My Community

I am calculating band structure for CNT structure. I order to do this, I relaxed the structutre and sel-consistently obtained the charge density successfully. In an attempt to caculate the band energies using this optimized charge density along the tube axis with greater mesh, the following warning shows up. "stress and forces are not correct". I am using Vasp 4.6
Can any one help. my INCAR and KPOINTS are:

Symmetry line k-points
ymmetry line k-points
15
Line-mode
Cartessian
0 0.5 0 ! X
0 0 0 ! Gamma
0 0 0 ! Gamma
0 0.5 0 ! X

SYSTEM = CNT13-10 ATOMS

NWRITE = 2;
ISTART = 0;
LCHARG = .FALSE;
ICHARG = 11;
PREC = HIGH; NGX = 110; NGY = 30; NGZ = 110;
LREAL = AUTO
LSCALAPACK = T;
ALGO = V;

! IONIC RELAXATION
NSW = 0;
EDIFFG = -0.005;
IBRION = -1;
ISMEAR = 0; SIGMA = 0.1;

!ELECTRONIC RELAXATION
EDIFF = 0.00001;

I remember this question have been discussed before.
1) if the k-mesh is changed, the read SCF charge density is not the self-consistent charge density for that new k-sampling set.

So, the solution is that you have to use same KPOINTS setting in relaxation and DOS calculation.
Try it.

if you use a self-consistent charge-density for a certain set of k-points, this charge density is not self-consistent with respect to a different k-set anymore. Hence the forces and stresses are of course not expected to be correct in this case.
the VASP output simply reminds you of that fact.