Hi;
If i want to see hydrodynamics effects on polymer translocation,what kind of lammps commnds can help me?
I think fix srd command can help me but after 6 months working with lammps, i realize it could not help me.
Could combination between fix srd and fix rigid commands help me? what is your suggestions?
Thanks in advance
Samira
There is a FENE coarse-grain model in LAMMPS (bond_style fene) and
100s (probably 1000s) of papers have been written on using it to
model all kinds of polymer dynamics.
Steve
It's true SRD neglects solvent-solvent interactions, so it will get
the hydrodynamics wrong. Getting dynamics right is very difficult
even if you use explicit all-atom simulations with expensive
force-fields. Are you sure you really need to worry about the
dynamics?
You can try to introduce hydrodynamics in a qualitative way by
representing the solvent explicitly using computationally cheap
coarse-grained solvent models. There are numerous single-particle
coarse-grained models of water, for example. Here are a few I'm
familiar with:
1) Martini water
J. Phys. Chem. B 2007, 111, 7812-7824 (Martini water)
2) Larini & Voth model
J. Chem. Phys. 128, 244115 (2008); http://dx.doi.org/10.1063/1.2938857
3) mW water:
Molinero, V. and Moore, E.B., J. Phys. Chem. B 2009, 113, 4008-4016
(There are also simpler 2-body Lennard-Jones approximations of water.
I think the Larini paper mentions some of them.)
You have to choose the solvent model very carefully. This will not
only effect the dynamics of the system. Interactions between the
solute and solvent introducing hydrophobic-effect-like forces which
are the dominant forces responsible for behaviour of the polymers in
your system. (Presumably this is what you are more interested in
studying.)
--- hydrodynamics with implicit water ---
I could not tell, but it sounded like perhaps you wanted to represent
the solvent implicitly. In that case if you don't care about getting
the dynamics right, then just use langevin dynamics. For more
realistic dynamics, perhaps you want to use Langevin/Brownian dynamics
with some kind of non-isotropic diffusivity tensor (for example, Oseen
or perhaps Rotne-Prager? See references below.)
To my knowledge, this particular approach has not been implemented in
LAMMPS. It would be neat of some ambitious LAMMPS user wants to add
this feature. (Of course, don't expect efficient scaling for any
simulation with implicit solvent and a non-uniform spatial
distribution of particles. ...For this reason I would not expect the
LAMMPS developers to add this feature. It's not what LAMMPS was
designed for.) I think there are other programs which do have this
feature. (Perhaps TINKER or DLpoly?)
Your situation sounds frustrating. I'm curious, so let us know how
you decided to proceed. Good luck.
Andrew
I don't know if you use a water solvent or not. If yes, then need to
correct some of the references I sent earlier. Otherwise, feel free to
ignore this email.
There are numerous single-particle
coarse-grained models of water, for example.
Here are a few I'm familiar with:
2) Larini & Voth model
J. Chem. Phys. 128, 244115 (2008); http://dx.doi.org/10.1063/1.2938857
I'm sorry this was the wrong link. It should be:
J. Chem. Phys. 132, 164107 (2010)
http://dx.doi.org/10.1063/1.3394863
1) Martini water
J. Phys. Chem. B 2007, 111, 7812-7824 (Martini water)
I think there are also several variations of this force field. One of
which uses a polerizable force-field. Consult with google.
(There are also simpler 2-body Lennard-Jones approximations of water.
I think the Larini paper mentions some of them.)
3) mW water:
Molinero, V. and Moore, E.B., J. Phys. Chem. B 2009, 113, 4008-4016
There's an example of how to mix a polymer in mW water in one of the
examples at moltemplate.org. The parameters for the mW water model
are in the file "watMW.lt". However the settings in this example not
been optimised (the forces between the water and the carbon have not
been optimised and Lennard-Jones forces used for the polymer forced us
to use an artificially small time-step. This could be fixed, perhaps
using respa.)
There are probably numerous other water models I don't know about.
(Consult the google oracle.)
Good luck.
Andrew
I should have read your email more carefully before I replied. Some
of the things I said (below) were clearly not relevant to your
question. Sorry about that.