HFSS Simulation error: "failure in processing solution for output.." and pml
I have an error in a HFSS simulation:
[error] Adaptive solution setup, process hf3d error: Failure in processing solution for output.. (7:02:12 PM Oct 06, 2013)
[error] Simulation completed with execution error on server: Local Machine. (7:02:12 PM Oct 06, 2013)
Does anyone have any insight into the error. I am exciting an antenna with a standoff layer over a ground plane with a plane wave, where the antenna is connected to a transmission line. There are pml on the sides parallel to the yz plane and radiation boundaries on the adjacent sides. The response of these structures is somewhat directional towards and away from the PMLs.
More specifically, I am wondering how thick should I make my pml??? 1/4 wavelength? Also, since the response is directional, should I make all of the pmls for absorbing guided waves (with minimum propogation constant equal to some value or set to 0?).
Do you have any reference for what all the different error messages actually mean? I am usually totally confused by most of them. Thanks for your feedback.
-Eric
"The response of these structures is somewhat directional towards and away from the PMLs." - please clarify
PML - Represents an open boundary condition using several layers of specialized materials that absorb outgoing waves.
Why not use radiation boundaries on all sides instead of PML?
Here is what my simulation looks like. So, more specifically, the antenna absorbs the radiation and "channels" it down the x axis where I have PML touching all three layers (the Air box, the metal layer, and the stand-off. On the other side I have two PMLs for the air box and standoff. Radiation boundaries are used on the adjacent sides.
Can I use a radiation boundary to absorb the surface wave at the end of the transmission line? It should be that the PML will be much better for absorbing highly directional radiation or waveguide/surface waves, right? I just assumed a radiation boundary would not work on the end where the transmission line terminates.
Here is another picture where I have highlighted the pml for the transmission line. The surrounding pml for the airbox is nonrectangular so I made that PML manually by drawing the box then assigning the box as a pml.
Radiation boundaries shouldn't be used very close to conductors, so PML is better in those cases. The termination of the line should be a port. From the image I think the box (-x) on the right can be a radiation boundary instead of the PML box.
You can touch conductors with radiation boundaries if the energy is about 30dB down where it touches. You could just increase the size of the radiation boundary box so your patch circuit is isolated from it. You can put radiation at ports. The port will overide it. If that is a waveport, you might want to draw a smaller rectangle, like 5W, around the microstrip and make that the port. If the entire entrance to the box is the port you are probably getting a waveguide mode instead of a microstrip mode. Experiment removing the PML.
Hey Reidintransit, thanks for your input. So, if I understand you correctly, I should keep the PMLs towards +x, except I should convert the one PML that extends the transmission line into a port? Does it matter that I am not going to be using the port to excite the structure (it is being excited by a plane wave through the antenna)? I will think about removing the pmls towards -x as well. Thanks!
Hey Steve,
You are saying I can put a port at the end of the transmission line in place of the PML, right? There is no waveport in the simulation, the structure is being excited by a plane wave (from the far-field) according to the red arrows and I want to look at the microstrip mode that propogates from the patch that is excited. Can I use this port that is 5*w to just absorb the microstrip mode?
Thanks for your suggestions.