I am working on transmission lines, so i decided to simulate the real delay of electromagnetic waves in a cylindrical transmission line. base on my calculations and formulas, the delay of a wave in a 1 meter oil filled transmission line is 5ns , but when I simulate this simple structure in CST MICROWAVE STUDIO, I see a wrong time delay about 1 ns. so what do you think about this bug?
I do not know the answer, but the simulation results may contain a precursor field that will travel up to the speed of light; em/optical precursors are e.g. very high frequency components that essentially do not interact with the structure and just travel at c. However, your 1m/1ns arrival seems to be in excess of c
But I could imagine that there might be numerically generated precursor-like artifacts, depending on the integration scheme and/or the meshing; which for iterated or high order schemes integration might create small numerical artifacts faster than c.
So (a) is this 1ns signal you see likely to be some tiny numerical artifact? and, if not, (b) are you sure you have not made a mistake? CST is fairly widely used in quite demanding contexts, and anything that would produce such an error would, I would have thought, been already noticed.
I've just run a CST time domain simulation for a coaxial line with lossy polyimide as dielectric (er 3.5) and the port signals show a delay just over 6 ns S21, which is what I expect.
Outer radius 9 mm inner radius 3 mm, 0 to 10 GHz. lossy polyimide and annealed copper.
You may have been launching a higher-order mode which was cut-off but not far from its cut-off frequency, so the exponential fall-off length was quite long, resulting in strong evanescent (inductive) fields at the far end.
- Badges
- Science topic