Basically, Horn is made up of metallic material which can radiates from inner and outer side if electric form signal is fed. Probably, your analytical tool assuming feed as an electrical signal.However in actual case, Horn antenna is fed an energy which is in the form of wave (through wave guide) & horn shape confines the radiation direction. 

The radiation pattern looks similar to that of dipole antenna. Figure shows horn structure without the feed details.

Horn requires a feed where it excites electromagnetic waves. Horn will guide the electromagnetic waves.

Standard feed designs are available for horn antenna. You may try again with proper feed.

Hi AR and Amipara,

Thanks for your reply.

Actually, I used a waveguide port for feeding. There is a rectangular part that came out of my antenna structure (right side of Fig) it is a waveguide and I connect a waveguide port to that waveguide for feeding.

This feed excite the electromagnetic waves perfectly, because I checked the E_fields and it looks fine.

if the problem is feeding system , what should I do ? I am more suspicious to boundries or some software settings, But I could not figure it out  

If you are feeding wave energy to the horn, where at center it should have maxima and at surface zero. The opposite side lobe shows that horn is acting as a primary radiator. Check horn dimensions. It seems, Horn is excited by wave energy, and it is possible only if at inner surface of horn wave non zero condition occurs. Back lobe is as high as primary one, that means lot of energy is striking to the horn surface metal and hence it is acting as a radiating element. OR

Something is wrong with your software. Have you used wave port?

Thank you Amipara for explanation. I should check all dimensions again. I designed the dimension of antenna by using Balanis book. I used a waveguide port for feeding in CST I also choose open add space boundary.

Dear Sara

Another possibilities that could be added:

1. There might be some wrong symmetry planes set  - if you have them on, pls. disable.

2. The pattern may  be actually correct - only the orientation of your cutplane (I mean the angles phi and theta from your plot)is wrong. Check the 3D farfield while visualizing the structure too (there is an option in CST farfield ribbon for that). Then you can be sure if there is something wrong in the simulation.

3. Check the Near field inside the structure to see how the wave moves, if it is what is expected. Sometimes there is a hole in the geometry, or the materials were not correctly set or the background material was set to PEC (therefore there is no propagation at all). By checking on the fields inside the volume one is able to realize what's wrong.

good luck!

marcelo

Dear Marcelo,

Thank you very much for your help.

Actually, I checked the near fields, they look OK. But I could not find farfield Ribbon  for checking the 3D farfield in CST. The only problem is the back lobe of radiation pattern. Is it usual to have a back lobe as big as the main lobe ? I think it is supposed to be smaller than the main lobe.

I also have another question, when I have a wrong radiation pattern can I trust to s11 result?

I attached a 3D view of my radiation pattern. a polar view and also near field.

Cheers,

Sara

I think that the simulated radiadiation pattern is not wrong in your designed antenna.If you want to check a main lobe and back lobe, see the radiation pattern after changing the axies of radiation pattern. For example, to compare the diffferenc of the main lobe and back lobe, fix the phi axis as 90 degree and the theta rotate from 0 to 360 degree.

good luck !!!!

Jeongho

Ref to your another question about S11, it is to be seen judiciously, there may not be clear cut generalization answer of yes or no. Basically, right or wrong are decided:

1) Based on acceptable limit

2) In case of high insertion loss, you may not have reflected power which will show you very good S11 but still pattern may not be acceptable

you need to observe the H-plane radiation performance of the antenna, fix phi at 90 degree and then observe theta vs gain (dB) plot.

Here you have observed the radiation performance for theta = 90 degree. 

By using SIW technique the aperture size of horn has been reduced, so it may be the reason of getting back lobe.

You have designed a corrugated H-plane SIW horn antenna. The flaring should be corrugated. In your design at first linear tapering (flaring) has been observed after that you have designed the stair case flaring. 

You need to start the corrugated flaring immediate after the waveguide.

I have attached some papers and link based on SIW H-plane horn antenna. Go through those papers, I hope it will help you.

http://www.jpier.org/PIERB/pier.php?paper=14022603

http://onlinelibrary.wiley.com/doi/10.1002/mmce.20237/abstract;jsessionid=02DE8FB1E84A94D0E5D99F0DB82F7649.f03t04

https://www.google.co.in/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&uact=8&ved=0CCsQFjADahUKEwjwj-L73vjIAhXBHJQKHTOXAm8&url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel7%2F7086256%2F7150604%2F07150698.pdf%3Farnumber%3D7150698&usg=AFQjCNE_lgEkGyiwXEgdq8G6w6-3JVDfqA&sig2=D39orjuLFFIAUQxhH5avag&bvm=bv.106923889,d.dGo

Thank you very much for all helps .

I really appreciate it.

Hi Sara,

What are the aperture dimensions of your antenna with respect to the wavelength you're using? If the aperture is small or comparable to the wavelength then this could lead to a significant back-lobe. Another thing to check - when computing the far-field pattern, have you ensured that the surface of integration fully encloses all the equivalent electric and magnetic current sources?

Hi Sara,

I would agree with the colleagues saying that the pattern is fine. The radiating aperture of the antenna is rather small and hence the dipole like pattern in E-plane resulting in the large backlobe. As a test exercise I would suggest to increase the thickness of the substrate. As long as it does not allow TE01 mode you can do this safely as your structure has only H-plane steps and teh S11 should be the same. The aperture will increase in the vertical plane and you should have your pattern squeezed and the backlobe suppressed. Hope this helps.

Thank you Ray and Dmitry,

I solved my problem. I really appreciate all your helps guys.

Good luck

Sara