[lammps-users] Elastic constant

Hi,all,

I want to calculate the elastic constant of crystal polyethylene at 0 K, it has orthorhombic unit cell. I ran the simulation according to the input script provided by Dr.Thompson in the following link.
http://sourceforge.net/mailarchive/message.php?msg_id=C7A7FF8C.1D6EC%25athomps%40sandia.gov

But I got very big elastic constants which are several orders bigger than the DFT values.I don’t know where is wrong.

And I also have a few questions concerning the script.
Why the mass is so small, 1.0e-20? And why for C11 and C66 there are no postive/negative deformation?

Can anyone help me with these?

Thank you

Chenchen

I'll let Aidan respond to this.

Steve

The mass is set to an obviously incorrect value because LAMMPS sometimes
requires that masses be specified, even though they are not used by the CG
minimizer.

Some of the postive/negative deformations are absent because for a 2 atom
HCP cell, they exceed LAMMPS limits for xy, xz, or yz.

I can't tell you why your numbers are so far off. Since your potential was
probably not fit to 0 K elastic constants, I would not expect it to be very
accurate for these. Also, you should check that you have specified the
potential correctly, perhaps by calculate properties that have been
previously published for this potential.

Aidan

Hi Aidan,

Thank you for your reply. I use the reax force filed and I have checked my potential,it works well when predicting the lattice constants at 0K. I have add the negative/postive deformation for C11 and C66.But still I can’t get a correct answer.

The attached is my input file.Could there be any other reason that cause this problem?

Thank you!

Chenchen

in.polyethylene (20.1 KB)

Have you checked what units the lattice consants from the script are in? It
looks like eV/A^3.

Hi Chenchen,

In your input script, you used reax force field, in which the pressure
unit is atmosphere. So the pfac1 in your script should correspondingly
change.

BTW, can you also show me your data.pe file? Your simulation box is a
triclinic or orthogonal? It looks the Adian's script cannot be used
for the orthogonal box.

Dongshan

Hi,Dongshan,

Than you for your reply. I have attached my data.pe file.
I have verified the units and made some changes to the correct unit,but the result is still not right.I think if setting the tilt factor to 0,the script can be used for orthogonal box.

Chenchen

data.pe (650 Bytes)