RE: how to apply a random force on a atom

From: Blake Charlebois (bdc_at_mie.utoronto.ca)
Date: Mon Jun 06 2005 - 16:06:44 CDT

Hi,

The Langevin temperature control scheme will do something similar to what you are trying to do. You can apply Langevin forces to one or more atoms. If I am not mistaken, the random forces have no time correlation and have a Gaussian distribution with a mean of zero and a mean squared value that depends on the temperature and the damping constant. However, the scheme also applies a viscous drag force. Using the Langevin scheme would be less work than trying to implement random forces in tcl. It would also guarantee that the random forces do not heat up the system. Why do you want to apply random forces to one atom?

Blake

-----Original Message-----
From: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] On Behalf Of ÎâºéÃ÷
Sent: June 6, 2005 1:32 PM
To: namd-l_at_ks.uiuc.edu
Subject: namd-l: how to apply a random force on a atom

i am a green hand to namd,now i want to do a simulation withing a random force
applying on a atom in a peptite in the whole simulation process.For example,i
want the force random in every step ,not only the size ,but also the
direction.Is there anyway ??? Can the tclforce meet this require???

-- 
lsp04whm_at_zsu.edu.cn
whm_zsu_at_hotmail.com
020-84035639 (宿舍)
020-84110039 (实验室)

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:18:48 CST