Re: Tcl_Forces = pressure shifts while Langevin Piston is ON

From: Arturas Ziemys (ziemys_at_chbmeng.ohio-state.edu)
Date: Tue Dec 05 2006 - 13:47:08 CST

>> Must excuse - I found my bug concerning Tcl force! Just a stupid mistake in
>> analysis I found after a several days. Since nobody answered I thought
>> there was no reason to inform about my own mistake.
>>
>> But it is not clear about getbond, still. I get coordinated from loadcoord.
>>
>> Arturas
>>
>> >A possible reason why the pressure drops would be if forces from the Tcl
>> >interface were not included in virial calculations. It would be good to
>> > have confirmation from NAMD people on that, since many uses of TclForces
>> > could be affected.
>> >
>> >About getbond: it just computes the distance between two 3d vectors, it
>> > has no clue what PBCs are. The code of that function essentially reduces
>> > to (r2-r1).length(). So the relevant question is, where do you get the
>> > coordinates you feed to getbond? Presumably loadcoords...
>> >
>> >Jerome
>> >
>> >On Friday 01 December 2006 17:53, Arturas Ziemys wrote:
>> >> Hi,
>> >>
>> >> I have the questions on Tcl Force still.
>> >>
>> >> I applied the "soft" repulsive force from the center of the water box to
>> >> some water molecules. It was very strange thing to find that the
>> >> pressure dropped significantly (to -80 bars), while I used Langevin
>> >> Piston targeted at 1 bar .
>> >>
>> >> How that effect could be explained ??
>> >>
>> >> I could understand that if I would apply the attractive force acting to
>> >> the center of my systems, the pressure could drop. Any comments are
>> >> welcome !
>> >>
>> >> Best
>> >> Arturas
>> >>
>> >> P.S. However I still got no answer about the Tcl function 'getbond' and
>> >> how it deals with PBC ... :) ...
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:42:54 CST