Hello,
I am using LAMMPS to study polymer network, in which bead-spring chain model is applied. Recently I met a problem that I am not sure if it is realistic to use LAMMPS pizza toolkit (chain tool) to build equilibrium polymer melt containing long bead-spring chains (250~500 beads). My previous experience dealt with short bead-spring FENE chains (10~30 beads), in which I used the chain tool to construct some chains (1000~1500), and minimize the system in LAMMPS and make it run for a few million steps before turning on polymerization using fix bond/create. In this way the polymer melt should have reached equilibrium state before reaction. However, if the chains are very long, this simple method may not be applicable as I have learned the relaxation time increases significantly with chain length, but I am not sure.
If the usage of the chain tool, combined with the method mentioned above, is not able to build well-equilibrated long-chain polymer melt, does any one know how to do this? I think some methods should have been proposed regarding this problem. Some discussion on the mailing list focus on building all-atom polymer melt, but I think my case may be easier than that.
Any help will be appreciated.
Thanks,
Shimiao
Hello,
I am using LAMMPS to study polymer network, in which bead-spring chain model
is applied. Recently I met a problem that I am not sure if it is realistic
to use LAMMPS pizza toolkit (chain tool) to build equilibrium polymer melt
containing long bead-spring chains (250~500 beads). My previous experience
dealt with short bead-spring FENE chains (10~30 beads), in which I used the
chain tool to construct some chains (1000~1500), and minimize the system in
LAMMPS and make it run for a few million steps before turning on
polymerization using fix bond/create. In this way the polymer melt should
have reached equilibrium state before reaction. However, if the chains are
very long, this simple method may not be applicable as I have learned the
relaxation time increases significantly with chain length, but I am not
sure.
If the usage of the chain tool, combined with the method mentioned above, is
not able to build well-equilibrated long-chain polymer melt, does any one
know how to do this? I think some methods should have been proposed
regarding this problem. Some discussion on the mailing list focus on
building all-atom polymer melt, but I think my case may be easier than that.
the first place to look for enlightenment on such a type of topic
should be the published literature, i.e. what have other people done
and how do they discuss the ease that they methods provide to reach
equilibrium and "quality" of that equilibrium.
second, i think you may want to check out using fix adapt in
combination with a soft potential to unoverlap and gently equilibrate
your system by allowing atoms to "move through each other" for a
while. also, fix bond/swap might be worth looking at as an alternate
approach. but no experiment can replace a thorough survey of the
literature.
axel.
You can also look at fix bond/swap and the references therein.
Steve
Hello,
I am using LAMMPS to study polymer network, in which bead-spring chain
model is applied. Recently I met a problem that I am not sure if it is
realistic to use LAMMPS pizza toolkit (chain tool) to build equilibrium
polymer melt containing long bead-spring chains (250~500 beads). My
previous experience dealt with short bead-spring FENE chains (10~30 beads),
in which I used the chain tool to construct some chains (1000~1500), and
minimize the system in LAMMPS and make it run for a few million steps
before turning on polymerization using fix bond/create. In this way the
polymer melt should have reached equilibrium state before reaction.
However, if the chains are very long, this simple method may not be
applicable as I have learned the relaxation time increases significantly
with chain length, but I am not sure.
Oh yes, be very sure that the longer the chain the longer it will take to
reach equilibrium (at least performing bare bones MD simulations). There is
a reason why people use MC-based simulations to drive the system to
equilibrium instead of going the MD route no matter how big is the cluster
where you intend to run your jobs.
If the usage of the chain tool, combined with the method mentioned above,
is not able to build well-equilibrated long-chain polymer melt, does any
one know how to do this? I think some methods should have been proposed
regarding this problem. Some discussion on the mailing list focus on
building all-atom polymer melt, but I think my case may be easier than that.
Ok I am a bit outdated when it comes to knowing how capable are the codes
out there that deal with the problem of building a reasonably good initial
guess for the equilibrated chain. Yet, I would suggest you look into
software and tools developed by the MC community that
deals with such type of problems. I am thinking about codes like MCCS
Tohwee for example. Additionally, browse the lammps archives as I remember
email exchange about a year ago addressing similar challenges. Cannot
really remember the names of the contributors..., just dig and dig and you
may find a treasure 
Carlos