Jump to content

Torben Voigt

  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by Torben Voigt

  1. Hi @Kurt Knox, the attached script could be interesting for you:Advanced_Fieldpostprocessing.zip
  2. Hi @ebdaculan, Your model seems to exceed the limitations you have with the Student Edition, see here. https://altairuniversity.com/feko-student-edition/
  3. Hi @mgmv7 You may use the attached *.ffe file (FEKO far field result) of an isotropic radiator for an RX far field antenna: The *.ffe file represents an antenna with 0dB Gain in any direction. see above similar to receiving antennas you can use *.ffe files for Far Field Point Sources: You will also find examples in the ExamplGuide.pdf:
  4. Hi @Farai, I assue this was done in post-processing with given results for differnet rotor seetings. You may either post-process the results in MatLab (or Altair Compose!) or in POSTFEKO using lua scripting. Unfortunately I don't have any examples on this.
  5. Hi @sangwoo.kang, the ""Electric dipole" is a Hertzian dipole meaning that it's an ideal dipole. A wire dipole is a real dipole, incl. spatial extends.
  6. Hi @lostarmour, Use the attached model: 08262019_Feko_two_blades_alt_2018.cfx What you probably didn't do is coarsening the mesh.
  7. Hi @lostarmour, You should make all regions PEC and then use "Simplify". This deletes 24 (small) redundant faces: Then I found out that these areas are the tricky parts: I closed these two and now it works 9,072 triangles (compared to 51,554 in the original model): 08262019_Feko_two_blades_alt.cfx It solves in a couple of minutes. Please make sure to refine the mesh slightly and compare the results. This is just an example.
  8. Hi @Marcus Chang, 20 lambda is just a rough estimation for RL-GO, it always depends on the model. Your model seems ideal for PO or LE-PO (only one interaction on the reflector). The solution should be quite accurate. RL-GO would work but it will require much more time because the near field source will be transformed into a spherical modes source first (and same for the Rx antenna). Just apply PO (full ray tracing) to the reflector face and you should be good. Also compare with LE-PO (full ray tracing).
  9. Hi @Haj Hassan, This might work if you do a Grid search (optimisation method). It's similar to a parameter sweep, but without automatically merging the results. In the graph you would then have 1,2,3,...
  10. Hi @Haj Hassan, interpreting the far field in a lossy halfspace unfortnately isn't easy. When using the Reflection Coefficient Approximation for example, FEKO deactivates the computation of far fields below z=0 completely. In principle the far field in FEKO is computed in an infinite distance by neglecting the term exp(-j*beta*r)/r. bet however is different for the two media.. I would suggest to rather calculate near field values in a certain distance if you're really interested in the field below ground.
  11. Hi @Kuzi, To save time it would be good to also have these files: fek, bof, out.
  12. Hi @Kuzi, this doesn't seem to be the correct model. In your pics I see an antenna structure and also a polar plot of far field results. In the attached model only a plane wave excitation is present and the far field request is set to "Calculate fields in plane wave incident direction". Could you please check?
  13. Hi @bouvy, Does running a simulation from CADFEKO work?
  14. Hi @Kuzi, I think there are methods to calculate the RCS based on radiated near field results (near field to far field transformation). I don't know details about it, but Dr. Andre Rennings may have some knowledge here.
  15. Hi @designer, I don't see a problem here. When you select all your mesh parts in CADFEKO and then go to Export -> Mesh -> Gerber mesh, the exported *.gbr file seems to be correct:
  16. Hi @SUG, FEKO assumes wires to be thin. This means that a segment (= mesh part of a wire) should not be shorter than 5 times the radius of the wire. Else using cylinders is recommended. Looking at your screenshot, you should clearly model the coil as a cylinder. The attached model shows the idea. coil.cfx
  17. Hi @Kuzi, I can only see a small part of the polar plot. Are you sure that the main lobes are in the theta = 180° direction? Could you attach the files (cfx, fek, bof, out)?
  18. Hi @SUG, New FEKO user will find very helpful examples in the ExampleGuide.pdf and basic knowledge in the GettingStarted.pdf:
  19. Hi @Dimon, Presently diffraction effects from edges and wedges of RL-GO PEC faces meshed with curvilinear triangles are supported.
  20. Hi @Yasaman, the layered dielecrics are very thin comared to the wavelength: I don't think you will be successful wit MoM. It may be feasible to use FEM instead. I attach a FEM version of the model but you will need to see if reducing the mesh size of the thin dielectric layers changes the results. full_turnstile_fixed_FEM.zip
  21. Hi @Kuzi, You may use a Parameter Sweep for the varying height. In POSTFEKO the horizontal axis could then be height.
  22. Hi @Marcus Chang, No, in FEKO you will always have one mesh (= one geometry) per model. You can of course use the Parameter Sweep to automatically create different models.
  23. Hi @Lydia He, these are just comments (not warnings). They say that you can increase the speed of your simulations by activating parallel computation and also that the former *.str file has been re-written. FEKO stores the currents in *.str files so that they can be re-used in later simulations of the same model, if e.g. different requests are made.
  24. Hi @Marcus Chang, I don't think that the extension (efe / hfe) are needed. However, you can do it like this: properties.EFieldFilename = "02_NF source_PEC corner reflector_190406_sphere placed 72 lambda_alt_NearField_"..i..".efe" properties.HFieldFilename = "02_NF source_PEC corner reflector_190406_sphere placed 72 lambda_alt_NearField_"..i..".hfe"
  25. Hi @Marcus Chang, the code temp.lua seems to work. It successfully adds 3 NearFieldData in CADFEKO: What's still missing is the "Start reading from line" which is 1 in all cases. You should change line 15 properties.ReadFromLine = "1" to properties.ReadFromLine = "1+(20*20)*"..(i-1) 20 is the number of field points in x and y, so this will ensure that the starting number will be like 1, 401, 801. Hope this helps!
  • Create New...