Hi ​@Asuku,

Yes, you are correct, you cannot use POP to propagate through a GRIN material.  

Think of POP as a Fourier Transform (FFT) propagator between 2 planes in the z-direction.  For a FFT to be valid, the medium needs to be homogenous such that the beam is not “refracting” at the same time as the beam is propagating. 

POP is a two step process that repeats itself for every surface in the LDE:

• Convert a wavefront to rays on the left side of a surface & refract through a surface
• Reconstruct the wavefront on the right side of the surface and FFT to the next surface

For GRIN ray tracing, OpticStudio essentially breaks the GRIN material in individual planes along the z-direction and recalculates the index and the gradient of the index at each deltaT step.  This GRIN ray tracing is incompatible with POP.

If you need to simulate diffraction through a GRIN material, I would look at using Lumerical.

Thanks ​@MichaelH That explains it. So, how can we model SM Fiber Coupling Efficiency using Grin Collimators in Zemax. Would that be possible?

-Asuku

​@Asuku, you can definitely use POP (or Analyze > Fiber Coupling > Single Mode Coupling) to calculate the coupling efficiency of a SMF.  For POP, you will simply have to use dummy surfaces to in the LDE to model the appropriate Start Surface.  Since you can’t have POP start at S0, if you want to model the full system, you should make S0 thickness something really small (like 1e-5), add a dummy surface at S1 and make this thickness 2.54.  Then, you can select S1 as the starting surface & use POP to calculate the SMF efficiency.  However, what I mentioned before still holds that inside the GRIN lens, only discrete steps of geometric ray tracing is used to propagate the rays.  The FFT will only be calculated from S1 to the front of the GRIN and then from the back of the GRIN to the Image surface.  The SMF efficiency will be a function of the beam’s complex amplitude and the receiving fiber’s complex amplitude.