Back reflection analysis in Zemax

[Glasa...@gmail.com]

I have a refractive laser beam expander optical train that I have
designed. I would like to do a back reflection analysis to see how
much of the incident beam reflects back into the laser via surface
reflections from the AR coated lenses. I have contemplated doing this
in Zemax using Non-Sequential Surfaces without ports. I would put a
source at the laser location, then the lenses, and finally a Detector
Object back near the laser location. The hope would be to put AR
coatings on the surfaces and see how much energy reflects back onto
the detector. I can't seem to get any back reflection energy. I put
a coating file on a lens surface with some reflectance, even up to 50%
using the sample "I.50" file to just see if something gets back, but
nothing reflects, or so it seems. What boxes to I have to check to
make this work. i.e. "use polarization", "split rays", check any of
the "scatter" boxes, etc. I am using the proper "Face" on the "object
properites" dialog box, so I think that is at least right. Shouldn't
I see some back reflection on the NSS Layout plot? The detector
object gathers no back reflection rays.
This should be pretty easy to do, but even using the examples in the
online tutorial I can't seem to do it.

I mostly use CodeV so I don't have anywhere near as much experience
using Zemax.

Any help would be appreciated.

[Helpful person]

As you have only a few reflective surfaces I suggest using the old
fashioned way. Set up multiple ray traces using a different surface
as a mirror in each. Then calculate the amount of energy getting back
to the laser.

I suspect that using the method you are trying can be made to work,
but is it worth it? How would you check that Zemax is giving the
correct answer? (I've had many occasions when the Zemax calculation
for an unusual arrangement has given the wrong answer.)

By the way, any properly designed system should avoid reflection
getting back to the laser.

[Harvey Spencer]

Thanks for the suggestion. I considered the old fashioned way.. Actually I
think there are more than just a few surfaces. I don't have the actual
optical train in front of me, but I believe the guy said 16 surfaces. A
properly designed system should avoid a large back
reflection problem like you say, but let's face it, the gut ray and those
close paraxial rays around it always get back. It is a question of how
many of those rays. The issue of checking the correct answer from Zemax is
probably not too hard. Just set up the system
the "old fashioned" way with one retro-reflecting surface and compare with
what Zemax finds doing the NSS raytrace.
When you say that you have many cases of unusual systems giving the wrong
answer in Zemax, I would re-phrase that to say that the modeling
of the unusual system gave the wrong answer. Maybe the way you modeled it
was the only way possible or maybe there was another way.

"Helpful person" <rrl...@yahoo.com> wrote in message
news:02022ead-bb9e-47ad...@f63g2000hsf.googlegroups.com...

[Ron Gibbs]

>>What boxes to I have to check to
>> make this work. i.e. "use polarization", "split rays", check any of
>> the "scatter" boxes, etc. I am using the proper "Face" on the "object
>> properites" dialog box, so I think that is at least right. Shouldn't
>> I see some back reflection on the NSS Layout plot? The detector

Zemax can certainly do this job, although NSR can be frustrating at times.
"split rays" would be a good box to check, but who knows what else is wrong
without looking at your lens file. ZDC/Optima (in Europe) support is
actually pretty good, and worth paying for!

Ron
--
Gibbs Associates
Optical Design Consultant
www.gibbsassociates.co.uk

[Glasa...@gmail.com]

The ray splitting worked. I didn't read enough of the manual to know
that has to be checked, otherwise the refractive path is always chosen
and
surface reflections off the AR coated surface won't be active. If you
use "simple ray splitting" the raytrace is much more managable since
you don't split every ray, you just reflect the percentage of rays
appropriate to the coating. I.e. if you have a 98% transmitting
coating, 98 out
of every 100 rays refracts and 2 out of 100 reflect.

[Ron Gibbs]

"Ron Gibbs" <ron....@physics.org> wrote in message
news:48294d01$0$560$c5fe...@reader.usenet4all.se...

OP (some...@gmail.com) wrote:
*The ray splitting worked. I didn't read enough of the manual to know
*that has to be checked, otherwise the refractive path is always chosen
*and
*surface reflections off the AR coated surface won't be active. If you
*use "simple ray splitting" the raytrace is much more managable since
*you don't split every ray, you just reflect the percentage of rays
*appropriate to the coating. I.e. if you have a 98% transmitting
*coating, 98 out
*of every 100 rays refracts and 2 out of 100 reflect.

I take that as a "thank you". In which case, you're welcome! Even though my
usual news server filters your posts (I guess because of your gmail
address). Hmmmmm.

Everything is somewhere in the manual, but it's not always easy to locate
what you need when you need it.

[James R (Jim) Lynch III]

ZEMAX has a sequential mode "Tool" made to order for this analysis:

As an example of how to use this tool on a simple lens system:

Left click on Files ==> Open ==> C:\program
files\zemax\samples\sequential\objectives\Doublet.zmx

Left click on: Tools ==> Miscellaneous ==> Ghost Focus Generator...

Bounces: Double Bounce
First Surface: 1
Last Surface: 3
[x] Save Files
Ghost Reflector Coating: I.99 (It didn't like a lower case i.99)
Left click "OK"

ZEMAX creates and saves three ZMX files for the three possible double-bounce
cases and does an analysis of each of the created lenses in a text window.

James R (Jim) Lynch III

[Ron]

<Glasa...@gmail.com> wrote in message
news:8b0a3b29-2db5-4b0d...@f24g2000prh.googlegroups.com...

I take that as a "thank you". In which case, you're welcome!

Everything is somewhere in the manual, but it's not always easy to locate
what you need when you need it.

Ron