I have a problem about PAW_GGA

[zzzhong]

Hello, everyone
I want to calculate GaMnAs system. I want to use PAW_GGA pseudopotential. First, I want to validate PAW_GGA pseudopotential. I calculate bulk GaAs using Ga and As potpaw_GGA. I get that bulk GaAs lattice constant is about 5.76 angstrom, but 5.65 in experiment. Is the LDA resulted? Next, I use Ga and As PAW pseudopotential. I get that bulk GaAs lattice constant is about 5.64 that is very near exmperimental value. Is the PAW_GGA something wrong? Any replies and explains are appreciated. Thanks

Best regard.
zzzhong

Below is my input file
INCAR
general:
System = Zinc Blende GaAS
ISTART = 0 ; ICHARG=2
PREC=HIGH
ISMEAR = 0; SIGMA = 0.001;
ALGO= V

GGA=91
VOSKOWN=1

POTCAR
PAW Ga
3.00000000000000000
parameters from PSCTR are:
VRHFIN =Ga: s2p1
LEXCH = 91
EATOM = 59.0602 eV, 4.3408 Ry

TITEL = PAW Ga
LULTRA = F use ultrasoft PP ?
IUNSCR = 1 unscreen: 0-lin 1-nonlin 2-no
RPACOR = 2.170 partial core radius
POMASS = 69.723; ZVAL = 3.000 mass and valenz
RCORE = 2.600 outmost cutoff radius
RWIGS = 2.650; RWIGS = 1.402 wigner-seitz radius (au A)
ENMAX = 134.733; ENMIN = 101.050 eV
ICORE = 3 local potential
LCOR = T correct aug charges
LPAW = T paw PP
EAUG = 320.264
DEXC = -.088
RMAX = 3.972 core radius for proj-oper
RAUG = 1.300 factor for augmentation sphere
RDEP = 2.684 radius for radial grids
QCUT = -3.147; QGAM = 6.294 optimization parameters

KPOINTS
K-Points
0
Monkhorst Pack
11 11 11
0 0 0

loop.sh
#! /bin/bash
BIN=vasp
rm WAVECAR
for i in 5.4 5.45 5.5 5.55 5.6 5.65 5.7 5.75 5.8 5.85 5.9 ; do
cat >POSCAR <<!
Zinc Blende
$i
0.0 0.5 0.5
0.5 0.0 0.5
0.5 0.5 0.0
1 1
Direct
-0.125 -0.125 -0.125
0.125 0.125 0.125
!
echo "a= $i" ; $BIN
E=`tail -1 OSZICAR` ; echo $i $E >>SUMMARY.dia
done
cat SUMMARY.dia

[doris]

GGA potentials tend to over-estimate the lattice costant a little bit
sometimes, but 1.7% seems rather large
Please check if your single steps were electronically converged (EDIFF ) .