Re: Momenta output format

From: Jim Phillips (
Date: Tue Feb 03 2004 - 16:54:28 CST


On Mon, 2 Feb 2004, Neema Salimi wrote:

> 2 questions: Is the linear (P) momenta output as X, Y, Z in amu * (A / s)?
> What is the format of the angular (L) momenta output?

Linear momentum is the sum of mass * velocity.

Angular momentum is the sum of mass * position X velocity. This is raw
position and velocity, not relative to the center of mass. Angular
momentum isn't really useful if you have periodic boundary conditions.

Mass is in amu, position is in Angstroms, and velocity is in internal
units. In short, multiply by 20.45482706 to get amu*A/ps or amu*A^2/ps.
Here's a long and nasty LaTeX explanation of the velocity units:

Energy is expressed as kcal/mol and hence force as kcal/mol/\AA. In order
to obtain \AA\ after multiplying forces by the timestep twice, the
timestep is divided by TIMEFACTOR = 48.88821. This number is the inverse
of the square root of the converstion factor for changing from units of
(kcal fs$^2$)/(mol \AA\ amu) to \AA\ (which is {\em exactly} $4.184 \times
10^{-4}$ assuming thermochemical calories).

The timestep in DCD files is already divided by TIMEFACTOR, so converting
it to ps is simply a matter of multiplying by 48.88821.

Velocities may be converted to something useful like \AA/ps by first
multiplying by TIMEFACTOR to obtain (kcal fs)/(mol \AA\ amu) and then
converting this quantity by multiplying by $4.184 \times 10^{-1}$. If
TIMEFACTOR were calculated exactly, the total conversion factor would be
20.45482828. With the given value of TIMEFACTOR as used in the code, the
value is 20.45482706 instead. This conversion is applied when velocities
are written to a text PDB file, but not for binary restart files.


This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:37:19 CST