ABINIT Discussion Forums

Dear all,

I want to caclulate the bulk module for a cubic zinc Blende crystal structure of the binary compound CdS by method of increasing the "acell", i tried firstly through the toturespon (elastic toturial 6) and i got alot of warnings, actually i don't know how to calcuate it so please can you explain for me how i do that step by step and in below is my input file :


Best regards...



ndtset 12 # Total number of datasets (3*4)
udtset 3 4 # Double loop for k-sample convergence study

nspden 2
nsppol 2
nspinor 1

# Set 1 : Initial self-consistent and lattice optimization run

getwfk?1 0
ionmov?1 2 # Broyden lattice optimization scheme
ntime?1 5 # Maximim lattice optimization steps
optcell?1 1 # Optimize cell volume only
strfact?1 100 # Test convergence of stresses (Hartree/bohr^3) by
# multiplying by this factor and applying force
# convergence test
tolmxf?1 1.0e-6 # Convergence limit for forces as above
tolvrs?1 1.0d-18 # Need excellent convergence of GS quantities for RF runs

# Set 2 : Additional iteration to print density just at converged acell

prtden?2 1 # Third dataset needs density
tolvrs?2 1.0d-18

# Set 3 : Converge unoccupied wave functions

getden?3 -1 # Use density from previout set
tolwfr?3 5.0d-19 # Only wave function convergence can be used with
# non-self-consistent calculation
tolwfr23 1.0d-30 # This is simply for a reason of portability of automatic tests
nstep23 25 # This is simply for a reason of portability of automatic tests
nstep33 35 # This is simply for a reason of portability of automatic tests

# Set 4 : response-function calculations for all needed perturbations

kptopt?4 2 # Time-reversal only for RF calculation
nqpt?4 1
qpt?4 0 0 0 # By symmetry, only need one direction
rfdir?4 1 0 0
rfstrs?4 3 # Need both unaxial and shear strains
tolvrs?4 1.0d-12 # Need reasonable convergence of 1st-order quantities

#Common input data

#Double loop data passing

getcell -1 # Start from optimized (datasets ?2-?4) or previously
# optimized (datasets ?1) acell
getwfk -1 # Use last set of wave functions (except datasets ?1)


#Lattice definition
acell 3*10.96 # Starting value
dilatmx 1.05 # Allow for optimization
rprim 0.0 0.5 0.5
0.5 0.0 0.5
0.5 0.5 0.0

#Definition of the atom types and atoms
ntypat 2
znucl 48 16
natom 2
typat 1 2

#Atomic position
xred 0.0 0.0 0.0
0.25 0.25 0.25

#Definition of the plane wave basis set
ecut 20.947118 # Maximum kinetic energy cutoff (Hartree)
ecutsm 0.5 # Smoothing energy needed for lattice parameter
# optimization. This will be retained for
# consistency throughout.

#Definition of the k-point grid - loop over 3 k-point densities

ngkpt1? 6 6 6
ngkpt2? 8 8 8
ngkpt3? 10 10 10

nshiftk 4 # Use one copy of grid only (default)
shiftk 0.0 0.0 0.5 # This gives the usual fcc Monkhorst-Pack grid
0.0 0.5 0.0
0.5 0.0 0.0
0.5 0.5 0.5

#Definition of occupation numbers and number of bands
nband 10 # With metallic occup
occopt 4 # Femi-function smearing
tsmear 0.02

#Definition of the self-consistency procedure
iscf 5 # Use conjugate-gradient SCF cycle (datasets 1 & 3)
nstep 50 # Maximum number of SCF iterations
# This might not be enough for the very demanding tolwfr?3 above,
# but was chosen for portability reasons.
# enforce calculation of forces at each SCF step
optforces 1

Best regards
Jasim Q. Ali
+9647504571118

Hi!

If you got a lot of warnings, that doesn't have to mean that your calculation (or your result) is wrong. Often a lot of these warnings are insignificant ot the user.

Besides, if you want to calculate only the bulk modulus, you don't have to use elastic constants. You can get it by fitting your E=E(V) data with Murnaghan-Birch equation of state. It will save you a lot of time.

In the end, if you want help, we need know your exact problem. Tutorial on elastics is very self-explanatory.

Best regards,

Igor L.

Dear ilukacevic,

Thank you very much for your cooperation, actually i want to calculate Bulk modulus (B) and its pressure derivative (B') i know they can be calculate by fitting Birch Murnghan but really i don't know how i can do this fitting and what should i calculate to do fitting, so if you can tell me how i can do that step by step i will be very thankful.


Best Regards

Hi!

Well I suggest that you first do a little homework. First, find the answer to your question on this forum. I am sure that someone already posted how to do this step by step. Second, read some literature, for start maybe, R. Martin's book "Electronic structure..."

Best regards,

Igor

Dear ilukacevic,

Thank you for your helping, now i know how to fit but my problem is the initial guess, i know Bp we can put it 4 and Eo is the minimum energy and Vo is the volume at minimum energy but hw we can find Bo that equal Bo=-V * dp/dv so V: which volume? and how we calculate dp/dv?

best regards

I see prof Zwanziger already answered your question. I can only add a quote from prof. Verstraete about the same thing another way.



And last but not least you can also use xmgrace, which does a least
squares fit to any function you give it:

* load your data E(V) (I presume below you use atomic Hartree for
everything, otherwise you will have a hard time with units)

* go to menus Data->Transformations->Non-linear curve fitting

* set the number of parameters to 4 (E_min, B0, B', and V_min in that order=
)

* enter the following (for Murnaghan EOS) in the Formula:
y = A0 + (A1*x)/(A2*A2-A2) * ( A2*(1.-A3/x) + exp(ln(A3/x)*A2)-1.)

* you may also have to set some reasonable starting values, eg
A0,1,2,3 =3D typical etot, 0.001, 4, typical volume
otherwise it may not converge (due to the sensitive dependency on the power=
A2)

* hit Apply (eventually several times if 5 iterations of least-squares
convergence is not enough)

* read off converged values of parameters:
A1* 2.9421e+04 is B_0 in GPa
A2 =3D B' is unitless

* you can also extract the equilibrium energy and volume by hand and
fit with only 2 parameters (I find this works better sometimes).



M-B from wiki:
A0+(9*A3*A1/16)*((((A3/x)^(2/3)-1)^3)*A2+(((A3/x)^(2/3)-1)^2)*(6-4*(A3/x)^(2/3)))


Matthieu