Dipole/cut

Dear Lammps users,

I am trying to use lj/cut/dipole/cut pair_style, but getting the following error

Pair dipole/cut requires atom attributes q, mu, torque (…/pair_lj_cut_dipole_cut.cpp:362)

I have used atom_style dipole, but still it gives the error.

LAMMPS version : 17Jan14

Atom_style sphere provides those properties. So you
need to use

atom_style hybrid sphere dipole

for example.

Steve

Thanks Very much Steve,

I was trying with atom_style hybrid full dipole…

Even that gives the same error… Wouldn’t we use combination of full and dipole…

I am trying to represent AA with CG model via. beads(full) and few beads(full as well as dipole)…
Can this be allowed in lammps using hybrid/overlay lj/sdk/coul/long 15 lj/cut/dipole/long 15

Thanks
Karteek

Thanks Very much Steve,

I was trying with atom_style hybrid full dipole..

Even that gives the same error.. Wouldn't we use combination of full and
dipole..

no. you also need "sphere" to provide the torque property. please keep
in mind that the dipole styles are written for the general case and
thus require all relevant atom properties to be present, even if they
are not used in your specific case.

Thanks Axel,

In read_data command, below sentences are mentioned

Note that if a non-standard value is defined by multiple sub-styles, it must appear mutliple times in the atom line. E.g. the atom line for atom_style hybrid dipole full would list “q” twice:

atom-ID atom-type x y z q mux muy myz molecule-ID q 

Can't we just use atom_style hybrid full dipole?

If not can we use three atom_styles,    atom_style hybrid full sphere dipole ?

Thanks
Karteek


Thanks Axel,

In read_data command, below sentences are mentioned

Note that if a non-standard value is defined by multiple sub-styles, it must
appear mutliple times in the atom line. E.g. the atom line for atom_style
hybrid dipole full would list "q" twice:

atom-ID atom-type x y z q mux muy myz molecule-ID q

Can't we just use atom_style hybrid full dipole?

NO! how many more times do you need people to tell you this?

If not can we use three atom_styles, atom_style hybrid full sphere dipole
?

YES! (or four or five)

Thanks Axel,

Its working but I am worried about the following thing.

1) pair_style hybrid/overlay lj/sdk/coul/long 15.0 lj/cut/dipole/cut 15.0

first step energy

TotEng = 15.6565 KinEng = 9.7760 Temp = 298.1500
PotEng = 5.8805 E_bond = 8.6908 E_angle = 4.6912
E_dihed = 0.0000 E_impro = 0.0000 E_vdwl = -8.8402
E_coul = 1.8710 E_long = -0.5323 Press = 1.2092

2) pair_style hybrid/overlay lj/sdk/coul/long 15.0 lj/long/dipole/long off off 15.0

first step energy

TotEng = 14.9916 KinEng = 9.7760 Temp = 298.1500
PotEng = 5.2156 E_bond = 8.6908 E_angle = 4.6912
E_dihed = 0.0000 E_impro = 0.0000 E_vdwl = -8.8402
E_coul = 1.2061 E_long = -0.5323 Press = 1.1944

Coulombic energy seems to be different in two situation. Is this beacuse

lj/cut/dipole/cut involves Elj + Eqq + E qp + Epp and Eqq being evaluated twice?

And also dipole components mux, muy, muz are not being changed in second situation, but changes in first situation?

Files are being attached

Thanks
Karteek

parm.BTA.CG (1.53 KB)

in.BTA.CG (718 Bytes)

data.BTA.CG (3.53 KB)

Thanks Axel,

Its working but I am worried about the following thing.

I have no time to debug everybody’s inputs. You have the source, you have the manual, you can do your “homework” by yourself.

It is straightforward to build small inputs where you can check everything step by step.

Thanks Steve and Axel in setting up the system with point dipole as sphere and remaining beads as point particles.

I have used the following attributes in setting up.

Thanks Steve and Axel in setting up the system with point dipole as sphere and remaining beads as point particles.

I have used the following attributes in setting up.


atom_style full dipole sphere

group NCO type 2
group rest type 1 3 4

fix 1 NCO langevin 298.15 298.15 100.0 23482341 omega yes # thermostat for angular velocities

fix 2 NCO nve/sphere update dipole # integrate spherical particles & update dipole orientation from torque evaluation

fix 3 rest nvt temp 298.15 298.15 100.0 # integrate point particles


Are the above sentences fine?

I don’t know. Do they produce the trajectory that you can expect from your model?

If i could predict the result of complex simulations from a small input fragment, I would have long ago started a business and made a huge amount of money by telling people the outcome of simulations without them having to spend large amounts of money on computers.

Thanks Axel,

I am able to produce the simulations what I expected and want to be sure of them.

I have attached the input files.

Thanks
Karteek

CG_BTA.tar.gz (19.7 KB)

Axel’s real point is that the purpose of the mail list
is not to validate people’s simulations or help them
write complex input scripts. That is the user’s job.

The mail list is really to answer specific questions
about how to do a task in LAMMPS or identify a problem
that LAMMPS seems to be having.

Steve