Uncovering the Truth: Examining the Existence of Gravitational Waves

[Spin_Network]

A great recent article details the problems of GW's?

http://www.physicstoday.org/vol-58/iss-9/p43.html#ref

So do Gravitational Waves exist?

This was meant to be placed in General forum!

 

[SpaceTiger]

I would be very surprised if gravitational radiation didn't exist. It appears in weak field GR -- a limit I think most physicists are confident with -- and we even have observational evidence for them:

http://astrosun2.astro.cornell.edu/academics/courses//astro201/psr1913.htm"

It's indirect evidence, certainly, but pretty solid nonetheless. There will be some serious head-scratching if none of the upcoming gravitational wave experiments has a detection.

 

[Garth]

I agree that gravitational waves most probably do exist, especially as we have detected the loss of orbital energy from the PSR 1913+16 system as predicted.

However there is a question about the detection of GWs. It is similar to the question, "If the universe is expanding what exactly expands with it? Do rulers co-expand?" If so there would be no detectable expansion. The standard answer to this question is No - rulers do not expand with the universe and therefore they can detect that expansion.

However if this understanding is mistaken, and everything embedded in space-time does co-expand with the expansion of space, then any physical apparatus would not be able to detect GWs either, for they would simply 'wash over' the detectors leaving no signal. Could this be a possible explanation if the non-detection of GWs continues?

If in fact this is the case then another explanation is required for cosmological red shift, such as a mass field effect (Hoyle: F. 1975 ApJ 196:661-670 "On the origin of the microwave background") - But that is definitely not "mainstream"!

Garth

 

[Spin_Network]

I agree that gravitational waves most probably do exist, especially as we have detected the loss of orbital energy from the PSR 1913+16 system as predicted.
However there is a question about the detection of GWs. It is similar to the question, "If the universe is expanding what exactly expands with it? Do rulers co-expand? If so there would be no detectable expansion." The standard answer to this question is No - rulers do not expand with the universe and therefore they can detect that expansion.
However if this understanding is mistaken, and everything embedded in space-time does co-expand with the expansion of space, then physical apparatus would not be able to detect GWs either, they would simply 'wash over' the detectors leaving no signal. Could this be a possible explanation if the non-detection of GWs continues?
If in fact this is the case then another explanation is required for cosmological red shift, such as a mass field effect (Hoyle: F. 1975 ApJ 196:661-670 "On the origin of the microwave background") - But that is definitely not "mainstream"!
Garth

I do comply with the fact that GW's exist, I also agree that their existence is not an observable fact?

Gravitational Waves exist, but their existence is not Observation Dependent.

 

[Haelfix]

I don't understand this line of reasoning. You better believe the measuring apparatus expands as the universe expands, its just its so vanishingly small (98 orders of magnitude small for instance) that its for all intents and purposes a nonfactor.

However we aren't measuring the expansion of space, we are presumably measuring ripples from linear perturbation series from some object (like a neutron star). As they pass us, you should see a kink in the machine.

If nothing is detected as far as I am concerned something is very, very wrong. In fact i'd put the blame on the experiment rather than the theory, so confident I am that it has to exist. If however it persists through generations of LIGO and others, well it would be deeply embarrasing for many of us.

 

[Garth]

I don't understand this line of reasoning. You better believe the measuring apparatus expands as the universe expands, its just its so vanishingly small (98 orders of magnitude small for instance) that its for all intents and purposes a nonfactor.

No - this is not a matter of measurement but principle.

In GR if the measuring apparatus expands as the universe expands then there would be no detectable expansion. That is why all standard textbooks carefully explain that if the expanding universe is modeled by a balloon being blown up, with galaxies modeled as spots on it, then the spots themselves do not expand, this is where that analogy breaks down as a model of the standard theory. It is normally explained that a better model is one in which the galaxies are represented by pennies/cents being glued onto the expanding balloon.

Having said that the question as to whether galaxies (and by extension rulers) really do expand with the universe is also a observational matter. But, as you say, the rate is too small to observe.

One clue that all is not right with the standard model is the Pioneer Anomaly in which the Pioneer spacecraft , now well away from the perturbations of the planets in the solar system, appears to exhibit an anomalous extra acceleration towards the Sun of
(8.74+/-1.3) x 10^-8 cm/sec^-2
this is very nearly equal to cH and could therefore be cosmological in nature.

[Note a simplistic equating of PA = universe expansion does not work as the acceleration is the towards the Sun, you have to think it through consistently in an alternative gravitational theory.]

However we aren't measuring the expansion of space, we are presumably measuring ripples from linear perturbation series from some object (like a neutron star). As they pass us, you should see a kink in the machine.

Not if everything is perturbed equally. Suppose a GW causes a line of atoms to expand in one dimension (Weber detector), in GR the atoms remain of fixed size but an increase in the gap between them passes down the detector as a wave. If however the atoms, and any ruler, also themselves expand then the wave is undetectable by those atoms and ruler. I believe it was this reasoning that led Einstein to deny GWs in his rebutted paper

If nothing is detected as far as I am concerned something is very, very wrong. In fact i'd put the blame on the experiment rather than the theory, so confident I am that it has to exist. If however it persists through generations of LIGO and others, well it would be deeply embarrasing for many of us.

I concur; then we would have to see alternative gravity explanations for the PA!
Garth

 

[EL]

Comment about if "a ruler" expands when the universe expands:

The ruler does not expand simply because there are forces between the molecules in the ruler keeping it together at a fixed size, and the physical laws determining its size do not change with the expansion of the universe! As soon as the space between the molecules are expanding, the forces will pull them back to their initial distances from each other.
The same holds for galaxies.

As Garth said, if the detectors would expand with the expansion of the universe, there would be no detectable expansion, and hence the concept of an expanding universe would be meaningless. The crusial point is that the space expands, but "the physical laws" do not.

 

[Garth]

Comment about if "a ruler" expands when the universe expands:
The ruler does not expand simply because there are forces between the molecules in the ruler keeping it together at a fixed size, and the physical laws determining its size do not change with the expansion of the universe! As soon as the space between the molecules are expanding, the forces will pull them back to their initial distances from each other.
The same holds for galaxies.
As Garth said, if the detectors would expand with the expansion of the universe, there would be no detectable expansion, and hence the concept of an expanding universe would be meaningless. The crusial point is that the space expands, but "the physical laws" do not.

The question is situated on the problematic interface of QM and GR!

Are 'forces', and the force carrying virtual particles fundamental (QM) or are accelerations by which those forces are measured fundamental (GR)?

Normally QM and GR are kept well apart in their respective domains of influence, however when the BB itself is approached the two become of equal importance, so which is to be the most fundamental?

In the normal expanding cosmological solution of GR what exactly is it that is expanding? If it is space-time itself, as demanded by the theory, then what expands with it?

As the Schwarzschild solution for gravitational orbits is embedded in that space-time should not its solutions co-expand?

Also as the Bohr/Schrödinger/Dirac equations of atomic physics are also so embedded then should not their solutions expand?

If, as a consequence, gravitational orbits and atoms together with the physical rulers constructed of those atoms so co-expand with the universe, then surely there would be no detectable expansion?

As I said above in that case gravitational red shift would have to be interpreted by another cosmological effect, such as a secular increase in atomic masses.

Garth

 

[EL]

In the normal expanding cosmological solution of GR what exactly is it that is expanding? If it is space-time itself, as demanded by the theory, then what expands with it?

I have always thought of it like that every distance between every pair of objects expand. Then for objects close to each other the distance between them increases quite slow, and also the force between them usually are quite strong, and hence the forces can keep up with the expansion rate rather easy. For objects further away from each other, the distance between them increases faster and also the forces are weaker, and hence cannot keep up with the universal expansion.
So this is my explanation of why galaxies and rulers don't coexpand with the universe, while larger structures do.

As the Schwarzschild solution for gravitational orbits is embedded in that space-time should not its solutions co-expand?

But the Schwarzschild solution is time independent and does not expand. Of course in reality you have some kind of overlap between the SS and the expanding universe solution, but for small enough systems the SS clearly dominates. Hence neither the solarsystem, or not even the galaxies, coexpand. So it really works in the same way as for forces.

Regards /EL

 

[Labguy]

EL;
The highly accredited site http://www.astro.ucla.edu/~wright/cosmology_faq.html#SS" tends to agree with you.

 

[Garth]

But the Schwarzschild solution is time independent and does not expand. Of course in reality you have some kind of overlap between the SS and the expanding universe solution, but for small enough systems the SS clearly dominates. Hence neither the solarsystem, or not even the galaxies, coexpand. So it really works in the same way as for forces.
Regards /EL

The Schwarzschild solution is a static spherically symmetric solution embedded in a Minkowsian space-time, the GR solution tends to SR as r tends to infinity.

However in the real universe the spherically symmetric solution should be embedded in a R-W metric, whether this makes a significant difference or not is a matter of debate, but as Labguy's link to Ned Wright says the consensus opinion is that it does not. The problem is that an accurate model of the solar system gravitational field has to be embedded in the galactic field which has to be embedded in the cluster field etc. up to the cosmological R-W field. Whether or not the GR cosmological expansion affects the solutions embedded hierarchially within it or not is not clear.

But certainly in the standard model, GR, because of the equivalence principle and constant particle masses, atomic rulers do not expand with the universe. However, it may be informative to consider conformal gravity theories in which they do.

Garth

 

[Labguy]

The problem is that an accurate model of the solar system gravitational field has to be embedded in the galactic field which has to be embedded in the cluster field etc. up to the cosmological R-W field. Whether or not the GR cosmological expansion affects the solutions embedded hierarchially within it or not is not clear.

And THAT is a very good point that must be considered.

 

[EL]

Whether or not the GR cosmological expansion affects the solutions embedded hierarchially within it or not is not clear.

Well my intuition strongly tells me the expansion would not affect the solar system very much. But, of course it's just a guess...
So do you know some papers about this?

 

[pervect]

I seem to recall someone has looked at this issue (fitting the Schwarzschild solution to an expanding universe).

Google finds

http://arxiv.org/abs/astro-ph/0112320

which I've only skimmed, but looks good so far. The solution I find of interest is the one which assumes a cosmological constant (i.e. a vacuum energy density and an associated negative pressure) for "empty" space.

 

[Garth]

The more basic problem was examined by no less than Einstein and Straus "The Influence of the Expansion of Space on the Gravitation Fields Surrounding the Individual Stars" Rev. Mod. Phys. 17, 120–124 (1945).
in which they embedded a Schwarzschild mass in a sphere of vacuum cut out of the background cosmological representative medium. This is because the Schwarzschild solution is for a spherically symmetric and static mass surrounded by vacuum whereas in the cosmological solution the mass is distributed as a homogeneous density.

Note that in the real universe the opposite is true; the density of the Solar System medium is higher than the ISM or IGM. It is not surprising therefore to read W. B. Bonnor's paper: "Local Dynamics and the Expansion of the Universe" GRG Volume 32, Number 6 June 2000 Pages: 1005 - 1007 .

Abstract A brief history is given of attempts to discover whether the cosmic expansion influences local dynamics. Early work, especially the Einstein–Straus model, suggested that there is no influence, but recently the issue has been reconsidered and now seems open.

Garth

 

[EL]

http://arxiv.org/abs/astro-ph/0112320

Nice reading, exactly what I was asking for.
Baker Jr concludes:

In conclusion, it appears that for larger scale structures composed of galaxies and inter-galactic space, the observed increase in the rate of expansion may be an important feature in determining the size of selfbound gravitating systems. For smaller structures like galaxies, globular clusters etc. other mechanisms are presumably dominant.

I find this very reasonable.

 

[Garth]

To go back to the OP - if GW's are detected in the next few years there will be nothing to worry about, however if they are not, and continue to be undetected even as sensitivities increase by OOMs, then the above discussion may be relevant in explaining their invisibility.

Garth

 

[Haelfix]

Im in complete agreement with the link, although there is a bit of a messy continuity problem. When you imbed local metrics into larger ones you really want to require asymptotia to match precisely, the details of this matching can be messy and linear perturbation theory can break down (the root and source of the problem).

At one point you really want to say, at what distance scale do rulers start to coexpand. Then take a local neighborhood of that point, and observe the differential structure. Since we are still working with manifolds here by assumption, everything has to be smooth in some localized coordinate system.

 

[pervect]

I have big semantic problems with this statement. Rulers are the means by which we define and measure distances, so they never expand or contract.

A pair of massive particles will either get closer together because of their mutual gravitational attraction (be bound), or move further apart because of the expansion of the universe (be unbound).

This doesn't have anything, in my book, with rulers expanding or contracting. In fact, we need the rulers to be able to *measure* the distance between the pair of interacting particles to determine whether or not the distance is increasing, or decreasing.

 

[Garth]

I have big semantic problems with this statement. Rulers are the means by which we define and measure distances, so they never expand or contract.

A pair of massive particles will either get closer together because of their mutual gravitational attraction (be bound), or move further apart because of the expansion of the universe (be unbound).

This doesn't have anything, in my book, with rulers expanding or contracting. In fact, we need the rulers to be able to *measure* the distance between the pair of interacting particles to determine whether or not the distance is increasing, or decreasing.

Absolutely; if rulers or clocks are said to vary then that statement makes no sense unless it is also stated against what standard are they measured to vary.

Notice that the size and de Broglie wavelength (inverse frequency) of an atom is inversely proportional to its mass and proportional to Planck's constant, so if either of these 'constants' should in fact vary then that variation itself would require a non-varying standard of mass, length and time.

Note: SCC suggests that such a standard is to be found in a free photon, traveling along a null-geodesic, (after Doppler effects are allowed for). As photons in the CMB and other cosmological radiation 'expand with the universe' [$$\lambda = \lambda_0\frac{R(t)}{R(t_0)}$$]
then as measured by such photons the universe is non-expanding and so local 'rulers'/orbits also do not co-expand. (In the SCC Jordan conformal frame) Cosmological red-shift is caused by a secular increase in atomic masses.

Garth

 

[Haelfix]

Umm, its fairly obvious the notion of 'distance' is parameter dependant. On cosmological scales, they can run proportional to things like the scale factor and so forth.

Either way, we can be precise about this. Are we all in agreement that gravitational waves should be seen as tidal forces on some local measuring apparatus (say an interferometer like LIGO)? Eg this is a case where the principle of equivalence is violated, as we now have acceleration affects that are second order in the geodesic eqn.

I think they call this geodesic deviation.

Anyway you should 'observe' a nonzero tidal tensor, which should make sense assuming we have gravitational wave energy lose from the source object (again here we have to go to much higher order in linearized perturbation series to see the 'loss' from say an oscillating relativistic object).

Assuming the interferometer is large enough, then the kink will manifest itself in errors between signal transfers as the wavefront passes through the arms.

So I still don't quite see how much smaller effects (like tiny distortions from scale factor effects should in any way effect the results gross order of magnitude)

 

[Garth]

So I still don't quite see how much smaller effects (like tiny distortions from scale factor effects should in any way effect the results gross order of magnitude)

Physical measurements are a comparison of an observable with a standard unit. If the standard unit varies with the observable then no variation is detected. This applies both to the size of the universe on the large scale and also to a fluctuation in a GR detector, or even just an atom, on the small scale.

Garth

 

[Haelfix]

Yea sorry I still don't see why that's the case, the distance between the two arms of the interferometer is inconceivably tiny relative to any FRW scale parameter effect, for all intents and purposes you can just assume that length is a 'constant' ruler. Otoh, a much larger tidal 'kink' from a spacetime ripple (which has strictly nothing to do with universe expansion rates) that is traversing it should register as a potential difference in xing times, assuming the accuracy of the lasers is good enough and so forth.

In the previous posts I mentioned that I still think this 'constant' ruler does expand with say the scale parameter, its just so vanishingly, vanishingly small relative to every other local effect (solar system Schwarzschild metric variables, that are also tiny), much larger stress energy tensor asymetries from EM effects and so forth that it makes strictly zero difference for an experiment (and heck LIGO has much larger local effects to deal with and damp out, like the local vibrations from train stations 30 miles away).

 

[Stingray]

However there is a question about the detection of GWs. It is similar to the question, "If the universe is expanding what exactly expands with it? Do rulers co-expand?" If so there would be no detectable expansion. The standard answer to this question is No - rulers do not expand with the universe and therefore they can detect that expansion.
However if this understanding is mistaken, and everything embedded in space-time does co-expand with the expansion of space, then any physical apparatus would not be able to detect GWs either, for they would simply 'wash over' the detectors leaving no signal.

And this is why people have rigorously extended Newtonian elasticity theory to curved spacetime. Everything works out according to the standard expectations, even if these justifications are rarely mentioned.

Arguments of how things are "dragged along" when space "expands" or "contracts," were some of the original reasons that gravitational waves were debated (as the original article here discusses). But these arguments have been conclusively settled for decades.

Anyway, here's a nice starting point for the discussion (though definitely not in the class of rigorous results I referred to above):
http://xxx.lanl.gov/abs/gr-qc/0508052"

 

[pervect]

Umm, its fairly obvious the notion of 'distance' is parameter dependant. On cosmological scales, they can run proportional to things like the scale factor and so forth.

The way I see it, the notion of distance requires two things. One is easy, a knowledge of the Lorentz intervals between any two points, which we get from the metric. The other is more subtle - we need a notion of time, relative to which the distance will always be orthogonal. We then find the distance by integrating the lorentz interval along the appropriate curve of "simultaneity".

There are at least two possible notions of time to use for an extended body. One is to base a time coordinate on Einsteinian clock synchronization (so that all clocks on the body will be synchronized with each other via the Einstein convention). This is the notion of synchronization that I associate with a ruler, though it appears to be little used in cosmology. The other notion of time is to use cosmological time. This gives us the usual notion of "proper distance" in cosmology.

Either way, we can be precise about this. Are we all in agreement that gravitational waves should be seen as tidal forces on some local measuring apparatus (say an interferometer like LIGO)?

I agree with the description in terms of tidal forces.

Eg this is a case where the principle of equivalence is violated, as we now have acceleration affects that are second order in the geodesic eqn.
I think they call this geodesic deviation.

This remark puzzled me for quite a bit. Do you mean that the region of space-time is too large to apply the principle of equivalence because of the curvature induced by the gravitational waves?

Anyway you should 'observe' a nonzero tidal tensor, which should make sense assuming we have gravitational wave energy lose from the source object (again here we have to go to much higher order in linearized perturbation series to see the 'loss' from say an oscillating relativistic object).
Assuming the interferometer is large enough, then the kink will manifest itself in errors between signal transfers as the wavefront passes through the arms.
So I still don't quite see how much smaller effects (like tiny distortions from scale factor effects should in any way effect the results gross order of magnitude)

My main objection was to the idea that rulers changed length. I will agree that the tidal forces due to the expansion of the universe are extremely low and not detectable with any sort of practical measurement.

 

[Haelfix]

"This remark puzzled me for quite a bit."

No nothing like that. Just a general remark about tidal forces, and the fact that we can't just throw away higher order terms in this expansion for this particular calculation as that's where the relevant physics lies.

 

[Spin_Network]

A great recent article details the problems of GW's?
http://www.physicstoday.org/vol-58/iss-9/p43.html#ref
So do Gravitational Waves exist?
This was meant to be placed in General forum!

If one reads the original article above, one can see that there is quite obvious problems with respect to the derived "form and function" of GW's.

This link is very interesting as it deals directly with Einstein-Rosen waves,

http://arxiv.org/abs/gr-qc/0002056

Now I have found that I have to give the E-P-R paper(original thought experiment), and its interpretations, a further reading, as it appears to have evolved historically from around this era, fascinating stuff.

 

[Spin_Network]

If one reads the original article above, one can see that there is quite obvious problems with respect to the derived "form and function" of GW's.
This link is very interesting as it deals directly with Einstein-Rosen waves,
http://arxiv.org/abs/gr-qc/0002056
Now I have found that I have to give the E-P-R paper(original thought experiment), and its interpretations, a further reading, as it appears to have evolved historically from around this era, fascinating stuff.

This just appeared from GW detection team:
http://arxiv.org/abs/gr-qc/0512078

 

[Stingray]

If one reads the original article above, one can see that there is quite obvious problems with respect to the derived "form and function" of GW's.

You still seem to be misunderstanding this. This was a history article. The relevant confusions do not exist anymore. A large portion of the article is even devoted to how Einstein accepted his errors.

 

[Spin_Network]

You still seem to be misunderstanding this. This was a history article. The relevant confusions do not exist anymore. A large portion of the article is even devoted to how Einstein accepted his errors.

It is obvious the article is Historical, the GW's predicted by Einstein in 1916?..the articles content deals with a peer review dispute.

I am not disputing the existence of GW's, but I do not think they will be detected (in forseeable future), for which reasons the Einstein-Rosen-Paradox must have a baring?

Something can exist, and still be beyond detection?

 

[Stingray]

It is obvious the article is Historical, the GW's predicted by Einstein in 1916?..the articles content deals with a peer review dispute.
I am not disputing the existence of GW's, but I do not think they will be detected (in forseeable future), for which reasons the Einstein-Rosen-Paradox must have a baring?
Something can exist, and still be beyond detection?

I don't understand your point. Yes, it was a peer review dispute, but Einstein later agreed (circuitously) that the referee was right.

On what basis are you claiming that gravitational waves won't be detected soon? And what is the Einstein-Rosen paradox?

Their result was only confusing before they realized that their chosen coordinate system was misleading. It's pretty easy to confuse yourself in GR with coordinate problems if you're not very careful. This was not as widely acknowledged then as it is today. People had little experience with recognizing and resolving these issues.

 

[Garth]

On what basis are you claiming that gravitational waves won't be detected soon?

As well as this being a discussion about the interpretation of theory, GWs have not yet been detected and, if this non-detection continues indefinitely, then that interpretation may have to be revised.

As an example this eprint was posted today on the physics ArXiv: Joint LIGO and TAMA300 Search for Gravitational Waves from Inspiralling Neutron Star Binaries

Yet again:

We find no evidence of any gravitational wave signals

Garth

 

[SpaceTiger]

Garth, that paper puts an upper limit of ~50 neutron star mergers per year in the Milky Way galaxy. Are you saying you would have expected it to be larger than that?

 

[Garth]

Garth, that paper puts an upper limit of ~50 neutron star mergers per year in the Milky Way galaxy. Are you saying you would have expected it to be larger than that?

I personally expect GWs to be eventually detected, however, after linking to an article describing Einstein's ambivalence over GWs, the OP question was "So do Gravitational Waves exist?" That question becomes more pertinent the longer their non-detection continues.

Now, as for that paper's expected limit for neutron star/BH mergers. If we, for the moment, accept that short GRBs are such mergers and they are detected about once a month, and they are relatively nearby, though not necessarily in the Milky Way, then we might expect ~ 10 neutron star mergers per year in the Milky Way galaxy. So I am not expecting it to be larger than ~50 , however the upper limit is now at last approaching the expected detection limit and the next few years should be interesting!

Of course, not all NS/BH mergers necessarily produce short GRBs, and there are other sources of GWs as well, so the expected GW detection rate is probably highter than this, but I have no expertise in this area.

Garth

 

[SpaceTiger]

I personally expect GWs to be eventually detected, however, after linking to an article describing Einstein's ambivalence over GWs, the OP question was "So do Gravitational Waves exist?" That question becomes more pertinent the longer their non-detection continues.

I think we're still in the stage where we would have been surprised if we had detected them. LIGO doesn't have the sensitivity to see most of the conventional gravitational wave sources, and those that it can see require a great deal of luck.

Now as for that paper's expected limit for neutron star/BH mergers. If we, for the moment, accept that short GRBs are such mergers and they are detected about once a month, and they are relatively nearby, though not necessarily in the Milky Way, then we might expect ~ 10 neutron star mergers per year in the Milky Way galaxy.

Our understanding of short GRBs is extremely crude and any failure to detect such a signal would almost certainly be due to a failure in those models, not in our theory of gravity. Given the high-quality data that came from PSR 1913+16, the only way you'll see astronomers/physicists seriously questioning the existence of gravitational waves is if we point our detectors at a source that we know is above our threshold of sensitivity and get no detection.

The only observations that tell us anything about gravitational waves have come out in favor of them. None of the direct detection experiments have been able to address GR, they've only been able to put limits on the frequency of certain astrophysical events.