I had design multi-element structure for MIMO antenna system with desire resonating frequency of 3.5 GHz. But if the same structure is simulated with adaptive mesh option the resonating frequency shift to 3.6.
And when I use same structure in HFSS the resonating frequency is 3.7.
I wonder why this shift of resonating frequency is occurring...
Or Their should be some proper manual or notes about accuracy of CST microwave result.
And When I fabricated my antenna the resonating frequency was 3.8...
Your suggestion in this regard will be highly appreciated.
Try using an adaptive mesh at a frequency partly off of the resonance, but not all the way off.
In HFSS what is your solution frequency? You are doing a frequency sweep, correct?
Thanks Mr. Raymond for your worthy advice and I did use the adaptive mesh (Energy Based Method) and now the antenna resonate at 3.56 so it is near to HFSS result.
So it mean that we had to use adaptive mesh for every simulation for correct result?.
Thanks Mr. Richard for you kind reply and my solution frequency is 3.5 and I am using fast sweep.
I noted as well a difference between both simulator. In CST i used a good mesh and simulating a stripline the results are quite close to theory, instead using HFSS i've problem with S parameters ( they are ocmpletly wrong). i Changes port, and i checked all the properties of the simulation.
Any suggestions?
Basically, because mesh shapes in HFSS (FEM; tetrahedron) and in CST MWS (FDTD; cubic voxel) are different, it cannot be totally identical between the results.
Although it depends on the antenna structure, generally in HFSS, radiation boundary size and analysis solution type and frequency can affect the result. Also, the solution frequency should depend on the solution type. Additionally, I think port set-up also can influence the result.
I recommend discrete sweep for a high accuracy. In this case, you can use any solution frequency.
Try using curved mesh elements in HFSS,. They usually give much accurate results, if the structure has curves or cylindrical elements.
In my experience with HFSS and frequency sweeps, the 'Fast Sweep' tends to yield less reliable results than the 'Discrete Sweep.' Also in the case of a frequency sweep, select your solution frequency to be higher than your stop frequency in the sweep. HFSS uses the solution frequency to generate an appropriate mesh size and so, if you select the solution frequency to be higher than the stop frequency you are ensuring your mesh size is adequate for the whole sweep. If you select a solution frequency lower than the stop frequency, the mesh may be (and usually is) inadequate for sweep frequencies above the solution frequency.
This is because of the different mesh sizes in HFSS and CST, but I think for frequency less than 5 GHz, HFSS should show comparatively better results compared to measurements.
Frequency sweep, solution frequency, radiation boundary, port-setup, everything counts too obviously.
There is a tutorial video on youtube explaining the whole process of exporting a design from HFSS and then importing it in CST and simulation of the same. The link of the video: https://www.youtube.com/watch?v=9BGVd2uiu-0
- Badges
- Science topic