From chemistry-request /at\server.ccl.net Wed Apr  3 12:56:30 2002
Received: from nori.rl.ac.uk ([130.246.135.154])
	by server.ccl.net (8.11.6/8.11.0) with ESMTP id g33HuTp24790
	for <chemistry #at# ccl.net>; Wed, 3 Apr 2002 12:56:30 -0500
Received: from kohn.nd.rl.ac.uk.nd.rl.ac.uk (kohn.nd.rl.ac.uk [130.246.48.218])
	by nori.rl.ac.uk (8.11.1/8.11.1) with ESMTP id g33Hu9V01493;
	Wed, 3 Apr 2002 18:56:15 +0100
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <15531.16953.867937.992101(+ at +)kohn.nd.rl.ac.uk>
Date: Wed, 3 Apr 2002 18:56:09 +0100
To: "Sue Heavner" <sheavner-: at :-hsc.wvu.edu>
Cc: <chemistry %-% at %-% ccl.net>
Subject: CCL:convergence of MD simulations
In-Reply-To: <scaaf80d.001 ^at^ MAIL.hsc.wvu.edu>
References: <scaaf80d.001#* at *#MAIL.hsc.wvu.edu>
X-Mailer: VM 7.03-rfhacked under 21.4 (patch 4) "Artificial Intelligence" XEmacs Lucid
From: "Keith Refson" <kr- at -isise.rl.ac.uk>

Sue Heavner writes:
 > Dear CCL members:
 >  I read that in NVE simulations, the ratio of RMS fluctuations of
 >  the potential and kinetic energies is an indication
 > of convergence of energy, with RMS(PE)/RMS(KE) = 0.01 indicating a
 > satisfactory level of global energy stabilization.  Is there such a
 > test for NTP simulations?

If you are using a Hamiltonian-derived set of equations for the NTP
simulation (such as Nose/Parrinello-Rahman or Nose/Andersen) then the
answer is yes -- just use the conserved Hamiltonian in place of the
energy.  

Note that you can't do this directly for the common Nose-Hoover
formulation, though Frenkel and Smit (p139) suggest a way of
reconstructing a conserved quantity with a little extra work in
solving for the s variable.

Keith Refson
-- 
Dr Keith Refson, 
Building R3
Rutherford Appleton Laboratory
Chilton
Didcot
Oxfordshire OX11 0QX
T: 01235 446652		K.Refson-: at :-
F: 01235 445720		:~at~:rl.ac.uk