Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


mel last won the day on September 19

mel had the most liked content!


About mel

  • Rank
    Super User

Profile Information

  • Gender
  • Are you University user?

Recent Profile Visitors

2476 profile views
  1. It seems the API does not have this function, yes. See the bottom of the .pfg file created during optimisation. You could just parse this file. In the below example analysis number 11 with goal 1.024278 was the optimum. (This is also printed out at the end of the optimisation, and after each iteration in the screen output.) <valueSeries autoLegend="0" label="global goal" legend="global goal" lineColour="6" markerColour="0" markerSymbol="-1" unwrap="0" lineWeight="1" smithMode="0" leftOffset="0" bottomFactor="1" dataScale="0" smithChartOffset="0" lineStyle="0" isdB="0" dbFactor="10" markerSkip="0" bottomOffset="0" independentVar="9" leftFactor="1" visible="1"> <coord0>1.000000000000E+00 2.660984357729E+00</coord0> <coord1>2.000000000000E+00 5.576534763817E+00</coord1> <coord2>3.000000000000E+00 7.111736604099E+00</coord2> <coord3>4.000000000000E+00 3.930480833089E+00</coord3> <coord4>5.000000000000E+00 5.432063754701E+00</coord4> <coord5>6.000000000000E+00 4.588627015879E+00</coord5> <coord6>7.000000000000E+00 6.504855382087E+00</coord6> <coord7>8.000000000000E+00 6.810322213765E+00</coord7> <coord8>9.000000000000E+00 1.334521460692E+00</coord8> <coord9>1.000000000000E+01 4.004993085123E+00</coord9> <coord10>1.100000000000E+01 2.134784482058E+00</coord10> <coord11>1.200000000000E+01 1.024278584819E+00</coord11> <coord12>1.300000000000E+01 3.037102051344E+00</coord12> <coord13>1.400000000000E+01 4.972591349106E+00</coord13> <coord14>1.500000000000E+01 1.639715252672E+00</coord14> <coord15>1.600000000000E+01 2.643155741597E+00</coord15> <coord16>1.700000000000E+01 1.518139618721E+00</coord16> <coord17>1.800000000000E+01 1.737249594941E+00</coord17> <coord18>1.900000000000E+01 1.312758545407E+00</coord18> <coord19>2.000000000000E+01 1.183687323123E+00</coord19> <coord20>2.100000000000E+01 1.283535425539E+00</coord20> <coord21>2.200000000000E+01 1.182512813433E+00</coord21> <coord22>2.300000000000E+01 1.251935038082E+00</coord22> <coord23>2.400000000000E+01 1.181573426912E+00</coord23> </valueSeries>
  2. I don't have any ideas on alternate ways to model this (yet). Could you give some more background on what you are trying to model?
  3. If you want to define the PEC faces of the patch, use "Perfect Electric Conductor". The dielectric faces would be "Dielectric Boundary".
  4. It would be best if you contact the Altair representative in your region for these details.
  5. The internal error is given due to extreme dimensions of the model. The multilayer substrate is extremely thick in terms of the wavelength. The multilayer substrate is modeled using interpolation tables and these appear to break down for these extreme dimensions. I don't think Feko is going to be able to model this, even if the internal error message is addressed.
  6. Looks like you broke Feko The developers will address this.
  7. You have the model setup correctly already. A "coarse" mesh is perfect for PO with a plane wave as the source. If the model is symmetrical you can save time by not computing from 0 to +45 degrees.
  8. The POSTFEKO file is about 6.8 GByte and this is due to the excessive number of modes (194 038) included in the spherical modes source. Using just one source and running the solver shows in the out file: EXCITATION BY AN IMPRESSED SPHERICAL MODE Name: so_0 Excitation index: 1 Frequency in Hz: FREQ = 3.50000E+10 Wavelength in m: LAMBDA = 8.56550E-03 Source position in m: x = -9.42857E-02, y = -9.42857E-02, z = 0.00000E+00 Orientation of source: THETA = 0.00 PHI = 0.00 Number of modes: 194038 (of which 194014 suppressed) If you reduce the number of modes and thus the sph file the POSTFEKO file will be reduced to a size that you can open/view.
  9. Have a look at Example Guide example H3 "Horn feeding a large reflector"
  10. Feko does not do static analysis and thus the DC components are estimated. Results could differ based on the lowest frequency simulated. You could try to simulate from a lower frequency. Also make sure you are using Feko 2018 or later where options for the DC estimation was added/improved. This is accessed as seen below.
  11. Login to support.altair.com and download the Characteristic mode synthesis and design script.
  12. See this post with script.
  13. mel

    ERROR   49000

    If you have multiple nodes, use only say 8 processes per node, so maybe 64 processes spread out over 8 nodes. If there is symmetry in the model, apply electric or magnetic symmetry. The MoM memory will reduce by a factor 2 or 4 per symmetry plane. You can also try to increase the size of the FEM part (to allow using more processes) by meshing the FEM finer (I see there are warnings about the tetrahedra sizes) Or you could add air dielectric around the FEM part which would also reduce the size of the FEM/MoM boundary and thus the memory. See videos here and here.
  • Create New...