I'll let Paul answer this - if there is any inconsistency between
the code and the doc page for these units, then we should change it.
What should be the case, is that you can use fix ttm with any
LAMMPS units choice (e.g. metal, real, LJ), so long as
you get the inputs in the correct units as specified on the doc page.
Steve
Yachao,
Note that the table you reference gives the "electronic specific heat" per mole, but fix ttm in LAMMPS requires "electronic specific heat" per electron. So it appears to me that you've left out a factor of Avagadro's number. Also, it is unclear to me why that table has K^2 units in the denominator --- seems like it should be K. Perhaps Nicolas can comment. If I assume it really should be K, not K^2, from the table I calculate a value of 6.43e-9 eV/K for the electronic specific heat of copper. But it is possible that I've misinterpreted the data in that table.
Paul
thank you for your reply, I really appriciate your help!
Yachao