[Pw_forum] cutoff convergence of sapphire for elastic properties

[Michael Mehl]

The elastic constants of sapphire are all in the 100-500 GPa range (same 
reference as before).  A 1 kbar error in your calculation, or even a 10 
kbar (the pessimist in me says 100 kbar) is going to be small compared 
to DFT errors.  In general, trying to get elastic constants accurate to 
1 kbar (0.1 GPa) with current DFT functionals and techniques is futile.  
(Consider the change in the elastic constants due to volume, noting that 
DFT does not get the equilibrium volume correct to parts per thousand, 
much less parts per million.)

In any case, Hooke's law usually works for strains up to a few percent, 
though I'm not sure what the range is for sapphire.   My personal choice 
for doing elastic constant calculations is to take relatively large 
strains (1%, 2%, ... , even up to 10%) and see how long stress is 
proportional to strain.  Then the effects of relatively small errors in 
stress are less important.  (I should say I do this mostly with energy 
versus the strain squared, rather than stress versus strain, but the 
principle is the same.)



On 05/08/2012 12:42 PM, Jörg Buchwald wrote:
> Am Mon, 7 May 2012 14:22:58 -0400
> schrieb Mike Mehl<rcjhawk at gmail.com>:
>
> But to get the elastic constants in the elastic regime, i would
> like then apply strains of serveral per mill, which is of the same
> order of magnitude, i.e. also corresponds to changes of the stress
> tensor of the order of 1 kbar, which means that errors in the kbar
> range would be too high.
> An alternative could be the calculation of the elastic constants using
> the second derivative of the energy. But this won't work for big
> supercells due to the computation time and the number of measuring
> points needed.
>
>


-- 
Michael J. Mehl
Head, Center for Computational Materials Science
Naval Research Laboratory Code 6390
Washington DC