Re: how work: harmonic position "constraints" module + "WrapAll on"?

From: Sebastian Stolzenberg (
Date: Mon Jul 20 2009 - 12:52:06 CDT

Good job guys!

I stopped being lazy and made a test run, and it exactly works as
described below.

Thank you,

Joshua Adelman wrote:
> Hi Sebastian,
> The 'wrap' commands should only impact the way the outputs are
> written. As far as I can tell from the description of your problem,
> the constraints should work correctly. The thing to do to make sure
> everything is working as you expect is to either:
> 1. Run a small test system and look at the distribution of the ion
> position and see if it looks like you would expect. You can do this in
> vacuum, so it would take very little time to test.
> 2. Look at the code to see if NAMD is wrapping the forces (or more
> likely just the positions); The code in this section is fairly well
> commented so it shouldn't be difficult to follow what is going on.
> Best of luck,
> Josh
> On Jul 20, 2009, at 5:12 AM, Sebastian Stolzenberg wrote:
>> Dear All,
>> How does the harmonic position "constraints" module (with "consref"
>> and "conRef") deal with periodic conditions that use "WrapAll on"?
>> Conceptionally, let us assume we have an ion very close to the box
>> boundary, focusing on one dimension only:
>> - box boundaries are $x-$a and $x, where $x and $a are positive floats
>> - an ion has position $x-$eps, to which it is weakly harmonically
>> "constrained" with 0<$eps<<$a.
>> Now, let us assume, having ended a first run, whose final restart
>> file is such that the ion is at $x-$a+$eps, the opposite side of the
>> box.
>> The next run has to use the original "consRef" file with the ion
>> reference set to $x-$eps (see *note in the bottom).
>> My question:
>> In which direction does the "constraints" module "pull" the ion? Is
>> it smart enough to pull it into the negative direction, or will it
>> pull it into the positive direction?
>> Many Thanks,
>> Sebastian
>> *note: I know that I could simply change my "consRef" file to the
>> last restart file, but that would complicate the way my wrapper code
>> works.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:53:03 CST