relaxation with fix nvt

Dear Lammps users

I have a cluster with definite size and shape of copper,i want to minimize the system and find the potential energy of system.
in temperature 0 kelvin i used minimize command,but in other temperature for example 300 kelvin,what command should i use?

i used this command but the results are nod in good agreement.

fix 1 all nvt temp 300 300 0.02

fix 6 all box/relax iso 0 vmax 0.1

run 10000

Does anybody know how to have a cluster in desire temperature and minimum the structure?

Thank you very much indeed
Malihe

Dear Lammps users

I have a cluster with definite size and shape of copper,i want to minimize
the system and find the potential energy of system.
in temperature 0 kelvin i used minimize command,but in other temperature for
example 300 kelvin,what command should i use?

i used this command but the results are nod in good agreement.

fix 1 all nvt temp 300 300 0.02

fix 6 all box/relax iso 0 vmax 0.1

run 10000

Does anybody know how to have a cluster in desire temperature and minimum
the structure?

this doesn't exist. please grab a text book on statistical mechanics
and learn why.

only at 0K you have a minimum structure, at finite temperature you
have ensembles of structures with equivalent total energy. the best
what you can do is to determine/approximate that distribution, but
structures at the same state (i.e. total energy) may be very different
(and have very different probability).

as a minimal example, look at a harmonic oscillator. at a given total
energy, it will have a certain average(!) kinetic energy, but there
are many states of different pairs of kinetic and potential energy
behind that, and you can derive the probability of each state and
their distribution. thus at finite temperature, you cannot have a
minimum structure but you can determine a most likely structure or an
average structure or some estimate of the distribution of structures
and so on.

apart from that, using fix nvt doesn't make sense, either. what kind
of heat batch is your cluster coupled to and how (microscopically).
thermostat algorithms are generally derived for bulk systems to
approximate the statistical mechanical ensemble of an infinitely large
bulk system that is coupled to a system of finite temperature.

perhaps, at this point it would be a good idea to have a talk with
your adviser and discuss these issues.

axel.