Possible bug in MWS5.1?

I am simulating a structure with periodic boundary conditions along y and z directions and open boundary conditions along x direction. A plane wave is incident upon the structure from the negatif side of the x axis. There is a probe in front of the structure. The distance between the probe and the structure is far enough that once the pulse passes through the probe, still the front of the pulse wouldnt have hit the structure. So basically I shouldnt have probe magnitude larger than 1. Interestingly, the probe magnitude turns out to be larger than 1. Things get even wierd. To get probe magnitude equal to 1, I had to use a very very high mesh density and I had to turn of PBA and to switch to Staircase meshing.I also had to turn of automesh on all features.

Isn't it possible that the reflected wave is in phase with the incident wave so they are superimposed to obtain a magnitude of greater than 1? You should also find that at a certain frequency and multiple of this first frequency that the magnitude is greater than 1. Your curve of a wide frequency sweep should look like ripples. Also if you changed the location of your probe, the frequency at which the magnitude is observed to be greater than 1 will change.

Dear Sassyboy, I am talking about the probe signal in time domain. Not in frequency domain. If the light doesnt travel back from the feature faster than the speed of light probe signal can not be greater than 1. Note that there is enough distance between the probe and the structure. So during the first passage of pulse there is no reflected light. A better way to illustrate my point: Assign all the features to vacuum. And run the simulation. There shouldnt be any reflected light. Because everything is vacuum. In the time domain, the probe signal amplitude must not be greater than 1. I hope this clarifies the point for you.

Hi Irfan,

I just tried a model like this (periodic boundaries, Plane wave from on side...) in my case the results look very reasonable. If the probe is far away from any dielectric surface, the incident signal travels without perturbation. (Signal strength is one) Only if the probe gets closer to the surface I get a overlap of reflected and incident wave.

Could you post an easy example for which you see this problem?


F

Dear RFSimulator,
It seems that I am having this problem only with this structure. You can simplify the structure by deleting some of the layers along the propagation direction. But it is still odd. I mean the structure shouldnt matter.

Hi Irfan,

jep, looks like a bug. From what I see, the plane wave is already launched into the structure with an amplitude slightly larger then 1. I place some additional probes. The amplitude does not change inside thee structure. The calculation itself seems to be correct. The pulse propagates undistorted till it hits the structure. Since this does not happen with other structure it would be interesting to see what is special in your case.

Have you contacted the CST support? You should send the model to them. From my experience the CST developers fix bugs pretty fast as soon as these problems are knows?.


F.

Dear RFSimulator,
Thank you very much for your help
I am sure CST guys will fix it vey soon.