Thickness of substrate of microstrip antenna in cst

Hi, everyone. Please, help me to solve my problem. I'm designing a microstrip patch antenna, using inset feed, freq - 2 Ghz, substrate - fr4.
I got a task to define the thickness of the substrate, that would give the best results in simulating. I've read the specifications for fr4 - there is siad that the thickness of the substrate to be from 0.05 to 3.2 mm, but i can't define the perfect thickness of the subs to get best gain, efficiency, directivity and bandwidth. So, can anyone tell me how to do that?

with FR4 with 1.6 mm height you can find good result

Thanks very much. I know that i can't increase the height infinitely, there must be some size to achieve when results would start to get worse. But when i increase the height of the substrate(more than 1.6) results get better anyway. Maybe i'm doing somethin wrong?

not wrong with higher thickness your antenna bandwidth will increase but i think your antenna is not match perfectly

But how can i see bad matching? Cause i can make input imp. approximately 50 homs(desired impedance), increasing the width of microstrip line. Or does it express in return loss maybe? sorry for my incompetence)

what will be the thickness of the substrate if i increase the frequency ?
as i am working on 10 Ghz frequency and i am wondering about FR-4 thickness on 10GHz freq
- - - Updated - - -
what will be the thickness of the substrate if the operating frequency is 10GHz ? as i am looking for the thickness of FR-4 substrate at operating frequency of 10GHz. please is there any method to find the the thickness of the different substrate

FR4 working up to 12 GHz so u dont have problem for 10 GHz and the thickness is related to your application and structure

While it may technically work, I'd just like to mention that I've had an extremely large number of troubles modelling FR-4 correctly (i.e., getting the measured and simulated to match), especially at higher frequencies (> 2 GHz). Some others may have had better experiences, but I would avoid FR-4 at all costs if possible when working in the X-band. Other substrates are much higher performing with not much more cost.

@planarMetamaterials. what will you suggest other substrates working on 10GHz ?

Pretty much anything else. I use Rogers 3003 most often.