Segmentation fault (core dumped) during NVT

GROMACS version: 2020.2
GROMACS modification: Yes

Hello everybody,

I am simulating C60 (all atom) and TAPC (United atom) molecules in a box with a graphene sheet at the bottom. I am using gromos force field (The reason I am suing gromos is that I am reproducing a paper which simulated the same molecules using gromos).

I tried to simulate 1500 molecules (mixture from both C60 and TAPC molecules) and the simulation was completed successfully (all the steps, energy minimization, NVT, NPT and md run was completed). I tried to increase the number of molecules into 2000 molecule, the simulation was not successfully completed. In this later simulation I receive Segmentation fault (core dumped) at step zero in the nvt equilibiration. In the 2000 molecule case, the minimim force was not less than 2 * e6 and the potential energy is 2.8 * e11. I understand that the energy minimization is not essential to reach the minimum required force and it is normal in some systems if the energy is positive.

I don’t know what is exactly the problem, specially that it is dependent in the number of inserted molecules. I don’t think that the problem is that the 2000 molecule is too much with respect to the box size because I can insert in the same box up to 3800 molecule.

I have added my minim.mdp, nvt.mdp and my log files in the below link, can anyone help me recommendations and also can anyone tells me if my .mdp files need any modification.

https://drive.google.com/drive/folders/1yioYtutWU7DgNZqxV02nV4i4jxjXD9Yw?usp=sharing

Thanks