surface stress

Message

vasp · #1 Post by **vasp** » Fri Nov 09, 2007 2:33 pm

Hello everyone! I would like to calculate surface stress. In the OUTCAR stress in x-, y- and z-directions is given. Are these stress components (Total) in eV/Ã… or eV/(Ã…*Ã…*Ã…)? Often the surface stress is presented in the form eV/(Ã…*Ã…). How is surface stress obtained in this unit (eV/(Ã…*Ã…)) from the x- and y-components given in the OUTCAR?
I have been pondering these questions, because I get significantly different results (x,y,z stress components) for the same system, when I use slabs with different surface sizes. Therefore, I have been thinking, whether these components should be multiplied or divided by the lattice vectors to get comparable values.

Any advice would be great and most welcome. Thank you!

vasp · #2 Post by **vasp** » Tue Nov 13, 2007 8:47 am

I realize that the surface stress is very sensitive to the details of the calculation, and that the results are comparable between systems of different surface areas. Furthermore, the unit surface area and strain should be taken into account in the units.

oja · #3 Post by **oja** » Thu Dec 13, 2007 1:09 pm

To my understanding, stresses are in units eV/reduced length, as stated in the OUTCAR file. Therefore, they are dependent on your unit cell size and have to be divided by the lattice vector in the direction in question to obtain results in eV/Ã….

vasp · #4 Post by **vasp** » Thu Dec 27, 2007 8:03 am

'To my understanding, stresses are in units eV/reduced length, as stated in the OUTCAR file. Therefore, they are dependent on your unit cell size and have to be divided by the lattice vector in the direction in question to obtain results in eV/Ã….'

Yes, I understood this. Thank you for your answer!