# How to get liquid metal at appropriate temperature?

Dear all
It’s known that the melting point of iron is about 1800K. But when I simulated a 101010 iron box from 2 to 3000K, I found that whatever the potential file I choose, the iron change into liquid state at around 2200K. It’s much higher than its melting point. Thus I wander how to get the liquid iron at about 1800K?

The horizontal ordinate of the picture below is temperature, and the vertical coordinate is volume.

Dear all
It’s known that the melting point of iron is about 1800K. But
when I simulated a 10*10*10 iron box from 2 to 3000K, I found that whatever
the potential file I choose, the iron change into liquid state at around
2200K. It’s much higher than its melting point. Thus I wander how to get
the liquid iron at about 1800K?​

​for starters, you could use a better method to determine the melting
point. try to get it from solid-liquid coexistence.​

axel.

Dear axel

Thanks for your reply! But what I want is liquid iron at 1800K, not solid-liquid coexistence iron. Did you mean that I can get the liquid iron by heating the solid-liquid coexistence iron slightly?

Best regards!

Zhuang

Dear axel

Thanks for your reply! But what I want is liquid iron at 1800K,
not solid-liquid coexistence iron. Did you mean that I can get the liquid
iron by heating the solid-liquid coexistence iron slightly?

​this is a *completely* different question. ​you definitely need to learn
to formulate questions better and find some better guidance about how to do
MD in general by studying textbooks on MD simulations more. what you are
asking about is a trivial process and has been followed by people
simulating liquids with MD for at least 30 years:

just take your original structure, then simulate it at a *much* higher
temperature than the melting point. this will *definitely* give you a
liquid. then cool it down to the desired temperature. mind you, that it is
generally not a good idea to do simulations at a point very close to an
expected phase transition.

determining the melting point, by starting at a temperature of 3000K and
then cooling it down. ...and if you are *really* curious, you should
compare the results from running this at different cooling and heating
rates. mind you, this is nothing new and has been well studied and is well
known.

axel.

Dear axel

Thanks for your kindly reply and sorry for my poor English. I will improve my way of expression. And now I know how to get the liquid iron. Thank you very much!

Best regards!

Zhuang

Dear axel

Thanks for your kindly reply and sorry for my poor English. I
will improve my way of expression. And now I know how to get the liquid
iron.

​let us get one thing straight here: this is *not* about poor command of
the english language. that is a cheap and in this case invalid excuse. your
initial description was pretty clear and well formulated in proper english.
what you *did* get wrong, was the understanding of what you want and how to
get it. if you just cared about getting liquid iron, this whole experiment
about determining the melting point was a waste of time and thus is sign of
limited understanding of the procedure. that, however, has nothing to do
with your command of the english language, but is a result of insufficient
studies of MD methodology and common practices. so if there is anything
that needs improving, it is that. your english is quite good compared to
what other non-native speakers sometimes use here.

axel.

Dear axel

Thanks for your reply. I have learnt some basic theory about molecular dynamics, and study the manual of LAMMPS. But when come to the simulation, I still have some troubles about the details. So it’s really kind of you for explaining me so detail about the question.

Best regards

Zhuang

Dear axel

molecular dynamics, and study the manual of LAMMPS. But when come to the
simulation, I still have some troubles about the details. So it's really
kind of you for explaining me so detail about the question.

​ everybody makes mistakes, all the time. but i consider it a fundamental
part of research and part of the learning process. people learn best, when
they make mistakes and then understand why this was a mistake. but that
requires to own up to them and not hide behind excuses. excuses rarely work
anyway, because experienced people have seen and heard pretty much all of
them.

so if you really want to improve your understanding of the situation, you
*really* should now also do the simulations that i suggested, where you run
the temperature gradient from high temperature to low and then do a couple
more of those runs with different heating rates, i.e. shorter or longer
simulations.
and if you then dig through the literature about studying phase transitions
with molecular dynamics, and round this up with reading up on the related
text book material on statistical thermodynamics, you will have learned an
awful lot from this little mistake and will do future simulations with much
more confidence and understanding.

axel.

Dear axel