Weird idea for optics... one way mirror sphere

OMcGovern

Hi,

I just wanted to bounce this idea off ye... what do you reckon it'ld do, if anything ?

If you could construct a glass sphere, with a sort of one way mirror inner surface.... then all light entering the sphere would be trapped inside.

Maybe if a tiny hole or lense were on some part of the sphere, it would focus the light back out ?

Maybe the "one way mirror" bit is infeasible, because I think one way mirrors only work if the hidden side is in darkness.

If I let my imagination run away with itself, then this device might focus the light into a beam which is directly out through the hole / lense.
A bit like a laser....
( this would happen because the light would bounce randomly inside the sphere until it aligned perfectly to escape through the hole )

any thoughts, or is there a simple rule of optics to blow this out of the water ?

regards,
Owen

Gurgle

OMcGovern wrote:

If you could construct a glass sphere, with a sort of one way mirror inner surface.... then all light entering the sphere would be trapped inside.

1. The best of mirrors reflect approx 90% of light, so considering the amount of reflections each 'ray' would go through before finding its way out through the hole, you wouldn't get light out the hole of significantly higher intensity than the incident light. You certainly wouldn't be able to store light for any measurable time.

2. To focus the light from a large area to a small point at higher intensity, use a convex lens. Thats the effect you're getting when you start a fire with a magnifying glass.

BrayHead

What you are imagining already exists! It's called an integrating sphere and is used for calibration measurements on various light sources - LED's, laser diodes etc.

Check out

http://www.gigahertz-optik.de/pdf/catalogue/Integrating_Spheres.pdf

and

http://www.gigahertz-optik.com/database_en/html/applications-tutorials/tutorials/iii.-measurement-of-light-with-integral-detectors/iii.1.-the-detector-8217-s-input-optics-and-its-directional-sensitivity.html

The output intensity can of course not be higher than the incident intensity unless some gain material is added i.e. think LASER action

OMcGovern

Cheers Gurgle & BrayHead, for very informative replies.

My 15 year old leaving cert level physics weren't enough for me to answer this myself :-)