- #1
Petr Mugver
- 279
- 0
Let's take a system, for simplicity with only one degree of freedom, described by a certain lagrangian
[tex]L[x,\dot x][/tex]
I define the momentum
[tex]p=\frac{\partial L}{\partial\dot x}[/tex]
Now, if I make a change of coordinates
[tex]x\longmapsto y\qquad\qquad\qquad(1)[/tex]
I obtain a second lagrangian
[tex]M[y,\dot y]=L[x(y(t)),\partial_t x(y(t))][/tex]
and I can define a second momentum
[tex]q=\frac{\partial M}{\partial\dot y}[/tex]
My question is, if instead of the transformation (1) I want to consider the transformation of the momenta
[tex]p\longmapsto q\qquad\qquad\qquad(2)[/tex]
How can I find the corresponding transformation (1)? In other words, given that I know how to do (1)-->(2), how can I do (2)-->(1)?
[tex]L[x,\dot x][/tex]
I define the momentum
[tex]p=\frac{\partial L}{\partial\dot x}[/tex]
Now, if I make a change of coordinates
[tex]x\longmapsto y\qquad\qquad\qquad(1)[/tex]
I obtain a second lagrangian
[tex]M[y,\dot y]=L[x(y(t)),\partial_t x(y(t))][/tex]
and I can define a second momentum
[tex]q=\frac{\partial M}{\partial\dot y}[/tex]
My question is, if instead of the transformation (1) I want to consider the transformation of the momenta
[tex]p\longmapsto q\qquad\qquad\qquad(2)[/tex]
How can I find the corresponding transformation (1)? In other words, given that I know how to do (1)-->(2), how can I do (2)-->(1)?
Last edited: