Polymer flow stop due to solidification

Dear all,

As you might know from other mails, I am simulating how a polymer gets into a nanocavity until it freezes due to the heat transfer with the walls. It is part of my PhD.

I have successfully modeled the heat transfer and the polymer cools until the temperature of the walls.

What I want to model now is the polymer stopping its movement when it reaches the solidification temperature.

I am using the command

fix 8 polymer_group addforce 0 0 -3.598e-3

to model the force that pushes the polymer into the cavity.

I have thought the following approach:

If Temperature_Polymer > No_Flow_temperature

fix 8 polymer_group addforce 0 0 -3.598e-3

else

unfix 8

fix 9 polymer_group setforce NULL 0.0 0.0

endif

Do you consider this a good approach?

Would it be possible to model the real solidification of the polymer? I don’t really know how to solve it. I have simulated the polymer at temperatures far below its no flow temperature and it still moves due to the addforce command, which is physically not possible because the polymer would stay still.

Thanks.

Do you need some information more? I am really interested in knowing if my approach is correct and how could be improved.

Thanks in advance!

You’re asking if your model is “good”. That’s not
really a LAMMPS Q - you should search the literature
and see if others have implemented similar models.

Your use of addforce and setforce fixes seems fine,
in terms of LAMMPS syntax. Whether those commands
will “solidfy a polymer” in a meaningful way is a question
for you model, again really not a LAMMPS Q.

Steve

You are right Steve. I did the question in this list because people from here don’t only know about LAMMPS itself but also about MD.

Thanks!

Dear all,

As you might know from other mails, I am simulating how a polymer gets into
a nanocavity until it freezes due to the heat transfer with the walls. It is
part of my PhD.

I have successfully modeled the heat transfer and the polymer cools until
the temperature of the walls.

What I want to model now is the polymer stopping its movement when it
reaches the solidification temperature.

I am using the command

fix 8 polymer_group addforce 0 0 -3.598e-3

to model the force that pushes the polymer into the cavity.

I have thought the following approach:

If Temperature_Polymer > No_Flow_temperature

fix 8 polymer_group addforce 0 0 -3.598e-3

else

unfix 8

fix 9 polymer_group setforce NULL 0.0 0.0

endif

Do you consider this a good approach?

steve is always trying to be very polite on these issues. when he says
"check out the literature" that can be translated into "do not abuse
the mailing list to obtain scientific advising". people may be willing
to drop a hint here and there and particularly for tricky details, but
what you are asking for is the kind of advice (and training!) that you
should receive from your personal adviser. that takes time and effort
and expecting to receive this from a mailing list is asking for too
much. keep in mind that it is not about whether people here are
qualified to give than kind of advising, but rather that they prefer
to spend their time and effort on LAMMPS or their own students and
collaborators.

to not only send you away with a lecture on what is good mailing list
behavior, here are a few personal thoughts on your suggestions and
conclusions.

i don't think that your suggestion is a good approach, because what
you are suggesting is effectively creating an animation and not doing
science (i.e. you are preparing yourself for getting a job at Pixar
and not for the nobel prize). the key step when doing computer
simulations is not: "i don't see what i expect to see, so i have to
change my input until the simulation does what i expect". rather, if
you don't see what you expect, you have to investigate and find out
the reason for it. only then can you decide whether a change to the
input falls into the pixar or nobelprize category. the fact, that you
are making such a suggestion is a typical indication of lack of
competent and effective advising and training (we see this regularly
on this list, but it is hard to say who exactly is to blame, the
adviser or the advisee, probably both)

Would it be possible to model the real solidification of the polymer? I

it should be, but it may not work the way you seem to expect. you are
currently neglecting finite size effects and the fact that you are
talking about an activated process. for an as tiny volume as you are
simulating, you can get very far into the supercooled regime,
specifically for materials that don't form well defined crystals and
don't have a (macroscopically) sharp melting temperature.

also, consider the following: a macroscopically stiff material (take a
carbon nanotube for example) may still appear to be fairly "floppy" in
atomic simulations. if those "floppy" motions are fast enough, they
don't show macroscopically (same as you cannot hear sounds beyond a
certain frequency or see UV light). ... and this is just one example
how things can be very different on the nanoscale from how materials
act macroscopically. in fact, most of what is commonly referred to a
"nanotechnology" depends on these differences and taking advantage of
them.

don’t really know how to solve it. I have simulated the polymer at
temperatures far below its no flow temperature and it still moves due to the
addforce command, which is physically not possible because the polymer would
stay still.

these arguments don't convince at all. they seem motivated by an
idealized macroscopic picture and ignore microscopic behavior. a) as
already mentioned, you are disregarding finite size effects and
possibly being in a supercooled regime and b) even if your material
has become solid, it may still have some elasticity or plasticity and
respond to pressure. this will even apply to the macroscopic
situation. thus just turning the driving force off makes no sense at
all. if there is an external pressure, it will remain there. however,
that external pressure will not be applied to the entire polymer
material in the first place, but only to the outward facing parts.

with all that in mind, let me repeat, that this kind of inquiry into
fundamental concepts of nanoscale research is indeed off-topic for a
mailing list about a specific software package, as it is really a
question about how to conduct nanoscale science and not about how to
implement a specific model with the features available in that
software.

axel.

Dear Axel,

thanks for your time, I really appreciated your answer.