Hi,

This is very helpful, Mark.  

I also would like to model a system which I am analyzing with a shearing interferometer.  My design includes a black box of a Powell lens, plus a number of other lenses.  I would like to introduce decenters, tilts, etc. to the design to determine likely error sources in the fringe pattern I am getting, which is not as good as I expect.  I don’t know if the errors are due to defects in the lenses, the lens positions, or a combination.  

It seems like I have a couple of options, neither of them ideal:

1. I can make a sequential model of the Powell lens and convert the entire lens to non-sequential.  There is a risk my model of the Powell lens won’t be accurate.
2. I can leave the model sequential and try to simulate the shear plate in this mode.  I have more confidence in the lens design model but don’t know if it is possible to model the shear plate in sequential mode.

Is there a better solution which I am not aware of?  

Thanks

David Aziz

I looked at Powell Lenses in the Thorlab website and saw this:

I would make this model using a Polygon object, in either pure sequential or hybrid nonseq if you really need a sequential analysis that’s not available directly in sequential mode.

One caveat: don’t align it so that the gut ray hits exactly on the edge of the prism faces (i.e. right along the roof of the prism). There’s no meaningful derivative at an infinitely sharp edge. There’s no problem ‘in real life’ of course as no light would refract at such an edge for the same reason.  

This link may be of some help: 

Regards,

Jeff