Hi,
I want to use LJ potential in conjunction with REBO potential, in a way that LJ starts functioning when REBO stops functioning (at r=2 A). Is there any specific pair_style other than “pair_style table” or any specific command to do this?
Thanks in advance.
Hi,
I want to use LJ potential in conjunction with REBO potential, in a way
that LJ starts functioning when REBO stops functioning (at r=2 A). Is there
any specific pair_style other than "pair_style table" or any specific
command to do this?
pair_style airebo 3.0 1 0
please see: http://lammps.sandia.gov/doc/pair_airebo.html
Thanks Axel,
As far as I know AIREBO and REBO are slightly different. I think turning off the torsion term makes the potential, “AIREBO without torsion” not “REBO+LJ”. As you remember, there was a bug fix regarding REBO potential few months ago, and the paper by Favata et al (An analytical benchmark and a Mathematica program for MD codes Testing LAMMPS on the 2nd generation Brenner potential) addresses this. I have already used AIREBO potential for simulating CNTs, but, the geometries result from this potential have some discrepancies. This is one of the reasons that I want to use REBO+LJ instead of AIREBO. Surprisingly, most of the reported works for CNTs in the literature have gone this way, maybe coincidentally.
The other reason is that I want to define LJ interactions only between different layers of a multi-layer CNTs.
Obviously, these issues are not related to the LAMMPS. I just want to know how to define REBO+LJ in LAMMPS in a way that these two potentials don’t interfere at distances less than 2A, by a command other than pair_style table.
Now, I see a new fixed bug dated as “7 March 2017” regarding AIREBO potential which might be related to the above-mentioned issue. Unfortunately, I am using windows executable version of LAMMPS that dose not have this bug fix.
Hi,
I want to use LJ potential in conjunction with REBO potential, in a way that LJ starts functioning when REBO stops functioning (at r=2 A). Is there any specific pair_style other than “pair_style table” or any specific command to do this?
pair_style airebo 3.0 1 0
please see: http://lammps.sandia.gov/doc/pair_airebo.html
Thanks Axel,
As far as I know AIREBO and REBO are slightly different. I think turning
off the torsion term makes the potential, "AIREBO without torsion" not
"REBO+LJ". As you remember, there was a bug fix regarding REBO potential
few months ago, and the paper by Favata et al (An analytical benchmark and
a Mathematica program for MD codes Testing LAMMPS on the 2nd generation
Brenner potential) addresses this. I have already used AIREBO potential for
simulating CNTs, but, the geometries result from this potential have some
discrepancies. This is one of the reasons that I want to use REBO+LJ
instead of AIREBO. Surprisingly, most of the reported works for CNTs in the
literature have gone this way, maybe coincidentally.
The other reason is that I want to define LJ interactions only between
different layers of a multi-layer CNTs.
Obviously, these issues are not related to the LAMMPS. I just want to know
how to define REBO+LJ in LAMMPS in a way that these two potentials don't
interfere at distances less than 2A, by a command other than pair_style
table.
there is no such command.
Now, I see a new fixed bug dated as "7 March 2017" regarding AIREBO
potential which might be related to the above-mentioned issue.
Unfortunately, I am using windows executable version of LAMMPS that dose
not have this bug fix.
if you want the latest bugfixes (or pair styles that don't yet exist), you
will have to teach yourself to compile LAMMPS from source. that is
typically easier to do with Linux than windows, so the recommended approach
would be to install a linux virtual machine.
pre-compiled windows executables for newer releases will eventually happen,
but a) this is not a high-priority item, and b) this requires porting the
build infrastructure from the now obsolete LAMMPS-ICMS branch to upstream.
axel.
Thanks. So, I will try pair_style table, also, I have to work on compiling LAMMPS from the source.
Thanks Axel,
As far as I know AIREBO and REBO are slightly different. I think turning off the torsion term makes the potential, “AIREBO without torsion” not “REBO+LJ”. As you remember, there was a bug fix regarding REBO potential few months ago, and the paper by Favata et al (An analytical benchmark and a Mathematica program for MD codes Testing LAMMPS on the 2nd generation Brenner potential) addresses this. I have already used AIREBO potential for simulating CNTs, but, the geometries result from this potential have some discrepancies. This is one of the reasons that I want to use REBO+LJ instead of AIREBO. Surprisingly, most of the reported works for CNTs in the literature have gone this way, maybe coincidentally.
The other reason is that I want to define LJ interactions only between different layers of a multi-layer CNTs.
Obviously, these issues are not related to the LAMMPS. I just want to know how to define REBO+LJ in LAMMPS in a way that these two potentials don’t interfere at distances less than 2A, by a command other than pair_style table.
there is no such command.
Now, I see a new fixed bug dated as “7 March 2017” regarding AIREBO potential which might be related to the above-mentioned issue. Unfortunately, I am using windows executable version of LAMMPS that dose not have this bug fix.
if you want the latest bugfixes (or pair styles that don’t yet exist), you will have to teach yourself to compile LAMMPS from source. that is typically easier to do with Linux than windows, so the recommended approach would be to install a linux virtual machine.
pre-compiled windows executables for newer releases will eventually happen, but a) this is not a high-priority item, and b) this requires porting the build infrastructure from the now obsolete LAMMPS-ICMS branch to upstream.
axel.
The in-house REBO code that the Brenner, Harrison, Stuart, and Sinnott research groups use to publish papers using the REBO potentials include the short-range REBO term and longer-ranged LJ. The two interactions are combined via a spline, which you cannot realize with the LAMMPS implementation of pair_style rebo, even if you hybrid it together with LJ.
So the answer for REBO+LJ is: you can use “pair_style hybrid/overlay rebo lj/cut {cutoff}”or use pair_style airebo with no torsion – the two are going to give you different results but neither will be the same as the in-house REBO code.
The good news is: for carbon nanotubes the LJ interactions are zero so you can safely use “pair_style rebo” for CNTs. If you want to include hydrocarbons with CNTs that is another story.
Ray