[gmx-users] Constraint algorithm error

[Anton Feenstra]

nicolas Dinter wrote:

> Dear Gromacs users,
> 
> I am simulting a 64 diphenyl carbonate molecule box, at 1bar and 298K. After
> building a big box (15*15*15) I am reducing the box size by NPT. So far it
> runs well but now the calculation crash after several steps.  In the md.log
> file I found the error: "constraint error in algorithm step 84910...".
> I wonder if it's not a Gromacs problem, because in my mdp file I specified:
> constraints= none, and I tried to use the shake algorithm with a 0.0001
> shake_toll and I also tried the lincs algorithm with lincs_order = 4.
> Any idea?

Do you have explicit constraints defined in your topology file? The mdp
option only tells grompp if it should converg bonds and/or angles into
constraints. You should probably look carefully at your conformations to
see if terrible clashes occur. Also, look at how your pressure is developing;
the pressure coupling might be shrinking the box too fast causing the
molecules to clash.

It would help us if you show what your molecule looks like in more detail.
Do you have (other) solvent as well?


-- 
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)                               |
|_____________|_______________________________________________________|