Re: [ESPResSo] MMM2D

[aarnold]

Owen Hickey wrote:
> Hello,
>
> I have a few quick (at least I hope) questions concerning implementing
> espresso to study a system in a simple slab geometry with periodicity 1 1
> 0.  First of all using this periodicity where should I build my system?
> e.g.
> should it be from 0 to box_z or from -box_z/2 to box_z/2?

MMM2D itself does not care about that, but the cell system used to
calculate any other short ranged interactions will work much faster if you
keep the particles in the specified box, i.e. the coordinates should be
between 0 and box_z.

>  I am getting
> some
> strange behavior from espresso and I was wondering if this might be the
> source of my problems.  Most notably when I try to implement MMM2D I get
> the
> following error
>
> /home/ohickey/Espresso/Espresso/obj-Pentium_4_64-pc-linux/Espresso: line
> 45:
> 17156 Segmentation fault      $ESPRESSO_CALL

Can you send me the script? MMM2D works normally quite reliably, but I am
not sure what happens if some parameters are not reasonable. Seems that I
don't check that properly...

>
> Also In order to keep my particles initially in the box I have to add some
> forces which I later discard.  Currently for my test system when I want to
> take away the forces I simply set the constants for the depth of the
> potential equal to zero but I worry that Espresso is still looping over
> all
> the particle pairs and calculating the forces between the particles.  Does
> anybody know if Espresso checks the parameters to see if there is indeed
> and
> interaction or if there is a way to delete forces from the inter list?

>From the interaction list, interactions cannot be deleted, because they
specify how particle types interact. However, if you constrain your
particles using a constraint, you can (and should) delete the constraints,
using "constraint delete". Otherwise, the forces are calculated, even if
the constant is zero. If you need the constraints later again, you can
readd them.

Axel