Skip to Content.
Sympa Menu

forum - Re: [abinit-forum] rprim

forum@abinit.org

Subject: The ABINIT Users Mailing List ( CLOSED )

List archive

Re: [abinit-forum] rprim


Chronological Thread 
  • From: Xavier Gonze <gonze@pcpm.ucl.ac.be>
  • To: forum@abinit.org
  • Subject: Re: [abinit-forum] rprim
  • Date: Tue, 29 Apr 2003 20:09:14 +0200

Dear Martin,

M.G.A.Tijssens wrote:
Dear abiniters,
I have been calculating elastic constants for NiTi. One deformation
mode of the unit cell that I use is a simple shear. To accomplish this
I include in the input file the lines:



rprim: 1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0

rprim+ 0.0 0.02 0.0
0.0 0.0 0.0
0.0 0.0 0.0

As I understood, these give the directions of the vectors spanning
the unit cell and they are multiplied by the dimensions given in
acell (which is three times 5.718 in my case, corresponding to the
minimum energy lattice parameter found for Troullier-Martins
potentials).

The results that I anticipated is that the energy will increase, and
reach a maximum when (the first line of) rprim equals 1.0 0.5 0.0
since at that moment the unit cell has been sheared by 45 degrees.

However, the maximum is reached when the first line reads 1.0 1.0 0.0.
Can anybody tell me if this is a mistake in abinit (doubtfull...)
or my misinterpretation of rprim. If so, how then should I interpret
rprim (and I did read the help file for this variable!).

To complement the answer given by Razvan, a long time ago ...
(if you are still interested ;-)

Let's say that your Ni atom is at the origin, and your Ti atom is
at 0.5 0.5 0.5 (or the reverse, no importance).
When the rprim reach 1.0 1.0 0.0, note that the
Ti atom located in 0.5 0.5 0.5 is not at all at the
BCC site of the simple cubic sublattice of Ni atoms , but at the center
of squares with four Ni atoms.
(Suggestion : look at the atomic coordinates xcart !)
In order to obtain the periodicity, you need to go to 1.0 2.0 0.0.
(Still, the periodicity will not be perfect, because the FFT
grid is not the same, but this is a 0.00001 Hartree effect, typically).

Xavier






  • Re: [abinit-forum] rprim, Xavier Gonze, 04/29/2003

Archive powered by MHonArc 2.6.16.

Top of Page