0
$\begingroup$

I'm having trouble understanding the examples Sethna uses in this section to illustrate the microcanonical ensemble.

First he talks about the probability density $\rho(Q)$ that $N$ ideal gas particles will be in position space configuration $Q\in R^{3N}$ inside a box of volume $V$. He says that since $\int \rho dQ=1$ and integrating over the positions gives a factor of $V$ for each of the $N$ particles, $\rho (Q)=1/V^N$.

I am confused by a few things here. First, what does it mean to integrate over position space, and where does the factor of $V$ come from for each position? This is confusing to me because $Q$ is a position vector and the volume is continuous, so why can we give any answer besides there are an infinite number of spatial configurations? Second, what is $R^{3N}$?

$\endgroup$

1 Answer 1

0
$\begingroup$

A spatial microscopic configuration of a set of $N$ non-interacting point-like particles in 3D is given once we have all the positions, i.e. the set of all the position vectors $Q=\{{\bf r}_1, \dots , {\bf r}_N \}$.

If the particles are confined in a finite volume, each cartesian component of the position ${\bf r}_i$ is an element of a subset of the real numbers $R$, each position vector ${\bf r}_i$ is inside a finite volume $V$, and the $3N$ position $Q$ is an element of a finite volume subset of $R^{3N}$, the $3N$-fold cartesian product of $R$, of volume $V^{N}$.

Notice that the continuity of the values of the positions has nothing to do with the spatial dimensionality. When we say that the continuous values of the position ${\bf r}_i$ belong to a 3D space, we mean that we need $3$ real numbers to uniquely assign a position in the space.

$\endgroup$
4
  • $\begingroup$ So the position space has 3N dimensions. But why is the volume of the subspace V^{3N}? Shouldn’t it be V^N since V is 3D? Also, while your point about the continuity is well taken, why does that make the probability distribution nonzero? In 3D space each particle could be anywhere! $\endgroup$ Commented Aug 1, 2022 at 16:00
  • $\begingroup$ It was my mistake. Of course, the volume in the $3N$ dimensional space is $V^N$. $\endgroup$ Commented Aug 1, 2022 at 16:30
  • $\begingroup$ @idiot-physicist For continuous probability distributions, the probability of finding a point in a volume $dV$ is $\rho({\bf r}) dV$. In the case of a uniform distribution it is $\frac{dV}{V} $. Therefore it is zero for a single point ($dV=0$) but different from zero for any finite subvolume. $\endgroup$ Commented Aug 1, 2022 at 16:36
  • $\begingroup$ I see. Thank you for clarifying. I was a little confused because our expression for the probability density doesn't depend on Q (but that is how it should be, because the probability distribution is constant). What would it look like for us to integrate this expression from Q to Q+dQ? After all, Q is a vector. $\endgroup$ Commented Aug 1, 2022 at 16:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.