yukawa/colloid energy

In pair_yukawa_colloid.cpp ::compute lines 105-110, the energy is attributed to evdwl and then sent to ev_tally as:

ev_tally(i,j,nlocal,newton_pair,evdwl,0.0,….

Should this be ecoul and sent to ev_tally as

ev_tally(i,j,nlocal,newton_pair,0.0,ecoul,…. ?

This is also in pair/yukawa and the GPU and OMP packages which means several eyes have seen this - so I suspect I am simply misunderstanding the way the code in LAMMPS categorizes this potential. Any clarification would be greatly appreciated.

Kurt Sjoblom

Drexel University

Although the Yukawa potential is a “screened Coulomb” potential that approximates the Coulomb potential, its functional form does not depend on charges. Therefore in the eyes of LAMMPS it is considered as a charge-independent potential, and the energy is summed to evdwl.

I am not familiar with yukawa/colloid, but I suppose this applies to the colloid version as well.

Ray

Although the Yukawa potential is a "screened Coulomb" potential that
approximates the Coulomb potential, its functional form does not depend on
charges. Therefore in the eyes of LAMMPS it is considered as a
charge-independent potential, and the energy is summed to evdwl.

yes. steve and i had a discussion about this a while ago, when
implemented the openmp variant and noticed that it would tally into
ecoul, but use the special bond settings for LJ/vdwl. so i suggested
to make it consistent and - like ray said - the ultimate choice was
made because because of yukawa not using the charge property. one can
argue either way. this is a case where the "one man, one vote"
principle applies: steve is "the man", so he's got to have the (final)
vote. :wink:

ciao,
    axel.