Some statement regarding Transient simulation, termination criterion.

HI guys,

Transient simulation. Termination criterion.

" To obtain accurate broadband in Frequency domain the Electromagnetic energy in the computational domain needs to be sufficiently decay"

"If the energy is not decayed sufficiently the accuracy of the results suffer from the truncation error"

I am abit blur.

Anyone understand this statement?
This statement is made on CST transient solver.

Please advice.

Thank you.

If the width of the pulse is T secs, then the sidebands occur at 1/T Hz. So the number of sidebands are inversely proportional to the pulse width, so if the pulse gets long enough, you run out of space to compute the addition of a very large number of side bands.
frank

The simulation uses a time domain pulse (gaussian) that has a wide, continuous spectrum. From simulation, you get the reflection and transmission voltages and currents.

To get the frequency domain S-parameters, you need to do an FFT. If the time period used as input to the FFT is too short, and the voltage/currents are not down to zero, the resulting S-parameters are inaccurate. You will see a ripple on the S-parameters in this case.

The time reqired until the energy in the systems is zero (multiple reflections are down to zero) depends on the simulated object. For a matched transmission line, the signal decays very fast. For a high Q resonator, it decays very slowly.

I guess, the confusion is caused by a misunderstanding. The termination criterion doesn't apply to all transient analyses. Only those that are transformed to a frequency domain representation, e.g. FDTD electromagnetic solvers.