Can light be trapped indefinitely in a mirrored sphere?

[alkammy]

hi every body, I'm new here abouts

i have a question that has been puzzling me...

if a flash of light was 'let off' in a mirrored sphere, how long would the light be there for? would it be forever reflected internally? if not why not? would you be able to crack open the ball some time later and see the flash?
after all we know that light can travell for billions of years, but what about in a confined space?

 

[DaveC426913]

No mirror is perfect. A percent will be lost as heat with every reflection.

also, whatever "let off" your flash would interfere with light passage, and would absorb it as heat.

BTW, note that the container need not be spherical at all. As long as the container bounces the light around, it doesn't much matter the angles.

In fact, a spherical mirror is the *worst* design, since the light will be impacting perdendicular to the surface, catching it at its worst reflectivity. Light reflects better at a highly oblique angle (such as a low angle off water).

This is how fibre optic cables work. Your experiment could be rearranged as a loop of fibre optic cable, letting the light race around in a circle for an arbitrary length of time. (and considering how cheap fibre optic cable is, you could build and experiment with this yourself.)

If you could manage to get a reflective enough surface so that the light could travel 300,000 metres before dissipating, then yes, I suppose your light flash would last one second.

 

[pervect]

OK, for the best dielectric mirrors around, I would estimate from what I've read that one can expect losses on the order of one part per million, i.e. 99.9999 percent reflectivity.

source: http://www.rp-photonics.com/encyclopedia_s.html#k_supermirrors

(and I'm open for better estimates).

With this ultra-high efficiency (I would assume the bandwidth over which this level of refelctivity would be maintained would be very low, requiring the light being reflected to be a laser pulse)

the light could reflect about 693,000 times before it halved in intensity. If you know the mean free path between reflections (this would be proportional to the radius of the sphere, you can figure out the time it would require for the pulse to halve in power.

Let's say that the sphere was room-sized, and that the mean free path was 10 meters, the intensity would halve every .02 seconds. If you have a huge room, 100 meters, you could get that up to .2 seconds.

 

[DaveC426913]

Note though, that you would never be able to *see* this light flash. Since it's in a prefectly reflecting container, no light will escape to be observed, and if you really had to take a peek, the simple act of observing it would "use it up".

 

[lawtonfogle]

Lets say, theorectically, we had a perfect mirror. We also had a fiberoptic cord so long it would take five minutes for the light to go all the way though. We let a flash light shine for one minute, and then we connect the cords perfectly. How long will that light last.