# fix nvt and rdf

I have 2 questions. I am actually modelling some nanoparticles in a polymer melt.

Question1 :Firstly, I want to melt the polymer from by changing the temperature 5.0lj units to 2.0 lj units and also equilibrate it at 2.0 lj units temperature.

So I am writing

fix 1 all nvt temp 5.0 2.0 100.0(to melt)

then i am restarting it and writing,
fix 2 all nvt temp 2.0 2.0 2.0(to equilibrate).

While running the simulation why is it showing Temp 0 at the 0th timestep or anything else other than 5.0 lj units.

Question 2: Rdf. I read from the documentation and i understood that lammps caculate the rdf till the cutoff radius.But i need it till half the box length. So what shall I do? Is VMD of any help in this?

I have 2 questions. I am actually modelling some nanoparticles in a polymer
melt.
Question1 :Firstly, I want to melt the polymer from by changing the
temperature 5.0lj units to 2.0 lj units and also equilibrate it at 2.0 lj
units temperature.
So I am writing
fix 1 all nvt temp 5.0 2.0 100.0(to melt)
then i am restarting it and writing,
fix 2 all nvt temp 2.0 2.0 2.0(to equilibrate).
While running the simulation why is it showing Temp 0 at the 0th timestep or
anything else other than 5.0 lj units.

please inform yourself how a nose-hoover thermostat works and how
quickly it can exchange energy with its heat reservoir. if you have no
initial temperature assigned, of course your system will show no
temperature. also, i don't quite understand your choices of
parameters, neither the order of the temperatures, nor the magnitude
of the time constant.

Question 2: Rdf. I read from the documentation and i understood that lammps
caculate the rdf till the cutoff radius.But i need it till half the box
length. So what shall I do? Is VMD of any help in this?

the g(r) calculation in VMD is brute force O(N**2) and thus can go
over all pairs. its normalization is written in such a way, that you
can even go *beyond* half the box up to about 0.8*the box. LAMMPS can
handle that, too, if you insert a potential with a large enough
cutoff, but that will slow down the calculation massively, so it
really only makes sense in combination with the "rerun" command as a
post-processing step.

axel.

Actually sir, I want to melt the polymer by changing temperature from 5 epsilon/kb to 2 epsilon/kb in the first simulation and then I want to equilibrate the system at a fixed temperature of 2 epsilon/k. I want to do all these stuffs applying a nose hover thermostat. So what shall I need to do?

Assuming the Tstart to be 5 and Tstop to be 2 i wrote
fix 1 all nvt temp 5.0 2.0 100.0(to melt)

Actually sir, I want to melt the polymer by changing temperature from 5
epsilon/kb to 2 epsilon/kb in the first simulation and then I want to
equilibrate the system at a fixed temperature of 2 epsilon/k. I want to do
all these stuffs applying a nose hover thermostat. So what shall I need to
do?

as i already stated. you first need to *get a better understanding*. i
already gave you an explanation and recommendation. if you don't
understand it, you have to figure it out. i don't have the time to do
*your* thinking for you. furthermore, what is the point of simply
questions that *i* raised.

also, i don't quite understand how you can melt something by
*lowering* its temperature. in my experience it would work the other
way around.

Sir, what I want to do is:

i) to melt the polymer at high temperature of 5 e/k

ii) to do a slow temp. tunneling from 5 to 2 e/k

iii) then i want to equilibrate system at 2 e/k

all using nvt. So i dont understand from the documentation how I will do the second step. It will be great if you can giove me a little clue.

Sir, what I want to do is:
i) to melt the polymer at high temperature of 5 e/k
ii) to do a slow temp. tunneling from 5 to 2 e/k
iii) then i want to equilibrate system at 2 e/k
*all using nvt*. So i dont understand from the documentation how I will do
the second step. It will be great if you can giove me a little clue.

we are moving in circles here. please *do* pay attention to the advice
you are given and do *not* simply repeat your question in variations.
you still have not explained *why* you made the choices that you did
and this description of steps conflicts with your earlier statement
(it actually makes more sense). that second step was actually given
more-or-less correctly, except for the unexplainedly large time
constant.

again, i have to remind you that a manual is not a tutorial, i.e. if
you don't understand it, you have to go and teach yourself or have
somebody teach you what you don't understand. a/this mailing list is
not the proper place for it. read the proper literature, discuss with
how to use fix nvt and there is nothing that i have to add to that.
thousands of other LAMMPS users have managed to understand it, and so
will you if you stop repeating the questions and start thinking and
exploring.

personally, i disagree with your choice of fix nvt for the melting and
cooling down. and even for the first part of equilibration, there is
no requirement for having a thermostat that correctly samples the NVT
ensemble, but instead i would prefer an arrangement that guarantees
equipartitioning (e.g. fix temp/csvr or fix langevin in combination
with fix nve), which in turn should help to get to equilibrium more
effectively.

axel.

Try setting the initial temp of the system
to 5.0 using the velocity command. Then
the thermostat will do better at keeping at
there (initially).

Steve