Whats wrong with 2D electromagnetic field solvers available with some general-purpose 3D tools?
These solvers are designed to support extraction of S-parameters in some 3D tools and usually do not have possibilities to extract frequency-dependent RLGC parameters of multiconductor lines. They do not compute characteristic impedance with 3D definition – the impedance definition closest to experimentally measured so far. In addition these tools usually do not have causal dielectric models and simulate metal with simplified surface impedance boundary conditions with incorrect low and mid-frequency asymptotes and without roughness.
