Physics Stack Exchange is a question and answer site for active researchers, academics and students of physics. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Problem statement: The Hamiltonian of a system is given by the formula:

\begin{equation*} H = \frac{p_r^2}{2m} + \frac{p_\theta^2}{2mr^2} + V(r,\theta). \end{equation*}

Under what condition is $f=p_\theta^2$ an integral of motion?

Attempted solution:

In order for $f$ to be an integral of motion, according to Poisson theorem, is that:

\begin{equation*} [f, H] =0, \end{equation*}

where $[,]$ is the Poisson bracket.


\begin{align*} \left[\frac{p_r^2}{2m} + \frac{p_\theta^2}{2mr^2} + V(r,\theta), p_\theta^2\right] &= 0 \Leftrightarrow \\ \left[\frac{p_r^2}{2m}, p_\theta^2\right] + \left[\frac{p_\theta^2}{2mr^2}, p_\theta^2\right] + \left[V(r,\theta), p_\theta^2\right] &= 0. \end{align*}

But how do I proceed from here on?

EDIT1: Regarding the first term:

\begin{align*} \left[\frac{p_r^2}{2m},p_\theta^2\right]&=\frac{1}{2m}\left[p_r^2,p_\theta^2\right]=\frac{1}{2m}[p_r p_r, p_\theta^2]=\frac{1}{2m}\left(p_r[p_r,p_\theta^2]+p_r[p_r,p_\theta^2]\right)\\ &=\frac{1}{m}[p_r,p_\theta^2]= \frac{1}{m}\left(p_\theta[p_r, p_\theta]+p_\theta[p_r,p_\theta]\right) = \frac{2p_\theta[p_r,p_\theta]}{m} \end{align*}

So it boils down to what is $[p_r, p_\theta]$ equal to ? I'm not sure whether it can be answered directly without resorting to the definition of Poisson bracket.

The coordinates in our problem are $q_i = \{r, \theta\}, p_i = \{p_r, p_\theta\}$. Therefore:

\begin{align*} [p_r, p_\theta] &= \sum_{i=1}^2 \left(\frac{\partial p_r}{\partial q_i}\frac{\partial p_\theta}{\partial p_i} - \frac{\partial p_r}{\partial p_i}\frac{\partial p_\theta}{\partial q_i}\right)\\ &= \left(\frac{\partial p_r}{\partial r}\frac{\partial p_\theta}{\partial p_r} - \frac{\partial p_r}{\partial p_r}\frac{\partial p_\theta}{\partial r}\right) + \left(\frac{\partial p_r}{\partial \theta}\frac{\partial p_\theta}{\partial p_\theta} - \frac{\partial p_r}{\partial p_\theta}\frac{\partial p_\theta}{\partial \theta}\right)\\ \end{align*}

And now I'm stuck again.

The way I understand it $p_r, p_\theta$ are two components of momentum along $r, \theta$ respectively. Does that imply that $p_r$ does not depend on $\theta$ and that it only depends on $r$ ? If the answer is yes, then the result for the first term is zero.

EDIT 2: Based on the answers in Hamiltonian formalism all the canonical variables are taken independent to each other, therefore $[p_r, p_\theta]=0$. Similarly it may be shown that the 2nd term is also zero.

share|cite|improve this question
up vote 2 down vote accepted

I think it is pretty obviously in your last line. So what are the results of the commutator $[p_r,p_\theta]$ and $[p_\theta,p_\theta]$? Note that $p_\theta$ is not commute with $\theta$. Then you should get the answer.

share|cite|improve this answer
And "What is the identity to get from $[A^2,B^2]$ to $[A,B]$?" :) – Alex Nelson Feb 6 '14 at 21:46
:) Heh, I meant it as another question the OP should ask themselves while pursuing your outlined approach...(although, square-brackets means Poisson brackets here, not commutators, so your relationships need a little revision) – Alex Nelson Feb 6 '14 at 21:55
Regarding $[p_r, p_\theta]$ please see my latest question. Regarding $[p_\theta, p_\theta]$ how did you get rid of $r^{-2}$ ? Shouldn't it be $[p_\theta/r^2, p_\theta]$ ? – Zet Feb 7 '14 at 2:17
@Zet In Hamiltonian formulation, $p_r$ and $p_\theta$ are treated as independent variables, so the derivative with repest to others canonical variable would be zreo. – hwlau Feb 7 '14 at 3:15

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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