The ABINIT Users Mailing List ( CLOSED )

["Anglade Pierre-Matthieu" <anglade@gmail.com>]

To get reallistic dynamic you only need energy conservation then
absolute convergence is needed only for energy. You need only a nice
real SCF convergency for MD (nnsclo and nline increased for large
cells). And having your forces converge to 10^-5 (or -6) Ha/bohr in
the scf is already enough (unless your workin with H, He, Li).

On 3/30/06, Nuno Galamba <ngalamba@cii.fc.ul.pt> wrote:
> Dear Dr Xavier
>
> That is definitely the ideal situation. The problem is, to get significant
> results one needs a reasonable time of simulation. That requires a
> significant amount of computational time. I am therefore trying to come up
> with a reasonable set of input parameters before I start such simulation. In
> addition this may even decide wether I do have the computational resources
> for even thinking about doing such a simulation. That is why I could benefit
> from someone else's experience on this problems.
>
> Thanks
>
> N. Galamba
>


--
Pierre-Matthieu Anglade