The ABINIT Users Mailing List ( CLOSED )

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

Dear Zhang,

Using the published space group R-3c for alpha-Al2O3, and the elastic 
compliance symmetry properties for class -3m (see J. F. Nye, Physical 
Properties of Crystals, Clarendon Press, Oxford, 1985, p.299) the 
symmetry properties of your elastic tensor look correct except for the 
fact that your crystallographic axes are apparently  rotated 30 degrees 
from the conventional orientation, thereby effectively interchanging x 
and y, so C15 <-> C14.

I can't comment on discrepancies with respect to experiment, except to 
note that in previous correspondence addressed to you on the Forum, 
Lallet François indicated that he got very good results for the Al2O3 
bulk modulus.  Elastic constants may converge considerably more slowly 
with ecut than other properties, and choices of pseudopotentials and 
density functionals could have a significant effect on your results.

There's not much I can add to what I've previously written to you on 
another thread about your problems with N2O4.

Regards,
Don Hamann

waigen zhang wrote:
> Dear all,
> Still the strange elastic constant problem.
> I use abinit to calculate the elastic constant of alpha-Al2O3 as a 
> check(What I need is elastic constant of N2O4).In N2O4 calculation,I 
> tried over 20 times,but I always got negative diagonal element. What I 
> did was,
> 1st,Optimize the structure.
> 2nd,Use the output of xred,acell,angles from the optimization step and 
> other variables such as ecut,ngkpt,iscf...kept the same in the elastic 
> constants calculation.There shouldn't be any mistake in the input 
> files.If needed,I will post it.
>
> Some of the output files of Alpha-Al2O3 are,
> The first derivatives are following,is it good enough or the elements 
> are too big?
> 1st derivatives              - # elements :      36
>    1   1  0.25869898402403D-03  0.00000000000000D+00
>    2   1  0.25869898402403D-03  0.00000000000000D+00
>    3   1  0.25869898402403D-03  0.00000000000000D+00
>    1   2  0.56251931526327D-03  0.00000000000000D+00
>    2   2 -0.56251931526327D-03  0.00000000000000D+00
>    3   2  0.00000000000000D+00  0.00000000000000D+00
>    1   3  0.00000000000000D+00  0.00000000000000D+00
>    2   3  0.56251931526327D-03  0.00000000000000D+00
>    3   3 -0.56251931526327D-03  0.00000000000000D+00
>    1   4 -0.56251931526327D-03  0.00000000000000D+00
>    2   4 -0.42072581611787D-29  0.00000000000000D+00
>    3   4  0.56251931526327D-03  0.00000000000000D+00
>    1   5  0.25869898402403D-03  0.00000000000000D+00
>    2   5  0.25869898402403D-03  0.00000000000000D+00
>    3   5  0.25869898402403D-03  0.00000000000000D+00
>    1   6 -0.56251931526327D-03  0.00000000000000D+00
>    2   6  0.56251931526327D-03  0.00000000000000D+00
>    3   6 -0.21036290805894D-29  0.00000000000000D+00
>    1   7  0.00000000000000D+00  0.00000000000000D+00
>    2   7 -0.56251931526327D-03  0.00000000000000D+00
>    3   7  0.56251931526327D-03  0.00000000000000D+00
>    1   8  0.56251931526327D-03  0.00000000000000D+00
>    2   8 -0.84145163223575D-29  0.00000000000000D+00
>    3   8 -0.56251931526327D-03  0.00000000000000D+00
>    1   9 -0.25869898402403D-03  0.00000000000000D+00
>    2   9 -0.25869898402403D-03  0.00000000000000D+00
>    3   9 -0.25869898402403D-03  0.00000000000000D+00
>    1  10 -0.25869898402403D-03  0.00000000000000D+00
>    2  10 -0.25869898402403D-03  0.00000000000000D+00
>  3  10 -0.25869898402403D-03  0.00000000000000D+00
>    1  13 -0.16403457711541D-02  0.00000000000000D+00
>    2  13 -0.16403457711541D-02  0.00000000000000D+00
>    3  13 -0.15138176513697D-02  0.00000000000000D+00
>    1  14  0.00000000000000D+00  0.00000000000000D+00
>    2  14  0.00000000000000D+00  0.00000000000000D+00
>    3  14  0.00000000000000D+00  0.00000000000000D+00
>
> The calculated elastic constants are following,which are far away from 
> the well-known results.Also there seems to be some mistake about this 
> matrix analysied from symmetry.C14 shouldn't be zero and........
>
> Elastic Tensor (relaxed ion) (unit:10^2GP):
>   (at fixed electric field boundary condition)
>
>    9.0146811   3.0170874   1.2282716  -0.0000014  -2.5594517   0.0000081
>    3.0170874   9.0146790   1.2282701  -0.0000021    2.5594490  -0.0000043
>    1.2282716   1.2282698   7.7354829  -0.0000015   0.0000067  -0.0000015
>    0.0000003  -0.0000005  -0.0000003   1.5197380   0.0000032   2.5594511
>   -2.5594517   2.5594490   0.0000002   0.0000087    1.5197329   0.0000042
>    0.0000083  -0.0000049  -0.0000003   2.5594511   0.0000003   2.9988020
>
> Regards,
> WG

--
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