[gmx-users] Problem with user-defined potentials

[Mark Abraham]

David Waroquiers wrote:
> Hello,
> 
> I'm getting problems to simulate amorphous silica with potentials I made
> myself : table_Si_Si.xvg, table_Si_O.xvg, table_O_O.xvg and table.xvg
> (which should not be used as all the possible non-bonded interactions
> are given in the three previous tables).
> 
> The generation of the topology file with grompp works well but when I'm
> try to simulate my system (mdrun), Im getting a segmentation fault.

It's vanishingly rare for GROMACS to segfault without an error message. 
Consult all of stdout, stderr and your .log file for messages.

Probably, your combination of starting configuration and model physics 
are broken. If you're confident your starting configuration is good 
(e.g. you know it works with some other Si-O force field) then you'll 
have to go over your table files, I suppose.

Mark

> Am I
> missing something in the topology file or grompp.mdp file ? I just want
> to have the non-bonded interactions defined in my tables to be applied.
> My topology and input file are hereafter. I can give you my tables.xvg
> but they are too big to be posted on the forum.
> 
> Thanks a lot
> 
> My topology file TOPOL.TOP is the following :
> 
> [ defaults ]
>  2 1 0
> [ atomtypes ]
> ; name      mass   charge       ptype           c6            c12
>  O      15.99940     -0.955209      A           1.0           1.0
> 0.0000;
>  Si     28.08550      1.910418      A           1.0           1.0
> 0.0000;
> [ nonbond_params ]
>   O       O      2      1       1       1
>   O      Si      2      1       1       1
>   Si     Si      2      1       1       1
> [ moleculetype ]
>  O      0
> [ atoms ]
>  1       O      1       O        O      1       -0.955209     15.99940
> [ moleculetype ]
>  Si     0;
> [ atoms ]
>  1       Si     1       Si      Si      2        1.910418     28.08550
> [ system ]
>  Silica
> [ molecules ]
> O        384
> Si       192
> 
> and my mdrun input file GROMPP.MDP is :
> 
>  title                   = 576-atoms a-SiO2 model 01
>  cpp                     = cpp
>  define                  =
>  integrator              = md
>  dt                      = 0.001
>  nsteps                  = 25000
> 
> ; Coulomb type
> 
>  coulombtype             = User
>  vdwtype                 = User
>  rvdw                    = 1.0
>  rlist                   = 1.0
>  rcoulomb                = 1.0
>  fourierspacing          = 0.1
>  ewald_rtol              = 1e-5
>  energygrps              = Si O
>  energygrp_table         = Si Si Si O O O
> 
> ; Temperature Coupling
> 
>  tcoupl                  = nose-hoover
>  tc_grps                 = System
>  tau_t                   = 0.02
>  ref_t                   = 7000
>  gen_vel                 = yes
>  gen_temp                = 7000
>  gen_seed                = -1
> 
> ; Simulated Annealing
> 
> ; annealing               = single
> ; annealing_npoints       = 3
> ; annealing_time          = 0 56 150
> ; annealing_temp          = 5000 5000 300
> 
> ; Output parameters
> 
>  nstxout                 = 100
>  nstxtcout               = 1000
>