My Community

When calculating the bandstructure the system always notes me that the stress and force is not correct? Can anyme help me to find out the question? The INCAR is like this,
SYSTEM=Pt-fcc
ENCUT=500
ISTART=0;ICHARG=2
ISMEAR=2; SIGMA=0.2
NSW=100
POTIM=0.2
IBRION=1
ISPIN=2
ALGO=V
LVTOT=.TRUE.
LELF=.TRUE.
GGA=91
EDIFF=1E-5
EDIFFG=-1E-4
PREC=Accurate
LREAL= Auto

The results is
TOTAL-FORCE (eV/Angst)
--------------------------------------
0.127604 0.127604 0.127604
0.127604 0.127604 0.127604
0.127604 0.127604 0.127604
0.127604 0.127604 0.127604
-0.000485 -0.000485 -0.000485
-0.000485 -0.000485 -0.000485
-0.000485 -0.000485 -0.000485
-0.000485 -0.000485 -0.000485
-0.065171 -0.065171 -0.065171
-0.065171 -0.065171 -0.065171
-0.065171 -0.065171 -0.065171
-0.065171 -0.065171 -0.065171
-0.000069 -0.000069 -0.000069
-0.000069 -0.000069 -0.000069
-0.000069 -0.000069 -0.000069
-0.000069 -0.000069 -0.000069
-0.061879 -0.061879 -0.061879
-0.061879 -0.061879 -0.061879
-0.061879 -0.061879 -0.061879
-0.061879 -0.061879 -0.061879
--------------------------------------
-0.004213 -0.014628 -0.004096


FREE ENERGIE OF THE ION-ELECTRON SYSTEM (eV)
----------------------------------------------
free energy TOTEN = -82.747996 eV

energy without entropy= -82.672625 ener

external pressure = -38.28 kB Pullay stress = 0.00 kB

what do you do, exactly? (slab, bulk,. have you optimized the lattice structure before?)
-- I think you must have mixed up INCAR: the error message only shows up if the charge is kept constant during the run (ICHARG>10). In such a case, the forces and stress tensors are not expected to be correct.
-- the forces are far from being converged, which ionic step do they come from?

Thanks Sir. I want to study the adsorption on metal surface. The above data are the optimization results for a 20 atom Pt slab containing five layers. The lattice constant of the bulk is calculated as 3.9765.
I still cann't find the problem in INCAR file. Please help me !!

thank you for your additional information.
concerning your INCAR everything seems ok (except that I would not set LVTOT and ELF in a standard relaxation run), therefore the message that "stress and force is not correct" puzzles me:
-- usually, this shows up if you start a run from a converged CHGCAR with ICHARG=11 or 12 and change the k-mesh (then, the charge is of course not the scf-charge for that k-mesh and the forces and stress are not correct.
-- the external pressure (-38.28 kB) seems ok: Due to the vaccum you have in your slab, it will not go to 0, even if you work with relaxed bulk lattice parameters.
-- I would check the POSCAR file (maybe there is some printing error in the positions of the atoms, it seems strange that layers
2 and 4 are almost perfectly converged whereas 1,3 and 5 are not (neither in x,y nor z) . It may also be helpful to check the distances between the nearest neighbors as written in OUTCAR (lines after nearest neighbor table ).

Does relaxation under the condition of fixed ionic positions give such error information? I will try it.

no, it does not

Pt-fcc
3.9765
1.0 -1.0 0.0
0.0 1.0 -1.0
2.785239 2.785239 2.785239
20
Selective Dynamics
Direct
0.000000 0.000000 0.119679 F F F
0.000000 0.500000 0.119679 F F F
0.500000 0.000000 0.119679 F F F
0.500000 0.500000 0.119679 F F F
0.000000 0.000000 0.478714 T T T
0.000000 0.500000 0.478714 T T T
0.500000 0.000000 0.478714 T T T
0.500000 0.500000 0.478714 T T T
0.333333 0.166667 0.000000 F F F
0.333333 0.666667 0.000000 F F F
0.833333 0.166667 0.000000 F F F
0.833333 0.666667 0.000000 F F F
0.333333 0.166667 0.359036 T T T
0.333333 0.666667 0.359036 T T T
0.833333 0.166667 0.359036 T T T
0.833333 0.666667 0.359036 T T T
0.166667 0.333333 0.239357 F F F
0.166667 0.833333 0.239357 F F F
0.666667 0.333333 0.239357 F F F
0.666667 0.833333 0.239357 F F F

thanks Sir, this is my poscar file