The reliable of radiation pattern!?

With the same structure, if we use about x-meshcell and accuracy of -30 dB and use about 2*x meshcell and acauracy of-50dB?!...
which radiation pattern case is more reliable? (at same operating fre.)
It only chages a little bit, I mean, the radiation PATTERNS will be same!
Am I confusing?

r u interstd in the x pol level down to 50 dB or so

hi khangkhang,
Firstly, meshcells do have great effect on the accuracy of the simulation results, but this parameter depends largely on your structure and usually you could trust the initial value automatically determined by the "expert system" if your structure is not a very special one.
Secondly, i suppose the term "accuracy" you referred to is the first option in "!T"--transient solver. This parameter is used by the system to evaluate to what extent the energy in the structure has decayed, so it is different from the ordinary meaning of the word "accuracy" we usually use.
So it is not strange that you didn't find any obvious differences between your 2 simulations after different parameter settings if the number of the meshcells "expert system" originally adopted was already appropriate for your structure, and your following changes didn't mean very much--in my opinion.