[gmx-developers] A different MD integrator and temperature conservation

[David van der Spoel]

On 2015-03-15 04:46, Andrew AbiMansour wrote:
> Hi everyone,
>
> I have been experimenting with modifying the do_md integrator in Gromacs
> 5.0. If I keep the function do_md intact and only change its name to
> do_somethingelse (and update the enum list for the integrators as well,
> etc.) and then I specify the integrator in the mdp file to be
> 'somethingelse'.  This technically works *but* in NVT simulations I
> realized that after some time has passed, the temperature of the system
> is slowly but steadily increasing. The thermostat is not working
> properly. Any ideas why?

  % cd src/gromacs
  % grep integrator */*cpp */*c | wc
       35     414    3851

>
> Thanks!
>
> *Andrew Abi-Mansour*
>
>


-- 
David van der Spoel, Ph.D., Professor of Biology
Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205.
spoel at xray.bmc.uu.se    http://folding.bmc.uu.se