[Wien] Molecular dynamics using Wien2k

[Peter Blaha]

As mewntioned before, the   -so option is wrong. One cannot calculate 
forces with SO in wien2k. Maybe you meant:  -sp    for a spin-polarized 
calculation.

case.finM will be enerated automatically from the forces of a 
scf-calculation. Using   -so you did not calculate forces, ....

MOLD is in my opinion fairly useless, as the average kinetic energy 
(Temperature) will solely depend on your starting geometry and is 
basically "undefined". "Every" MD simulation in literature uses a 
thermostat, mostly the NOSE method and a predefined 
"temperature"(although I admit that we have not followed better 
thermostat developments in recent years).

I also admit, that I have not done any MD run for years, mainly because 
MD in dense packed solids is often of quite limited use because in 
solids very often "nothing happens" at realistic temperatures and 
accessable time scales. Please think carefully about your intended 
application and what you really want to obtain from a MD simulation. 
Otherwise you just burn cpu hours.

And finally, as MD runs would be very expensive, think about your system 
and if an all-electron method like wien2k is necessary and efficient. 
wien2k can compete in speed, if the spheres are large, but can become 
rather expensive if small spheres are needed (like with organic 
molecules, short bonds like in CO, OH, ...)

Am 18.12.2015 um 06:50 schrieb Gavin Abo:
> I haven't tried MD with min_lapw.
>
> The case.finM should be created by the script min_lapw [
> http://zeus.theochem.tuwien.ac.at/pipermail/wien/2006-May/007135.html ].
>
> One of the other users might know better, but it looks like your
> "min_lapw -p -so" command does not look right.
>
> I suggest you run in the terminal the command: min_lapw -h
>
> I'm not seeing an option for '-so' in the output from that command.
>
> Since you did not use '-j', it looks like the script uses "run_lapw -I
> -fc 1. -i 40" by default (whereas '-sp' should use "runsp_lapw -I -fc 1.
> -i 40" by default).
>
> For SO, you probably need to use the "-j" option:
>
> min_lapw -p -j "runsp_lapw -I -so -fc 1. -i 40"
>
> or it looks like there is the '-m' option that will extract force-input
> and execute mini (without JOB) and exit.
>
> On 10/23/2003 8:03 PM, Guo-ping Zhang wrote:
>> Dear Wien2k users,
>>
>> I am interested in using Mini to run MD. After reading the manual and
>> old mailing list, I could not find a workable example. For instance,
>> the manual has an example for NOSE (case.inM) not for MOLD. In the
>> structure optimization section,  all the examples are used for
>> minimization (which works well for my problem using MSR1a option in
>> case.inm though the mixer is very slow).
>>
>> I tried to use min_lapw -p -so (after a prior successful SCF)  but
>> failed, since it needs case.finM (but the manual does not say how to
>> get case.finM). I already looked at haupt_cad.f, which seems to work.
>>
>> Has anyone succeeded using Wien2k to run MD?
>>
>> If so, is it possible to share with me some details how this is done?
>>
>> I would appreciate any hints.
>>
>>
>> Thanks a lot in advance!
>>
>> Guoping
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
> SEARCH the MAILING-LIST at:
> http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html

-- 
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WIEN2k: http://www.wien2k.at
WWW:   http://www.imc.tuwien.ac.at/staff/tc_group_e.php
--------------------------------------------------------------------------