HFSS Interpolating vs Discrete Sweep (S11 match but not S21)

Hi everyone,

I am doing a simulation of two Vivaldi antennas working from 4-8 GHz separated by a distance of 500mm with PEC cylinder in between (Screenshot attached). I want to obtain the scattering parameters for them.

While interpolating sweep is giving close results to discrete sweep for Return Loss(S11), the Insertion Loss(S21) has a awkward curve at lower frequency range. I have attached a screenshot of the scattering parameters ( Dashed Lines - Interpolating sweep, Solid Lines - Discrete Sweep).

If anyone can shed some light as to why by using interpolating sweep I am getting such a good match with Discrete sweep (which is accurate) for S11 while I don't get it for S21. Correct me if I am wrong but aren't both parameter calculations supposed to be dependent on one another ?

I am using Iterative Solver with solution frequency at 8 GHz.

Thanks in advance

Hi. It may be because the phase of S21 varies much faster than the phase of S11.