From: Hyonseok Hwang (danggi_at_northwestern.edu)
Date: Tue Sep 21 2004 - 23:39:41 CDT
Dear all,
I did a couple of test calculations for the pair interaction energies
and forces by changing the 'pairlistdist'. The 'pairlistdist' of the
first one is '13' and that of the second is '15'. Below is two different
outputs.
-----------------------------------------------------------
When pairlistdist = 15
ENERGY: 1 0.0000 0.0000 0.0000
0.0000 -190.2772 11.4117 0.0000 0.0000
0.0000 -178.8654 0.0000 -178.6936
-178.6936 0.0000 4.1851 -91.6689 37705.5000
4.1851 -91.6689
PAIR INTERACTION: STEP: 1 VDW_FORCE: -3.0951 19.1297
-4.1225 ELECT_FORCE: -0.1429 -22.0072 -2.1688
---------------------------------------------------------------
When pairlistdist = 13
ENERGY: 1 0.0000 0.0000 0.0000
0.0000 -190.2772 11.4117 0.0000 0.0000
0.0000 -178.8654 0.0000 -178.6936
-178.6936 0.0000 4.1851 -91.6689
37705.5000 4.1851 -91.6689
PAIR INTERACTION: STEP: 1 VDW_FORCE: 44.4982 -8.8246
-15.0844 ELECT_FORCE: -66.5043 19.8897 40.5486
I just showed the first step and if you look at the forces, you will
find that the two results are very different. I thought the pair
interaction forces as well as pair interaction energies should not
depend on the 'pairlistdist' if pairlistdist is longer than 'cutoff'
dist (here cutoff =12). If you look at the energies output, they do not
depend on the pairlistdist, and I think that energy calculation is OK.
In case of force calculations, however, I'm wondering what's wrong with
the pair interaction force calculations and if this is associated with
the 'sign bug' in the pair interaction force calculation.
One more note : If I increase the pairlistdist, the forces converges to
some values.
Thanks a lot!
-Hyon
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:38:53 CST