Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by JIF

  1. Hello Mohamed Hamdalla, FEKO has an option to export a .rhs file (Right Hand Side) that represents the source information. Unfortunately I don't have any experience with reading that file, but it is listed in the documentation, but it looks like it simply shows the code for reading the file in Fortran. I think you can read it in any language that reads in the values in the same format, but I have never tried to do this myself.
  2. Hello pmcardle, Are you modelling the platinum as a sheet or a volume? If you can model it as a sheet, the meshing requirements for the sheet would be similar to those for PEC. If you model it as a volume (dielectric with equivalent material properties), the meshing requirements with auto meshing could be quite high (the mesh size is determined by the wavelength in the medium for dielectrics, but the equivalent wavelength in the medium will be very small (infinite for PEC). The better way to model this would be to model the bounding region with sheet metal faces.
  3. Hello shashank19, I understand that you can't share the model, but are you able to reproduce the same problem in a smaller, simplified model that you are able to share? What version of FEKO are you using? We have never seen an error like this and would really like to investigate it further, but without a model, we can't seem to reproduce the problem and thus we can't fix it or give you advice. Are the RL-GO faces in contact with the MoM faces? This is not allowed (or should not be allowed). Maybe that is what is causing the problem?
  4. Hello shashank119, Are you using a full commercial version or a student version? Maybe you ran into a limit of the EDU version (https://altairuniversity.com/feko-student-edition/)? Are you running on a cluster? Maybe your time on the cluster ran out? These are all guesses, since FEKO would usually give more detailed information in addition to the SIGABRT message that indicates what caused it. Do you have more info in the output that you didn't include in your post? Are you able to send the model so that we can try to reproduce it?
  5. Hello Mohamed, Yes, the configuration in your image is what I would expect when adding loads close to the end of a wire (in this case you have a single conductor above a ground). It does not have to be an infinite ground as in your example and could be any geometry (as an example, think of a loaded wire connected between cars or any other geometry).
  6. Hello Mohamed, If there is nothing connected to the end of the wire, it will be floating. Think about what the model would like if you built it in real life - soldering a resistor to the end of a wire will result in a floating resistor. The port simply allows you to add a resistor anywhere along the wire (think of it as cutting the wire and soldering the resistor between the two wires).
  7. From the error message, it sounds like D:\Temp does not exist. Have you tried creating that directory? Or does it already exist?
  8. Hello FieldForcer, If you don't define any transform, then the copies will all be at the same place. The idea is to use a translate / rotate transform that moves the objects. See the gif below. If you are trying to model an array (only an array with no other geometry), you sould also look at the finite array capabilities (see screenshot below) that allows linear, cylindrical and custom arrays to be created and allows faster simulation using the finite array modelling feature. When using that feature to construct the model, you can also perform a solution without using the finite array feature (then performing the simulation using standard solution methods such as MLFMM or MoM). When using this feature, you create only one element of the array and the other copies are created by the array definition.
  9. Hello FieldForcer, There are a couple of options in CADFEKO. I don't have a good mental picture of what your structure looks like, but your options are: Copy and translate: The "copy and translate" or "copy and rotate" feature allows one CAD part to be duplicated and moved / rotated. It allows multiple copies to be created at the same time. Scripting: The scripting features in CADFEKO are quite powerful. Almost anything that a user can do manually in CADFEKO, can be done via scripting. Scripting then allows the user to add FOR loops and conditional statements so that complex models can be created that would have been tedious to create without scripting. CADFEKO also has a script recording feature that makes script creation quick and easy. Simply activate script recording, perform the action that you want to perform once, stop script recording and then customize the recorded script by adding loops and conditional statements. I hope one of these options provide you with the functionality that you are looking for.
  10. Hello cpech, The problem is that you are connecting to the ground directly from a pin, while the shield is in the air. Have a look at Pin2 for the Front connector, it connect via a capacitor to ground, but in both cases the shield is not connected in any way to ground. We have not been able to prove that the formulation works for these kinds of configurations and thus we have added an error to prevent users from incorrect results. Usually the shield will be connected to your structure in some way (very low resistance or a capacitance). If you add such a connection from the shield to the ground, you can measure the voltage of the pin relative to the shield and the error would be resolved. Note that I'm not an EMC / cable expert, so maybe some others on the forum can also advise.
  11. Hello Wilson, It depends on what you want to model and thus it is not easy to advise. Most people would simply union, but creating holes could also work.
  12. Hi Wilson, I'm glad the problem has been solved. The warning simply states that you have requested a near field calculation on the axis of a segment, but this is not possible (singular field). I do see that you used an assembly for your wires. This might not be a problem if your wires only connect (touch) at the start and end of wires, since the segments will still align correctly. If your wires and your box never touch, all should be find. In general, you should union to ensure correct connectivity. I suspect you know this, I'm just making sure.
  13. I have replied in the related topic, since I think they are indeed related.
  14. Hello Wilson, Without a model and without pictures, it is quite difficult to follow what you are doing and thus difficult to comment if what you are doing is correct. Are you able to reproduce the problem with a simplified model of send screenshots of the tree. Why do you unlink the mesh? When you mesh, a mesh representation will be associated with the geometry. When you unlink, you create a disconnected mesh instance. The source will still be on the geometry port (not port1_1) as far as I know and that is probably not what you want. The error stating that there is no excitation means exactly that. You somehow created a model and there is not source (voltage source, plane wave, etc.). From the steps you mention, it seems to be correct, since you added load, but no source. Adding a port is not the same as a source. A port is simply a connection point where sources or loads can be connected (or serve as a port for S-parameter calculations).
  15. Hi Wilson, That error usually happens when you have a port, but have not meshed the model yet. Basically what happens is that CADFEKO has a source or a load that it needs to write to the *.pre file, but there is no mesh instance for the port. Simply meshing and saving should resolve the problem. But, since you stated that you did mesh, maybe something else is causing the problem. Are you able to provide the model so that I can take a look? I would recommend that you upgrade to the latest version of FEKO (version 2018).
  16. I understand, many of our customers can't share the models. I would really appreciate it if you could replicate the problem in a model that you can share. We are aware that the mesher does not always adhere to the meshing requirements and we are working with them to find a solution to the problem. It has thus happened on occasion that a few elements would result in a warning, particularly if the coarse auto meshing has been used.
  17. Hello shashank119, Coarse segmentation of the model can have an effect on the results. The exact amount of the effect depends on the model and what you are calculating. As an example, performing RCS calculation (or far field only calculations in general) is much less sensitive to coarse segmentation due to the integration nature of the far field calculation. Coarse segmentation can have a big impact on nearby near field values and port impedance values. In general, if FEKO gives you a coarse segmentation warning, it is advised to resolve it by meshing finer (possibly only where the segmentation is coarse). Thus, your solutions include meshing the entire model finer or locating the face(s) causing the warning and applying a local mesh refinement. If you use the auto meshing in FEKO, you should not run into these problems (if that is the case here, please provide the model so that we can investigate it further). For extremely coarse segmentation, FEKO will give an error since the answers could then be completely wrong.
  18. Hello Mohamed Hamdalla, You can add capacitive loads on the curved wire in the same way that you would add them on any other wire. Define your ports and add your loads. You can define a segment or vertex port to be located anywhere along the wire. There are restrictions for the vertex port and you won't be able to add a vertex port at the end or start of a wire when it is not connected to another wire or a face.
  19. Thanks for the file. Please change the file by replacing (line 2) setEnv([[FEKO_USER_HOME__SETUP]], [[%APPDATA%\feko\2018.0_]]..getEnv([[FEKO_INSTALLPATH_HASH__SETUP]]), true); with setEnv([[FEKO_USER_HOME__SETUP]], [[c:\temp\feko\2018.0_]]..getEnv([[FEKO_INSTALLPATH_HASH__SETUP]]), true); This will avoid the use of %APPDATA% that I suspect has a non-ASCII character in it (probably your user name on the PC includes a non-ASCII character). You have installed FEKO into the "Program Files" folder (nothing wrong with that and it is the default) and Windows (not FEKO) requires admin rights to edit files in the "Program Files" folder. I think that is what you are referring to when you say that you don't have permission to change the file.
  20. Hello Pawel Sorry, I requested FEKOenvironment.lua, but that was a mistake on my side. I should have asked for FEKOenvironmentFromSetup.lua instead (located in the same folder). Why? Line 66 is the following: setEnv("FEKO_USER_HOME", (getEnv("FEKO_USER_HOME__SETUP") or [[%APPDATA%\feko\%FEKO_VERSION%]]), OVERWRITE) As you can see, the line is trying to set the value of FEKO_USER_HOME (where configuration settings and logs are stored) by either using the value of FEKO_USER_HOME__SETUP (set in FEKOenvironmentFromSetup.lua) or using the value of %APPDATA%. Thus, I need to see FEKOenvironmentFromSetup.lua to see if something is not correctly. The alternative could be to change the line in FEKOenvironmentFromSetup.lua that sets FEKO_USER_HOME__SETUP to point to a folder that definitely does not contain unicode characters. Below is an example line that you can use instead of the one that you currently have in FEKOenvironmentFromSetup.lua (you may have to create the C:\temp folder). setEnv([[FEKO_USER_HOME__SETUP]], [[c:\temp\feko\2018.0_]]..getEnv([[FEKO_INSTALLPATH_HASH__SETUP]]), true); Let me know if this solves your problem. I hope it helps.
  21. Hello Pawel, This is probably the same problem as described in: The problem most likely that %APPDATA% points to a directory that uses a unicode (non-ASCII) character that we don't currently support. If you don't manage to fix the problem with the details on the forum topic mentioned, please attach the FEKOenvrionment.lua file so that we can help you (the path to the file is shown in your screenshot).
  22. There is no easy way to just "not sample on segment vertices", but the near field request has an option to "sample on edges" or not and this moves the sample points by half a increment. Depending on your model, that could be the quickest way to avoid these points.
  23. Hello EMCanuck1, Voltage sources can be added on segments (vertex port or segment port) or on the edge between triangle faces (edge port) or on microstrip ports (not applicable for your question). Are you referring to a segment / wire port or a edge port? For wires, there are segment ports and vertex ports. A segment port applies the voltage over the length of the segment that has the voltage source, while the vertex port applies the voltage between segments with a "zero length gap". In both cases FEKO does not add additional capacitance. The difference between the two will be dominated (when applicable) by the difference in phase due to the difference in the voltage source position (difference in the length of the gap). For edge ports (if you are modelling the antenna as a cylinder), no additional capacitance is added either and the voltage source is applied on the boundary of the two faces. If you have more questions, please add some pictures or models so that we can better understand what you are referring to with your question.
  24. Hello shashank119, Warning 88 has been removed since FEKO now detects these triangle singularities and changes the integration strategy to use a singularity cancellation routine. Thus, warning 88 is only applicable to older versions. Regarding warning 81, the text description is "Singular field on the axis of a segment" and thus it applies to segments, not triangles. I assume you are requesting the a near field value on the axis of a wire segment and there the field is singular. I would suggest that you move your near field request point away from the segment to avoid the warning. Again, I would not sample too close to the segment either (a distance equal to the length of the segment is safe, but I have not done an accuracy study in order to provide more details regarding the recommended sample distance).
  • Create New...