[Pw_forum] vc-relax newbie question

mbaris at metu.edu.tr mbaris at metu.edu.tr
Sun May 27 14:57:27 CEST 2007

Hello to all,
  This is my second post to this list today, sorry about that, but I 
think this requires another thread.  As I have mentioned before, I am a 
newbie, and some of the questions may not make sense, however I would 
be glad if you answer them.  In the project I am assigned, I am 
required to find lattice constant of the Pd crystal several times. As 
suggested, due to simplicity, this can be done by interpolating a 
volume vs. E curve, but my previous experience with other programs 
suggest that this is rather inefficient if not inaccurate. At this 
point vc-relax comes to help. This is the input (please notice &cell 

--------begin input card --------------

    calculation = 'vc-relax'
    verbosity = 'minimal'
    pseudo_dir = 'somewhere'
    ibrav = 2,
    celldm(1) = 7.353,
    nat = 1,
    ntyp = 1,
        ecutwfc = 25,
        ecutrho = 250,
    occupations = 'smearing',
        smearing = 'mp',
        degauss = 0.001,
        wmass=425.68   <- I am ok with the defaults mostly, however 
running the #pw.x it complains about missing wmass, bug? /
Pd 106.42 Pd.pbe-nd-rrkjus.UPF
Pd 0 0 0
K_POINTS {automatic}
6 6 6 0 0 0

--------end input card --------------
If I understand correctly, this (the defaults) should be sufficient for 
the calculation of ideal (theoretical) lattice constant (any 
help/comments is appreciated here).

The problem is, I can not locate the resulting lattice parameters in 
the output file!

in reporting the stress on the system, there is an output like

----- begin output -------
     entering subroutine stress ...

          total   stress  (Ry/bohr**3)                   (kbar)     P=  107.95
   0.00073383   0.00000000   0.00000000        107.95      0.00      0.00
   0.00000000   0.00073383   0.00000000          0.00    107.95      0.00
   0.00000000   0.00000000   0.00073383          0.00      0.00    107.95

     Damped Dynamics: convergence achieved, Efinal=   -76.06544689
------ end output ----------

The stress seems to be minimized but high

in the previous non-converged step, there was volume reported, but in 
the last step it is missing and volume information may not be relevant 
for i.e. oblique cells in finding angles etc.

the other output is pretty useless to me as I don't expect anything 
other than fcc:

----- begin output -------

  -0.500000000   0.000000000   0.500000000
   0.000000000   0.500000000   0.500000000
  -0.500000000   0.500000000   0.000000000

Pd       0.000000000   0.000000000   0.000000000
------ end output ----------

but where is a(1)?

I looked through the documentation and could not found anything (is 
there something like OUTPUT_PW? If not, it should be greatly 

and digging thorough various xml files in tmp directory yielded nothing 
up to now.

The most interesting thing is, when I change the lattice dimension to 
8, the stress increase and the system converge immediately in the first 
step. The error seems to be most fundamental but being new to the 
topic, I can not locate it.

P.S. increasing verbosity does not help.

any help is appreciated. O. Baris Malcioglu

More information about the users mailing list