Learning Forum

[maximilian]

Hello there!

I encountered a problem of deviating S21 results for the simulation of two dipole antennas facing each other. If the antennas are placed in free space, I only achieve consistent results for different box sizes by utilising FEBI. This aligns with the conclusions of the following post:
https://innovationspace.ansys.com/forum/forums/topic/what-is-the-radiation-box-size-in-hfss-if-we-have-two-antennas-2/ 

However, if I place the same antennas inside a block of dielectric material, S21 results are not consistent for both ABC and FEBI. The minimum distances of ABC (153.8mm) and FEBI (57.7mm), starting from the block surfaces, are the preset paddings for HFSS's open region command (operating frequency is 650 MHz). I extended these values up to 300mm, thus exceeding lambda/4 (pushing the paddings >300mm is possible but takes very long simulation times).

Has anyone encountered similar problems or by chance knows a solution to the problem?

PS: It's worth mentioning that S11 results are very consistent in all cases.

Thank you and best regards

Maximilian

[DMARATHE]

Hi Maximilian,

I think you should consider results for large box size 300 mm. Additionaly check the mesh on dieletric region and antennas geometry. If its corser you can apply 

Manual mesh seeding operations for denser mesh. Look for number of adapative passes and converage criterion from solution data.

Inside Setup, you can apply additional convergence for S21, using Expression cache

Hope this helps.

Even if this does not help, submit a support ticket if you have access to Ansys Technical Support. Ansys support engineer can connect with you and help you out with settings. Thanks.