MoM singularities problems

Hello oussama1,

If you would like us to take a look, the best is to provide the model so that we can do investigations.

Since I don't see the model on the issue, I'll make suggestions for you to investigate.

• Zoom in on the area that you suspect has a problem (around 150 MHz, 440 MHz), etc. - you only need to investigate one of these for now. Reduce the frequency range around the 'problem area' and increase the number of frequency points. Keep zooming in in by reducing the frequency range more and more. If there seems to be a jump in the results from one frequency to the next, it could be a solver problem, but if the results remain the same, it might not be a solver problem at all. (We have seen cases in the past where users think there is a problem with the solver and then it turns out to be a physical phenomenon.)
• Have you performed a mesh conversion study? Refine the mesh and see if the results remain the same.

As I said, if you provide the model, we can have a closer look.

Hello, JIF

the parameters of the design are : 

h=50e -3 , sb=1.5e-3 , s=0.7e -3 , where the z components for each wire is h+sb+s, h+sb-s, h-sb+s, and h-sb-s . the wire lengths L=1  (units is ( m ))   the loads  ZL=100 ohm  (all of them)

i have discovered one thing, if there is only two transmission lines, no distortions occurs. however, if another single transmission line is add to the first two, this distortion will occur again. 

i want to thank you about your replay. i m very happy about your kindly replay, thank you so much.

i want to say that                                                                                                                                                       ___________________________________________________________                    

     __________________________________________________________                                                                                      ___________________________________________________________

Load                                                                                    Load                                                                           Load                                                                                                     Load

     __________________________________________________                                                                                      ____________________________________________________________

                  1-  no distortion                                                                                                                                                        2- there is distortion

note that, the electric field is parallel to the line as was shown in the first post.

thank you so much for your help

i m waiting your response.

i really i become so happy when i saw your post. It is very useful, thank you

Could you add the model(s) to the forum? That saves me time, since I don't have to rebuild the models (and I could make an error while building them). Once you have added the models, I'll have a closer look.

hello

this is the model, it is constructed in FEKO suite 7.0

thank you in advance

Unable to find an attachment - read this blog

note that, i have sent you the model, where the frequency range is only 425e6 to 453e6 , please extend it to  1e6 to 1e9 Hz to see the whole desired range 

oussama1,

I think these resonances are 'real' and this is not a problem in the MoM. I can't think what would be causing them, but if I think of something, I'll let you know. Maybe someone else on the forum has an explanation for you.

hello

thank you so much for your help. the problem is that this distortion occurs only at the case when the electric field is parallel to the line, secondly the transmission line model cannot predict this resonance (the graph is smooth), also this resonance does not occur when there is only  two wires above ground as explained above. i tried to modify the transmission line equations to obtain this resonance but i couldn't.

one question: is it possible to use FEM (Finite Element Method) to solve this problem ? or FDTD ?

thank you so much

i will try to find a solution

and i hope some one give us a complete explanation

thank you .

Have refined the MoM mesh and also used the FDTD solver (not necessarily the best mesh solved, since server time is limited) but results are in reasonable agreement. 

I concur with JIF. This is a characteristic of the model. See attached image.

<?xml version="1.0" encoding="UTF-8"?>

thank you so much

can you explain how could i use the FDTD method to simulate the results ?

Remove the infinite ground. For the FDTD boundary set a PEC boundary at the bottom and the other ends of the bounding box keep on default.

You will need a fine mesh to ensure enough voxels due to the small wire-wire spacing.

I used a GPU, else the solution time would be a few hours. Memory was about 2.5 to 3.5 GB depending on the mesh.

thank you so much mel.

hello@JIF

I want to tell you that the software FEKO and your suggestions have helped me to solve the problem that we have discussed above(about one year ago). As consequence, I have written a paper about this problem. this paper is just published in IEEE Transactions on Electromagnetic Compatibility, which is a very good journal.  the paper link is

O. Gassab, L. Zhou, Z. Zhao and W. Yin, 'Effect of the Scattered Field on the Terminal Risers in Twisted-Wire Pairs With Floating Loads,' in IEEE Transactions on Electromagnetic Compatibility.  https://ieeexplore.ieee.org/document/8600362

the resonance frequencies that appear in the DM response is due to the Scattered field effects that were neglected in the transmission line model. Now with this paper, the tranmission line model is improved. 

thank you for your help, and your software FEKO is very amazing.

Oussama Gassab

@mel