Jumps in Cons Qty

[Ondrej Kroutil]

Hi CP2Kers.
I'm doing simulations of cisplatin in water. System was equilibrated by 
classical MD, then I switched to the BOMD.
It's NVT simulation with PBE functional, deuterated hydrogens, time step 1 
fs (and more in the input file). I use CP2K version 2.4 (Development 
Version), SVN source code revision 12755.
After i did first simulation (100 ps) I realized that I observed something 
like "ice-cube flying" when whole system looks like frozen and all atoms 
moves one direction. So I used COMVEL_TOL 1e-8 to remove COM movement and 
started simulation again.
Now (after 10 ps) trajectory looks fine, but if I plot time evolution of 
conserved quantity I can see strange jumps there (see picture 
"cisplatin-problem.jpg", big graph - whole trajectory, inset - first 2000 
steps).
Do you think this can be connected to (maybe tight) COM remove?

Thank you all for suggestion and keep on rockin'

Ondrej Kroutil

Btw. Do you plan for this year similar workshop you held last year in 
CECAM? I'd like to join it, because i don't want to give you (maybe) stupid 
question I posted now...

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20140212/fe1847b7/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cis+0-no_ci.inp
Type: chemical/x-gamess-input
Size: 3729 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20140212/fe1847b7/attachment.inp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cis+0-no_ci-1.ener
Type: application/octet-stream
Size: 1310261 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20140212/fe1847b7/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cisplatin-problem.jpg
Type: image/jpeg
Size: 105007 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20140212/fe1847b7/attachment.jpg>