04.08.2013 Views

Statistical Mechanics - Physics at Oregon State University

Statistical Mechanics - Physics at Oregon State University

Statistical Mechanics - Physics at Oregon State University

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

1.3. STATES OF A SYSTEM. 7<br />

<br />

g(N, M) = 2 N<br />

M<br />

(1.8)<br />

Note th<strong>at</strong> if we can only measure the values of N and M for our Ising system,<br />

all the st<strong>at</strong>es counted in g(N,M) are indistinguishable! Hence this multiplicity<br />

function is also a measure of the degeneracy of the st<strong>at</strong>es.<br />

Thermodynamic limit again!<br />

Wh<strong>at</strong> happens to multiplicity functions when N becomes very large? It<br />

is in general a consequence of the law of large numbers th<strong>at</strong> many multiplicity<br />

functions can be approxim<strong>at</strong>ed by Gaussians in the thermodynamic<br />

limit. In our model Ising system this can be shown as follows. Define the<br />

rel<strong>at</strong>ive magnetiz<strong>at</strong>ion x by x = M<br />

N . The quantity x can take all values between<br />

−1 and +1, and when N is large x is almost a continuous variable. Since g(N,M)<br />

is maximal for M = 0, we want to find an expression for g(N,M) for x ≪ 1. If<br />

N is large, and x is small, both N↑ and N↓ are large too. Stirling’s formula now<br />

comes in handy:<br />

Leaving out terms of order 1<br />

N<br />

and with<br />

we find<br />

N! ≈ √ 2πNN N 1<br />

1<br />

−N+<br />

e 12N +O(<br />

N2 )<br />

and smaller one finds<br />

(1.9)<br />

log(N!) ≈ N log(N) − N + 1<br />

log(2πN) (1.10)<br />

2<br />

log(g(N, x)) = log(N!) − log(N↑!) − log(N↓!) (1.11)<br />

log(g(N, x) ≈ N log(N) − N↑ log(N↑) − N↓ log(N↓)<br />

+ 1<br />

2 (log(2πN) − log(2πN↑) − log(2πN ↓)) (1.12)<br />

log(g(N, x) ≈ (N↑ + N↓) log(N) − N↑ log(N↑) − N↓ log(N↓)<br />

− 1<br />

1<br />

log(2πN) +<br />

2 2 (2 log(N) − log(N↑) − log(N↓)) (1.13)<br />

log(g(N, x)) ≈ − 1<br />

2 log(2πN) − (N↑ + 1<br />

) log(N↑<br />

2 N ) − (N↓ + 1<br />

) log(N↓ ) (1.14)<br />

2 N<br />

Next we express N↑ and N↓ as a function of x and N via

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!