forum@abinit.org

Subject: The ABINIT Users Mailing List ( CLOSED )

List archive

Re: [abinit-forum] meaning of tolmxf?

From: matthieu verstraete <matthieu.jean.verstraete@gmail.com>
To: forum@abinit.org
Subject: Re: [abinit-forum] meaning of tolmxf?
Date: Wed, 29 Jul 2009 20:34:38 +0200
Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type:content-transfer-encoding; b=CkXWJo6c0TdGoeSyiHOpuHa0D7NZuqbIRF0+HDLc6giY7yYO3jsX88+riuOmhwinO4 Fm+wzKT/U3q+TFdtLTLVgOBji1ZRxPcz8/E0TWNak8tkoHmalNCB+ZeqvsDcupVvaKxn y9uFom0wLiL77SirKzl4y3eCZUraC7zvNqpwc=

which optcell have you used? Check the source file
src/67_common/fconv.F90 for more specifics on how the fmax is
calculated in each case. Your forces are clearly converged, so it's
the stress, eventually subtracting the trace or something like that,
which gives fmax (3.5201E-04)

Matthieu

On Wed, Jul 29, 2009 at 7:41 PM, <bmeredig@gmail.com> wrote:
> Hi all,
>
> Hunted around here for a while but couldn't find exactly what I'm looking
> for--apologies in advance if this is a very simple question.
>
> I ran a BFGS minimization of a cell, which converged after 9 steps. The code
> printed
>
> At Broyd/MD step 9, gradients are converged :
> max grad (force/stress) = 3.5201E-04 < tolmxf= 5.0000E-04 ha/bohr (free
> atoms)
>
> I was under the impression that the convergence criterion (here 3.5201E-04)
> ought to relate to the largest single Cartesian force component on any of
> the
> atoms, or the largest stress component multiplied by strfact (the default,
> 100). Yet my forces are
>
> cartesian forces (hartree/bohr) at end:
> 1 0.00000022196000 -0.00000026184815 -0.00000036891432
> 2 0.00000014872891 -0.00000019322115 0.00000038078365
> 3 0.00004108870399 0.00007711205062 0.00000161583881
> 4 -0.00004129394759 -0.00007690312473 -0.00000162123890
> 5 -0.00005079422183 -0.00008924105679 0.00000990463720
> 6 0.00005062877651 0.00008948720020 -0.00000991110645
> frms,max,avg= 4.5074698E-05 8.9487200E-05 -1.466E-08 1.384E-08 -4.976E-10
> h/b
>
> and my stresses are
>
> Cartesian components of stress tensor (hartree/bohr^3)
> sigma(1 1)= 2.33910376E-05 sigma(3 2)= 3.67303937E-06
> sigma(2 2)= 1.91354379E-05 sigma(3 1)= 1.59285145E-06
> sigma(3 3)= -5.30868121E-05 sigma(2 1)= -5.32246843E-05
>
> Could someone kindly explain the origin of the 3.5201E-04 that caused the
> code
> to reach convergence?
>
> Thanks much,
> Bryce
>

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dr. Matthieu Verstraete

European Theoretical Spectroscopy Facility (ETSF)
Dpto. Fisica de Materiales,
U. del Pais Vasco,
Centro Joxe Mari Korta, Av. de Tolosa, 72, Phone: +34-943018393
E-20018 Donostia-San Sebastian, Spain Fax : +34-943018390

Mail : matthieu.jean.verstraete@gmail.com
http://www-users.york.ac.uk/~mjv500

[abinit-forum] meaning of tolmxf?, bmeredig, 07/29/2009
- Re: [abinit-forum] meaning of tolmxf?, matthieu verstraete, 07/29/2009
  - Re: [abinit-forum] meaning of tolmxf?, Bryce Meredig, 07/30/2009