[pstricks] pst-optexp: Custom component with multiple reflections?

[Lars O. Grobe]

Dear all,

I am a beginner when it comes to pstricks, and the main reason I started 
looking into it is the pst-optexp package. I have to illustrate rays 
passing through geometrical structures frequently, and getting this 
integrated into my latex documents sounds too good to be true. I am 
looking into modeling of reflective surfaces, like venetian blinds (a 
stack of repeated profiles), and started implementing them as custom 
components with \newOptexpTripole. The diffuse reflection from a rough 
surface in the manual was my starting point.

I have been defining one profile as custom component. Then I am placing 
several instances of that component to build the blinds.
I need to trace rays so that a) the upper profile shades part of the 
lower profile, and b) the rays reaching the lower profile and getting 
reflected upward must stop at the upper profile. So a typical ray would 
be A->profile1->profile2->profile1->B. Some rays end on the outside of 
the system (A), others get transmitted to the inside (B). Is this 
actually possible with pst-optexp? At the moment, I use

\multido{\r=+2}{20}{%
\psset[optexp]{raytrace=true}
\drawbeam[beamangle=-45, beampos=\r](A){}{profile2}
}%end multido

This reflects the shading (start->profile1->profile2), but does not show 
reflections from profile2.

\drawbeam[beamangle=-45, beampos=\r](A){profile2}(B)


This shows reflections from profile 2 correctly, but rays are simply 
passing through profile1.

Another way would be to define several profiles as one combined custom 
component. In this case, I would need a way to define multiple 
reflections within a component.

Has anyone used pst-optexp for such an application so far? I guess it is 
the drawbeam concept that is confusing me, but I am also not sure wether 
it is possible at all to have rays not simply sequentially passing 
through a chain of components, but getting reflected back and forth.

I will try to set up an example as soon as possible, but my more general 
doubts made me ask this right now...

Thank you, regards, Lars.



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://tug.org/pipermail/pstricks/attachments/20160701/a194b650/attachment.html>