- Home
- Polarization and Thin Film Coatings
- How To Design Birefringent Polarizers
How To Design Birefringent Polarizers
- By Mark Nicholson
- Published 24 June 2005
- Polarization and Thin Film Coatings
-
Rating:
Summary and References
This article has demonstrated the basic techniques needed to model birefringent components in ZEMAX. In summary:
- Birefringent materials require ZEMAX to trace two rays in each piece of birefringent material to compute the ordinary and extraordinary ray-trace
- This is most easily handled for a general component by using a configuration for each ray, and using the Mode parameter of the Birefringent-In surface
- A two-crystal component therefore requires four configurations, a three-crystal component will need eight, etc
- The ray amplitudes, not intensities, need to be added across configurations to compute the total transmitted energy.
Further Reading
1. Saleh and Teich, Fundamentals of Photonics, Wiley Interscience
2. Quan-Ting Liang, "Simple ray tracing formulas for uniaxial optical crystals", Applied Optics Vol. 29, No. 7, (1990).
Related Articles
Attachments
6 Responses to "How To Design Birefringent Polarizers" 
|
said this on 25 Aug 2005 10:07:22 PM PDT
excellent article!
|
Author)|
said this on 16 Sep 2005 8:39:19 AM PDT
This is excellent demonstration of the unique abilities of ZEMAX for birefringent materials. One question is does this method of sum of amplitude E-field components work for rays off-axis in angle. For that case the two rays of interest (conf 1 and 3) have a very small laterial displacement at last surface eventhough the rays have same direction cosines. Does this small displacement introduce a optical path length different similar to a shear plate interferometer
|
|
said this on 16 Sep 2005 9:16:34 AM PDT
Hi
Its a good question. The answer is yes the method is still valid. The ray represents some field and as long as the field is constant over the displacement then the coherent addition of fields is still valid.
So at the very edges of the beam where the two fields do not over lap it may be wrong but over the vast area of the beam it will be correct.
This is similar to the interferometer feature in which two configurations of a system are overlapped and interference fringes between the two wavefronts are computed. This requires the two exit pupils to have the same position and the same magnification. What if they're different by 1 micron Probably not important. What about 10 cm Probably important.
The issue is does the field have a significant slope over the displacement region If it does not the coherent addition is still valid.
- Mark
P.S. I realize now that in order to respond to your comment I also have to give the article a rating. As the article author that puts me in a difficult position!
|
|
said this on 30 Nov 2005 11:10:26 AM PDT
A great article but two questions please:
Modes 2 and 3 of birefringent surfaces are not discussed. For those modes:
1) Is the summing of S and P handled automatically in a manner equivalent to the approach discussed in the article and 2) Do the "Polarization Pupil Map" and "Polarization Ray Trace" tools provide an absolutely correct "Phase difference between X and Y" or is some approximation involved. Thanks.
|
|
said this on 01 Dec 2005 10:20:20 AM PDT
To Comment 4: I'll be writing a follow up on designing birefringent waveplates soon. If you have an immediate question please email support@zemax.com
On the second question the phase shown in the Pol Pupil Map is exactly what you'll get from the Polarization Ray-Trace calculation. Again if you have a specific question please email support@zemax.com
|
|
said this on 17 Feb 2008 11:51:44 PM PDT
Excellent article!
|