Approximation to simple harmonic motion.

[Angelos]

[SOLVED] Approximation to simple harmonic motion.

Homework Statement

A small mass m, which carries a charge q, is constrained to move vertically inside a narrow, frictionless cylinder. At the bottom of the cylinder is a point mass of charge Q having the same sign as q. Show that if the mass m is displaced by a small amount from its equilibrium position and released, it will exhibit simple harmonic motion with angular frequency
\omega = \sqrt{\frac{2g}{y_0}}
where y_0 is the height, when the mass is in equilibrium.

Homework Equations

Simple harmonic motion:
F=-kx

Coulomb's Law(one dimension, same sign of charges):
F= \frac{kq_1q_2}{y^2}

Newton's Second Law:
F=ma

The Attempt at a Solution

Force acting on the mass m at a height y is:
F=\frac{kqQ}{y^2} - mg

I found the potential energy by:
U=-\int F dy= \frac{kqQ}{y} + mgy

I found the equilibrium point:
mg = \frac{kqQ}{y_0^2}
y_0 = \sqrt{\frac{kqQ}{mg}}

Near (stable) equilibrium I can approximate the potential energy by some parabola with equation:
U_n = U(y_0) + B(y - y_0)^2 = 2\sqrt{kqQmg} + B(y - \sqrt{\frac{kQq}{mg}})^2
where B is some (positive) constant.

Than I can find the approximate force by taking the negative derivative of U_n. That is
F = -2B(y - y_0)

That is the same as simple harmonic motion and so we can say that
\omega = \sqrt{\frac{2B}{m}}

Well my problem is that I have no idea how to find the B or how to show that the B = \frac{mg}{y_0}. The thing is I don't know how to mathematically describe that I want the parabola of the approximate potential energy function to be the most similar to the original function near equilibrium.

I will really appreciate any help from you.
Thanks.

 

[CompuChip]

You said
U(y) = \frac{kqQ}{y} + mgy
(which I didn't check, but I assume it is correct).

I would consider making a Taylor expansion around the equilibrium point, i.e.
U(y) = U(y_0) + U'(y_0)(y - y_0) + \frac12 U''(y_0) (y - y_0)^2 + \mathcal{O}((y - y_0)^3).
Since it's an equilibrium point, the linear term will vanish. Then approximate the potential by throwing away the (y - y0)^3 terms and you will have a parabola, from which you can easily read off the coefficient.

I quickly tried this and it will give you the B you stated, so it should work. But there might be another way.

 

[Angelos]

Hi...I had never heard about Taylor's polynomials, but I looked it up, when I read your reply and it's a great stuff. Exactly what I needed and how you said - it works. So thanks a lot. I really appreciate it.