Talk:Tests of general relativity

From Wikipedia, the free encyclopedia

This article is within the scope of the following WikiProjects:

There are plenty of additional tests of GR:

  • Pound-Rebka, which confirmed the existance of the gravitational red-shifing of light.
  • Haefele-Keating Experiment, which used atomic clocks in aircraft to test GR and SR together.
  • GPS, which was found to be inaccurate unless the effects of GR were taken into account.

See the USENET Relativity FAQ experiments page for more information of these and other relativity experiments.

also Gravity Probe B satellite recently launched, to detect "frame dragging". Should have enough data in a couple years. GangofOne 04:36, 12 Jun 2005 (UTC)

Contents

[edit] MERGER OF PAGES

I think this page should be merged with classical tests of general relativity

done

[edit] Precession of Mercury's perihelion

Numbers of 570 and 5600 seconds of arc per century are both correct, and they are quoted equally often. Mercury's perihelion precesses 574 seconds per century with respect to the inertial frame of the Sun. Earth's perihelion also precesses at much higher rate of approx. 5000 seconds per century in the opposite direction. As a result, orbit of Mercury precesses 5600 seconds per century as seen from Earth. --Itinerant1 19:11, 20 May 2005 (UTC)

The 5600 arc-seconds comes not from the Earth's orbital precession, but from its axial precession. Becuase of how the direction the axis of rotation points changes, the right ascension coordinates (which are stated in relationship to the position of the Sun at the vernal equinox) are changing. So the 5600 arc-seconds per century is mostly a coordinate precession.

I have been sitting on an update to the Mercury section that describes this. I will post it this weekend. --EMS | Talk 05:27, 21 May 2005 (UTC)

I already put a comment on the page of Mpatel [1] about this - yes the 5000 sec/cy is due to the precession of the Earth's axis, which changes the Vernal Equinox relative to a celestial reference frame fixed in distant quasars (used to be fixed using stars). The Earth's perihelion is not used as a reference point for many reasons. One is that it is poorly determined, because the orbit is nearly circular. More importantly, it is a secondary quantity; the Earth's rotation axis and equatorial plane, which are used to find the Vernal equinox, are directly determinable by observing stars (for amateurs) and quasars (for the IERS). Because the axial precession is east to west, it adds into the perihelion precession of Mercury; it does not subtract.Pdn 14:15, 21 May 2005 (UTC)

Sorry, I'm new at this, but it wasn't clear what the source was for the numbers related to Mercury orbit precession. It isn't clear to me what the sources of error in the GR contribution are, or where the solar oblateness contribution comes from. I've been reading up on the subject, and it seems like the quadrupole moment of the sun has been a source of ongoing debate. I consider putting a 'citation needed' flag on the section, but I'm not sure if that's proper etiquette. If the original author of this section is still around, I think direct citations in the table would be helpful. And4e (talk) 03:27, 22 April 2008 (UTC)

[edit] Special relativity

I suspect that all the effects listed here could be explained by special relativity + the assumption that mass and energy are equivalent also in the gravitational sense (i.e., the "gravitational charge" is in fact energy, thus the photon feels gravitation although it has no rest mass). Is there a real need for general relativity? I haven't found any reference that could convince me of that need. --Philipum 11:41, 31 May 2005 (UTC)

If one calculate the perihelium of the mercury using SR instead of GR, one get a third of the GR-value.
(comment belately added misformatted by careless Swedish anon 194.47.215.48 on 16:18, 20 March 2006)
As one might suspect simply from the fact that this anon has been blocked for vandalism, this above comment is incorrect in several respects. Philipum, if you are still reading this, you didn't mention your background, but assuming you have a first year graduate knowledge of math/physics, try Will's review and then see citations in that paper. If that doesn't work for you, try his popular book cited by Pdn back on 6 Jun 2005 (see next section).---CH 15:54, 21 March 2006 (UTC)

[edit] reply to Philipum

There have been innumerable attempts to mimic GR or replace it with flatspace theories. Early ones are discussed in the text on Relativity and Cosmology by H. P. Robertson and T. W. Noonan, (Saunders, 1968.) Other variants are covered by Clifford Will in Was Einstein Right?: Putting General Relativity to the Test: New York, New York, U.S.A.: Basic Books, 1986, or Theory and Experiment in Gravitational Physics also by Clifford Will (ISBN 0521439736, Cambridge University Press, 1993.) Also see my comments on clock comparisons as contrasted to photon frequency change. A simplistic concept that photon frequencies change when the photon travels "uphill" or "down" in the potential fails to explain clock synchronization results. Pdn 02:49, 6 Jun 2005 (UTC)

[edit] Picture

Shouldn't the left and right in the diagram caption be switched? (Question was apparently from User:Shanedidona who did not sign.) (else the light signal got deflected)

No. The apparent position is shifted away from the Earth and the straight-line prolongation of the observer's line of sight is the fictitious position! This is an easy one to mix up. The light is bent towards the axis of symmetry from earth center to observer but the illusion is that the source is deflected away from that line. Pdn 01:05, 6 Jun 2005 (UTC)

[edit] The illustration does not make sense

Image removed

It seems to me this picture is wrong. The curved surface with the sphere weighing it down is supposed to represent the space within which the rays of light move. But the rays of light are distinctly shown moving above that surface. Michael Hardy 02:08, 6 Jun 2005 (UTC)

You seem to be right, but the picture is perhaps only intended to be suggestive, with the bent sheet indicating a deformation of spacetime, but not necessarily a surface within which the rays travel. In fact, that surface in three dimensions is a plane, though the metric within it is non-Euclidean. Yet the timelike part of the metric is needed to give the correct answer, I believe, so a faithful diagram would be hard to construct. I don't see this suggestive picture as being so bad, but the Sun looks more like a planet. It is probably not hard to obtain an actual diagram of the outward stellar displacements; I've seen one but not sure where. Of course, then you have error circles and you do not see the deflected and undeflected ray paths. Pdn 02:49, 6 Jun 2005 (UTC)

I understand all that, but if the light is not showing moving within the surface, then the surface will not explain anything. I understood it only because I already knew this before I saw the picture. Michael Hardy 02:50, 6 Jun 2005 (UTC)
Then perhaps this illustration without the rubber sheet would be better. Or even better would be a 2-D view of this effect, which I have seen. If you like, we can ask Cleon to look for and/or make such an illustration. He seems to have knack for both finding internet resources and for doing computer graphics. --EMS | Talk 14:31, 6 Jun 2005 (UTC)

[edit] Clean up

Hi all, shouldn't the list of topics at the beginning of the article correspond exactly to the sections below? E.g. periastron precession is listed, then light-bending,... and the sections should be "Periastron precession", "Light-bending" and so forth.---CH (talk) 15:54, 29 August 2005 (UTC)

Hi, I made a first start at a major revision. I grossly underestimated the amount of work it would take, so I will have to come back to it, hopefully tonight. –Joke137 19:26, 22 September 2005 (UTC)

There it is. Let me hear your comments. It's still pretty technical, and the cosmological tests and strong-field tests need real improvement, but there are only so many hours in the day... the language probably gets too technical in parts, as well, although I have tried to keep it as comprehensible as I can in a quick draft like this. –Joke137 03:27, 23 September 2005 (UTC)

Only had time for a quick glance but noticed two or three misstatements (which I hastily corrected):
  1. other than supermassive black holes do not occur only in centers of galaxies, in fact stellar mass black holes and neutron stars are known to sometimes be ejected from galaxies do to proper motion imparted (apparently sometimes by the "kick" of massive directed gravitational radiation!) at the time of formation,
  2. Brans-Dicke theory is certainly not 'simpler' than gtr, in fact it is more complicated than gtr, because it's description of gravitational field includes both Riemann tensor plus an additional scalar field, and it does have an adjustable scalar, ω (see for example MTW); this parameter is contrained by observations to be quite large, but it is certainly adjustable in the standard sense of that word, in fact it is an iconic example of an adjustable parameter.
  3. There is not one Mach principle but many Mach principles, most not very well-defined, or if well-defined, not defined in a covariant manner. This makes it difficult to say that a given metric theory of gravitation is or is not 'compatible with Mach's principle'. Yes, such claims have appeared in the literature, but wiser heads (notably the late Hermann Bondi) debunked them. One can argue that general relativity may be compatible with some Mach principles, and Bran-Dicke is not compatible with other Mach principles.
One other thing: when I have time, I plan to write more detailed articles on the four classical tests, and also on Lense-Thirring precession, deriving the formula at least for gtr and perhaps, using PPN formalism, in greater generality.---CH (talk) 21:59, 23 September 2005 (UTC)
  1. Agreed.
  2. Yes, I misstated that. I ought to have said no dimensionful parameters. Again, some people thought it was simpler, whatever you might think, since it doesn't contain a coupling which has to be adjusted to an artificially small value.
  3. OK, but the comment was made for its historical interest. I'm not making any claims about how Brans-Dicke theory is seen now, only in 1959 when it was published.
–Joke137 22:10, 23 September 2005 (UTC)
Sorry, I don't follow. What theory has a "a coupling which has to be adjusted to an artificially small value"? Which coupling constant (yes?) do you have in mind? Can you rewrite what you recently added to reflect the verifiable facts (as stated in the previous version, which I wrote) that it is not universally accepted that any theory is Machian or even what "Machian" means? That it is certainly not universally accepted that Brans-Dicke is "simpler" than GTR? TIA---CH (talk) 18:41, 21 October 2005 (UTC)

GR does. The Planck mass is much larger than, say, the mass of the proton. This is called the hierarchy problem and it keeps a lot of people up at night, still. I think that "arguably simpler" and "a version of Mach's principle" is adequate qualification. The point of the article is not to describe Mach's principle or evaluate which theory of gravity is more elegant. It is to describe tests of general relativity, and an important point to be made was that there were competitors to GR that people took seriously, and the sentence as it is does a decent job of that, I think. –Joke137 22:37, 21 October 2005 (UTC)

[edit] reds need blued

binary pulsar and parameterized post-Newtonian formalism. I will take a stab at the second one day, but if someone else can look at the first, I would appreciate it. It's really abominable that nobody has written an article on the Hulse-Taylor binary pulsar yet. –Joke137 16:42, 22 September 2005 (UTC)

[edit] Tests possible but not realized yet

How about tests of GR that are mentioned in popular books but not realized yet? Will you refer to them in this article?

More specifically, I was sure to remember something about three-body effects, especially in the Moon orbital motion, and didn't find anything here in the Wikipedia. Later I've found out a remark about them in (Misner, Thorne, Wheeler Gravitation, 1973) (along with references to (Brumberg, 1958), (Baierlein, 1967), (Krogh, Baierlein, 1968), (De Sitter, 1916)), where it is said that they are about 10 to 100 cm, and they would be hard to find due to peculiarities in Moon orbit. Nevertheless it would be rather curious to read about these effects here, because they show the level of our knowledge and limits of observations; and they are also one of most beautiful consequences of GR to my mind. (Misner, Thorne, Wheeler, 1973) says that these effects appear already in the PPN case, and they depend not only on γ, but as well on such PPN parameters as β, β2, ζ, Δ1, Δ2 (in their own notation). (Sorry for not quoting (Misner, Thorne, Wheeler, 1973) literally, but I don't have english version of this book.) -- fir-tree 83.237.184.72 05:59, 24 November 2005 (UTC)

This is a belated response, but I think you may be referring to the Nordvedt effect constrained by the Lunar Laser Ranging Experiment. –Joke 04:32, 2 February 2006 (UTC)

[edit] Lenard's motivations for attacking Einstein

Hi, 24.255.46.150 you removed the characterization of "antisemite". (And you know Eugene Dynkin?). You might be correct that this word is merely a distraction in the context of this article, but the characterization is not unfair. Lenard's poisonous political views are very well attested from his own words and from contemporary evidence. See for example Berlin in lights: the diaries of Count Harry Kessler, 1918-1937 (translated and edited by Charles Kessler) and the biography Suble is the Lord by Pais. (Kessler knew Einstein socially, had similar political sympathies, and followed the controversy with interest.) It does seem fairly clear that antisemitism did play a major role in motivating the famous denunciation of Einstein in an open letter signed by Lenard and other scientific and cultural figures in 1920s Berlin. Have you read this letter? You should be able to find the text on the web someplace, and it might be wortwhile finding an appropriate place to quote from some of it. (If memory serves, the poster announcing the first anti-Einstein meeting in Berlin is reproduced in volume I of Kessler's diaries.) See also Johannes Stark for a bit more information.---CH 05:35, 19 February 2006 (UTC)

[edit] Death by a Thousand Edits?

Hi all, I have tried to draw attention to a fundamental drawback of wikis: even well intentioned editors who are inexperienced writers or otherwise insensitive to

  • the requirement to stick new additions into a place which makes sense, or else to adjust adjacent paragraphs to maintain a graceful and clear flow of ideas
  • long-range order generally

can degrade the quality of an article by making careless or ill-considered edits. Case in point: someone stuck on Also, they should propagate via space with the speed equal c to 1% accuracy to the end of a preexisting paragraph describing quadrupole radiation, just after the sentence Thus, although the waves have not been detected, their effect is necessary to explain the orbits. I hope it is obvious why I reverted to the previous version and ask that editor to either reconsider adding that at all (wouldn't it make more sense to add a new paragraph mentioning the Kopeikin controversy and linking to the Speed of gravity article?), or else to find a better way of adding this. TIA ---CH 16:21, 14 March 2006 (UTC)

[edit] Students beware

I had hoped to greatly improve this article and been monitoring it for bad edits, but I am leaving the WP and am now abandoning this article to its fate.

General relativity attracts many cranks and despite the good work until now of various knowledgeable editors in removing slanted information, misinformation, and in reverting vandalism, it is likely that at least some future versions of this article will be seriously misleading, so I'd urge students to be cautious in using anything they might find here (or elsewhere on the web).

Good luck in your search for information, regardless!---CH 18:07, 1 July 2006 (UTC)

[edit] Another test...

... is reported here - the Shapiro effect. MP (talk) 16:38, 20 September 2006 (UTC)

[edit] New reference format

I've tried to incorporate a new reference format to match the one I'm using in general relativity. It allows one to skip back and forth between the reference and the main text in which the reference is embedded. Hope this is ok. MP (talk) 19:03, 22 September 2006 (UTC)

[edit] Hafele-Keating experiment ?

Shouldn't the Hafele-Keating experiment be mentioned somewhere in the article, as it did test an effect of general relativity, namely the modified time dilation effect ? MP (talk) 17:07, 4 October 2006 (UTC)

My mistake - it is mentioned, but the references for it are not given. Will add them in. MP (talk) 17:09, 4 October 2006 (UTC)

[edit] Relativity Wikiproject

I've suggested at the proposed wikiprojects page that a relativity wikiproject be created. If interested, you can add your name to the list and check out the plan for the project at WikiProject Relativity. MP (talk) 13:08, 29 October 2006 (UTC)

[edit] Proposed addition; Winterberg's experiment

In editing GPS, I pulled the following text out:

Friedwardt Winterberg predicted in 1955 that when observed from the Earth's reference frame, satellite clocks would be perceived as running at a slightly faster rate than clocks on the Earth's surface [an experiment that could demonstrate the effect of general relativity.]

While its interesting, and was in GPS because they experience general relativity as his experiment might have, it doesn't really seem appropriate to keep in the GPS article. I thought it might be a better fit here. Was this the first such proposal to demonstrate general relativity? If not, maybe this isn't even notable. Thanks! - Davandron | Talk 04:33, 3 March 2007 (UTC)

[edit] No test on Earth

I think the tests of general relativity are really extremely poor. The theory predicted a very small value for very perturbations in the orbit of mercury exactly of the same size as stated by astronomers in the 19th century. The gravitational deflection of light is predicted only twice as large as by Newton's laws (also expected by Einstein in 1911). The gravitation red shift can't be reagarded as proof, since the predicted value is exactly the same as derived by special relativity. All proofs are based on observation on far astronomic objects and can't be reproduced in experiments on Earth. 84.169.248.150 19:05, 31 March 2007 (UTC)

General relativity, like any other scientific theory worth its salt, predicts phenomena which have been tested extremely accurately (check out the references and links). No one is claiming that the theory is 'proved' by any of the experiments you mentioned above (or any other experiment, for that matter). Experiments only lend support (heavy or otherwise) to the theory. Also, nothing in the philosophy of science says that Earth-based experiments take precedence over non-Earth-based ones. Think about this last sentence and you will realise how ridiculous your last comment was. Cheers. MP (talk) 19:17, 31 March 2007 (UTC)
Since we live on Earth and not anywhere else in the universe, a theory that can't predict anything that can be directely observed on Earth looks not very useful. Therefore, I think a theory not predicting anything proven in Earth-based experiments are not proven to be useful. 84.169.248.150 19:45, 31 March 2007 (UTC)
We can't directly 'observe' many things. If by 'observe' you mean 'experimentally test', then you are taking a rather narrow view of how we perform experiments. Are you suggesting that the coronal mass ejections emitted by the Sun, and photographed by the SOHO satellite, are not real, as we can't see the ejections from Earth ? As you seem to have missed an important point in my last comment, let me reiterate again that no theory can ever be proved by any experiment. MP (talk) 20:30, 31 March 2007 (UTC)
This discussion should be continued offline, as it is not relevant to the article. Michaelbusch 20:37, 31 March 2007 (UTC)

[edit] Dipole gravity test

Hey Silly rabbit,

Is there any reason you are deleting the "dipole gravity test article"? —Preceding unsigned comment added by 70.244.206.145 (talk • contribs)

In my edit summary, I indicated that there are WP:COI and WP:OR concerns. These have already been raised on the issue of much the same content in this AfD discussion with another anonymous contributor (also from the Austin, Texas area) and this discussion as well. Other issues include WP:Notability and WP:Undue weight, since only one reference is provided, and the topic lies far from mainstream thought on the issue. Finally, the exact same material was added to multiple articles: Mach's principle, Gravitomagnetism, and Dark matter. If this research truly does meet the criteria for inclusion in the encyclopedia, and is significant enough to be duplicated like this, then it obviously must deserve its own article. But the AfD discussion above already addressed this. Silly rabbit 09:21, 25 June 2007 (UTC)
Agree with sillyrabbit; the dipole gravity discussion is completely off-topic for the tests of general relativity page, since it is not a test of general relativity, and it sounds like at best marginally relevant to the other topics where it was added listed above. If it should be on wikipedia anywhere (and it's not clear it should), it belongs in the article on alternatives to general relativity. Geoffrey.landis 15:27, 26 June 2007 (UTC)

[edit] Deleted subsection on angular deflection by stars

I deleted the material since it was a very long table with no context, and more important, no citation. Without a better reference than "compiled by AJB", it counts as original research. Perhaps the editor who originally inserted this table (AJB?) or another editor could put back in a much shorter table, with more context and explanation, and with a citation of a reliable source. I would also suggest fewer decimal places. -- Spireguy (talk) 17:22, 16 February 2008 (UTC)

Thanks for putting in a citation, but the table is still way too long and still without much context. Why are these the "interesting" entries? Why is there a need for more than a few of these, in a general encyclopedia article? And I'll repeat, most of those decimal places are unnecessary, if not completely misleading. I'm not going to edit this right now, since I strongly dislike edit wars. But a response here on the talk page to my concerns would be helpful. Thanks. -- Spireguy (talk) 03:59, 18 February 2008 (UTC)

I agree with Spireguy here; this table seems to be out of place. I deleted the table, but left the citation.
In fact, from the text I can't even figure out exactly what you're referring to, and I don't have a copy of the reference cited. The angular position of the stars cited in the table are deflected by the gravitational field of what object? Geoffrey.landis (talk) 18:57, 19 February 2008 (UTC)