215 (1961). UFN 49, 634 (1953).
^L. D. Landau and E. M. Lifshitz, Teoriya ^L L. Bernshtein, DAN SSSR 75, 635 (1956).
polya (Theory of Fieliife) Fizmatgiz, 1960. ®A. Schawlow, Scientific American 204, 52
^Ya. N. Fel'd and L. S. Benenson, Antenno- (1961).
fidernoe ustroistvo (Antenna and Feeder ^ s t e m )
(VVIA P r e s s . 1959).
^L. D. Landau and E. M, Lifshitz, Kvantovaya
mekhanika (Quantum Mechanics) Gostekhizdat, Translated by R. T. Beyer
1948. 105
