Kepler problem: flows generated by constants of motion

Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
1 reply · 1K views
Clausius
Messages
3
Reaction score
0
Consider the Hamiltonian of Kepler problem
[tex]H(\boldsymbol{r},\boldsymbol{p})= \frac{|\boldsymbol{p}^2|}{2\mu} +\frac{\alpha}{|\boldsymbol{r}|}, \qquad \mu>0>\alpha,[/tex]

where [tex]\boldsymbol{r}\in M=\mathbb{R}^3\setminus\{ 0 \}, \ (\boldsymbol{r},\boldsymbol{p})\in T^*M[/tex]
and
[tex]|\boldsymbol{r}|=\sqrt{r_1^2+r_2^2+r_3^2}.[/tex]

The quantities
[tex]\boldsymbol{m}=\boldsymbol{r}\times\boldsymbol{p}, \qquad \boldsymbol{W}=\boldsymbol{p}\times\boldsymbol{m}+ \mu\alpha\frac{\boldsymbol{r}}{|\boldsymbol{r}|}[/tex]

are constants of motion, as is well known.

My question is: how can I prove that the flows generated by the functions m_i and W_i, i=1,2,3 are canonical transformations?
Moreover, are such transformations point transformations?

A canonical transformation [tex]\Phi: T^*M\to T^*M[/tex]
is a point transformation if it is induced by a transformation [tex]\phi:M\to M,[/tex]

so that
[tex]\Phi(\boldsymbol{r},\boldsymbol{p})= (\phi(\boldsymbol{r}),\phi^{*-1}), \ \phi^*_i= \frac{\partial\phi_i}{\partial r_j}p_j.[/tex]
 
Physics news on Phys.org
In this case, the flows generated by m_i and W_i are given by\boldsymbol{r}(t)= \boldsymbol{r}_0 + t\boldsymbol{m} \qquad \boldsymbol{p}(t)= \boldsymbol{p}_0 + t\boldsymbol{W}.To prove that these flows generate a canonical transformation, we need to show that\frac{d}{dt}\Phi(\boldsymbol{r},\boldsymbol{p})= \{H,\Phi\}.Letting \Phi(\boldsymbol{r},\boldsymbol{p})=(\boldsymbol{r}(t),\boldsymbol{p}(t)), we have\frac{d}{dt}\Phi(\boldsymbol{r},\boldsymbol{p})= (\dot{\boldsymbol{r}}(t),\dot{\boldsymbol{p}}(t)) = (\boldsymbol{m},\boldsymbol{W}).On the other hand, the Poisson bracket is given by\{H,\Phi\}= \frac{\partial H}{\partial r_i}\frac{\partial \Phi_j}{\partial p_i} - \frac{\partial H}{\partial p_i}\frac{\partial \Phi_j}{\partial r_i}.For the Hamiltonian of the Kepler problem, we have\frac{\partial H}{\partial r_i}= -\frac{\alpha r_i}{|\boldsymbol{r}|^3}, \qquad \frac{\partial H}{\partial p_i}= \frac{p_i}{\mu}.Substituting these expressions into the Poisson bracket and using the fact that \Phi(\boldsymbol{r},\boldsymbol{p})=(\boldsymbol{r}(t),\boldsymbol{p}(t)), we obtain\{H,\Phi\}= -\frac{\alpha}{|\boldsymbol{r}|^3}\left(\boldsymbol{r}\cdot\boldsymbol{m}\