Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Loop Antenna design - Internal error in geometry decomposition

Sathya Sai Seetharaman

Please login with a confirmed email address before reporting spam

Hello all,
I am trying to design a loop antenna, which I want to do microwave experiments with. I wanted to do a model first with only the loop antenna, to compare with the reflection parameter S11 that I recorded in my experiment. The loop antenna is to be fed by a coaxial cable and grounded to the outer sheath of the coax (I haven't got to the second part of the design, i.e. grounding). I set the coaxial cable's end to be the input port. I don't set an output port however, since I am not worried about the transmitted signal. Please find my model attached.

The problem however is with the geometry. I thought of using a parametric curve to describe, quarter of a circle and sweep the geometry to join the faces of the coaxial cable's tip to the loop. I seem to be getting an error message saying "Internal Error in Geometry Decomposition". I'm wondering, if despite careful allocation of co-ordinates for the objects, I'm somehow still not joining the faces of objects correctly somewhere.

I also wanted to ask if there might be a neater/more elegant way of doing this design, rather than sticking to a design that would be very true to my experimental design. Without the coaxial cable, the loop antenna is only as good as a split-ring resonator and I'll be studying split-rings in my experiments, so I don't want my model to end up studying coupling between split-ring resonators. Many thanks in advance for answers/advice.
Regards,
Sathya.


2 Replies Last Post 08.03.2017, 04:54 GMT-5
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 04.03.2017, 19:34 GMT-5
The sweep tool is more difficult to use than most other geometry building tools. But in this case, you can accomplish the same 90 deg bend by using a simple revolve operation instead. Use a revolve face operation. I have opened your file and done so, just to show you how it works. See the attached file.

You still have a lot more work to do. And, as always, don't forget to include a reasonable volume of space around your objects, to contain the fields around them. Good luck.

CORRECTION: You tried to sweep a non-circular face into a circular one. You can't do that easily. Although the revolve operation I did succeeded, it will likely fail in the meshing to come. You need to slice your torus differently, so that the face it revolves (or sweeps) from is a true circle. See the rev2 file I attached, which shows one way to accomplish that. I simply moved your block over. You could do it other ways too, of course. Good luck.
The sweep tool is more difficult to use than most other geometry building tools. But in this case, you can accomplish the same 90 deg bend by using a simple revolve operation instead. Use a revolve face operation. I have opened your file and done so, just to show you how it works. See the attached file. You still have a lot more work to do. And, as always, don't forget to include a reasonable volume of space around your objects, to contain the fields around them. Good luck. CORRECTION: You tried to sweep a non-circular face into a circular one. You can't do that easily. Although the revolve operation I did succeeded, it will likely fail in the meshing to come. You need to slice your torus differently, so that the face it revolves (or sweeps) from is a true circle. See the rev2 file I attached, which shows one way to accomplish that. I simply moved your block over. You could do it other ways too, of course. Good luck.


Sathya Sai Seetharaman

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 08.03.2017, 04:54 GMT-5
Hi Robert,
Thank you very much for your timely help. I have the model working now, which I'm attaching here for anyone's future reference.

Your suggestion worked much better than using the parametric curve to join the faces of the torus and the cylinder. Also, instead of using a block to subtract a portion of the loop, I just changed the revolution angle of the torus to build an incomplete one that could still work well enough as a loop. I revolved the other ending face of the torus similar to the previous one you built and extruded its end face further to form another cylindrical portion. I did something clumsy in the end. I just put a little metallic block to connect the outer sheath of coax to the cylinder to ground the loop. I did not expect it to work, but it did.

As for the meshing, there weren't any problems as long as I chose only the outer faces of the grounding block (bits not inside the coax material), I did not have any computing errors. The S-parameter shows a resonance where I expected it be. I have also tested the antenna closer to a split-ring and see that its resonance is excited correctly.
Cheers,
Sathya.
Hi Robert, Thank you very much for your timely help. I have the model working now, which I'm attaching here for anyone's future reference. Your suggestion worked much better than using the parametric curve to join the faces of the torus and the cylinder. Also, instead of using a block to subtract a portion of the loop, I just changed the revolution angle of the torus to build an incomplete one that could still work well enough as a loop. I revolved the other ending face of the torus similar to the previous one you built and extruded its end face further to form another cylindrical portion. I did something clumsy in the end. I just put a little metallic block to connect the outer sheath of coax to the cylinder to ground the loop. I did not expect it to work, but it did. As for the meshing, there weren't any problems as long as I chose only the outer faces of the grounding block (bits not inside the coax material), I did not have any computing errors. The S-parameter shows a resonance where I expected it be. I have also tested the antenna closer to a split-ring and see that its resonance is excited correctly. Cheers, Sathya.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.