From: Ana Celia Araujo Vila Verde (avilaverde_at_engr.psu.edu)
Date: Sun Apr 08 2007 - 14:57:29 CDT
Hi JC,
No, I wasn't doing SMD in initially; I was doing regular MD. I guess the easiest way for me to fix my problem is simply to restart my SMD, but this time setting the first time-step to 0.
Thanks again for the explanation; I really appreciate it!
Ana
________________________________
From: JC Gumbart [mailto:gumbart_at_ks.uiuc.edu]
Sent: Sun 4/8/2007 3:38 PM
To: Ana Celia Araujo Vila Verde; namd-l_at_ks.uiuc.edu
Subject: RE: namd-l: wrong velocity in steered molecular dynamics
The SMD code in NAMD calculates the reference position using the formula x = x0 + v*t (in one dimension) where x0 is the coordinate in SMDfile. This makes sense for restarting SMD simulations, so you don't have to constantly update SMDfile.
So in your case, your reference position is some distance away from the position in SMDfile AND the real position of the COM of the selected atoms. So there is a very large force on them trying to pull them to the reference position.
Why do the coordinates in the *coor file match those in SMDfile? Were you not doing SMD initially?
________________________________
From: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] On Behalf Of Ana Celia Araujo Vila Verde
Sent: Sunday, April 08, 2007 1:56 PM
To: namd-l_at_ks.uiuc.edu
Subject: RE: namd-l: wrong velocity in steered molecular dynamics
Hi JC,
Thanks for the suggestion. I checked and the reference coordinates in my SMDfile exactly match the ones in the *coor file that I use for the restart.
I'm not starting from time zero, but I can't understand exactly what you mean when you say that the coordinates should not match in this case. Is there a reference I can read to understand what you mean?
Thanks again,
Ana
________________________________
From: JC Gumbart [mailto:gumbart_at_ks.uiuc.edu]
Sent: Sat 4/7/2007 7:38 PM
To: Ana Celia Araujo Vila Verde
Cc: namd-l_at_ks.uiuc.edu
Subject: Re: namd-l: wrong velocity in steered molecular dynamics
Check that the reference coordinates in your SMDfile match those being loaded initially (either coordinates or binCoordinates in case you are restarting from a previous simulation). My bet is that they are very different.
If you are not starting from time 0, then they shouldn't match; instead, the real coordinate should be approximately v*t away from the reference.
On Apr 7, 2007, at 2:29 PM, Ana Celia Araujo Vila Verde wrote:
Hi all,
I'm running a SMD simulation, and I defined SMDVel as 0.00001 (angstroms/timestep). The *log file confirms this velocity: "Info: SMD VELOCITY 1e-05 ANGSTROM/TIMESTEP"
However, during my simulation I find that the SMD atoms travel 38 angstroms in 800 timesteps (0.0475 A/time-step or a whooping 5000 meters/second...)
Can anyone offer some insight as to what is going on?
Thanks,
Ana
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:33 CST