Please be careful that simple setting contain a lot of implicit parameters in ns-3.
Maybe you should check antenna orientation, LOS/NLOS condition and the existence of any fading.
I don't know the detail of your theoretical value, but in 4G/5G simulation, shadowing and multipath fading are considered in most cases.
So the SINR, simulated by implicit settings, can be much lower than the one calculated by FSPL.