Dear everyone:
The ecoul and elong are different in the manul. It mens the elong represents long-range interactions except coulombic interaction, is it right?
Meanwhile,in the manul,there is a note in the pe/atom command: The per-atom energy does NOT include contributions due to long-range Coulombic interactions (via the kspace_style command). It’s not clear this contribution can easily be computed.
These commands are used in my input:
compute peratom all pe/atom
compute pe all reduce sum c_peratom
thermo_style custom step temp etotal press pe c_pe
So I think the pe can be represent like this:
pe=c_pe+ecoul
But in fact, pe=c_pe+elong.
Who can give me some appreciate suggestion! What’s the relation between elong and ecoul? Thanks!
|
Dear everyone:
The ecoul and elong are different in the manul. It mens the elong represents long-range interactions except coulombic interaction, is it right?
no. elong are the potential energy contribution that are computed
in kspace, i.e. via ewald summation or PPPM.
Meanwhile,in the manul,there is a note in the pe/atom command: The per-atom energy does NOT include contributions due to long-range Coulombic interactions (via the kspace_style command). It's not clear this contribution can easily be computed.
that is correct.
These commands are used in my input:
compute peratom all pe/atom
compute pe all reduce sum c_peratom
thermo_style custom step temp etotal press pe c_pe
So I think the pe can be represent like this:
pe=c_pe+ecoul
that is wrong. ecoul is part of c_pe.
But in fact, pe=c_pe+elong.
that is correct and in keeping with the documentation.
Who can give me some appreciate suggestion! What's the relation between elong and ecoul? Thanks!
ecoul is the contribution of coulomb interactions that are
computed in real space.
elong is the corresponding lattice sum contribution.
... and before you ask, this contribution cannot be
partitioned into per atom contribution due to the way
it is computed.
this topic has been discussed many, many times on this mailing
list. if this answer is not satisfactory, i suggest you search through
the mailing list archive for previous discussions.
cheers,
axel.