The ABINIT Users Mailing List ( CLOSED )

["" <xzp02@mails.tsinghua.edu.cn>]

Dear All:
    Recently I have calculated the elastic constants of (5,0) carbon nanotube
ropes (van der waals's interaction between tubes), using one tube in a unit 
cell,
20 atoms, the RF calculation is based on the relaxed structure. But I met 
several
problems, and hope someone can help me.

1. Warnings:
#phonon modes warning- :
#accoustic sum rule violation met:the eigenvalues of accoustic mode
#are too large at Gamma point
#increase cutoff energy or k-points sampling.
#the three eigenvalues are:    6.917542E-01   -8.230006E-03   -6.150046E+00
The k-points sampling and cut-off energy is chosen basing on a convergence 
study
of ground state  calculation, need it to be re-checked during the RF 
calculation?
And I wonder how large eigenvalues are physically resonable in calculation.
 
#phonon modes warning- :
#unstable eigenvalue detected in force constant matrix at Gamma point
#the system under calculation is physically unstabale. 
Which is the reason of this unstability? Is it from the lower ecut and kpoints
sampling in the the warning above? Or some other paramters need to be changed?

2. using ANADDB, I got a strange result, as following lines, the relaxed 
elastic
tensor is quite  far from the clapmp constants. What can be the reason of 
this?

    Also the clamped one is nearer to the experimental value in bulk terms, 
but
the shear modulus is much larger that the experimental values. So maybe it's
reasonable for a LDA calculation in such van der waals binding structure? (I 
use
6c.pspnc).

# Elastic Tensor(clamped ion)(Unit:10^2GP,VOIGT notation):
#   3.6821201   0.9782996   0.8939238   0.0012755  -0.0004394  -0.0349566
#   0.9748925   3.9144465   0.9130225   0.0011397  -0.0003610   0.0757602
#   0.9655227   0.9767764  10.4703105  -0.0008429   0.0003368  -0.0754624
#  -0.0001281   0.0004466  -0.0002266   2.1834268  -0.0571842   0.0002460
#   0.0002052   0.0000761  -0.0000868  -0.0566541   2.1029358   0.0002131
#  -0.0516363   0.0605346  -0.0413156  -0.0008946   0.0001059   1.3313797
# Elastic Tensor(relaxed ion)(Unit:10^2GP,VOIGT notation):
#
#  85.7038665  60.6379229 -32.5085525   4.8423128  -1.8292667 -43.1631226
#  60.6345158  46.0598785 -23.1709481   3.4586340  -0.9408542 -31.6062256
# -32.4369536 -23.1071942  24.4083250  -2.2010627   0.6746097  17.8892018
#   4.8409092   3.4579409  -2.2004464   2.2461441  -0.0294756  -2.4656445
#  -1.8286221  -0.9404171   0.6741860  -0.0289456   2.2302799   0.7094024
# -43.1798024 -31.6214511  17.9233485  -2.4667851   0.7092952  24.3498322



--------------5_0.in----------------
GEOMETRY OPTIMIZATION:

optcell    2
ionmov     2 
ntime    100
iscf       5
nstep    100
toldff     5.0d-5
tolmxf     5.0d-4
diemac     1.0d0
dilatmx    1.2
ecutsm     0.5
ecut      35
kptopt     1
nshiftk    1
shiftk     0  0  0
ngkpt      2  2  4
...

RF CALCULATION:
ndtset    2
# jdtset=1
getwfk1  -1
kptopt1   1
nqpt1     0
iscf1     5
# jdtset=2
getwfk2  -1
kptopt2   2
iscf2     3
rfphon2   1
rfstrs2   3
rfdir2    1  1  1
rfatpol2  1 20

nstep    100
toldff     5.0d-6
tolmxf     5.0d-5
ecut      35
nshiftk    1
shiftk     0  0  0
ngkpt      2  2  4
......
-------------------------------------------

Best,
ZP
---------------------------------------------
Xu ZhiPing
Ph.D Candidate
Office: 3-405
Tel    : +86-10-627-83814
Department of Engineering Mechanics
Tsinghua University
Beijing, CHINA, 100084
---------------------------------------------