[Pw_forum] PHONON error with trans=.false. and elph=.true.

GAO Zhe flux_ray12 at 163.com
Mon Jul 25 12:21:22 CEST 2011


Were there any error report in your terminate, such like"Segment fault" ?

--
GAO Zhe
CMC Lab, MSE, SNU, Seoul, S.Korea


At 2011-07-25 13:20:19,"윤정훈" <yjh9898 at postech.ac.kr> wrote:

I am trying to calculate the electron-phonon coupling with PHONON 4.2.1.
I modified the program to work with elph=.true. and trans=.false., referring PW_forum. (i.e I fixed the bug that known)
The previous PHONON calculation for mode frequencies were done earlier, but I found the problem.


The input is written below:
---------------------------------------------------


biGr_Ca


&inputph
    amass(1)=40.078
    amass(2)=12.011
    outdir='./'
    iverbosity=0
    prefix='CaGr131scf'
    niter_ph=100
    tr2_ph=1e-12
    fildyn='131.dyn'
    fildrho='rhoresp'
    fildvscf='potvar131'
    trans=.false.
    ldisp=.true.
    elph=.true.
    nq1=6, nq2=6, nq3=1
/
----------------------------------------------------


AND, I've found the problem: The program stops WITHOUT error message, just before the calculation.
-------------------------------------------------------------------------
.........
     site n.  atom      mass           positions (a_0 units)
        1        Ca  40.0780   tau( 1) = (    0.00000    0.00000    0.00000  )


     Computing dynamical matrix for
                    q = (   0.0000000   0.0000000   0.0000000 )


     25 Sym.Ops. (with q -> -q+G )




     G cutoff =  553.0542  (  14740 G-vectors)     FFT grid: ( 48, 48,240)
     G cutoff =  201.1106  (   3234 G-vectors)  smooth grid: ( 30, 30,144)
     number of k points=    19  gaussian broad. (Ry)=  0.0300     ngauss =   1


     PseudoPot. # 1 for Ca read from file Ca.pbe-nsp-van.UPF
     Pseudo is Ultrasoft + core correction, Zval = 10.0
     Generated by new atomic code, or converted to UPF format
     Using radial grid of  845 points,  6 beta functions with:
                l(1) =   0
                l(2) =   0
                l(3) =   1
                l(4) =   1
                l(5) =   2
                l(6) =   2
     Q(r) pseudized with  6 coefficients,  rinner =    1.200   1.200   1.200
                                                       1.200   1.200




     Atomic displacements:
     There are   2 irreducible representations


     Representation     1      2 modes -E_1u To be done


     Representation     2      1 modes -A_2u To be done




     PHONON       :     4.38s CPU time,        6.23s WALL time


-----------------------------------------------------------------------------------------


Please help me for this unexpected error.


P.S I wonder the trans=.false. calculation with 16 cpus in a pool is compatible with phonon frequency calculation with 64 cpus in 4 pools. The fildvscf is splat in four files......
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.democritos.it/pipermail/pw_forum/attachments/20110725/d33fdd65/attachment.htm 


More information about the Pw_forum mailing list