[Pw_forum] phonon dispersion calculation abruptly stops

Eyvaz Isaev eyvaz_isaev at yahoo.com
Wed Jun 17 22:35:26 CEST 2009


Dear Mikiyas,


--- On Wed, 6/17/09, Mikiyas Tsegaye <mastermik at gmail.com> wrote:

Let us analyse your input files.

Here is a sketch of your scf-input:

> &system
> ibrav = 2, celldm(1) =10.56, nat=  2,
> ntyp= 1, ecutwfc = 22.0

That means you have a semiconductor (insulator). 

Now from ph-input:

> ldisp = .true.,
> nq1 = 6, nq2 = 6, nq3 = 6,

I suggest ph.x starts for "Electric Field Calculations", as it does not see the keywords specific for the metallic case, but can not do it as you did not specify epsil=.true., i.e. calculate electric field corrections.
By default epsil=.false., i.e. do not calculate "Electric Field Calculations" (sorry for some tautology) and then ph.x stops (most likely).

A short note on the mail subject: it is not connected with "phonon dispersion" calculations, this script produces only dynamical matricies for the q-mesh specified by nq1,nq2,nq3. 

Bests,
Eyvaz.

-------------------------------------------------------------------
Prof. Eyvaz Isaev, 
Theoretical Physics Department, Moscow State Institute of Steel & Alloys, Russia, 
Department of Physics, Chemistry, and Biology (IFM), Linkoping University, Sweden 
Condensed Matter Theory Group, Uppsala University, Sweden 
Eyvaz.Isaev at fysik.uu.se, isaev at ifm.liu.se, eyvaz_isaev at yahoo.com


--- On Wed, 6/17/09, Mikiyas Tsegaye <mastermik at gmail.com> wrote:

> From: Mikiyas Tsegaye <mastermik at gmail.com>
> Subject: [Pw_forum] phonon dispersion calculation abruptly stops
> To: pw_forum at pwscf.org
> Date: Wednesday, June 17, 2009, 10:40 PM
> Hello,
> 
> I'm trying to calculate phonon dispersion for Germanium.
> I have the pw.x and ph.x inputs prepared as follows:
> *********************************************************************************************************
> &control
>     calculation='scf',
>     restart_mode='from_scratch',
>     prefix='FF',
>     tprnfor = .true.,
>     pseudo_dir = '$PSEUDO_DIR/',
>     outdir='$TMP_DIR/'
> /
> &system
>     ibrav = 2, celldm(1) =10.56, nat=  2,
> ntyp= 1,
>     ecutwfc = 22.0
> /
> &electrons
>     mixing_beta = 0.7
>     conv_thr =  1.0d-8
> /
> ATOMIC_SPECIES
> Ge  72.64  Ge.pz-bhs.UPF
> ATOMIC_POSITIONS
> Ge 0.00 0.00 0.00
> Ge 0.25 0.25 0.25
> K_POINTS
> 10
>    0.1250000  0.1250000 
> 0.1250000   1.00
>    0.1250000  0.1250000 
> 0.3750000   3.00
>    0.1250000  0.1250000 
> 0.6250000   3.00
>    0.1250000  0.1250000 
> 0.8750000   3.00
>    0.1250000  0.3750000 
> 0.3750000   3.00
>    0.1250000  0.3750000 
> 0.6250000   6.00
>    0.1250000  0.3750000 
> 0.8750000   6.00
>    0.1250000  0.6250000 
> 0.6250000   3.00
>    0.3750000  0.3750000 
> 0.3750000   1.00
>    0.3750000  0.3750000 
> 0.6250000   3.00
> 
> EOF
> $ECHO "  running pw.x to calculate scf...\c"
> $PW_COMMAND < Ge.scfK.in > Ge.scfK.out
> check_failure $?
> $ECHO "done"
> 
> # try the whole dispersion
> cat > Ge.phG.in << EOF
> phonons of germanium
> &INPUTPH
> amass(1) = 72.640,
> outdir = '$TMP_DIR/',
> prefix = 'FF',
> tr2_ph = 1.0d-14,
> fildyn = 'FF.dynG',
> ldisp = .true.,
> nq1 = 6, nq2 = 6, nq3 = 6,
> /
> EOF
> 
> $ECHO "  running ph.x to calculate gamma
> phonons...\c"
> $PH_COMMAND < Ge.phG.in > Ge.phG.out
> check_failure $?
> $ECHO "done"
> 
> *******************************************************************************************
> Now, the scf step finishes succesfully, but the ph.x step
> only  
> finishes up to a certain point: It crashes with
> "segementation fault".  
> The output file Ge.phG.out shows the following:
> 
> ************************************************************
> 
> 
>      Program PHONON   
> v.4.0.5  starts ...
>      Today is 17Jun2009 at 10:43:58
> 
>      Ultrasoft (Vanderbilt)
> Pseudopotentials
> 
> 
>      Dynamical matrices for ( 6, 6, 6,)
> uniform grid of q-points
>      (  16q-points):
>        N   
>    xq(1)   
>    xq(2)   
>    xq(3)
>        1 
>    0.00000 
>    0.00000     0.00000
>        2   
> -0.16667     0.16667   
> -0.16667
>        3   
> -0.33333     0.33333   
> -0.33333
>        4 
>    0.50000    -0.50000 
>    0.50000
>        5 
>    0.00000 
>    0.33333     0.00000
>        6   
> -0.16667     0.50000   
> -0.16667
>        7 
>    0.66667    -0.33333 
>    0.66667
>        8 
>    0.50000    -0.16667 
>    0.50000
>        9 
>    0.33333 
>    0.00000     0.33333
>       10 
>    0.00000 
>    0.66667     0.00000
>       11 
>    0.83333    -0.16667 
>    0.83333
>       12 
>    0.66667    -0.00000 
>    0.66667
>       13 
>    0.00000    -1.00000 
>    0.00000
>       14 
>    0.66667    -0.33333 
>    1.00000
>       15 
>    0.50000    -0.16667 
>    0.83333
>       16    -0.33333   
> -1.00000     0.00000
> 
> 
> 
>      bravais-lattice index 
>    =           
> 2
>      lattice parameter
> (a_0)   =      10.5600 
> a.u.
>      unit-cell volume   
>       =     294.3959
> (a.u.)^3
>      number of atoms/cell   
>   =            2
>      number of atomic types 
>   =            1
>      kinetic-energy cut-off 
>   =      22.0000  Ry
>      charge density cut-off 
>   =      88.0000  Ry
>      convergence threshold 
>    =      1.0E-14
>      beta       
>               = 
>      0.7000
>      number of iterations used = 
>           4
>      Exchange-correlation   
>   =  SLA  PZ   NOGX NOGC
> (1100)
> 
> 
>  
>    celldm(1)=   10.56000 
> celldm(2)=    0.00000  celldm(3)= 
>   0.00000
>      celldm(4)=   
> 0.00000  celldm(5)=    0.00000 
> celldm(6)=    0.00000
> 
>      crystal axes: (cart. coord. in
> units of a_0)
>            
>    a(1) = ( -0.5000  0.0000  0.5000
> )
>            
>    a(2) = (  0.0000  0.5000 
> 0.5000 )
>            
>    a(3) = ( -0.5000  0.5000  0.0000
> )
> 
>      reciprocal axes: (cart. coord. in
> units 2 pi/a_0)
>            
>    b(1) = ( -1.0000 -1.0000  1.0000 )
>            
>    b(2) = (  1.0000  1.0000 
> 1.0000 )
>            
>    b(3) = ( -1.0000  1.0000 -1.0000 )
> 
> 
>      Atoms inside the unit cell:
> 
>    Cartesian axes
> 
>      site n.  atom   
>   mass       
>    positions (a_0 units)
>         1       
> Ge  72.6400   tau( 1) = (   
> 0.00000    0.00000     
> 0.00000  )
>         2       
> Ge  72.6400   tau( 2) = (   
> 0.25000    0.25000     
> 0.25000  )
> 
>      Computing dynamical matrix for
>                
>     q = (    0.00000   
> 0.00000    0.00000 )
> 
>      49 Sym.Ops. (with q -> -q+G )
> 
> 
>      G cutoff =  248.5712 
> (   4111 G-vectors) 
>    FFT grid: ( 24, 24,  
> 24)
>      number of k
> points=   10
>                
>        cart. coord. in units
> 2pi/a_0
>         k(    1) =
> (   0.1250000   0.1250000   0.1250000),
> wk =    
> 0.0625000
>         k(    2) =
> (   0.1250000   0.1250000   0.3750000),
> wk =    
> 0.1875000
>         k(    3) =
> (   0.1250000   0.1250000   0.6250000),
> wk =    
> 0.1875000
>         k(    4) =
> (   0.1250000   0.1250000   0.8750000),
> wk =    
> 0.1875000
>         k(    5) =
> (   0.1250000   0.3750000   0.3750000),
> wk =    
> 0.1875000
>         k(    6) =
> (   0.1250000   0.3750000   0.6250000),
> wk =    
> 0.3750000
>         k(    7) =
> (   0.1250000   0.3750000   0.8750000),
> wk =    
> 0.3750000
>         k(    8) =
> (   0.1250000   0.6250000   0.6250000),
> wk =    
> 0.1875000
>         k(    9) =
> (   0.3750000   0.3750000   0.3750000),
> wk =    
> 0.0625000
>         k(   10) =
> (   0.3750000   0.3750000   0.6250000),
> wk =    
> 0.1875000
> 
>      PseudoPot. # 1 for Ge read from
> file Ge.pz-bhs.UPF
>      Pseudo is Norm-conserving, Zval
> =  4.0
>      Generated by new atomic code, or
> converted to UPF format
>      Using radial grid of  291
> points,  2 beta functions with:
>                
> l(1) =   0
>                
> l(2) =   1
> 
> 
>      Atomic displacements:
>      There are   2
> irreducible representations
> 
>      Representation 
>    1      3 modes - To be
> done
> 
>      Representation 
>    2      3 modes - To be
> done
>      PHONON   
>    :     0.51s CPU
> time,    0.58s wall time
> 
> 
>      Alpha used in Ewald sum
> =   0.9000
> 
>      Electric Fields Calculation
> 
> *********************************************************************
> 
> Can anyone tell me why this is happening?
> 
> ---------
> 
> Mikiyas Tsegaye
> University of Virginia
> Department of Electrical and Computer Engineering
> Charlottesville, VA, USA
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
> 


      



More information about the users mailing list