If I'm given a potential, how do I determine its frequency?

675 Views Asked by At

Let's say I'm given a potential $U(x)=\frac{\epsilon}{2}\left ((\frac{x}{a})^2 +(\frac{a}{x})^2\right )$ and that I need to find the frequency associated with it. I'm given that the particle of mass $m$ moves in one dimension. How would I do this?

My attempt:

Since this is just a simple oscillating motion problem, I took a look at Taylor's Classical Mechanics and reviewed its section on such motion. I saw that he did

$$F=ma=-kx$$ $$a=-\frac{k}{m}x=-w^2x \quad ; \quad w=\sqrt{k/m}$$

He called $w$ the "angular frequency" with which the cart (in his example) moves. Is this the frequency I'm looking for? If it is, how do I find it, given the potential above?

2

There are 2 best solutions below

11
On BEST ANSWER

The "ordinary" frequency is $\omega/(2\pi)$. That is, the frequency of oscillation of a harmonic oscillator with mass $m$ and stiffness constant $k$ is $\frac{k^{1/2}}{2 \pi m^{1/2}}$.

To understand the frequency of small-amplitude oscillations about the minimum of your given potential, first find its minimum. Take the second derivative of the potential at this minimum. This will correspond to $k$ in the harmonic oscillator. Nonlinear effects of course mean that the real frequency is not exactly the one given by this linearization; but these effects become less and less significant as you send the amplitude of the oscillation to zero.

It will turn out that this frequency is (up to a multiplicative dependence on a dimensionless function of the amplitude) the only way to combine $\epsilon,a,m$ to get the units of frequency. Hence how you could also do this using dimensional analysis.

But that just boils down to figuring out the units of $\epsilon,a,m$, which I guess are energy, length, and mass, and choosing powers of them that give you units of frequency. That's routine: energy is $M L^2 T^{-2}$, length is of course $L$, mass is of course $M$. So you want $\epsilon^{k_1} a^{k_2} m^{k_3}$ where $k_1+k_3=0,2k_1+k_2=0,-2k_1=-1$.

1
On

Well, the question is not much well posed. I suppose that you are required to find the frequency of small oscillations around the minimum of $U(x)$, because in other positions the particle would make ample oscillations subject to a force that is not linear.
Once you found the minimum, then around that you should be able to develop $F=dU/dx=kx+O(x^2)$. And once you have found the value of $k$ then everything goes as you know.