The ABINIT Users Mailing List ( CLOSED )

["D. R. Hamann" <drhamann@mat-simresearch.com>]

Dear Zhang,

The elastic constants are in Cartesian coordinates.  If your Cartesian 
atom positions are the same using both rhombohedral and hexagonal unit 
cells, the elastic tensors should be the same.  If the one of the cells 
represents a rotation with respect to the other, the elastic tensors 
will be different but related by a geometrical transformation that is 
probably in some textbook on continuum elastic theory.

Don Hamann

waigen zhang wrote:
> Dear Lallet,
> Thank you for your replies.
> Have you tried to calculate the elastic constant? There are several 
> ways to define the structural geometry for alpha-Al2O3 and it seems 
> the nonzero elements in the elastic tensor are different if we choose 
> different geometry.
>
> W.G.Zhang
>
>
> On 6/9/07, *Lallet François* <francois.lallet@umontreal.ca 
> <mailto:francois.lallet@umontreal.ca>> wrote:
>
>     Dear Zhang,
>     I followed the instruction on webpage
>     http://cst-www.nrl.navy.mil/lattice/struk/d5_1.html. and I studied
>     the relaxation of the alpha Al2O3 rhombohedral unit cell with
>     ionmov=2; optcell=2, ngkpt=2 2 2 and ecut was fixed to 15Ha (I
>     used PAW pseudopotentials produced with atompaw) in LDA and GGA.
>     The energy was converged within 0.2eV/atom and the forces relaxed
>     below 10^-4Ha/(Bohr.atom), which was sufficient to produce relaxed
>     lattice parameters and internal ionic positions (see Wyckoff
>     "Crystal Structure" is you have any doubts) in very good agreement
>     with experiments. I calculated also the bulk modulus by fitting
>     E(V) with a Murnaghan equation of state and the results are very
>     good... but I don't know whether the hexagonal unit cell is better
>     than the rhombohedral one for this purpose...
>     Best,
>
>     F. Lallet
>
>     -----Original Message-----
>     From: waigen zhang [mailto:waigen.zhang@gmail.com
>     <mailto:waigen.zhang@gmail.com>]
>     Sent: Fri 6/8/2007 3:52 PM
>     To: forum@abinit.org <mailto:forum@abinit.org>
>     Subject: [abinit-forum] structural geometry of alfa-Al2O3
>
>     Dear all,
>     Dose anybody have idea about the structural geometry about
>     alfa-Al2O3 ?
>     Should we use the rhombohedral or hexagonal primitive cell for ABINIT
>     calculation of elastic constant?
>     I followed the instruction on webpage
>     http://cst-www.nrl.navy.mil/lattice/struk/d5_1.html.
>     But still not sure about that. Your suggestion is welcome.
>
>     Best,
>     W.G.Zhang

--
D. R. Hamann
Mat-Sim Research LLC    | Deptartment of Physics
P.O. Box 742            |  and Astronomy
Murray Hill, NJ 07974   | Rutgers University
phone: 908-370-8079     | 732-445-4381

email: drhamann@mat-simresearch.com