[gmx-users] Water molecule can not be settled

[Anton Feenstra]

David van der Spoel wrote:
> On Mon, 2003-03-17 at 14:15, Ester Chiessi wrote:
> 
>>Dear all,
>>
>>I was equilibrating my system (a polymer in a water box, at 303 K ) and the total
>>energy drift was decreasing as expected.
>>After about 50 ps of MD, my calculation stopped with the following message (the
>>timestep was 1 ps):
>>
>>t = 48.064 ps: Water molecule starting at atom 1865 can not be settled.
>>Check for bad contacts and/or reduce the timestep.Fatal error:
>>t = 48.064 ps: Water molecule starting at atom 1865 can not be settled.
>>Check for bad contacts and/or reduce the timestep.
>>
>>I verified a short contact between atoms of two water molecoles, not
>>involved in the hydration shell of the solute.
>>
>>How may I go on?
> 
> 
> this usually happens when you have one water without hydrogen bonds
> (i.e. in vacuum, or a completely hydrophobic surrounding), that starts
> spinning for whatever reason. You may indeed reduce the timestep, or
> otherwise regenerate velocities (with grompp). In both cases your
> trajectory will become discontinuous...

Alternately, you might 'tweak' your simulation a bit upstream. What
I mean is this: the chance of this water molecule starting to spin
so fast is very small. If you would take for example your last .trr
frame (x+v for restart) at a few ps before the crash, and introduce
slight rounding errors in the 6th decimal place for coordinates and
trajectories
using "trjconv -s topol.tpr -f traj.trr -ndec 6 -dump 45 -o frame.gro"
and "trjconv -f frame.gro -o frame.trr"
and "tpbconv -s topol.tpr -f frame.trr -o new.tpr"
you will have a nearly continuous trajectory (single precision is
only 7 decimal places, you round at 6), but the chance of getting that
same water to spin again is virtually zero.


-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610           |
|             | Feenstra at chem.vu.nl - www.chem.vu.nl/~feenstra/       |
|             | "If You See Me Getting High, Knock Me Down"           |
|             | (Red Hot Chili Peppers)                               |
|_____________|_______________________________________________________|