Zemax surface roughness

[Richard Mitchell]

Does anyone know if it is possible to apply the effects of surface
roughness to an imported object in Zemax? It would be useful if I
could use values from a surface roughness standards set.

Thanks,

Richard Mitchell
e2v Technologies
Chelmsford
UK

[Jim Klein]

My feelings on ZEMAX aside, use of "surface roughness" in a ray trace
code must be done with one's eyes widely open (even in non-visible
wavelength systems.)

Why? Because surface roughness and surface quality are many times
uses interchangibly and if the roughness model has a functional form,
and if that functional form is not the same as what happens in the
manufacturing process, then the model may not predict the true
performance of the system.

Jim Klein

[Richard Mitchell]

I understand and agree with the point you are making. The trouble I am
having is that the models that I am creating are using "ideal"
surfaces. My aim is to get some idea of how these models will perform
with imperfect surfaces.

To quantify how imperfect something is will always be difficult (if
not impossible), but to be able to "blur" the optics a little would
help me gain a perspective of how my system may perform.

Richard Mitchell
e2v Technologies
Chelmsford
UK

Jim Klein <acmeo...@earthlink.net> wrote in message news:<tbce0vkucqhk7hfa1...@4ax.com>...

[Jim Klein]

richard....@e2vtech.com (Richard Mitchell) wrote:

>I understand and agree with the point you are making. The trouble I am
>having is that the models that I am creating are using "ideal"
>surfaces. My aim is to get some idea of how these models will perform
>with imperfect surfaces.
>
>To quantify how imperfect something is will always be difficult (if
>not impossible), but to be able to "blur" the optics a little would
>help me gain a perspective of how my system may perform.
>

I performed a little study for OSL (Orbiting Solar Lab study) at TRW
some 12 years ago.

Using CODE-V's FRINGE ZERNIKE surface, we put surface error on the
primary so that we got a 0.1 drop in MTF at 1/2 the cut-off frequency.

The amount of RMS surface error for each of the 37 FRINGE ZERNIKIES
(one at a time) was all over the map. There were some clusterings but
nothing useful.

You would be best advised to consult with a potentail optic vendor and
get their estimate for the type of surface structure to expect, then
try to model it.

A purely random error is not what I would expect. The error would have
a strong connection to the size of the optic and the manufacturing
process.

Good luck,

Jim Klein