pair_style hybrid: which version do i need?

Hi,

im trying to simulate a water (tip3p)/Al(eam) system using the pair_style hybrid command. For the interaction Al-Water Im using a tabulated potential.

Depending on the lammps-version I use, i get different errors. Im not sure if an update will solve my problem and I first want to ask you because most updates create more problems than they solve…

the essential commands of ny input-script are as follows:

pair_style hybrid lj/cut 12.0 eam table linear 10001
pair_coeff 2 2 lj/cut 0.102 3.188 #O-O
pair_coeff 1 1*2 lj/cut 0.0 0.0 #OH and HH
bond_coeff 1 450 0.9572
angle_coeff 1 55 104.52
pair_coeff 3 3 eam Al_jnp.eam # Al-Al
pair_coeff 2 3 table pot.tmp MYPOT

LAMMPS (8 Aug 2014) reports “Illegal pair coeff command” in the last line
LAMMPS (27 Jul 2015) reports “illegal pair style command”

how can I include my tabulated potential?

Thanks for your attention,

regards,
frank.

Hi,

im trying to simulate a water (tip3p)/Al(eam) system using the pair_style
hybrid command. For the interaction Al-Water Im using a tabulated potential.
Depending on the lammps-version I use, i get different errors. Im not sure
if an update will solve my problem and I first want to ask you because most
updates create more problems than they solve...

​when updates to the LAMMPS code create (undocumented) problems, we would
*very* much like to hear about it.
making this kind of claim without examples to back it up is quite unfair to
the effort that a lot of people put into making LAMMPS work correctly and
consistently. specifically over the last two years a sizable effort has
been put into regression and continuous integration testing as well as
using various code analysis methods regularly to make sure that added
features do not break existing functionality. however, given the size of
the LAMMPS source code and the wide variety of features, these tests are
obviously incomplete. thus it is particularly important to get to know of
any kinds of issues that are not yet covered, so that more test cases can
be created to avoid this kind of breakage in the future.

the essential commands of ny input-script are as follows:

...
​​
pair_style hybrid lj/cut 12.0 eam table linear 10001
pair_coeff 2 2 lj/cut 0.102 3.188 #O-O
pair_coeff 1 1*2 lj/cut 0.0 0.0 #OH and HH
bond_coeff 1 450 0.9572
angle_coeff 1 55 104.52
pair_coeff 3 3 eam Al_jnp.eam # Al-Al
pair_coeff 2 3 table pot.tmp MYPOT

​this section of input works just fine with my LAMMPS executable which
follows the LAMMPS-ICMS git repo.​

...

LAMMPS (8 Aug 2014) reports "Illegal pair coeff command" in the last line

​for older versions of LAMMPS you have to check the bugfix history posted
at: ​LAMMPS Molecular Dynamics Simulator
as you may be encountering a bug that has been fixed or a change that
required an adjustment to the syntax that would break backward
compatibility.
steve takes great care to document these kind of changes whenever we
encounter them.

LAMMPS (27 Jul 2015) reports "illegal pair style command"

​"​
illegal pair style
​" almost always means that you are using an executable that is missing an
optional package and thus the corresponding pair style.

how can I include my tabulated potential?

​just do it right and it will work. :wink:

axel.​