Antenna Placement on Aircraft
Altair EmployeeHello there, currently i'm an aerospace student and still working for some task about antenna placement on aircraft and i'm still newbie about this topic. I never use software like CST or FEKO before and im lack of basic antenna knowlegde.
The situation is when i get the result, this warning pop up
WARNING 31096: ADAPTFEKO has reached minimum frequency stepping, results might not have converged
WARNING 31097: ADAPTFEKO has at least once reached the minimum frequency stepping and terminated prematurely (possibly no convergence)
WARNING 31062: ADAPTFEKO: Frequency subinterval 1 of 1 has not converged
Is my result is good enough? I upload my result below. And i want to ask, what else i can explore about antenna placement on aircraft that FEKO can provide me other than far field pattern?
My processor is intel i-3 @1.8GHz and 8Gb RAM
I really need your help. Thank you so much
Answers
-
The warning states that during the adaptive frequency run the minimum frequency range that you set ((freq_max-freq_min)/60) was reached.
In a first step I would suggest that you check plot the results as a function of frequency and make sure that they make sense - e.g. S11(f) should not be greater than 1.
If you want to investigate further, you can try rerunning the simulation using a smaller minimum frequency increment and compare the results to see if they are the same.
Some other thoughts:
- I would suggest that you union the antenna geometry with the fuselage, at the moment there may not be a proper galvanic connection
- you may also want to try using the MLFMM solver, which may be more efficient to solve this problem
0 -
Hello Alberth Prasetya,
The warnings are due to settings that you (or the previous person who created the model) applied on the frequency dialog. See the image below:
The warnings can be removed by trying one of the following:
- Use discrete frequencies. This will definitely avoid the problem, but you then have discrete results only (probably not an issue since it looks like you are running many frequencies anyway).
- Split your configurations to look at input impedance and S-parameters (no far fields or near fields) over a continuous frequency range, but use separate configurations that look only at the far fields at discrete frequencies.
- You could try removing 'Power data' from the list of quantities for the adaptive frequency sampling.
- Remove the settings for the minimum frequency increment and / or maximum number of samples.
Altair Forum User said:Is my result is good enough?
Without running the model, I can't tell you if the convergence warning is a problem or not. You can compare to a discrete frequency run to see if the results match.
Altair Forum User said:And i want to ask, what else i can explore about antenna placement on aircraft that FEKO can provide me other than far field pattern?
This is a rather strange question. You should know what you need to investigate and find software (or other means) to do the investigation and not the other way around. ;-) When doing antenna placement, engineers are usually interested in the following:
- Antenna patterns (usually comparing the pattern on the platform and without the platform and comparing the coverage area to the desired coverage area)
- S-parameter coupling (influence of one antenna on other antennas - usually you want them to not couple)
- Possibly coupling into cables etc., but that would be more advanced and usually inside the aircraft (really depends on the aircraft and requirements)
Altair Forum User said:I really need your help.
I see that your antennas are not 'mounted' correctly. This could also be part of the reason for the convergence issues, but will definitely cause incorrect results.
You have intersecting triangles from the antenna with the plane. These parts need to be unioned to become electrically connected, but it won't be as simpl as simply unioning. You need to mount your blade antenna onto the plane structure. You could remove the flat face that the antenna is mounted on and connect the blade directly or embed the flat face into your plane. You need to fix this before you continue with your simulations.
I would also recommend that you use the latest version. I'm assuming that you are not, but I could be wrong.0 -
Thank you so much @Peter Futter and @JIF for your responses.
But Peter, when I try to run the model with MLFMM, I always get this kind of error . I think I dont have enough memory
And for JIF, now I have already repaied and unioned the antennas and the aircraft , also I try the discrete solution that you gave me. I will run this model immediately and inform you of the result.
By the way i am using PO full tracing because somehow MoM and MLFMM always give me error notif.
0 -
Altair Forum User said:
By the way i am using PO full tracing because somehow MoM and MLFMM always give me error notif.
Based on the model in a different post (see below), I don't think you are using PO correctly and since you are new, I would recommend that you stick to MLFMM initially. I can't seem to open the model that you attached after downloading it (seems to be corrupt and I tried twice), so I can't take a look at what the memory requirements would be.
0 -
Yeah I think it is corrupt. Here is the file below
0 -
Hi Alberth Prasetya
8 GByte is really not that much memory nowadays...
I've had to do some acrobatics with your mesh but I think you can solve it now.
To save memory and runtime I have:
- changed the preconditioner to SPAI
- Used a default mesh size of lam0/5.5, but applied a local mesh size on the faces between the antennas of lam0/7 (to help the antenna coupling calculation accuracy)
- applied the combined field on all the faces bounding the volume of the aircraft (so this excludes the antenna surfaces, else you will get an error)
- meshed a small area around each antenna with lam0/15 for accuracy of the antenna impedance and local radiation in their respective areas
This mesh should give good 'ball park' results, but one would generally want to mesh a little finer than this.
I don't know why you have so many configurations listed on the top left - you could do with only 3 I think:
- SparameterConfiguration for the antenna coupling
- StandardConfiguration to get the pattern of antenna 1 (this will also give the input reflection coefficient for antenna 1when you plot the Source data in POSTFEKO)
- StandardConfiguration to get the pattern of antenna 2
I have attached the model. It will take around 7 GByte of memory. If still not enough, adjust the global mesh size slightly coarser.
0 -
-
For the small area, create an ellipse some small distance just above the surface of the fuselage.
Then click project.
0 -
@mel just 2 more question , what the difference between project and align? and so your step is split the aircraft then make an elips and then project the elips?
0 -
Altair Forum User said:
what the difference between project and align
Align is slimple a 3D translate + rotate all in one. The Project tool projects a face onto another face. You can read more about these tools in the manual or just experiment with the tools.
0 -
alright @JIF Thanks
0 -
hey @JIF why cant I use mesh_fine in local mesh size setting?
I got error 16449 :variable not found at position 0 for expression :
local mesh size : mesh_fine
0 -
If the variable exists, you should be able to use it. You are going to have to provide more detailed stems of what you are setting where for me to be able to reproduce the problem. I suggest a new forum post.
0 -
oh okay @JIF
I want to ask something again with my result. I ran another antenna placement model ( I put VOR antenna on the vertical tail) and I got the result but I dont know whether this result is corrent or not
Here my antenna and my result
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
0 -
I don't know if the answer is correct either. ;-) What I can see from the results and image:
- This antenna (VOR) is not as well matched as the other antenna that you used. This could indicate that you didn't connect the antenna correctly, but it could just be that the antenna is not well matched or is matched at a different impedance.
- Both antennas behave the same and thus you did the same thing correct or incorrect with both.
- They are well isolated (-30 dB) and that is probably what you want.
0 -
:):):) well @JIF I mean both S1,1 and S2,2 show a constant value. is it normal or not? Because back then if Im not mistaken, one of my friend said that S1,1 should show a curve, not a straight line. and you also said that the antenna is not well matched. how can you know it?
0 -
S11 and S22 can be close to a straight line and it definitely does not have to be some curve, but most antennas around the operating frequency will have some sort of 'dip' in the refelection coefficient where it has a better match and that is probably what your friend meant.
When S11 is very small, it means that there is a good match and a large percentage of the power going in at the port goes into the antenna and not reflected back. As an example, if S11 is 0dB, it means all the power is reflected back (terrible match) and when S11 is -10 dB, it means that only 10 % is reflected back (better match) and -20dB translates to 1% being reflected (good match). An operational antenna would usually have an S11 of at least less than -10dB (lower is better).
In your image, it looks like you are close to -3dB and that means that half the power is being reflected - not good for an antenna.
You need to know what you expect from the antenna. Have you simulated the antenna on its own? Placing it on the structure (as large as a plane) should influence the results, but you should still see similar behaviour. Thus, the reason I said that I can't say if the answers are correct, is that I don't know anything about your antenna and its expected performance. You need to know that. Start small and build your model up to be more complex and include more details and check your results with what you expect at each step.
Good luck with your project.
0 -
Hello @JIF I come back hehehe. Thank you and also goodluck with your job.
I want to ask something again. Is it possible for gain of mounted antenna bigger than gain of free-space antenna?
Because it seems that my result show that gain of mounted antenna is bigger than free-space antenna.
0 -
Hi,
Yes, this is possible. You antenna in free space could have a very small ground plane and when placing it on the platform (plane in this case), the larger (better) ground plane could increase the antenna's directivity in one direction (obviously reducing it in other directions). It could also couple to nearby geometry and that could increase the gain / directivity in some directions.
0 -
Oh i get it. Thanks @JIF
Can you help me explain why polar chart for phi configuration give me constant value for all degree? My theta (wrapped) configuration give me similar shape if I compare to 3D plot and I think it is correct but the phi configuration not.
Here is my result :
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
0 -
Oh i get it. Thanks @JIF
Can you help me explain why polar chart for phi configuration give me constant value for all degree? My theta (wrapped) configuration give me similar shape if I compare to 3D plot and I think it is correct but the phi configuration not.
Here is my result :
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
0 -
Altair Forum User said:
Can you help me explain why polar chart for phi configuration give me constant value for all degree?
You are plotting at theta=0 and probably want to plot at theta=90. At theta=0, all phi points result in the same point.
0 -
-
Hi Alberth,
Try the 3D view legend range settings:
0 -
Wow. Thanks for your respond @Torben Voigt. It really help
0
