Is there direct evidence photons attract gravitationally?

[ensabah6]

Not sure where better to place this thread, here or gr.

Is there direct evidence photons attract one another gravitationally? I am aware of the arguments that GR couples to the energy-stress tensor so photons should attract one another gravitationaly but I wonder if Gravity needs an inertial mass to act as a kind of charge analogous to electric charge in E&M.
So just as photons do not have electric charge, they don't have "gravitational" charge in the form of mass?

 

[javierR]

Not sure where better to place this thread, here or gr.

Is there direct evidence photons attract one another gravitationally?

No.

... I wonder if Gravity needs an inertial mass to act as a kind of charge analogous to electric charge in E&M.
So just as photons do not have electric charge, they don't have "gravitational" charge in the form of mass.

But there's evidence that light and massive objects gravitate (light if deflected by stars e.g.). If photons weren't "charged", how would this occur?

If they do self-interact due to gravity then why isn't this make QED non-re normalizable.

Because the gravitational interaction is not part of QED. QED is the interaction of photons
and particles that interact via photon exchange.

 

[hamster143]

That would be a vanishingly tiny effect that is impossible to detect at this level of technology (a 100 watt lightbulb emits 10^-12 g worth of photons every second ...) We do know that photons experience gravitational pull from massive objects, just like everything else.

Photons' gravitational self interaction is an important component of the Big Bang theory, the universe would evolve differently if they weren't attracting each other. And since the Big Bang theory seems to work well, that could be evidence in favor of this assertion. One would have to redo the math and see what predictions come out differently ...

 

[Naty1]

Photons' gravitational self interaction is an important component of the Big Bang theory, the universe would evolve differently if they weren't attracting each other.

Can anyone explain this comment??

 

[hamster143]

Can anyone explain this comment??

Really simple ... for the first 50 000 years or so after the Big Bang, radiation density exceeded matter density. Expansion during this period is modeled using GR equations, which assume that photons gravitationally interact with each other. In a sense, photons are the driving force behind this expansion (which is a curious paradox - gravity is an attractive force, but universe nevertheless expands!) If photons don't REALLY interact, GR is not applicable and our expansion timeline is off.

And expansion timeline affects numerous other things, from CMB temperature, to relative densities of basic isotopes, to large-scale galaxy structure ...

 

[humanino]

for the first 50 000 years or so after the Big Bang, radiation density exceeded matter density.

Which quantum aspect of the radiation is necessary in this period of the Big Bang ? I mean, why can I not treat the radiation classically ? If I can, is it fair to say that "photons" have been indirectly tested to attract each other ?

 

[Vanadium 50]

I should point out that there is no direct evidence that marbles attract gravitationally. However, nobody doubts they do. Why should photons be different? We know that they respond to a gravitational force, and conservation of momentum requires that this force be balanced. Isn't that good enough?

 

[hamster143]

I should point out that there is no direct evidence that marbles attract gravitationally. However, nobody doubts they do. Why should photons be different? We know that they respond to a gravitational force, and conservation of momentum requires that this force be balanced. Isn't that good enough?

Gravitational attraction of macroscopic objects ("marbles") was directly observed in the lab as early as 1798 by Cavendish.

It is a good point that photons respond to a gravitational force from macroscopic objects ... HOWEVER ... we could imagine a world where photons gravitate to baryons, and baryons gravitate to other baryons, but photons don't gravitate to other photons.

Or, in GR language, maybe we could write a complicated version of Einstein's field equation that replaces the full stress-energy tensor with the part that only contains massive fields. There's an obvious problem that such a stress-energy tensor would no longer be divergenceless .. but maybe there are ways around it.

But that would most likely be in vain due to the Big Bang argument above.

 

[Vanadium 50]

Gravitational attraction of macroscopic objects ("marbles") was directly observed in the lab as early as 1798 by Cavendish.

No, gravitational attraction of a marble by a very large mass has been directly observed in 1798. That tells you that the marble is attracted to the large mass, not the other way around. The marble-marble force is also unmeasured.

Of course, I don't doubt that marbles have gravitational fields - it's the OP's requirement of direct evidence that I think is unreasonably strong. Isn't conservation of momentum good enough to close the loop? It is for marbles.

 

[Naty1]

I wonder if Gravity needs an inertial mass to act as a kind of charge analogous to electric charge in E&M.

An interesting question: but as others have implied gravity seems to be a unique "force"...it effects ALL known particles and energies (kinetic, potential,heat,etc) and even pressure. Photons on the other hand seem to mediate only the electromagnetic force between charged particles while the strong and weak forces also have their unique, less wide ranging effects...

I don't believe anything is immune from gravity...not even spacetime!
Is the Higgs boson(s) immune...I'm not sure...there are lots of different theoretical models.

 

[njinear]

i have some questions about this, maybe you folks can educate me.

einstein says energy carries gravity, but he makes the exception for gravity itself, that the tensor has an entry of zero for gravity's gravity. since gravity does not have gravity itself and it can carry energy away (seems pretty well proven by binary star decay) then the gravity in the universe is not a conserved quantity, with significant cosmological implications since energy is more often emitted than absorbed.

the fact that photons are deflected by gravity doesn't mean they generate gravity. the deflection is due to the shape of spacetime, the photons are following the geodesic. we would expect gravity waves to do the same thing, because if for no other reason, there is no straight line EXCEPT the geodesic.

further, if a photon for example has gravity, it seems like a bit of a problem getting "gravitons" in some quantum version of gravity to be "entrained" by the photons.

as i understand it the gravity of photons is still not established. we know for sure that the nuclear force causes gravity from decay weight differences in fission and fusion. not sure about electromagnetic static fields, whether chemical energy has been proven for example to have mass.

perhaps massless particles do not carry gravity, just massive particles and static (nongravity?) fields. we already have an example of this with gravity waves themselves. if photons don't have gravity, then when they are absorbed for example as chemical energy, one would presume they must "reinstate" the gravity.

seems like there are a lot of loose ends here.

 

[Drakkith]

So is the overall view in science that Photons have their own gravity, or that they do not but are attracted by it?

 

[cragar]

Look at this thread and this article titled on the gravitational field of light.
http://authors.library.caltech.edu/1544/1/TOLpr31a.pdf
https://www.physicsforums.com/showthread.php?t=174805
And it also depends on if they are parallel or anti-parallel if they will be attracted .
As you will see from the paper

 

[Dr Lots-o'watts]

Is there direct evidence photons attract one another gravitationally?

Suppose we send two light pulses traveling parallel to each other towards the moon and they are reflected back. It they attract each other, the distance that separates them when they are detected back here on Earth should be smaller than the distance between the two sources.

Can anyone calculate the change in separation from an existing "photon-photon gravity" theory?

 

[Drakkith]

Suppose we send two light pulses traveling parallel to each other towards the moon and they are reflected back. It they attract each other, the distance that separates them when they are detected back here on Earth should be smaller than the distance between the two sources.

Can anyone calculate the change in separation from an existing "photon-photon gravity" theory?

Also, if light did attract itself, wouldn't light from far away stars and galaxies look...different after thousands, millions, or billions of light years?

 

[Researcher X]

Doesn't attraction necessitate a change in velocity, something impossible for light?

 

[russ_watters]

Light most certainly can show a change in velocity (via change in direction) - just not a change in speed.

 

[njinear]

i have a problem with the 1931 paper that it assumes what it is trying to prove, that is, that the mass energy tensor applies to photons. and really, it doesn't deal with photons, it deals with a "beam", which to say the least is not descriptive of what is really going on. the proof would have to be for an individual photon.

photons are deflected by gravity but this doesn't men they generate gravity. they have to "deflect" because space is curved.

so it seems that the question of photons gfenerating a grav field is an open question, and if they do it results in a number of problematic consequences.

 

[cragar]

In relativity it says that mass energy or pressure bends space-time .
And if i have an electron positron annihilation and it produces photons , It seems to me that the gravitational fields of the electron and the positron would be transferred to the photons .

 

[Pythagorean]

What about N3? If a photon is being pulled by a large mass, shouldn't it be pulling back [strike](even if orders of magnitude lighter?)[/strike]

EDIT: scratch "orders of magnitude lighter"

 

[njinear]

cragar you are right that generally any mass/energy should bend space time. but i am questioning if this is right. for example it is not clear what the gravitational field from a light speed particle would be like. also, imagine a 2 slit interference experiment where the photon could go through either slit-- and maybe the detector is a mile away. which way does the gravitational field go?

i mean basically there are fundimental problems with photons or any free particle cqarrying gravity (except maybe gravitons). that's my problem.

 

[Redbelly98]

Doesn't attraction necessitate a change in velocity, something impossible for light?

No, attraction necessitates a change in momentum. Also, as Russ pointed out...

Light most certainly can show a change in velocity (via change in direction) - just not a change in speed.

I'll just point out that if a photon or beam of light is heading directly towards or away from the object that is attracting it, then even its velocity would not change. However, the momentum (and frequency, and wavelength) will change.

 

[JDługosz]

Or, in GR language, maybe we could write a complicated version of Einstein's field equation that replaces the full stress-energy tensor with the part that only contains massive fields.

An early test of GR was the displacement of stars during a solar eclipse. Now, we have much more accurate measurements using radar, grazing the surface of the sun.

Meanwhile, GR supplied a correction to the precession of Mercury, due to more subtle effects than just the mass. So, do the radar measurements show that photons, not just planets, are affected in the full GR manner?

 

[njinear]

no, the deflection of light does not mean photons generate gravity. same as iron is attracted to a magnet but in itself is not particularly magnetic. weak but instructive metaphor. but light gets deflected because the star warps space. the photon, as far as it is concerned, just travels in a straight line. no evidence the photon makes gravity itself

 

[Dr Lots-o'watts]

The issue of photon gravitation may be related to hypothetical photon mass, so using the classical gravitation law, and a hypothetical upper limit for an x-ray photon mass from Ef=mc^2, I get that two x-ray photons going to the moon and back (see my other post), initially separated by d1 = 1 mm, come back separated by d2 = (1mm minus about Planck length). Although these elementary equations most likely don't apply, I believe this could be considered as a crude approximation for the maximum deflection.

Not very encouraging for eventual experimental evidence.

 

[Redbelly98]

Not very encouraging for eventual experimental evidence.

Indeed. To minimize diffraction effects to smaller than a Planck length, the beam diameters would have to be orders of magnitude larger than the 1 mm separation of the two beams!

 

[njinear]

really hard to measure. but it has much wider philosophical and cosmological implications: since significant amounts of mass are evaporating from the universe, the gravity would be decreasing. and the fundamental energy-mass tensor is not right.

 

[Vanadium 50]

i have a problem with the 1931 paper that it assumes what it is trying to prove

When someone points out that they have found an elementary flaw in a paper that thousands of people have studied for decades, alarm bells go off. Which do you think is more likely - Einstein made a grade-school mistake and 1000's of people missed it? Or that there is something you don't understand?

If it helps, that paper doesn't attempt to prove anything. It's a derivation - it shows that the theory predicts how light behaves, a prediction that can be compared with experiment. (As it happens, successfully)

for example it is not clear what the gravitational field from a light speed particle would be like. .

Of course it is. G = 8pi T. T₀₀=T₁₁ and all other terms are zero. Now you know G.

no, the deflection of light does not mean photons generate gravity. same as iron is attracted to a magnet but in itself is not particularly magnetic. weak but instructive metaphor. but light gets deflected because the star warps space. the photon, as far as it is concerned, just travels in a straight line. no evidence the photon makes gravity itself

First, iron is magnetic, and indeed, must be magnetized itself to be attracted to a magnet.

Second, the fact that photons fall plus conservation of momentum requires that photons "make gravity".

since significant amounts of mass are evaporating from the universe

Evidence, please?

 

[inflector]

No, gravitational attraction of a marble by a very large mass has been directly observed in 1798. That tells you that the marble is attracted to the large mass, not the other way around. The marble-marble force is also unmeasured.

Of course, I don't doubt that marbles have gravitational fields - it's the OP's requirement of direct evidence that I think is unreasonably strong. Isn't conservation of momentum good enough to close the loop? It is for marbles.

and

Second, the fact that photons fall plus conservation of momentum requires that photons "make gravity".

This seems like a classical Newtonian perspective to me so I must be missing something. Classically, in the case of the marble, it seems that you'd have a force with equal and opposite action and reaction from the change in the momentum vector due to the curvature.

Can you explain how conservation of momentum applies to a photon following a geodesic under general relativity? Is it that you still measure the momentum vector in flat spacetime (Minkowski space)? Otherwise, how is there any change in the momentum vector if it is following what is analogous to a straight line through curved geometry?

 

[Vanadium 50]

This seems like a classical Newtonian perspective to me

What's wrong with that? GR doesn't dispense with momentum conservation.

 

[njinear]

vanadium is right, in the far field the change in momentum for the photon must be balanced against a very small change in the velocity of the body it bends around, so free photons have gravitational fields.

however it is very poor form to assume that since others didnt see something it must not be true.

since they have gravity then what happens when a photon wave function splits at a beamsplitter? one path could include bending around a star and the other just shoots into space. now what do you do with the gravity? or in any attempt to work "gravitons" into a quantized version of gravity, how does the graviton and the photon couple together? still seems like there are lots of loose ends.

 

[inflector]

What's wrong with that? GR doesn't dispense with momentum conservation.

I'm not suggesting that GR doesn't conserve momentum. What I'm trying to figure out is where the change in momentum comes from that needs to be conserved/offset.

If a photon goes through a geodesic where does it have a change in momentum? Isn't it just taking the unaccelerated path? How does that change the photon's momentum such that this change needs to be offset by a change in the momentum of something else?

 

[JDługosz]

When you start bending space and time, the classic definition is meaningless. See http://en.wikipedia.org/wiki/Momentum#Four-vector_formulation" as a starting point.

 

[njinear]

inflector good question. i am thinking that in the far field (away from significant gravitational fields) if the photon is moving in a different direction then the momentum difference between inbound and outbound directions has to be balanced against a comeasurate change in momentum from the large gravitating body, which implies the photon must have chaged the momentum of the larger body.. so it must generate gravitation. there does seem to be an issue here, that the photon is merely following a straight line in space, not accelerating in any normal sense. but i am not willing to give up conservation of momentum.

 

[inflector]

When you start bending space and time, the classic definition is meaningless. See http://en.wikipedia.org/wiki/Momentum#Four-vector_formulation" as a starting point.

That addresses a side issue, it seems to me, that one need consider the Lorentz factor when looking at relativistic momentum. How does the four-vector have anything to do with a change in momentum for a photon following a geodesic?

Is the issue here that momentum is conserved in a frame-dependent way? Just like energy is conserved in a frame-dependent way? i.e. that from the photons frame it is just going straight so there is no change in momentum but from the perspective of an observer the photon curves and therefore there is a need for an offset to the change in the photon's momentum vector?

This doesn't really make sense to me either, I'm just trying to get a handle on this concept.

The real conceptual problem here to me is that this all acts very much like a field like EM rather than what general relativity is supposed to be (a curvature in spacetime). If energy can be transferred to spacetime itself through the act of curvature, this means that spacetime acts like it is springy in some ways.

Or is it really the case that general relativity is just a field theory like EM and that all this talk of curved geometries is just a mathematical approach to describing the field equations?