Can current EM simulation tools accurately predict Antenna efficiency and radiation pattern compared with actual anechoic chamber measurements? Any references for such a research comparing both results?
Yes they can, Try searching for articles based on the software EM suite by CST .
I think the accuracy of simulation depends on several tricks that not every one knows!
I have the same problem myself. I simulated a Cassegrain antenna in CST which is said to be the best EM simulator. The result showed a significant back lobe in the radiation pattern and I could not find the origin.
Anyway I fabricated the antenna and the practical measurement did not show any back lobe! Still I have no exact idea about it. However I believe it is due to the number of cells the simulator uses. Actually I am using a student licence version and the maximum number of cells is limited to 2,000,000 cells.
You should have a look at
Evaluation of the Radiation Efficiency and the Noise Temperature of Low-Loss Antennas
Maaskant, R. ; Bekers, D.J. ; Arts, M.J. ; van Cappellen, W.A. ; Ivashina, M.V.
Antennas and Wireless Propagation Letters, IEEE
Volume: 8
Digital Object Identifier: 10.1109/LAWP.2009.2035148
Publication Year: 2009 , Page(s): 1166 - 1170
Cited by: Papers (6)
IEEE JOURNALS & MAGAZINES
Depends on what kind of antenna you are simulating and the accuracy of the simulation. I had to simulate a patch antenna using HFSS from Ansoft and measure it using Labvolt. The results simulated and measured were almost the same.
In my view and experience, it depends on the antenna you have to measure/evaluate.
If efficiency depends mainly on the illumination of the aperture, i.e. on the effective area, you can simulate that fairly accurately. If ohmic losses and mismatches are involved (such as in an array antenna), then the problem starts being more difficult, unless you have accurate information about the characteristics of the passive components you are using.
validation itself is a big topic for me. having matched results means that you need to understand every detail on both simulation and measurement side. for me, there are two kinds of people who complain about inaccuracy of either simulation or measurement. the first one is those who figure out all the details and find something can't be done, such as some measurement (simulation) condition can't be modeled on the other side. The other kind of people are those who simply are lack of knowledge on either side and the complaints make their life easier.
Simulation accuracy is only 90-99% if measurements are done within a good environment for a well fabricated antenna. These type of results do exist mostly for your patch antennas on a proper substrate and with exact feed point
I agree firmlly with Vasile Surducan! After being a physics-based modeler for years, I can honestly tell you that these so-called simulations and numerical packages are all a scam. They imply 'perfect' reproduction of the phenomenon being studied; but, virially, they are simply made up of the all the equations we know and love, e.g. Maxwell's equations. Furthermore, these simulations face all the trials and tribulations we all face if we attempted to solve such equations. The insidious thing about numerical packages is the purported 'exactness' these programs claim to provide - it's simply not true. You would much better looking up the field equations for the antenna you have and calling it a day! It's cheaper, more accurate and honest.
That last point, honesty, happens to be what this world is most deficient of...
It depends on the accuracy of your simulating software as well as the calibrations of your measuring equipment s. A close tie is suppose to be realized all things being equal
The issue is mostly one of near-field effects. The far field effects, generally considered over 2 wavelengths away, do not materially affect the complex drive-point impedance of the antenna, but may shadow and diffract the pattern.. Anything with an impedance different from that of air and less than roughly two wavelengths from the antenna affects the antenna pattern and drive-point impedance. For microwave antennas the impedance of the insulated substrate affects the far-field result. For a waveguide antenna the mouth of the antenna governs.
It all depends on how you define antenna efficiency. Any simulation (ex: comsol) will show you a good results if you are carefully with physics that is describing your antenna design. If your goal is to optimize antenna design, I suggest to follow up NASA's work in the following link: http://idesign.ucsc.edu/projects/evo_antenna.html
They used evolving algorithms to optimize antenna design. Pay attention on their definitions of antenna efficiency. Hope it helps. Good luck.
with a full wave CEM tool, one can simulate very accurate antenna radiation pattern as compared to that measured in a full anechoic chamber. There is usually no accuracy issue for in-band, co-polarization and main lobe of the antenna. For far out sidelobes, cross polarization and out-of-band measurement, it then depends on the accuracy of the modeling. This is especially difficult for the small components typically at the feed. And if the feed has dielectric material, the complex dielectric constant of the material at various frequency of interest must also be known.
Accuracy some time requires mitigation of simulation data. The measurement inaccuracy could be reasoned out based on measurement place. Hope some DRDO people or isro peopole
Respected scholar,
Accuracy is a burning issue among researchers for quality of research,
it is not depends on only the person.
if we having software tools but user not expert,
regarding hardware aspect-it is totally depends on skills and engineering.
as per various survey and results, I found that up to 95% accuracy results may match.
When using a commercial software the accuracy of the simulation can be pretty good if all the material parameters are introduced correctly, for simulations at long wavelengths such as RF the built in material parameters can give good results, however for antennas at infrared or optical wavelenghts the material parameters have to be introduced manually to get the same accuracy as with the RF simulations. Overall full wave EM simulations give you a pretty good starting point to calculate the performance of a designed antenna, however at the end the measurements are the results that count :)
Depends on what you call "efficiency". In free-space, most concern is losses, especially for resonant antenna (dipoles, for instance). Non resonant design (log-periodic, etc) are less sensitive. In real-life, efficiency highly depends on the environment: hight above ground, soil permittivity, etc. For a low altitude antenna, the Fresnel zone may extend on a large area, and accurate predictions may be hard to obtain.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Electromagnetism