[gmx-users] Flat-bottom potential SEGFAULT

[Erik Marklund]

Dear dynamicists,

I am using flat bottom potentials to prevent water molecules from evaporating from a droplet in vacuum. The manual suggests this is pretty straightforward, but we run into a segfault with our approach. Having never used flat-bottom potentials before, chances are we are doing something wrong without realising it. Let me describe what we do and perhaps you can tell me where things go wrong.

We make our own copy of tip3p.itp, in which we put a [ position_restraints ] directive for the OW1, HW2, and HW3 atoms:
[ position_restraints ]
; id func_type    g   r  k
1    2            1   10 1000
2    2            1   10 1000
3    2            1   10 1000

We then create a new gro-file, where we move all water atoms in the middle of the box. This is to be used for reference coordinates in the next step.

We run grompp, where we provide the new gro file using the -r flag.

We launch mdrun, which segfaults. If we prepare the system without the -r flag in the previous step everything runs fins, but the reference point for the flat-bottom potential is presumably wrong.

Where might the error be?

Kind regards,
Erik

Erik Marklund, PhD
Postdoctoral Research Fellow, Fulford JRF