I adopted the COMPASSIII force field. After I set rigid body constraints for h-BN and ran NVT molecular dynamics, why do the bonds inside h-BN drift? Besides, the NVT simulation finished normally without any error reports.