cavity backed antenna in cst

Hi,
I am a freshman for CST. Now I want to research the electric field distribution in the metal box under the excitation of antenna. I just use the dipole as the transmit antenna. When I use the discrete port as the excitaion, the MWS shows that the port can not be in the metal box. the same to the wave port.
I alternative method is that I can design a balun for antenna and the coxial transmission line and extend the line out of the metal box. Then I can use the wave port excitaion. But this method is very complex and long computational time.
Is there any other method? In fact, in HFSS ,the lumped port can be located in the metal box. Can the discrete port be located in the metal box in CST?
Thank you for all.

If I understand correctly, your dipole lies inside a computational metal box (not a real metal plates box), I've encountered this problem before and got pretty much the same respone from CST.
my suggestion to you is to clear the background box ( under the background material properties menu -> Layer type -> Normal ).
Now, build a real metal box around your antenna and run the simulation.
In this case, i believe you are interested in Standing Wave patterns inside your resonator and not in the radiation far field pattern. Maybe you should try the EigenMode solver instead of the Transient solver.
Good luck,
P.

I guess you get the message "the port is inside of perfect conducting..." something like that. Well, the background material has to be normal. If you say it is PEC, CST fills out everything (that is not defined by another material) with PEC, so your port is inside a conducting area. However, just change that into normal. Then, you can modify your boundary conditions to PEC, so the E-field would be zero. That is already your box. You don't have to create a seperate box or anything like that. You also don't have to change anything on your feed. Just create the dipol with lumped port, background material is normal, boundary box is PEC.
good luck.

Hi,
Thank everyone for your suggestions. I am sorry that I have not give an exact description about my problem. The model is described as follow:
(1) The thickness of the metal box is 2mm and it is full close.The material is copper.
(2) The cavity in the box is filled with the air.
(3) The antenna is located in the metal box.
(4)The excitation is the descrete port between the two elements of the dipole.
The backgroud out of the box is normal.
When I simulated the model, the problem occured.
It seem that the excited port must be in the backgroud of the computational domian. I am not sure.
Could you give more advices? Thank you!

Hi yuanzy97,
pardon me but do you mean something like the attached? If your box is very small, you'll have to be careful with the meshing and make sure the discret port is not inside the metal parts.
Hope this helps.
Cheers,
Element7k

Hi,
send me your *.mod file via e-mail. I will look what you did and check everything.
elomatic@gmx.ch