hi
I need some clarification on the use of “compute pressure” in the case of long-range coulomb forces and the pressure tensor. There is certainly an issue with using “compute stress/atom” due to not having a per-atom partition of forces as the web-page states but the “compute pressure” says :
…virial as the sum of pair, bond, angle, dihedral, improper, kspace (long-range), and fix contributions to the force on each atom
which I can see for the pressure as a scalar but the formula for pressure as a tensor implies the existence of pair-wise forces.
The context is : I want compute the viscosity of TIP4 water using a Green-Kubo formula that needs the system average stress. My question is: does compute pressure give me the correct pressure tensor ?
thanks,
Reese
ps
I took a look at this thread in the lammps-users archives
Computing heat flux in LAMMPS: “compute heat/flux” v/s “compute stress/atom”
which proposes a solution for stress/atom for the Ewald sum, and Aidan & Steve’s paper: Thompson JChemPhys 2009, which discusses and develops generalized virial expressions based on what looks like total forces (not pairwise).
Hi,
As I understand there is not virial per atom for coulomb forces.
There is long discussion about calculation of heat flux with Coulomb.
Sincerely,
German
hi German,
please read my entire email. I am aware of the discussion (and was minimally involved in it) and the question remains: does the pressure tensor (not the scalar) computed by “compute pressure” correctly include the Coulomb forces? If you have a definitive answer for this, I would very much like to hear it.
thanks,
Reese
yes, compute pressure calculates the global pressure tensor correctly,
including long-range Coulombics. Ask Ahmed Ismail (emails on this
list a lot), about the accuracy of the virial for TIP4P. I think he's
currently working thru some issues for that - but it has nothing
to do with the long-range part.
Steve
Hi Reese,
Yes, tensor computed by "compute pressure" correctly include the
Coulomb forces.
Sincerely
German