[Pw_forum] an error with do_ee

Chenghua Sun c.sun1 at uq.edu.au
Sun Feb 6 12:18:17 CET 2011


Dear Emine,
 
The optimized structure used in my work is:
Mn       0.000084580   0.139307868  15.221554432
Mn      -2.663136985   4.752155376  15.221514785
Mn       5.567587322   0.000140707  14.977486470
Mn       2.663321146   4.752154342  15.221590927
O        0.000079739   3.214534101  17.469943271
O       -2.818078799   8.016256977  16.820461708
O        5.567618105   3.254298111  16.820496706
O        2.818266725   8.016258532  16.820492041
O        2.761895432   1.620036457  13.178651487
O        0.000108628   6.403594576  13.178620118
O        8.373256671   1.619981921  13.178625538

The above coordinates were obtained without any correction. 
 
&CONTROL
  calculation  = 'scf',
  pseudo_dir = '/home/uqcsun1/code/pseudo/',
  outdir       = './tmp',
  restart_mode = 'restart',
  !max_seconds = 85000,
  prefix = 'Mn4O7'
  etot_conv_thr = 1d-6,
  forc_conv_thr = 1d-4,
  nstep = 2000,
/
&SYSTEM
  ibrav     = 4,
  celldm(1) = 11.135,
  celldm(2) = 1.0,
  celldm(3) = 2.70,
  nat = 11,
  ntyp = 2,
  ecutwfc   = 25.0,
  ecutrho   = 300.0,
  nspin = 2,
  starting_magnetization(1)= 0.0,
  starting_magnetization(2)= 0.5,
  tot_charge = -1,
  occupations= 'smearing',
  smearing= 'methfessel-paxton',
  degauss= 0.01,
  lda_plus_u = .true.,  Hubbard_U(1)= 1.d-8, Hubbard_U(2)= 4.0,
  assume_isolated = .true.
/
&ELECTRONS
  conv_thr = 1.0d-8,
  mixing_mode = 'plain',
  mixing_beta = 0.1,
  electron_maxstep = 1000,
/
K_POINTS {automatic}
5 5 1 0 0 0
ATOMIC_SPECIES
O   16.0  O-PBE.UPF
Mn  54.938  Mn-PBE.UPF

If possible, could you please help me and run a calculation with the Makov-Payne correction? 
 
Thank you.
 
(I runned with assume_isolated = .true., and the calculated energy is exactly same as that without the correction. it is indicated that the correction is only applied for cubic lattice)
 
Cheers,
Coy
 
 
-------------------------------------------------------------------------
Chenghua Sun, PhD

ARC Centre of Excellence for Functional Nanomaterials
 
Centre for Computational Molecular Science

Australian Institute for Bioengineering and Nanotechnology,

AIBN Building 75, The University of Queensland

Qld 4072, Brisbane, Australia

Tel (617) 33463972;                Fax (617) 33463992

email: c.sun1 at uq.edu.au <mailto:c.sun1 at uq.edu.au>          Web: http://web.aibn.uq.edu.au/cbn <https://exchange.uq.edu.au/exchweb/bin/redir.asp?URL=http://web.aibn.uq.edu.au/cbn> 
--------------------------------------------------------------------------

________________________________

From: pw_forum-bounces at pwscf.org on behalf of Emine Kucukbenli
Sent: Sun 2011-2-6 19:38
To: pw_forum at pwscf.org
Subject: Re: [Pw_forum] an error with do_ee



Dear Coy,

> I also tested the EE approach (six jobs), and all of those have the  
> convergence problem.
Have you put your system in the middle of the cell? The charge density 
should be fairly small at the boundaries for DCC to work. If this is 
not your case, we need to check your input and output to understand 
better.

> Just one more thing: I can only use the above key words (both  
> assume_isolated and EE), I can only run 'scf' job, and fail to run  
> 'relax' for version 4.0, 4.1 and 4.2. Is it normal?
I dont think it is normal, especially in the case of MP correction, as 
the correction is just on the energy. Post more details so we can look 
into it.

Btw, have you tried martyna-tuckerman?
cheers,
emine kucukbenli, phd student, sissa, italy
>


----------------------------------------------------------------
   SISSA Webmail https://webmail.sissa.it/
   Powered by Horde http://www.horde.org/


_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://www.democritos.it/mailman/listinfo/pw_forum


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/ms-tnef
Size: 8891 bytes
Desc: not available
Url : http://www.democritos.it/pipermail/pw_forum/attachments/20110206/0fb2cc51/attachment.bin 


More information about the Pw_forum mailing list