ABINIT Discussion Forums

Hello forum,

which tools do you use to fit E-V for determining the bulk modulus?
Do you have some hints or links for me in "How-To determine the bulk modulus for hcp and wurtzite crystals."

Thanks in advance.

Best wishes,
Marc

Well, for the bulk modulus I usually don't use the EOS. Instead, I run a series of calculations as follows:

ndtset 5

strtarget1 3*3.4D-5 3*0.0
strtarget2 3*1.7D-5 3*0.0
strtarget3 3*0.0 3*0.0
strtarget4 3*-1.7D-5 3*0.0
strtarget5 3*-3.4D-5 3*0.0

ionmov 2
optcell 2
ntime 20
ecutsm 1.0
dilatmx 1.1
tolmxf 1.0D-6
toldff 1.0D-7 (or tolvrs 1.0D-12 if all ions are on special positions)

after these 5 calcs have run, you will have the optimized unit cell volume at 5 pressures: -1.0 GPa, -0.5 GPa, 0.0 GPa, 0.5 GPa, and 1.0 GPa, and you can then fit these data to get B = -V*dP/dV

Thank you for this very elegant way.

Dear jzwanzig,

I just calculated and i optimized the volume at different pressures and i have to fit the data but plz can you tell me how i can put the initial guess Bo, and when i fit how i can calculate B = -V*dP/dV .. dP/dV? and V is which volume here.



best regards

What you are trying to compute is B = -V dP/dV at the relaxed (zero pressure) volume. Therefore do a linear fit of P as a function of V, using your calculated results, where V is the unit cell volume (ucvol in the output). The slope of the fit is dP/dV. Then multiply by -V, using the V of the zero pressure run. The resulting B will be in whatever units you used in your fit for P.

Dear Prof.

Thank you very much i calculated Bulk modulus and the result was very good but when am trying to find pressure derivative of bulk modulus B' i got a bad results so can you tell me is there a good method for calculating B' , first i calculated the bulk modulus for every volume and then i fitted it with the pressure then i used ( B'=dB/dP )


best regards

To get B' by this method you would have to do A LOT of calculations, it would be easier to just use the Birch Murnaghan EOS. But why do you need to know B'? Typically what's of interest is just B.

Dear Prof.


Thank you very much and i am agree with you, but another question plz when i calculated Bulk modulus the ucvol in all data set was the same but unitary cell volume are different and i calculates Volume through optcell (i.e i multiplied each optcell three times by it self to get volume) is it true ? and here is my input file:

best regards


ndtset 5

strtarget1 3*3.4D-5 3*0.0
strtarget2 3*1.7D-5 3*0.0
strtarget3 3*0.0 3*0.0
strtarget4 3*-1.7D-5 3*0.0
strtarget5 3*-3.4D-5 3*0.0

ionmov 2
optcell 2
ntime 20
ecutsm 1.0
dilatmx 1.1

#Definition of the unit cell
acell 3*10.96
rprim 0.0 0.5 0.5 # FCC primitive vectors (to be scaled by acell)
0.5 0.0 0.5
0.5 0.5 0.0

ntypat 2
znucl 48 16
natom 2 # There is only one atom per cell
typat 1 2 # This atom is of type 1
xred 0.0 0.0 0.0
0.25 0.25 0.25 # This keyword indicate that the location of the atoms
ecut 20.99 # Maximal kinetic energy cut-off, in Hartree
nstep 10 # Maximal number of SCF cycles
toldfe 1.0d-6 # Will stop when, twice in a row, the difference
# between two consecutive evaluations of total energy
# differ by less than toldfe (in Hartree)

Each of your 5 datasets was run with ionmov 2 and optcell 2. Therefore for each one, the cell volume changed as it was optimized. Look through the output file (by hand or with a text search utility like grep) and find the last ucvol for each data set. This will the final, optimized value.

Dear Prof.

Thank you very much for your help, i just want to tell you that i calculated B for a binary compound and the result was very good on LDA, but i tried for GGA i got a big error between my result and experimental or other theoretical values, also for another compounds i got very bad result with LDA too, so do you prefer i do that by another method or???


kind regards