Actually, your answer to my first question is quite explicit and based on it
I have got a further question, that is, if I want to perform a simulation
about the melt of a LJ cluster in NVE ensemble and will the temperature
set in the input file should equal aproximately twice the value of
the expected temperature?
-> No you can not generalize that. The final temperature can be twice or half of
the temperature that you input. If you want to run your simulation at particular
temperature use 'fix nvt'. With 'fix nve' you can not control the temperature
because fixing N,V and E automatically fixes the equilibrium temperature of the
I have the book you recommended in hand. An example of argon in the book
seems different in value of T (which is about 120K) with your caculated
result. I wonder what is the epsilon? Or in other word what does the epsilon
refer to? The total energy of a Ar atom? I am quite confused.
-> epsilon refers to the well depth of the LJ potential, it is just a parameter
that is needed to define LJ potential. It is not the total energy. Do a search
on google about lennard-jones potential and read about epsilon. With the reduced
units example I was just trying to illustrate the idea, the value for epsilon
that I wrote may not be accurate. Do a literature search and re-do the
calculation, it should match up with the result in the book.
Wuhan Univ., China
May 23, 2007
Chemical Engineering Dept.
Iowa State University