Compress a polymer melt to a target density

Dear all,

I have generated an initial configuration of confined polyethylene chains in a 200x200x200 nm^3 box by means of Packmol. Below it I have introduced a lattice of atoms and, hence, the simulation box is 200x200x400nm^3.

I am now interested in increasing the density of the polymer in this this initial configuration. For reaching my target, I need to compress the simulation box 4 times (50x50x100nm^3).

I have tried this by means of change_box or deform commands with remap option. However, I obtain a big increase in the temperature that leads to missing bond atoms and so on. I think that the problem is that when I change the dimensions and use the remap option, the distance between the atoms also get shorter. This way, elevated repulsion forces are obtained.

Am I right when I say that it is not right to compress a polymer melt by using change_box or deform commands?

My key question is: How should I compress the polymer melt to my target density?

Thanks in advance.

Dear all,

I have generated an initial configuration of confined polyethylene chains in
a 200x200x200 nm^3 box by means of Packmol. Below it I have introduced a
lattice of atoms and, hence, the simulation box is 200x200x400nm^3.

I am now interested in increasing the density of the polymer in this this
initial configuration. For reaching my target, I need to compress the
simulation box 4 times (50x50x100nm^3).

I have tried this by means of change_box or deform commands with remap
option. However, I obtain a big increase in the temperature that leads to
missing bond atoms and so on. I think that the problem is that when I change
the dimensions and use the remap option, the distance between the atoms also
get shorter. This way, elevated repulsion forces are obtained.

Am I right when I say that it is not right to compress a polymer melt by
using change_box or deform commands?

no, you are wrong. i think both commands are suitable, if used
correctly and the overall workflow adjusted for it.
it is quite obvious that any form of compression will increase the
potential energy, which will subsequently converted to kinetic energy.
there are multiple ways to deal with that, either by removing excess
potential energy or by dissipating & removing kinetic energy.

it don't understand your fixation with the "remap" flag and would like
to point out that it has a different syntax and semantics for
change_box and fix deform. also, if you suspect any issues with the
added atoms, then this is all self-imposed and can be easily avoided
through either anticipating the expected deformation or through
deferring the insertion until after the desired density is achieved.

My key question is: How should I compress the polymer melt to my target
density?

as has been indicated before, this kind of question is beyond the
scope of this mailing list. you should discuss how to design and plan
your simulations with your adviser and/or your co-workers.

axel.

Dear Axel,

first of all, thanks for your nice and very useful answer.

Just one thing, could you tell me which commands should I investigate for removing the excess potential energy or by dissipating & removing kinetic energy?

Thanks.

Dear Axel,

first of all, thanks for your nice and very useful answer.

Just one thing, could you tell me which commands should I investigate for
removing the excess potential energy or by dissipating & removing kinetic
energy?

you are confusing the mailing list (or me) with being your adviser
*AGAIN*. you have now fully exhausted my good will and i am not going
to answer this.
in fact, if i were your adviser, this question would get you into big
trouble, as you should know the answer and you are making the same
mistake again, even after being asked to not do this. both are things
that i find very irritating.

axel.

Dear Axel,

since other people in the list might know it and be willing to share it I understand that I am free to ask any LAMMPS or MD doubt in the list, although you might don’t want to answer.

You have spoken about the figure of the advisor, but the reality is that not everyone working with LAMMPS has an advisor, that’s why sometimes people ask here general MD doubts in the list. If you are the moderator of this list and want to restrict this list to LAMMPS doubts I respect it, even though I think it would be a mistake because (i) LAMMPS and MD are closely related and (ii) the list could serve both proposals perfectly. In addition, as far as I know, there is no general MD list, therefore, you expect the people without an advisor to go and ask his doubts to… a non existing place.

Best regards,

Jordi Pina

Dear Axel,

since other people in the list might know it and be willing to share it I
understand that I am free to ask any LAMMPS or MD doubt in the list,
although you might don’t want to answer.

You have spoken about the figure of the advisor, but the reality is that not
everyone working with LAMMPS has an advisor, that’s why sometimes people ask
here general MD doubts in the list. If you are the moderator of this list

please see: http://lammps.sandia.gov/guidelines.html

where it expresses that the mailing list is about LAMMPS and resolving
LAMMPS problems. we usually don't mind, when people throw in the odd
off-topic question, as there is often just a small bit of extra
explanation needed to get people to do productive work with LAMMPS.

and want to restrict this list to LAMMPS doubts I respect it, even though I
think it would be a mistake because (i) LAMMPS and MD are closely related
and (ii) the list could serve both proposals perfectly. In addition, as far
as I know, there is no general MD list, therefore, you expect the people

if you feel, there should be a forum where people who are beginners
with MD and needs to exchange their questions and solutions, then i
agree with you: this is something that is missing. but the fact, that
there are some people here that may be competent to answer your
questions (or those of people with similar problems) does not entitle
you to ask them freely. the mailing list guidelines were written to
express what we want to talk about here.

but nothing stops you to set up such a list or forum by yourself and
recruit others to build and maintain it. i am certain, there is a
large number of people with an interest in such a discussion forum
specifically aimed at people with basic MD problems. i for one would
be happily sending the many people your way that don't even bother
going through the mailing list and instead are bugging me personally
with their beginner MD problems and questions and clogging up my
inbox. i would also suggest to consider a different approach than a
mailing list. due to the repetitive nature of the questions, a scheme
like shown on the stackexchange servers (BTW: have you tried
physics.stackexchange.com ?) might be more suitable, and if you would
recruit MD beginners and intermediates (to discuss answers) not from
this mailing list, but also from other large (or not so large)
community software packages like Gromacs, Amber, NAMD, CHARMM, etc.
you might reach a critical mass to have a working community. from my
personal experience, such a forum would not only be helpful for
beginners, but also for people that have just left that group, as
there is nothing more educational than helping other people who
struggle with issues you have just overcome. on this list, this rarely
happens for fear of being corrected by more experienced people.

without an advisor to go and ask his doubts to… a non existing place.

i personally consider trying to learn and do MD without having proper
supervision and training a doomed project for all but a very small
number of people with the right mindset, the right training in
fundamental physics and approaching scientific problems. otherwise you
are destined to make and repeat many mistakes, that proper supervision
and training would prevent you from. i still remember how many
mistakes i made when i was starting with MD and how much my adviser
had to correct me, explain things to me, demonstrate where i
overlooked important details, and show me where my conclusions were
wrong. proper supervision and training is important, since many of the
issues are quite subtle and even if one gets beyond the basic hurdles
of constructing working input files, there is a high chance of errors
and inconsistencies. if you try to shoulder this on your own and
without any other support than a mailing list, you are ultimately
wasting your time and indirectly also the time of others that try to
help you. keep in mind, that tutoring/advising via e-mail is extremely
tedious and time consuming, especially for the person giving advice.

in summary, your many struggles with rather basic concepts are to me
personally a testament to the fact, that you should either find
yourself an adviser/training/collaborator or reconsider your choice of
research topic. this may sound harsh to you right now, but we've had
much more extreme cases in the past.

axel.