Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Detection of gravitational waves

  1. Dec 27, 2009 #1
    What is the state-of-the-art of detecting gravity waves. Are the instruments available now sufficiently sensitive to detect gravity waves? Have we identified 'low-hanging' sources from which gravity waves should have been detectable. Have we yet reached the point where doubts are arising as to the direct detectiblity of gravity waves?
  2. jcsd
  3. Dec 28, 2009 #2


    User Avatar
    Science Advisor

    LIGO is still basically the only gravitational wave observatory. So far all it's been able to do is place upper limits upon the magnitude of gravitational waves.

    The next-generation instrument is a space-based interferometer, LISA.

    I don't think anybody in the field has any doubts at all that it's possible to directly detect these waves. It's just clear that it's a challenging problem.
  4. Dec 29, 2009 #3
    Are there theory calculations of how strong the waves are expected to be? How does that compare to how sensitive the LIGO is and how sensitive the LISA will be?
    Last edited: Dec 29, 2009
  5. Dec 29, 2009 #4


    User Avatar
    Science Advisor

    Indeed, quite a few. Unfortunately I can't say much more than that, except that the types of events that we expect to eventually be able to detect are mergers of massive, compact objects (such as neutron stars, black holes, and especially the mergers of supermassive black holes that is expected as a consequence of galaxy mergers). Some also hope for some direct detection of the gravitational wave background (analogous to the CMB), but we really have no idea what the magnitudes of those are going to be, as that depends upon the specific model of inflation.
  6. Jan 18, 2010 #5
    There is a Gravity Wave Measurement faculty at Livingston,Louisiana.

    Joe in Texas
  7. Jan 19, 2010 #6
    chalnoth - it seems to me that even in the event of two neutron stars colliding, the actual COG of the system perceived from any distance would not change, and thus no significant alteration of the gravitational field would occur, and no GWs would be detected. same way with exploding stars, etc.

    in order to create detectable grav waves at any appreciable distance, the COG of some massive body/system would have to be significantly and rather quickly altered, and i cant really think of any way that could occur. local measurement of GWs could be possible if you were in the near vicinity of a pair of massive bodies in orbit about each other.
  8. Jan 19, 2010 #7


    User Avatar
    Science Advisor

    The alteration of the gravitational field comes from the time rate of change of the quadrupole of the field, not the center of mass. Another way of saying it is that it's the change in the shape of the field that produces the gravitational waves. And in order to transition from two neutron stars to just one requires a very large change in the shape of the gravitational field.

    That said, most of the time people talk about detecting gravitational wave signatures, they're thinking about colliding black holes, not neutron stars.
  9. Jan 20, 2010 #8
    Is there any sitution where the merger of two neutron stars results in another neutron star? I would imagine that would tip them both past their Schwarzschild Radii. Does the massive blast of energy carry away enough mass so that they won't collapse further?

    I also believe, that as sensitive as LIGO is, and LISA will be; they don't have a prayer of detecting a collision of anything less than binary BH's.
  10. Jan 20, 2010 #9


    User Avatar
    Science Advisor

    The maximum mass of a neutron star is somewhere in the range of 3 solar masses, while the typical masses for neutron stars are 1.35-2.1 solar masses (from Wikipedia). This would seem to indicate that two neutron stars colliding would usually become a more massive neutron star, and not a black hole.

    I am not sure we know just how much mass would be ejected in such a collision.

    That is quite possible. But bear in mind that our ability to detect these phenomena depends critically upon how far away they occur.
  11. Jan 20, 2010 #10
    Gotcha. So probably two large neutron stars colliding WILL collapse into a BH, and that would emit detectable GWs I think. On avg though, with the energy released in a GRB I have to imagine most neutron star mergers are just that... mergers and not collapsars. Thanks very much for the info.

    For the distance, is that a result of losing energy (aka a weakening effect) or a "redshift" effect that would elongate the waves making them harder to detect? Or both? Or neither... Damn this is an exciting time to be alive if you have an active interest in science.
  12. Jan 20, 2010 #11


    User Avatar
    Science Advisor

    Well, it's the same as it is with light waves. You get a falloff of 1/r^2 due to the fact that we have 3 spatial dimensions, and you get an extra drop in energy (and thus detectability) from any redshift that occurs.

    Of course, if neutron star-neutron star mergers are detectable at all, they're probably detectable only within a nearby region of our own galaxy, so the redshift effect would be completely negligible.
  13. Jan 20, 2010 #12
    Understood! Thanks very much again for the clear info. :)
  14. Jan 20, 2010 #13


    User Avatar
    Science Advisor
    Homework Helper

    Gravitational wave amplitude actually goes like 1/r. (If by amplitude you mean the metric perturbation)
  15. Jan 20, 2010 #14


    User Avatar
    Science Advisor

    Well, I believe the energy still drops off as 1/r^2, so I believe that's the relevant quantity for detection (though I suppose I could be mistaken). Electromagnetic wave amplitude, after all, also drops off like 1/r.
  16. Jan 20, 2010 #15


    User Avatar
    Science Advisor
    Homework Helper

    There is a great example in Sean Carroll's relativity book about just how sensitive gravitational wave detectors need to be.

    If we have a binary system of black holes (10 solar masses each), seperated by 10 times their Schwarzchild radius (~10^7cm), and they are a cosmological distance away from us (~100 Mpc), then the gravitational wave amplitude is
    h ~ Rs^2 / (10Rs) / r ~ 10^-21.

    This means that if we have an interferometer detector with a length scale L of order kilometres, we need it to be sensitive to scales dL, where dL/L ~ h. So dL ~ 10^-16 cm. That is, much smaller than a nucleus.
  17. Jan 20, 2010 #16


    User Avatar
    Science Advisor
    Homework Helper

    Hmmm... All the references I read or see seem to do everything in terms of h and not energy. But that could just be for simplicity. I'm really not too sure about how the detectors work to know what the relevant quantity is.
  18. Jan 20, 2010 #17
    The detectors that are earthbound use very VERY long LASER beams and try to detect deviations in them caused by a wave. LISA I believe uses high precision gyroscopes and highly finished spheres to produce a similar effect in a smaller package.
  19. Jan 22, 2010 #18
    Hello guys;

    Since no one mentioned it ...

    I think you should be aware that there are several operational GW detectors globally of the resonant mass ("Weber bar") variety based upon Weber's original design, except that they are generally cryogenic....and have been operational for a number of years with excellent sensitivity... h < 10^-18.

    There are several in Italy, one in Denmark, and Australia, and one at LSU in Baton Rouge.
    This type uses a large (Aluminum) cylinder whose strain is measured as the GW passes through. (some are listed here:http://www.astrophysicsspectator.com/topics/generalrelativity/ResonantBarDetectors.html )
    One recent resonant mass design (in Brazil and elsewhere) is spherical so as to make it omnidirectional .

    Last edited: Jan 22, 2010
  20. Jan 23, 2010 #19
    Thanks for the link and info, that's fascinating stuff. I only ever really heard about these in terms of the near-perfection of the sphere in the newer design in a "popular" science context. This is is very interesting, but I'm still putting my money on LIGO or LISA.
  21. Jan 24, 2010 #20


    User Avatar
    Science Advisor
    Gold Member

    Gravity waves are very difficult to detect, but very important in cosmology. I think we are less that 5 years away from the first 'hit'. The LIGO project is in constant calibration and likely to produce the first definitive result. LISA is not scheduled for launch until 2016.
    Last edited: Jan 24, 2010
  22. Jan 24, 2010 #21
    Believe me when I say that I hope to hell that you're right. That would be an amazing piece of confirmatory evidence, and hopefully would open the gates for more funding and fast-tracking of follow-up projects.
  23. Jan 24, 2010 #22
    Hi Frame Dragger; Thanks for your response. But before you drain your bank account I think you should be aware that besides these two methodologies, interferometric and resonant mass, there has been some recent development on a somewhat different and promising GW detection technique using coupled superconducting cavity detection.

    It involves coupling of 2 superconducting resonant cavities. A passing GW induces a transition between the two resonant frequencies. Predicted to be quite sensitive; at resonant freq., around h = 10^-21 or better, depending on the Q factor.
    Details Here:

    In all these various types there is no need to 'pick one', since each detector type covers a different frequency band, so they become somewhat complimentary.

    Resonant mass - around 700 to 900 hertz (the Brazilian Spherical ~ ?? I'll have to look it up ).
    LIGO for low frequencies, around 0.1 to 5 Hz , currently, if I'm not mistaken
    LISA is ultra low freq. - milliHertz to 0.1 HZ
    Superconducting Cavities Res. cover high freq.; can be designed from around 4 to 12 kiloHertz.

    I like the recent progress in the SCRF.

    Creator :))
    Last edited: Jan 24, 2010
  24. Jan 24, 2010 #23
    I really love this forum. Thanks for the info Creator... this is all fascinating, but I need to read more about this. You've given me some new directions for learning in this area, and I appreciate that. :)
  25. Feb 5, 2010 #24
    I recently shot a question to Professor Hughes about an idea I had for building an experiment to generate gravitational waves. I know its like impossible but he didn't just tell me I was nuts he gave me the equations to show that I'm nuts lol. I will share that with you here as you will be able to follow it much easier than I can.... I had to study this for days. And I am still having a hard time trying to plug in alternative parameters.

    To get any interesting GW power, the spheres would have to move at a
    significant fraction of the speed of light.

    To set an upper limit, I'll take the spheres to be 2 meters in
    diameter (a bit larger than your 5 feet). I'll take them to be
    filled with mercury (denser than gallium; I'm not suggesting you
    actually do this, just doing the calculation to prove the
    principle). Mercury has a density of a bit less than 14 grams per
    cubic centimeter, so the spheres would each have a mass of about

    M = (4/3) pi (14 grams/cm^3)(100 cm)^3 = 59,000,000 gm

    (At each step I've rounded up slightly, so my calculation will be an
    overestimate.) If I have two "stars" of mass M orbiting one another
    with their centers separated by a distance R and orbiting with a
    period T, then the power generated by gravitational waves is given by

    P = (8/5) (G/c^5) M^2 R^4 (2 pi/T)^6

    (As you'll see in a minute, the factor of G/c^5 --- where G is the
    gravitational constant, and c is the speed of light --- really kills
    us. This is why all GW experiments are based on astrophysical
    sources, where we can get masses that are stellar or larger.)

    For your experiment, M = 5.9 x 10^7 grams. The center to center
    separation R is 200 cm (well, 5 feet --- but I'm rounding up a bit to
    get an overestimate). The period is

    1/(2000 rpm) = 0.005 minutes = 0.03 seconds

    The numerical factor G/c^5 is 2.76 x 10^(-60) sec^3/(gm cm^2). Let's
    put all of this together:

    P = [8/5][2.76 x 10^(-60) sec^3/(gm cm^2)][5.9 x 10^7 gm]^2 [200 cm]^4
    [2 pi/(0.03 sec)]^6

    = [2.76 x 10^(-60)][7.52 x 10^(38)] gm cm^2/sec^3

    = 2.08 x 10^(-21) erg/sec

    Converting to Watts (1 Watt = 10^7 erg/sec), your proposed apparatus
    would generate a gravitational wave power of about 2 x 10^(-28)
    Watts. If we imagine that you can make the rotational frequency go
    arbitrarily high, it would be useful to see how this results scales
    with that frequency:

    P = 2.08 x 10^(-28) Watts (f/2000 rpm)^6.

    Doing a little bit of algebra, we see that if want this to get up to
    1 Watt, we need to dial the frequency up to about 82,000,000
    revolutions per minute. At this speed, the spheres would be moving
    at a speed of about 1,400,000 meters per second --- about 0.5% of the
    speed of light. (Of course, at this speed, the material out of which
    the spheres are made would not be able to hold together. This is why
    astrophysical gravitational wave sources are objects like neutron
    stars and black holes --- their enormous self gravity is what allows
    them to hold together while they whirl about one another at speeds
    which are an appreciable fraction of the speed of light.)

    I'm afraid there's no way to overcome the fundamental limits set by
    that factor of G/c^5 --- you just need enormous masses and enormous
    speeds. Anything you can make on earth will not do the trick.

    Please note I'm going to be away from my email for the holidays and
    am unlikely to answer any followup questions on a short timescale.

    Scott Hughes

    I also put this in a topic I started about this very experiment of mine. Maybe you geniuses could tell me if my apparatus would have any good uses at all lol.
  26. Feb 5, 2010 #25


    User Avatar
    Science Advisor
    Gold Member

    LIGO uses very long range, sensitive instruments. Detectability is not an issue, isolating the signals from background noise is an issue.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook