[gmx-users] [Fwd: energy conservation problem]

[Anton Feenstra]

From: Petr Slavicek
> 
> The problem is that when using shake (i.e. spc water with settle) there
> is a very signifficant drift in total energy for NVE simulation. What i
> am simulating is a cluster, so i have followed the reccomendation to set
> all cutt-offs to 0, and nstlist to 0.  I have switched off the
> temperature and pressure coupling. The energy then explodes (unlike
> when using flexible water).

Do I understand you correctly that the only difference between your
'working' simulation and your 'exploding' one, is that you turn on
constraints (i.e. Settles) for SPC? That sounds strange. Also, I
don't know exactly what setting cut-offs to 0 does. Have you also
set 'pbc = no' in your .mdp file? Perhaps 'pbc = no' in combination
with all cut-offs larger than the size of your cluster will work.

> Interestingly, the problem is more pronounced when smaller timestep used.

That is probably related to the single precision calculations, where
a small timestep leads to more pronounced round-off errors.


-- 
Groetjes,

Anton
 ________ ___________________________________________________________
|        | Anton Feenstra                                            |
| .      | Dept. of Pharmacochemistry - Vrije Universiteit Amsterdam |
| |----  | De Boelelaan 1083 - 1081 HV Amsterdam - The Netherlands   |
| |----  | Tel: +31 20 44 47608 - Fax: +31 20 44 47610               |
| ' __   | Feenstra at chem.vu.nl - http://www.chem.vu.nl/afdelingen/FAR|
|  /  \  |-----------------------------------------------------------|
| (    ) | Dept. of Biophysical Chemistry - University of Groningen  |
|  \__/  | Nijenborgh 4 - 9747 AG Groningen - The Netherlands        |
|   __   | Tel +31 50 363 4327 - Fax +31 50 363 4800                 |
|  /  \  | K.A.Feenstra at chem.rug.nl - http://md.chem.rug.nl/~anton   |
| (    ) |-----------------------------------------------------------|
|  \__/  | "A Man Needs a Maid" (N. Young)                           |
|________|___________________________________________________________|