[Pw_forum] about the degaussq in the calcualtion of phonon coupling

Wei Zhou zdw2000 at gmail.com
Thu Mar 31 15:11:56 CEST 2011


thank you for you replies.
In fact I believe must choose a degaussq so that the“lambda” euqal the
"int alpha2F" in the output file afer executing the lamdba.x, but I
found if meet this condition, the gaussq must larger than 1.
if I was wrong, then  how to choose the value of degaussq.

the samll q-points weight  are come from a scf calculation results. is
there anything wrong with me?
SCF INPUT FILE
. ../environment/environment_opt
 rm $TMP_DIR/*
cat > ba.scf.in<< EOF
&control
    calculation='scf'
    restart_mode='from_scratch',
    prefix='ba',
    pseudo_dir = '$PSEUDO_DIR/',
    outdir='$TMP_DIR/'
 /
 &system
     ibrav= 4
     celldm(1) =7.21969953
     celldm(3) =1.511267317
     nat = 2
     ntyp =   1
    ecutwfc =50,
    ecutrho=350
    occupations='smearing', smearing='methfessel-paxton', degauss=0.022
 /
 &electrons
    conv_thr =  1.0d-8
    mixing_beta= 0.7
 /
ATOMIC_SPECIES
Ba 137.327 Ba.ncpp
ATOMIC_POSITIONS {crystal}
Ba       0.333333330   0.666666000   0.250000000
Ba       0.666666660   0.333333000   0.750000000
K_POINTS {automatic}
7 7 4  0 0 0
EOF
$PARA_PREFIX $PW_ROOT/pw.x  $PARA_POSTFIX < ba.scf.in>ba.scf.out
OUT SMALL-QPOINTS

     number of k points=    24  gaussian broad. (Ry)=  0.0220     ngauss =   1
                       cart. coord. in units 2pi/a_0
        k(    1) = (   0.0000000   0.0000000   0.0000000), wk =   0.0102041
        k(    2) = (   0.0000000   0.0000000   0.1654241), wk =   0.0204082
        k(    3) = (   0.0000000   0.0000000  -0.3308482), wk =   0.0102041
        k(    4) = (   0.0000000   0.1649572   0.0000000), wk =   0.0612245
        k(    5) = (   0.0000000   0.1649572   0.1654241), wk =   0.1224490
        k(    6) = (   0.0000000   0.1649572  -0.3308482), wk =   0.0612245
        k(    7) = (   0.0000000   0.3299144   0.0000000), wk =   0.0612245
        k(    8) = (   0.0000000   0.3299144   0.1654241), wk =   0.1224490
        k(    9) = (   0.0000000   0.3299144  -0.3308482), wk =   0.0612245
        k(   10) = (   0.0000000   0.4948717   0.0000000), wk =   0.0612245
        k(   11) = (   0.0000000   0.4948717   0.1654241), wk =   0.1224490
        k(   12) = (   0.0000000   0.4948717  -0.3308482), wk =   0.0612245
        k(   13) = (   0.1428571   0.2474358   0.0000000), wk =   0.0612245
        k(   14) = (   0.1428571   0.2474358   0.1654241), wk =   0.1224490
        k(   15) = (   0.1428571   0.2474358  -0.3308482), wk =   0.0612245
        k(   16) = (   0.1428571   0.4123930   0.0000000), wk =   0.1224490
        k(   17) = (   0.1428571   0.4123930   0.1654241), wk =   0.2448980
        k(   18) = (   0.1428571   0.4123930  -0.3308482), wk =   0.1224490
        k(   19) = (   0.1428571   0.5773503   0.0000000), wk =   0.0612245
        k(   20) = (   0.1428571   0.5773503   0.1654241), wk =   0.1224490
        k(   21) = (   0.1428571   0.5773503  -0.3308482), wk =   0.0612245
        k(   22) = (   0.2857143   0.4948717   0.0000000), wk =   0.0612245
        k(   23) = (   0.2857143   0.4948717   0.1654241), wk =   0.1224490
        k(   24) = (   0.2857143   0.4948717  -0.3308482), wk =   0.0612245


2011/3/31, Eyvaz Isaev <eyvaz_isaev at yahoo.com>:
> Dear Wei,
>
> As I noticed before, you do not need such kind HUGE degaussq.
> I always used ~0.1 and never had a problem.
>
> I have a question: how did you find  q-points weight?
> Wrong weighting may affect your result.
>
> Best regards,
> Eyvaz
>
>  -------------------------------------------------------------------
> Prof. Eyvaz Isaev,
> Department of Physics, Chemistry, and Biology (IFM), Linkoping University,
> Sweden
>
> Theoretical Physics Department, Moscow State Institute of Steel & Alloys,
> Russia,
>
> isaev at ifm.liu.se, eyvaz_isaev at yahoo.com
>
>
>
> ----- Original Message ----
> From: Wei Zhou <zdw2000 at gmail.com>
> To: PWSCF Forum <pw_forum at pwscf.org>
> Sent: Thu, March 31, 2011 9:37:55 AM
> Subject: [Pw_forum] about the degaussq in the calcualtion of phonon coupling
>
> Dear users;
>          I rencently do some calculations about the electron-phonon
> calculation using pwscf. however,I feel puzzled about the input file
> of lamdba.x, I want to know how to choose the degaussq in the input
> file of lambda.x, and I found in my case, I only can let the lambda
> equal int alpha2F by using a very larger degaussq, larger than 1, I
> also have read the program of lambda.f90, found it is a gauss
> broadeding when sum the q.  it is reasonable to use a degaussq larger
> than1? does anything wrong with me?
> any help will be apprecialted
> --
> ZhouDawei
> JiLin Universiyt ,ChangChun ,China
> zdw2000 at gmail.com
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>


-- 
ZhouDawei
JiLin Universiyt ,ChangChun ,China
zdw2000 at gmail.com



More information about the users mailing list