Bending light?

[zeon23445]

So heres my conundrum- if light can bend, doesn't that mean its velocity is changing, and therefor accelerating?

 

My theory is that when it hits a gravity field the gravitons fragment the light, which makes it break up in multiple directions without slowing down.

 

Am i right? Im confused because if it was the idea that its following the curveture in space then wouldnt that mean their are perpendicular curvetures to it that cancel it out?

[xyzt]

So heres my conundrum- if light can bend, doesn't that mean its velocity is changing,

 

Yes. The direction is changing, the speed is not. Light follows its own geodesic in curved spacetime.

 

 

 

 

and therefor accelerating?

 

No, in GR geodesic motion is not accelerated motion, you are trying to force fit your notions of Newtonian mechanics into GR.

 

 

 

My theory is that when it hits a gravity field the gravitons fragment the light, which makes it break up in multiple directions without slowing down.

 

This seems to be a fringe misconception generated by your lack of understanding the mainstream physics.

 

 

 

Am i right?

 

No.

 

 

[Endy0816]

If light could talk, it would say that it traveled in a straight line the entire time.

[md65536]

If light could talk, it would say that it traveled in a straight line the entire time.

Light follows a null geodesic (but not all geodesics are null, eg. a planet's orbit). Is what you wrote equivalent to saying that if you transported a short enough ("local") straight ruler along a null geodesic, the ruler never bends? The ends are always tangential to the null geodesic. Is that a reasonably good way to say it?

 

That would mean there is no local acceleration of light anywhere along the geodesic. Light of course remains at a fixed local speed, but also does not change direction locally.

 

However, a null geodesic can be curved when measured remotely (eg. gravitational lensing). Is it then correct to say that the only acceleration of light is coordinate acceleration (which is just called "acceleration" anyway?), but not proper acceleration?

 

 

Edit: After thinking about it, I don't think the idea of a locally contained ruler makes sense. There must be a different way to say it...

 

Light always travels in a straight line in flat spacetime, and spacetime is locally flat, and light follows that local flatness. A non-null geodesic is generally not locally flat?

Edited May 28, 2014 by md65536

[Endy0816]

 

However, a null geodesic can be curved when measured remotely

 

Stated exactly what I was getting at.

Edited May 28, 2014 by Endy0816