Skip to main content

Hi!

I am trying to set a system made of two Fresnel lenses (that I don’t know anything about), capable of accepting +/- 45 deg of source incident angle. I am working in non sequential and I am getting familiar with the optimization wizard. For this specific application I have run an optimization (orthogonal descent algorithm), in order to minimze the RMS on the detector and to have a minimum power of 0.08 W (see Local Optimization Viewer pic attached). The variables are:

  • Thickness of both lenses
  • Radius of curvature of both lenses
  • Coeff r2 of both lenses
  • Frequency of both lenses
  • Conic constant of both lenses
  • z1, z2 (distance between lenses, distance between second lens and detector)

This optimization yielded a 25 h process, leading to the results I attach here in the format of pics as well (Merit function editor, Detector viewer and System view).

How should I interpret these results? Should I think it is not possible then? Can I trust this system? 

Thank you for your precious help,

best,

Giulia.

Hi Giulia,

The optimization has clearly worked, since you can see the starting and ending values of the MF. I think the stray light from the flat facets is complicating the RMS spot calculation however.

You can see rays TIR-ing from the flat facets (this is a real effect, of course) and causing rings of rays to be scattered around the central spot. These rings increase the RMS spot size, of course.

Often a first-pass optimization is done using the idealized Fresnel object, where the grooves are taken to have zero height and so there is no TIR from the inside. You can then replace this with the real Fresnel, and put an aperture in front of the detector such that only the central spot is detected. Re-optimize the lens, and then remove the aperture to see all the secondary rings.

  • Mark 

 


Reply