For a collimated beam (or one that is close to collimated with mild aberration), you should change the OPD reference designation to “Infinity” or possibly “Absolute”:

In general, if the collimated beam has aberrations, then technically speaking the Wavefront Map should depend on propagation distance to some extent (might be very weak though).  Spatially coherent beams diffract during propagation, so the spatial complex amplitude will vary.  Ray tracing, which doesn’t account for diffraction, should be okay for short propagation distances.

Regards,

Jeff

A wavefront at any wavelength continually changes shape as it propagates through the optical system and exits the at exit pupil.  It begins in the Rayleigh-Somerfeld region, morphs into the Fresnel region, then finally the Fraunhofer region near focus.  The transition distance from one diffraction region to the next is in part driven by the strength of the aberrations.  Zemax can approximate the changing wavefront anywhere along the Fresnel/Fraunhofer path, with accuracy dependent on sampling density.  Higher sampling (e.g., 1024x1024 vs. 128x128) gives better accuracy at the cost of increased execution time.  Diffraction modeling very near the exit pupil is much more difficult, as the interaction and polarization of electric and magnetic fields comes into play.  Codes such as rigorous coupled wave (RCW) and 3D method of moments (MOM) are required in extreme deep-field analysis.