A query

Dear Developers and Users,

I am simulating a reservoir connected to multilayer graphene. In my model, I considered all the carbon atoms (graphene layers) as “rigid” with zero lennard-jones potential interaction and also I have immobilized the graphene layers through applying the “fix setforce” and "fix move linear " commands.

Now, I’d like to do this simulation with flexible graphene layers. To this aim, I have used tersoff potential for intra-interaction plus LJ potential interaction as inter-interaction between C-C of the graphene layers.

I was wondering on making use of “fix spring/self” command instead of above procedure.

My question is that does it possible to model the graphene layers as " flexible" with using “fix spring/self” command, nevertheless the LJ potential interaction between C-C remains zero so? (same as the rigid model)

Thanks in advance

If there are Tersoff interactions wtihin the sheets, and you don’t

set forces to 0, then the sheet will be flexible (i.e. move due

to internal forces). I don’t understand the LJ part of your Q.

Steve

Dear Steve,

Many thanks for your answer. I think I was making the model complicated by defining the extra LJ interactions between C-C of the graphene layers. Actually, your comment satisfies my desire based on having the flexible layers with zero LJ potential interactions between C-C. So, I will follow your suggestion through bellow commands that works fine.

carbon 1
Hydrogen 2
Oxygen 3

pair_style hybrid/overlay lj/cut/coul/long 10.0 10.0 dpd/tstat 300.0 300.0 10.0 34387 tersoff
#pair_coeff 1 1 lj/cut/coul/long 0.0000 0.00000
pair_coeff 1 2 lj/cut/coul/long 0.00000 0.00000
pair_coeff 1 3 lj/cut/coul/long 0.093627 3.1900
pair_coeff 2 2 lj/cut/coul/long 0.000000 0.0000
pair_coeff 2 3 lj/cut/coul/long 0.000000 0.0000
pair_coeff 3 3 lj/cut/coul/long 0.155300 3.1690
pair_coeff * * tersoff /home/SiC_1994.tersoff NULL NULL C

I just wondering to know that in this case (no LJ interaction between C-C or its better to say rigid C-C) , both Tersoff potentials and fix spring/self command can be used to model a flexible structure, Right?

Thanks in advance

How would you hold C-C rigid?

I don’t know anything about your model.

Steve

Dear Steve,

Many thanks for your reply. To hold the C-C rigid it is just required that we do not consider no VDW interaction between C-C. For instance, the LJ potential interaction between C-C defined as zero.

In my model I have a reservoir filled with water connected to multilayer graphene. In my model I only have defined LJ interactions between oxygen-carbon and oxygen-oxygen. In this case by defining the LJ interaction between C-C as zero I made the grapheme layers as rigid.

To model a flexible graphene, I have used Tersoff potential for intra-interaction between C-C. Because Tersoff potential does not include VDW interaction, I think I have to define an extra LJ interaction between C-C too, Right?

Dear Steve,

Many thanks for your reply. To hold the C-C rigid it is just required that
we do not consider no VDW interaction between C-C. For instance, the LJ
potential interaction between C-C defined as zero.

In my model I have a reservoir filled with water connected to multilayer
graphene. In my model I only have defined LJ interactions between
oxygen-carbon and oxygen-oxygen. In this case by defining the LJ
interaction between C-C as zero I made the grapheme layers as rigid.

​this makes no sense. if you have forces on the C atoms, they will move
during time integration. see newton's laws of motion, especially law #2

assemblies of atoms stay rigid in MD simulations, a) when they are excluded
from time integration and thus immobilized b) when they have no velocities
and no forces (at all). and c) when a time integration method is applied
that moves the assembly as a rigid object.

​axel.​

Dear Axel,

Many thanks for your comment. You are right.

I have not completely explained how to make the model rigid. Sorry about that. Actually, in conjunction with zero LJ interaction between C-C, I have also immobilized the graphene layers through applying "fix setforce 0 " and “fix move linear 0” commands.

Dear Axel,

Many thanks for your comment. You are right.

I have not completely explained how to make the model rigid. Sorry about
that. Actually, in conjunction with zero LJ interaction between C-C, I have
also immobilized the graphene layers through applying "fix setforce 0 " and
"fix move linear 0" commands.

​that doesn't make much sense either. fix move ignores forces, so fix
setforce 0 will have no effect. with no displacement fix move has no
impact.
​you essentially trying to do a mix of all three cases i described, which
is really pointless. pick one and be done. case a) is the simplest.

if you are struggling with such trivial issues, it is no surprise that
doing are more complex setup is challenging, too. there is no fixed rule
for what you must do if you want your carbon atoms to be mobile, yet remove
any total drift or displacement of the center of mass of your carbon
layers. it all depends what you want to learn from this and what level of
detail and accuracy for the carbon model is needed. possibly, tersoff may
be overkill (but convenient). this is more a question of planning your
research and not so much about how to correctly use LAMMPS.

axel.

Dear Steve and Axel,

Many thanks for your comments and explanations.

Thanks in advance