FDTD problem
Altair Employee Hello,
I have the following problem.
I would like to solve my electromagnetic problem using the 'Time Domain Solver (FDTD).
First of all I activated the Finite Difference Time Domain from the 'Solver settings' menu.
After doing this, the problem I find is that I can't apply the mesh to the rectangular waveguide, giving me a problem on the ports I've applied to it.
Can someone help me?
<?xml version="1.0" encoding="UTF-8"?>
Answers
-
Hello @John1989
The FDTD solver currently only supports wire ports and edge ports, so you could use a pin feed to excite the waveguide if you want to solve it with the FDTD. You can refer to Example A-11 in the Feko Example Guide: The model that uses a pin feed to excite the horn antenna can be switched to and solved with the FDTD.
It is on the development roadmap to support waveguide ports for the FDTD solver. If this is of importance to you, or if you would like to be informed when it is supported, please request this through the official support channels - see Altair Connect for support contact details or to create a support request.
Kind regards
Madelé
0 -
I entered a pin feed to feed the guide, I managed to meshare. But when I started FEKO Solver I had these problems.
<?xml version="1.0" encoding="UTF-8"?>
0 -
Can anyone help me solve this problem?
0 -
Hi John
Have a look in the help folder of Feko for the ExampleGuide model Example-H04-Optimise_Waveguide_Pin_Feed_LocationIt contains a simple pin feed for a waveguide.
If you can use a different solver apart from FDTD I would still recommend the FEM since the modal port is an ideal port and applies the correct mode (fields) to the waveguide as if the waveguide stretches to infinity away from the port. Waveguide pin feeds are usually not ideal over bandwidths.0 -
Hi Mel,
Thanks for the reply. I'm using FDTD for the resolution for my model because using 4 GPU I decrease the resolution times compared to FEM.
I followed the guide she suggested, I managed to create a model that I wanted but in Feko Solver I have this error.Can you help me?<?xml version="1.0" encoding="UTF-8"?>
0 -
Yes, set a longer 'Maximum time interval'.
You can also add say 50 Ohm loads to the ports - lossless models tend to you require longer runtimes for the wave to propagate back and forth in the grid.
0 -
Thank you so much Mel.
I wanted to ask you one last thing.
I have this project (waveguide with dielectric) solved with both FEM and FDTD.
The scattering parameters obtained with the two methods are not coincident.
I would like to know how I can ensure that results obtained with FDTD coincide with those of the FEM.What can I change in the FDTD project?
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
0 -
Hi John
This is exactly what I alluded to earlier. The pin feed is not ideal and will introduce differences and standing waves.
And it could very well be that the 50 Ohm loads I suggested are not the best match for the waveguide which will also cause standing waves.
0
