# Classical mechanics, Hamiltonian formalism, change of variables

1. Dec 29, 2012

### c299792458

1. The problem statement, all variables and given/known data
This problem has to do with a canonical transformation and Hamiltonian formalism. Below is my attempt at solving it, but I am not too sure about it. Please help!

Let $\theta$ be some parameter.

And
$$X_1=x_1\cos \theta-y_2\sin\theta\\ Y_1=y_1\cos \theta+x_2\sin\theta\\ X_2= x_2\cos \theta-y_1\sin\theta\\ Y_2=y_2\cos\theta+x_1\sin \theta$$

Suppose the original Hamiltonian is $$H(x,y)={1\over 2}(x_1^2+y_1^2+x_2^2+y_2^2)$$
I wish to find solve for the motion in terms of the new variables. I am also given the restriction that $X_2=Y_2=0$

____
***Attempt:***

I believe we have $$H(X,Y)={1\over 2}(X_1^2+Y_1^2+X_2^2+Y_2^2)$$

Now the normal Hamiltonian formalism would suggest that $$\dot X_i={\partial H\over \partial Y_i }\\ \dot Y_i=-{\partial H\over \partial X_i }$$

Which gives $$\ddot X_1=-X_1\\ \ddot Y_1=-Y_1$$
Therefore, $$X_1(t)=A(\theta)\cos t+B(\theta)\sin t\\ Y_2(t)=C(\theta)\cos t+D(\theta)\sin t$$***Is this form of solutions right?***

We see that the $${\partial X_1\over \partial \theta}=-Y_2=0\\ {\partial Y_1\over \partial \theta}=X_2=0$$
So $A,B,C,D$ must be constants.

Are these arguments right? And can I get a better solution, say by getting a more specific set of $A,B,C,D$, given only the given information?

Thank you.
1. The problem statement, all variables and given/known data

2. Relevant equations

3. The attempt at a solution