0
$\begingroup$

Consider a Hamilton function $$H_0(x,p) = \frac{p^2}{2m}+ V(x).$$

The canonical equations then read $$\dot{x}(t) = p/m$$ and $$\dot{p}(t) = -V'(x)$$

Now imagine, we add an additional term $$\dot{x}(t) = p/m + g_0(x,p)\quad\text{and}\quad\dot{p}(t) = -V'(x) + g_1(x,p)$$

Then, we can see $g_1$ as an external force, but what is the interpretation of $g_0$? Is it meaningful to call it a friction?

$\endgroup$
0

2 Answers 2

1
$\begingroup$

$mg_0$ is apparently the difference between the kinetic momentum $m\dot{x}$ and the canonical/conjugate momentum $p$. However if the perturbation $g_0,g_1$ does not preserve the canonical structure, i.e. descend from an interaction Hamiltonian, then the notion of canonical/conjugate momentum is not well-defined.

$\endgroup$
0
$\begingroup$

You can't just arbitrarily add anything, Since the evolution equation for $(x,p)$ is derived from $H$.

In this case, $$\dot{x}=\partial_pH=\frac{p}{m}+g_0(x,p)$$ $$\dot{p}=-\partial_xH=-V'(x)+g_1(x,p)$$ Since you must have, $$dH(q,p)=\partial_qHdq+\partial_pHdp$$ $$\rightarrow -\partial_p\dot{p}=\partial_q\dot{q}$$$$\Rightarrow -\partial_p g_1(x,p)=\partial_qg_0(x,p)$$ If so you can write, $$H=\frac{p^2}{2m}+V(x)+f(x,p)$$ with $$g_0(x,p)=\partial_p f(x,p)\ \ \&\ \ \ g_1(x,p)=-\partial_xf(x,p)$$


One can regard the tern $f(x,p)$ as the term that couples particle's velocity and its position. A specific example of such a Hamiltonian can be, Charge particle in a uniform constant EM field. So if $\vec{B}=B\hat{k}$, Then you can choose $\vec{A}=-By\hat{i}$ so that $$H=\frac{(p_x+eBy)^2}{2m}+\frac{p_y^2}{2m}+\frac{p_z^2}{2m}+e\phi(\vec{r})$$ If you open up the first braket you get a term that will couple the velocity and position.

$\endgroup$

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.