[gmx-users] temperature explosion...

[Anton Feenstra]

David wrote:
> On Fri, 2003-05-09 at 17:50, Giorgos Karvounis wrote:
> 
>>Dear users...
>>I want to simulate a pentapeptide at 360 K without Tand P coupling. I
>>first do an equilibration run , starting from a 360K velocity and  T / P
>>coupling for 100 ps. After that, i switch the coupling and continue the
>>same run . I monitor the temperature and it seems that it explodes, from
>>358 to 491 to 651 K every 10000 steps. Is this suppose to happen? Shall
>>i perform the equilibration longer?
>>Is it possible to keep the temperature/energy constant without using
>>coupling?
> 
> yes, use PME.

You can have a look at J. Comput. Chem. 20, 786-798 (1999) for what
size of energy drift you can expect at different timestep. Note these
simulations were done with shifted potentials, to eliminate the cut-off
noise (hence Davids suggestion for PME, which does nearly as good),
and in double precision. It depends on how long you want the temperature
to remain constant, and why you want to run without temperature coupling.


-- 
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)                               |
|_____________|_______________________________________________________|