Getting "terminated abnormally" error with CST MWS

Hi folks,

I keep on getting this error when I try to simulate a liquid crystal cell's transparency for THz applications. Some more facts: the spatial dimensions are 4.525 mm x 4.525 mm x 3.9 mm. The liquid crystal layer is sandwiched between two silica glass plates. I want to simulate the S2,1 parameter for transmission. Therefore the boundary conditions I've chosen are:
xmin, xmax = conducting wall (in reality we will insert copper spacers on both sides)
ymin, ymax = magnetic tangential = 0
zmin, zmax = open
Also I chose the symmetry planes (in order to reduce the amount of mesh cells):
yz: electric tangential = 0
xz: magnetic tangential = 0
xy: none
When I look at the port modes, eveything seems to be correct -- but the simulation won't succeed! I have about 370 million meshcells and I even decreased the transient solver's stability factor by 0.2. I'm aware that this simulation might take a very long time but it throws the error before excitation even begins!
Does anyone have an idea, what the problem could be? If you're interessted in the model, please let me know.

Greetings from Germany,

thestranger268

A little update:
The simulation succeeded after I turned off the adaptive mesh refinement and decreased the number of mesh cells. Maybe it was a memory based error...
But -- now I have some ripples in the S parameters due to the high resonating structure. I postprocessed the simulation results with the AR-filter and re-ran it. Now the S parameters aren't that bad any more, but until that it was quite an effort... Is there any other possibility to simulate resonant structures? I also tried the frequency domain solver with the option "resonant: fast s parameter" but there the ripples were very intense and there is no AR-filter function...

I would appreciate any kind of help!

Thanks in advance.

HI
i have the same error.if you solve it,can help me?

have you tried to turn off electric shielding for all ports?

I suggest you to use the frequency domain solver instead of transient domain solver that is better for a resonant structures investigation.

Old thread and long solved, but I am surprised why this 1D problem (3 stacked materials) is analyzed with a 3D EM solver. The math to solve this in closed form isn't too complicated, and then the numerical solution would be some lines of Matlab code.