In may calculations I used the spot size calculated in the Zemax as a resolution limit. Does it is really good criteria for this if the influence of aberrations must be considered?
Zemax shows a RMS-radius value in the spot diagram, that is a realistic measure of the resolution to be expected and it includes all aberrations. It also shows an Airy-radius which is the limit in resolution that can theoretically be gained with the given aperture sizes and NA. It is however only a good measure for flat top illuminated apertures that create an Airy diffraction pattern. Zemax also offers a lot of other graphs (MTF, RMS,..) that perform a comparison between real aberrated imaging and theoretical limits.
If you expect the final rms wavefront error to be more than 2 waves, use spot diagram. Within 2 waves and closer to the diffraction limit, I would look at either wavefront error or some PSF/MTF-based metric.
Zemax shows a RMS-radius value in the spot diagram, that is a realistic measure of the resolution to be expected and it includes all aberrations. It also shows an Airy-radius which is the limit in resolution that can theoretically be gained with the given aperture sizes and NA. It is however only a good measure for flat top illuminated apertures that create an Airy diffraction pattern. Zemax also offers a lot of other graphs (MTF, RMS,..) that perform a comparison between real aberrated imaging and theoretical limits.
I agree with Alexander Dräbenstedt. Be careful not to underestimate your spot size: Zemax calculates the geometrical spot size. The lower limit of the real spot size is determined by the diffraction pattern (Airy radius).
I already simulated and optimized several imaging optics on ZEMAX. I think that the spot diagram is a good information to optimize the design but I saw that ZEMAX only consider perfect beams and often give spot sizes smaller than the Airy disk (which is as Alexander said the physical limit of the optics). However, you can clearly see the aberration by analizing the shape of the spot size on the edges of the imaging field.
The MTF function gives a better estimation of the limiting resolution of your optics since it considers the Airy limitations. The cutting frequency will be at around 30% contrast. Moreover, you can use the Seidel diagram to get a measurement of the aberrations.
Don't forget to define precisely all the wavelengthes and the full imaging field you will use. If you parameter a monochromatic wave in a paraxial configuration, you will not be able to see the aberrations of your optics.
Of course your ultimate limit is the diffraction limit. That said, I often find the RMS radius a poor indicator of the expected resolution of an imaging system. In the geometric trace, you can have a very tight central spot with a single ring with a large radius. For the same RMS radius you could also have a broad spot uniformly distributed. The first spot will have a much better resolution limit than the second one. Single outliers have a very large impact on RMS calculations that mayor may not be relevant. It is a very useful tool, but don't lend too much credence to it by itself.