I am trying to understand how the line impedance calculation works in CST Studio Suite, and for that I have tried to replicate the measurements done in [1], as in the image below. To do that, I created the model you can see below. It is simple 3 silver metal strips on top of a GaAs substrate. My goal is to replicate the results with the assumption of QTEM modes and thus validity of the transmission line model. In line with a previous discussion in 3DSwym CST - How to use CST as a 2D mode solver for quasi-TEM transmission line analysis, I use the time domain solver and use the "Calculate Port modes only" with the "Inhomogeneity port accuracy enhancement" enabled. I also make sure to use a fine mesh in the thin metals. I have made sure to define a waveguide port that has the mode well confined and also used a fine starting mesh to ensure convergence. I also used symmetry for the odd mode to speed things up. However as you can see the results are very odd and I cannot understand why. A big contributor I suspect has to do with not knowing how the 1D results depend on the waveguide port definition. Indeed, when I put a waveguide port enclosing all the plates as you can see below, I get a multimode solution and, in fact, the solver terminates abnormally when I let the port have 3 modes and with 2 modes it finishes but does not calculate line impedance for the mode of interest (the one in picture below). If anyone could chime in on this topic or point to resources I would be grateful.

The dimensions I'm using here are the same as in [1]

References:

[1] - E. Tuncer, Beom-Taek Lee, M. S. Islam, and D. P. Neikirk, “Quasi-static conductor loss calculations in transmission lines using a new conformal mapping technique,” IEEE Trans. Microwave Theory Techn., vol. 42, no. 9, pp. 1807–1815, Sep. 1994, doi: 10.1109/22.310592.