T*S could be negative？
Posted: Mon May 17, 2010 4:59 pm
Dear colleague,
I found the term T*S EENTRO was negative in OUTCAR when I tried to optimize the SIGMA value before a static total energy calculation. As the SIGMA value increased, the T*S changed to be positive. I was told to choose a SIGMA value when the corresponding T*S EENTRO value is less than 1meV.
In my physical sense, there is no way to appear a negative T*S term. So is it nonsense for vasp to calculate out a T*S value that is less than zero?
I really need your opinions.
Best wishes!
I found the term T*S EENTRO was negative in OUTCAR when I tried to optimize the SIGMA value before a static total energy calculation. As the SIGMA value increased, the T*S changed to be positive. I was told to choose a SIGMA value when the corresponding T*S EENTRO value is less than 1meV.
In my physical sense, there is no way to appear a negative T*S term. So is it nonsense for vasp to calculate out a T*S value that is less than zero?
I really need your opinions.
Best wishes!