Showing posts with label Statistical Physics. Show all posts
Showing posts with label Statistical Physics. Show all posts

Saturday, December 18, 2010

Microcanonical Ensemble

It is curious that one of the fundamental ensembles of Statistical Physics, the Microcanonical, has a very simple but tricky expression for the entropy

$S = k_Bln\Omega(E;\Delta E)$

$k_B$is the Boltzmann constant and $\Omega(E;\Delta E)$is the number of states within the compatible energies$\Delta E$. In a formal manner one can write that as
$$\Omega(E;\Delta E) = \int_{E \le \mathcal H (\alpha) \le E +\Delta E}d\alpha$$where$d\alpha = d\vec q_1d\vec q_2\ldots d\vec q_Nd\vec p_1d\vec p_2\ldots d\vec p_N$.

This integral is not easy to compute even in the most simple cases because one should guess which is they way you count your states. But, searching for a more systematic calculation of the latter, the another day I founded in a book of Schwabl a easier way to handle this. Because we are "just" counting states, in the end one can use another "representation" for the microcanonical ensemble

$$\Omega(E;\Delta E) =\sum^{\infty}_{n_1 = 0}\ldots \sum^{\infty}_{n_N = 0}\delta(E-\mathcal H(q,p))=\sum^{\infty}_{n_1 = 0}\ldots \sum^{\infty}_{n_N = 0}\int \frac{dk}{2\pi}e^{ik(E-\mathcal H(q,p))}$$.

And from there is much easier to compute the solution. I will try to do some examples next time.