PDA

View Full Version : Galactic rotation curves / speed of gravity

DyerWolf
2007-Jan-09, 07:48 PM
I know the speed of gravity hasn't been determined yet, and that there is a strong presumption that it propigates no faster than the speed of light...

but

If gravity propigated faster* could that explain the difference between the observed and expected curves seen in galactic rotation?

*(whether instantaneous, as Newton thought - or some speed between light speed and instantaneous)

Fazor
2007-Jan-09, 08:43 PM
I get confused by people asking what the "speed" of gravity is (mostly because I am not as studied on the subject as many of the members here). But the way I currently understand it is that gravity is a result on the distortion of space time by obects with mass. The famous (but perhaps too simplified) analogy is a bowling ball resting on a rubber sheet pulled taunt (think trampoline) how the sheet will begin to bow and streach under the bowling ball's weight. In that case, gravity is not a force that has a "speed" as it is not traveling, it is just a distortion of space.

But then you have theoretical "gravity waves" from my understanding these have not been detected yet but are thought to exist. I really don't know enough about them to comment on what they are or if they have a speed.

But as for a general gravity well (or whatever the propper term is) associated with a large object of mass, such as the earth, sun, moon, whatever, I wouldn't think gravity would have a speed. it's just there.

Altho as far as i see it, the only way a speed of gravity in this situation would be important (or measurable) is if gravity were to suddenly change. like, if it could be turned on and off. but because gravity is dependant on mass, and the mass isn't likely to suddenly change, then the radius of the gravity well will always stay (more or less) the same. And gravity will act upon a body if it's inside the radious.

Hope that at least sort of makes sense

DyerWolf
2007-Jan-09, 09:34 PM
I guess that's kind of the question, isnt it? I'm pondering the amount of time it takes for a change in the location of one massive object to affect another object.

As you point out - one test requires a sudden change, ala this thought experiment:
...if you were able to wink the sun entirely out of existence:

-If gravity propigates at the speed of light, it should take about 9 minutes for the earth to stop orbiting the no-longer-extant sun and begin wandering out into the cosmos.
-If gravity propigates instantly, the minute you wink out the sun, the earth should immediately go a'wandrin...

Put it another way:

Imagine a binary star system with the stars ten light hours apart. The stars are in orbit around their common center of gravity which is somewhere between the two stars. Would an arrow pointing from the center of one star through the binary system's center of gravity point directly at the other star (the effect if gravity propigates at the speed of light) or would the pointer "lead" the other star (the effect if gravity propigates faster than the speed of light)?

crosscountry
2007-Jan-09, 09:58 PM
I know the speed of gravity hasn't been determined yet, and that there is a strong presumption that it propigates no faster than the speed of light...

but

If gravity propigated faster* could that explain the difference between the observed and expected curves seen in galactic rotation?

*(whether instantaneous, as Newton thought - or some speed between light speed and instantaneous)

My thought is that no it would not affect nor explain the difference between observed and expected curves seen in galactic rotation.

Once the gravity from all other objects reaches you you will follow the laws of physics (also before:doh: ). So, even if gravity acted instantaneously our galaxies would rotate faster than expected.

Other than black matter some explanations have been given. One is that at long distances 1/r^2 doesn't hold and the g-law takes another form. Black matter is of course the preferred theory because it doesn't change any of our currently known laws of physics. - I say currently because barely over 100 years ago those laws all changed.

publius
2007-Jan-09, 10:32 PM
Yes, gravity does have to move. You've got one fixed lump of mass dropped into flat space time. Let that mass move relative to your origin. The space-time distortion, it's little "bowl" has to move relative to you. But, move with the mass, and that bowl becomes static. However, if that mass is undergoing acceleration, it gets even better. There is no non-accelerating frame against the flat background in which that mass is not moving any more.

And then put two masses in that space in relative motion. There is no coordinate system you can choose where no mass is moving.

Still thinking of the bowl, the little depression in the fabric, imagine what happens as that moves. It's a little ripple in that fabric that moves along with it. The speed of gravity is how the *changes* in the space-time propagate outward from the source causing them.

That's the speed of gravity. In GR, those changes are stipulated to move the same way EM field changes propagate against that space-time background. The *local speed* of gravitational changes is c. That is only a local speed, mind you, not a global coordinate speed. Roughly, gravitational changes get from A to B at the same time light would. Watching something deep in a black hole well, the coordinate speed of light is very slow to us outside, and so gravitational changes made down deep would take a long time to propagate outwards.

And conversely, if we were down deep in that well ourselves, we'd see the coordiante speed of light as very fast high up in the well. Graviational changes made out there would appear to propagate very fast.

That's the maddeningly subtle thing about gravity. Light and EM work against a background (to first order) determined solely by something independent of them. Gravity however, works against a background it itself determines. So when a change propagates, it changes how it propagates as it propagates. :) Emerson comes to mind:

"They know not the subtle ways/
I keep, and pass, and turn again.....

They reckon ill who leave me out/
When me they fly, I am the wings"

That poem is called "Brahma", but those lines above could be (GR) Gravity speaking to us.

That poem in its entirety would serve those well who think "reality" is only our science, too, but that's another thread I was reading that I should leave alone.

-Richard

publius
2007-Jan-10, 07:55 AM

The galactic rotation curve fails the Netwonian instantaneous propagation. Most simple models of this are based on Newtonian gravity as it is.

And so the question is, why does Netwon work so well with instantaneous propagation -- why does gravity appear to work instantaneously (for the most part) when GR says it propagates at c. The answer can be found in several threads here by searching on "speed of gravity", "gravity propagation", etc, etc. Put "Carlip" in there and that ought to zero in on the relevant threads.

The short answer is GR gravity has acceleration dependent terms in it as well as position and velocity and it does a darn good job of cancelling out the propagation delay. It's only changes in acceleration in the light-travel meantime that make gravity "miss".

A radiation field is a consequence of this missing due to propagation delay. That's what gravitational radiation is, actually, and observations of binary pulsars show agreement with GR's prediction about this. If gravity propagated faster than c, there would be much less radiation. If it truly propagated instantaneously, there would be no radiation at all.

-Richard

kzb
2007-Jan-10, 01:34 PM
I'd say it doesn't matter, to the question, whether it's light speed or faster. You would need significant changes in the mass of the galaxy over a time scale of the same order as it's size in light years, if you see what I mean.

If you had a significantly decreasing of mass at the centre, starting 10's of thousands of years in the past, and continuing to recently, that might go some way towards explaining things, if gravity propogates at finite velocity. But I can't think why that mass loss should be happening.

If gravity is instantaneous, I can't think of a mechanism that would explain observations. Any change in mass would be transmitted across the galactic disc instantaneously, so it can't affect different radii differently.

2007-Jan-10, 02:05 PM
As a thought experiment that is far beyond my capability, imagine that some currently unknown and unsuspected effect could be applied to the central black hole of the MW causing the conversion of its entire "mass" into two or more photons in less than 10^-30 seconds. How would the gravitational effect propagate? As an aside is there a measurable quantum effect from my having done the thinking to pose this question?

[Anthropically speaking: In the beginning was the word and the word multiplied and thoughts emerged and from them came light and all that has followed]

publius
2007-Jan-10, 02:11 PM
Kzb,

And it turns out such a mass loss doesn't make sense anyway. This is the essence of one of the standard ways the "speed of gravity" question is posed: "If the sun disappeared, would the earth continue to orbit for 8 minutes, or immediately start going off in a straight line?"

It sounds like a very reasonable way to pose this. However, it turns out that GR cannot answer that question the way it is posed because, mass-energy cannot just disappear. The stress-energy tensor has conservation of the sources built right in. Mass cannot just disappear, as that would be a *invalid* source term for the gravitational field equations.

The only thing you can do is have mass move around very quickly, and even that is not easy to do. If a source mass is going to accelerate, something has to accelerate it, and that energy and momentum source goes right in the stress-energy tensor and itself contributes to the field.

In EM, you can do things like specify a source charge accelerates, and not worry where the energy comes from. EM conserves energy and momentum, but it only "worries about" those in EM form, in the fields.

Think of the field energy as water in a tank. There is mechanical to EM conversion, and you can think of that as water flowing out or in the tank. EM is worried about the total water in the tank. Conservation of energy here is only that the the rate the water level is changing is equal to what you're putting in less what you're taking out. :) It does not care about that energy outside the tank in non-EM form.

GR gravity, however, very much cares about the whole complete circuit of that water, because any form of that water (mass-energy-momentum) contributes to the field.

If there's energy anywhere in this universe, it makes gravity, no matter what form.

-Richard

Nereid
2007-Jan-10, 05:51 PM
'Sun blows up now' thought experiment, wrt 'speed of gravity' ....

Instead of the Sun going 'oops' and suddenly winking out of existence, how about it splits in two equal halves, with one half heading 'directly towards us', at 100,000 km/s (say). Then the question becomes (something like) 'how much warning would we have, that we were about to be splatted by a half-sol blob?'

It's a little messier than 'Sun goes bye-bye!', but should give at least a better feel for gravity in GR ...

DyerWolf
2007-Jan-10, 07:49 PM
'how much warning would we have, that we were about to be splatted by a half-sol blob?'

An event that would certainly garner more excitement than the 'wink out' scenario!

It dawns on me that the second thought experiment

Would an arrow pointing from the center of one star through the binary system's center of gravity point directly at the other star (the effect if gravity propigates at the speed of light) or would the pointer "lead" the other star (the effect if gravity propigates faster than the speed of light)?
may actually be observable.

(Although as I rethink it, the experiment works as I described it - if the observer were on the surface of one object looking toward the distant other, but perhaps not from outside the system)

Is the orbit of Pluto and Charon sufficiently large to measure the phenomena?

(after some thought) or would it take the placement of synchronized pulsing transmitters on each to be able to measure signal delay?:wall: (this stuff is tough)

crosscountry
2007-Jan-11, 10:03 AM
I'd say it doesn't matter, to the question, whether it's light speed or faster. You would need significant changes in the mass of the galaxy over a time scale of the same order as it's size in light years, if you see what I mean.

If you had a significantly decreasing of mass at the centre, starting 10's of thousands of years in the past, and continuing to recently, that might go some way towards explaining things, if gravity propogates at finite velocity. But I can't think why that mass loss should be happening.

If gravity is instantaneous, I can't think of a mechanism that would explain observations. Any change in mass would be transmitted across the galactic disc instantaneously, so it can't affect different radii differently.

better said than I was able to.

kzb
2007-Jan-11, 01:04 PM
Thanks crosscountry and Publius. I don't pretend to understand GR, I'm on a much simpler level than that I'm afraid.

Now you mention it, I can see that mass can't just disappear from the scene on these galactic scales. The best it could do would be to turn into EM and propagate away at c. It would still produce gravity.

Even if you envisage something physically ejecting mass from the galactic centre, that extra kinetic energy would still produce gravity.

I said what I said because I was trying to illustrate the unreasonableness of the conjecture. You would need a significant mass loss rate towards the centre, plus a "slow" propagation of gravity, to explain the rotation curve. The rate of mass loss needed is so fast that the galaxy couldn't exist more than something of the order 10^5 or 10^6 years. Plainly it's much older than this.

Or this mass loss only started a few 10's of thousands of years ago. And at much the same time in all galaxies.

So the conjecture is doubly damned. To be viable, it would take a significant mass loss rate (which as Publius explained, can't actually happen) AND a "slow" propagation of gravity.

crosscountry
2007-Jan-11, 02:58 PM
one thing to keep in mind is that these galaxies are millions of light years away.

If mass conversion was possible those galaxies would nearly be gone "by now". of course we wouldn't see it except within our own galaxy. It's pretty safe to say that doesn't exist.

DyerWolf
2007-Jan-11, 03:57 PM
As I understand it, the speed at which gravity propigates shouldn't matter wrt the galactic rotation curve, because:

- the rotating galaxy (system) is in equilibrium.
In my three examples of possible propigation speeds, (speed of light, instantaneous, and some speed in between), because the propigation within the sytem is uniform, absent a sudden (highly unlikely) change in the system, the gravitational effects within the system will remain constant. Thus it shouldn't matter how long it takes a change in x to affect y within a balanced system. (once in balance the propigation speed should not affect the curve - nicht wahr?)