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

A question on Dark matter

  1. May 29, 2010 #1
    I know that Newtonian mechanics predicts dark matter due the the fact that the sun (and other stars) is orbiting around the centre of the galaxy much faster than expected

    But i was just wondering if Einstiens general theory of relativity predicted this aswell?
    Does it "say" there is as much dark matter as Newtons theory?
    or are they about the same?

    Just wondering
    Thanks : )
  2. jcsd
  3. May 29, 2010 #2
    They don't "predict" dark matter... they predict a certain result, which does not match what people observe. So people have come up with the idea that there must be hidden matter so make our observations fit the equations. The other possibility is that there is no dark matter and the equations are not perfect yet.

    However there is some other evidence for DM besides just the velocity of stars on the edges of galaxies.
  4. May 29, 2010 #3
    Do the 2 theories calculate the same orbiting velocties?
    i.e. as a result of having the same orbital velocities conclude that there is the same amount of DM in the universe (The Galaxy)?
  5. May 29, 2010 #4
    You probably want to look at http://www2.phys.canterbury.ac.nz/~dlw24/" [Broken]'s work:
    Last edited by a moderator: May 4, 2017
  6. May 29, 2010 #5
    Oh this is a prof from my uni .. Geuss i should go talk to him :O
    Thank you guys
  7. May 31, 2010 #6


    User Avatar
    Science Advisor

    Yes. Wiltshire's work has nothing to do with that.
  8. May 31, 2010 #7
    Thanks Ich.. i was looking for ages through his papers and journals at uni and couldn't find anything

  9. May 31, 2010 #8


    User Avatar
    Science Advisor

    Wow - quick reply. I had to google for cesiumfrog's http://www2.phys.canterbury.ac.nz/~dlw24/universe/summary.html" [Broken], and it's about something completely different, and very speculative, too.
    Gravity in Galaxies is relatively weak, so it doesn't matter for all practical purposes whether you use GR or the Newtonian approximation. Except that the latter is much easier to apply.
    Last edited by a moderator: May 4, 2017
  10. Jun 1, 2010 #9
    That's a great idea! Please do and tell us what you learn.

    The OP's question was to what extent does the inferred dark matter ratio depend on whether the observational data (galactic luminosity and redshift curves) are interpreted under the framework of Newtonian physics or GR. Ich, can you site other sources that have addressed this question? The fact is that there has been a difficulty in identifying the theoretically correct averaging procedure to describe grainy matter distributions in GR properly.

    Do you have any evidence to support your claim that the Newtonian approximation does not make any relevant practical difference from GR here?
  11. Jun 1, 2010 #10


    User Avatar
    Science Advisor

    There has been a paper a few years ago which claimed a difference between GR and Newtonian analysis. It has been shown to be errorneous. I forgot both the paper and the replies, maybe someone else remembers it?
    Sure. The total mass of a galaxy is ~ 5*10^11 Msun, with a Schwarzschild radius of ~0.1 Ly. So we're talking about a potential of order ~10-6 to 10^-5. OTOH, we have velocity measurements with an accuracy of order ~0.1. No way for GR to produce significant corrections.
  12. Jun 1, 2010 #11
    I couldn't get hold of him, he's vbusy (which makes sense :P )
    But, i read his journals and papers in the library and couldn't find anything refering to this? Not to say he didn't as all his journals werent there.. I'm not sure.. Hopefully he will be my PhD supervisor, when i answer this question tehehe

    So I will let you know, when i solve it xD
  13. Jun 1, 2010 #12

    D H

    User Avatar
    Staff Emeritus
    Science Advisor

    To add to what Ich already gave:

    The orbital velocities of stars, clusters, and satellite galaxies about a central galaxy should be tiny. Instead, they are just small (compared to c). For example, NGC3198 has orbital velocities that are about 150 km/s from 4 to 30 kpc from the central core. 150 km/s is 0.0005 c, which is too small for any significant relativistic effects to be involved.
  14. Jun 1, 2010 #13
    When you put it that way, I understand
    Thank you :)
  15. Jun 1, 2010 #14


    User Avatar
    Science Advisor
    Gold Member

    Compelling evidence for dark matter is provided by the bullet cluster papers. That is about as good as observational evidence gets. Zwicky noticed the missing mass problem almost a century ago. Virial theory confirmed his suspicions. MOND is merely a drive by shooting victim. Dark matter is not going away any time soon.
  16. Jun 1, 2010 #15
    Can any of you experts tell me the correct way to use GR to model a galaxy? Obviously one does not use a Schwazschild metric except in a trivial worst approximation, since that would treat the bulk of the galaxy as a vacuum. Since GR is nonlinear, we know that the field of two stars is not quite simply the sum of the individual fields of either star in isolation. Do any of you know how much difference this makes? Have any of you seen the calculation attempted?

    Isn't it obvious to every thinking person that one will arrive at different numbers for the quantity of dark matter depending on whether one interprets that data using one theoretical framework or another subtly nonequivalent framework? So how is it invalid (nay, threatening) to investigate how much those numbers differ? You know, actually bothering to check quantitatively how good the approximation is? I don't think performing such exercises warrants the label "speculative", if anything then "speculative" would be asserting a particular outcome of such exercises without bothering to have anyone perform them carefully. I'm certainly not claiming that Newtonian physics is a terrible approximation: the papers I've seen only reported about a factor of 2 discrepancy, which should be perceived as no threat to the dark matter dogmatists. I'm only advocating placing more weight on nuanced application of scientific method (in this case, reading of published calculations that exist on the topic) than on simple hunch.

  17. Jun 1, 2010 #16
    What you do is to expand things out into a power series

    GR = Newtonian + (something) x + (something) x^2 + (something) x^3 + .....

    Then you look at what x is and how big it is, and x is v/c and for the purposes of galaxy rotation it's not big enough to make any difference. The good thing about this sort of argument is that could apply even if GR is wrong.

    Also if you show that x is big enough to be important, but x^2 isn't, then you end up with what is called the post-Newtonian formalism and that happens to be linear.

    Except that you won't. As long as your theory of gravity approximates Newtonian gravity then it's not going to matter. Now you could be in a situation where your theory of gravity *doesn't* approximate Newtonian gravity in which case you get into the world of MOND models.

    What you can show is that as long as things are "subtly" different, things aren't going to matter. Things have to be very different for things to matter.

    People do that. The trouble with that is that it can be done in two paragraphs and it's so easy to do that it's not worth writing a paper about it. Now what *would* be worth a paper is if you could stare at the basic argument and show that it's flawed.

    Be my guest :-) :-) :-)

    That's interesting since there is a whole series of papers that argues that Newtonian physics *is* a terrible approximation to galaxy rotations. If you want galaxy rotations to be a gravity effect, you have to argue that Newtonian gravity is wrong at galactic scales. This isn't a crazy thing to do. I mean, we know that Newtonian gravity is very wrong at cosmological scales, and we know it's pretty good at solar system scales.

    You don't want nuance. This is something that you want to be blunt about. You don't want complex math, you want a simple straightforward argument, and that turns out to be that v/c << 1.

    Also the calculation to check if GR matters or not is a trivially simple one, and it's so simple that no one is going to publish a paper about it. Now if you can think of a reason why the basic argument is wrong, then *that* would be worth a paper.
  18. Jun 1, 2010 #17
    Also astrophysicists like these sorts of simply arguments because they are quick. If it turns out that GR makes a difference then you are going to be spending the next months/years/decades figuring out the details.

    So a five minute argument that says "it's not going to matter" is something that is very useful. It also tells you that there are three possibilities 1) dark matter 2) some theory of gravity that is totally unlike newtonian gravity or 3) some fundamental problem with the quick argument.

    The problem with these arguments is that they never quite get published because it's too easy and not worth publishing.
  19. Jun 1, 2010 #18


    User Avatar
    Science Advisor

    It strikes me as odd that you rather call us dogmatists than consider the possibility that the problem really is easy enough to be settled with a few numbers. There are weak fields (~10^-6), slow velocities (~10^-3), so perturbation theory will work. That means that the respective corrections are themselves of order 10^-6, 10^-3.
    That's not magic, it's science. And if you don't like the result, because there's a dogma or two in your thinking that DM must be non-existent, well, then come up with a better argument. This one won't work.
  20. Jun 1, 2010 #19
    That was good for a chuckle, thanks. This came immediately to mind:
    [PLAIN]http://www.acsu.buffalo.edu/~dpadgett/ackbar.jpg [Broken]

    I find the dark matter result disturbing, but as Ich has said it is because of previous dogmatic thinking on my part. The more I learn about it (often in such forums as this) the more I understand how limited the options are. I would put some money on WIMPs, but in general, if DM didn't exist it would be almost inconceivable at this point. GR is just too good, and observations all seem to agree on the basic issues.
    Last edited by a moderator: May 4, 2017
  21. Jun 5, 2010 #20
    I apparently can't repeat this enough: I do not think such a thing. I don't understand how you have misread so extremely.
  22. Jun 5, 2010 #21


    User Avatar
    Science Advisor

    You tried to explain our refusal of such ideas with psychology. Called us dogmatists. I think it's understandable that I mirrored that word.
    But it's ok, so I misunderstood something.
  23. Jun 5, 2010 #22
    Ich, I'll reiterate my previous post.

    If you try to measure a quantity using the same data but two (not exactly equivalent) theoretical frameworks, you will expect to get two not exactly equal results. This much is surely obvious to you, but tell me if you need an explicit example.

    Now, my understanding of the OP is that it raises the quantitative question, how much do those numbers differ in the particular case of whether full GR is used to interpret the dark matter evidence. Hence, one would be both technically incorrect and failing to answer the question if one were to react insisting there were literally zero difference.

    Now, a very simple argument has been put forward to say that the difference would be (bounded above by) less than one percent, and (twofish-quant might find that arguments which are too little to warrant a separate research article still often tend to be mentioned in passing, and detailed in textbooks or review articles) indeed I understand this to be the consensus view among astronomers. Fine. Even if that argument is flawless, I still think it is worth asking about the actual difference, because I think the process necessary to actually approach this problem fully within GR must be very interesting (indeed, twofish-quant implicitly agrees with this by guessing it could take decades of work for the details to be worked out the first time). But it would seem the expert-reviewed GR literature does already mention attempts to answer this question, so the obvious thing to do is to cite those for the OP.

    As it happens, the result I cited describes a difference of more than 1% (and less than 100%; I don't see why you keep referring to "such ideas" like complete nonexistence of dark matter). Since this conflicts with the very simple argument used outside of the hard core general relativity field, we could suggest that experts on GR have not heard the basics that are known of GR. Alternatively, we could suggest that the a person whose qualifications include presiding for a society of professional GR and gravitation researchers might have already encountered (and understood the possible flaws in) such a simple argument. Since this is not the same area of GR as my expertise, I haven't a great deal to add.
    Last edited: Jun 5, 2010
  24. Jun 6, 2010 #23


    User Avatar
    Science Advisor

    Exactly. You expect two not exactly equivalent results. The size of "not exactly equivalent" can be estimated by the procedures that twofish-quant and I applied. "not exactly equivalent" is not 10% (measurement uncertainty), and it's definitely not a factor of 5. OP answered.
    Yes. Nobody did so.
    I don't think so. It would be a waste of time, nothing more. You'll always do the perturbative approach in such a situation, anything else would be ... insane. Like doing the full quantum mechanical approach in automotive crash test simulations.
    Well, at least I said that I remembered one attempt to explain DM with GR, and its debunking. I couldn't find the references after a quick search, so I thought it'd be ok if I give an OOM estimation with the same result. Maybe someone with better memory will give you the links.
    No, it doesn't. It says:
    That's nothing to do with the ratio of visible to dark matter in galaxies.
    Or, as a third alternative: There's no conflict, because Wiltshire is not talking about galaxy rotation. He's talking about cosmology. (a citation: "Here we take a “dust particle” to be of at least the scale of statistical homogeneity, 100h−1 Mpc or somewhat
    larger". Much bigger than a galaxy.)
    Maybe you noticed that I gave you numbers related to the potential, even if the deviation due to velocity is bigger. I did so because it is exacly the condition that the potential be << 1 that Wiltshire violates when talking about cosmological averaging effects. The "simple results" do not necessarily apply cosmologically.
    I'm convinced that there are no vital averaging effects in cosmology, too, but that's something completely different.
  25. Jun 6, 2010 #24
    Ah, you are correct. It is by analysing other categories of data in the framework of GR that he is dismissing 40% of the dark matter otherwise normally expected to exist in the universe. Looking at PRD 78 084032 (2008), it seems he has not yet tackled rotation curves, but outlines reasons to not presume those to be inconsistent with the same result.

    We can agree to disagree on whether it is interesting. For what it's worth, though I wouldn't use QM for every routine crash test, I think it would be interesting to see QM applied once to a giant system (such as a computer - in the context of MWI this might much better test/demonstrate emergence of classicality), especially a mysterious system with proposed quantum explanations (for example, "a person operating a computer" - you can see how that would get the attention of the followers of Penrose). In the case of galaxies, the perturbative approach makes implicit assumptions about the surrounding geometry, the validity of which are questioned in the paper I just mentioned.
  26. Jun 6, 2010 #25
    Didn't I say this was a trap?! No one listens to the Mon Calamari. :cry:
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook