Any help would be much appreciated!  I have looked at the source code for variable.cpp and cannot figure out how my code might be causing this error.

Thanks!
Michal

On Feb 27, 2014, at 11:43, Ray Shan <[email protected]> wrote:

Yes, it is a variation of charge equilibration.  Since the charge
dependent terms are not quadratic, you can not use the linear
conjugate gradient (CG) method ReaxFF (fix qeq/reax) is using.
Instead you will have to use the damped dynamics (DD) method COMB (fix
qeq/comb) uses.

But you still can not just use fix qeq/comb right out of the box.  You
will have to write a pair style to supply the derivatives of energy
with respect to charge and the neutrality constraint to fix qeq/comb
(also you will have to make slight modifications to fix qeq/comb).

A generalized variable charge fix capability will choose either the CG
or the DD solver depending upon the pair style used, but your pair
style still have to supply necessary information.

Ray


On Thu, Feb 27, 2014 at 7:39 AM, Steve Plimpton <[email protected]> wrote:
Step 2 sounds like charge equilibration.  There are a couple
potentials in LAMMPS that do it (COMB, ReaxFF).  And
Ray (CCd) is working on a more general stand-alone
capability for it.  For step 3, look at other 3-body potentails
in the code, like SW or Tersoff.

Steve


On Wed, Feb 26, 2014 at 11:35 AM, Michal Plucinski <[email protected]> wrote:

Hello,

I'm looking to implement in lammps the potential found on page two of
the attached paper, and I want to make sure I have the best strategy for
this. This first two terms present in the potential are standard
Coulombic and Born-Huggins-Mayer potentials that I can just used out of
the box.

However, for the Coulombic portion, dynamically redistributed charges
are used based on the positions of nearby atoms in every step of a
minimization. How would I implement this sort of calculation inside
lammps? It appears to be similar to a pair-wise calc, but it is not
calculating an energy.

Secondly, the third term is a three body term with a cut-off, so similar
to a pair-wise potential but with an additional atom. Is there any way
you would recommend I implement this sort of potential in lammps that
leverages existing code as much as possible?

Thanks,
Michal Plucinski
Master's Student at Dalhousie University


------------------------------------------------------------------------------
Flow-based real-time traffic analytics software. Cisco certified tool.
Monitor traffic, SLAs, QoS, Medianet, WAAS etc. with NetFlow Analyzer
Customize your own dashboards, set traffic alerts and generate reports.
Network behavioral analysis & security monitoring. All-in-one tool.

http://pubads.g.doubleclick.net/gampad/clk?id=126839071&iu=/4140/ostg.clktrk
_______________________________________________
lammps-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lammps-users



------------------------------------------------------------------------------
Flow-based real-time traffic analytics software. Cisco certified tool.
Monitor traffic, SLAs, QoS, Medianet, WAAS etc. with NetFlow Analyzer
Customize your own dashboards, set traffic alerts and generate reports.
Network behavioral analysis & security monitoring. All-in-one tool.
http://pubads.g.doubleclick.net/gampad/clk?id=126839071&iu=/4140/ostg.clktrk
_______________________________________________
lammps-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/lammps-users