The characteristic impedance of the device Zo, propagation constant γ in HFSS

According to the transmission line theory, the characteristic impedance of the device Zo, propagation constant γ only should be related to the materials and the cross-sectional structure theoretically. And HFSS solve the 2D fields first, then solve 3D fields. For this, I want to ask three questions:
1 Choose “solve ports only”, HFSS only calculate the cross-sectional fields( port fields ) then get Zo, γ, so I think the cross-sectional fields have been calculated but the “plot filed” options are gray, why?
2 the Zo, γ is independent with the length of the device, but in my simulation, the length influence these results, why?
3 if the cross-sectional structures are similar, the Zo, γ should be as same. But in my simulation, they are different sometimes. why?

Re: The characteristic impedance of the device Zo, propagation constant γ in HFSS

1.
"“plot filed” options are gray, why? "
you forgot to save them, or you started an optimization sequence which can not be plot.
Try to solve CW only once to see the port distribution.
In my experience, making a non-obj. cross-section and plotting vector fields there gives more information.
2.
Seems, Zo is not that is told in the help. I believe, reflections and numerical errors affect the model matrix, where Zo is extracted. Anyway, extracting the model impedance is the real hell in HFSS. You need to simulate ports separately, then go to Matlab and multyply your matrix with inverse of port matrices... boring.
BTW, there are a lot of delicate moments that are explained wrong in the help. For instance, ferrite's high frequency permeability is NOT constant, but a function based on classic L.L. model. But what kind of magnetic loss factor do they use, no-one knows. (there are two formulations only in Gilbert model, where DH can be formulated as a constant multiplied by frequency internally, when substituting to complex frequency, or can be treated as e frequency dependent value... which one?). I have also found some peculiarities with antenna AF, at least when trying to reproduce results in Field calculator...