Deviation of Equation of State from external pressure
Posted: Tue Feb 08, 2022 9:29 pm
Hi all,
I have recently noticed a strange effect in one of my calculations. I plotted volume vs external pressure -- as printed out in the OUTCAR file. I have also plotted the pressure derived by fitting an equation of state of the energy vs volume, I get a significant deviation at low volumes. The fit uses the Birch-Murnaghan equation of state and the fit is excellent, and shows no systematic error. Furthermore, the energies themselves and the equations of state have a close agreement with similar fits made based on calculations using the all-electron Wien2k code with the same functional (PBE).
This deviation does not seem to be caused by Pulay stress. I get approximately the same EOS and external pressure when using a 50% greater ECUT, nor when increasing NBANDS.
kpoint density has been converged well, and PREC=Accurate
I was wondering if anyone has any suggestions for what could be the cause of this deviation? Is there some reason why the calculated VASP stresses should deviate so greatly from the stresses acquired by fitting the equation of state?
All the best,
GW
I have recently noticed a strange effect in one of my calculations. I plotted volume vs external pressure -- as printed out in the OUTCAR file. I have also plotted the pressure derived by fitting an equation of state of the energy vs volume, I get a significant deviation at low volumes. The fit uses the Birch-Murnaghan equation of state and the fit is excellent, and shows no systematic error. Furthermore, the energies themselves and the equations of state have a close agreement with similar fits made based on calculations using the all-electron Wien2k code with the same functional (PBE).
This deviation does not seem to be caused by Pulay stress. I get approximately the same EOS and external pressure when using a 50% greater ECUT, nor when increasing NBANDS.
kpoint density has been converged well, and PREC=Accurate
I was wondering if anyone has any suggestions for what could be the cause of this deviation? Is there some reason why the calculated VASP stresses should deviate so greatly from the stresses acquired by fitting the equation of state?
All the best,
GW