<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title></title>
</head>
<body>
<p>
Hi GROMACS user group!<br />
<br />
Maybe this should be aimed at the developers' mailing list... Anyway, I've been having some problems (blowing up) when I tried to simulate a molecule such that one part is coupled to a thermal bath using SD, and another part is not coupled to any bath. Effectively, the uncoupled part has infinite "relaxation time" (tau-T) which is the same as zero coupling (gamma); however, in the *.mdp file it is unrealistic to put in the value infinity for the tau-T parameter. The 4.0 manual specifically says, in section 7.3.14, "Temperature coupling":<br />
<br />
tau t: [ps]<br />
time constant for coupling (one for each group in tc grps), 0 means no temperature coupling<br />
<br />
However, when I tried putting in zero I got the following (debugging information log-file)<br />
<br />
line 267: SD const tc-grp 0: b nan c inf d -inf<br />
line 268: SD const tc-grp 1: b 2.08438e-14 c 8.33021e-11 d -2.5e-07<br />
<br />
where group 0 had tau-T[0] = 0 and group 1 had tau-T[0] = 1 ps . This shows that when I put zero in the temperature coupling timescale (for SD this is not exactly a first-order relaxation time), the SD parameters blow up.<br />
<br />
Note that when I put tau-T[0] = 1000 or tau-T[0] = 10000, the simulation ran properly.<br />
<br />
<br />
Where in the code can I check that what is written in the manual is, indeed correct?<br />
(i.e., I expect an if statement saying that if the tau-T value of some temperature coupling group is zero, then take the inverse of that to be effectively zero).
</p>
<p>
<br />
</p>
<p>
Thanks in advance,<br />
</p>
<p class="imp-signature">
--<br />
Inon   Sharony<br />
ינון     שרוני<br />
+972(3)6407634<br />
atto.TAU.ac.IL/~inonshar<br />
Please consider your environmental responsibility before printing this e-mail.
</p>
</body>
</html>