From 4fcbd58d5a32656bd4fedd7b1ac486b5b6b8f137 Mon Sep 17 00:00:00 2001
From: Steve Plimpton <sjplimp@sandia.gov>
Date: Fri, 23 Jun 2017 15:54:14 -0600
Subject: [PATCH] doc page clarifications for CHARMM energy and dipole
 pre-factors

---
 doc/src/pair_charmm.txt | 10 +++++++++-
 doc/src/pair_dipole.txt |  8 ++++++++
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/doc/src/pair_charmm.txt b/doc/src/pair_charmm.txt
index 9c5973c725..1e78607c08 100644
--- a/doc/src/pair_charmm.txt
+++ b/doc/src/pair_charmm.txt
@@ -104,7 +104,15 @@ charmmfsw"_dihedral_charmm.html command.  Eventually code from the new
 styles will propagate into the related pair styles (e.g. implicit,
 accelerator, free energy variants).
 
-The general CHARMM formulas are as follows
+NOTE: The newest CHARMM pair styles reset the Coulombic energy
+conversion factor used internally in the code, from the LAMMPS value
+to the CHARMM value, as if it were effectively a parameter of the
+force field.  This is because the CHARMM code uses a slightly
+different value for the this conversion factor in "real
+units"_units.html (Kcal/mole), namely CHARMM = 332.0716, LAMMPS =
+332.06371.  This is to enable more precise agreement by LAMMPS with
+the CHARMM force field energies and forces, when using one of these
+two CHARMM pair styles.
 
 :c,image(Eqs/pair_charmm.jpg)
 
diff --git a/doc/src/pair_dipole.txt b/doc/src/pair_dipole.txt
index a9622b32fd..985581cac8 100644
--- a/doc/src/pair_dipole.txt
+++ b/doc/src/pair_dipole.txt
@@ -71,6 +71,14 @@ and force, Fij = -Fji as symmetric forces, and Tij != -Tji since the
 torques do not act symmetrically.  These formulas are discussed in
 "(Allen)"_#Allen2 and in "(Toukmaji)"_#Toukmaji2.
 
+Also note, that in the code, all of these terms (except Elj) have a
+C/epsilon prefactor, the same as the Coulombic term in the LJ +
+Coulombic pair styles discussed "here"_pair_lj.html.  C is an
+energy-conversion constant and epsilon is the dielectric constant
+which can be set by the "dielectric"_dielectric.html command.  The
+same is true of the equations that follow for other dipole pair
+styles.
+
 Style {lj/sf/dipole/sf} computes "shifted-force" interactions between
 pairs of particles that each have a charge and/or a point dipole
 moment. In general, a shifted-force potential is a (sligthly) modified
-- 
GitLab