Goal 1: Geometrically show that u(x, y, z) = \frac{1}{\sqrt{x^2 + y^2 + z^2}}is harmonic.

To show u(x, y) = \frac{1}{\sqrt{x^2 + y^2 + z^2}} is harmonic, it suffices to show that div.grad(\frac{1}{\sqrt{x^2 + y^2 + z^2}}) = 0. The key is to study the function f(x, y, z) = grad( \frac{1}{\sqrt{x^2 + y^2 + z^2}}). Notice that in polar coordinates, f(x, y, z) = f(r) (where r = \sqrt{x^2 + y^2 + z^2}) depends only on the radial distance. Hence, the magnitude of this function on any point of a circle of fixed radius is a constant.

Given arbitrary R > r > 0. Let the region between the circles of radius R and r be D. We now compute the flux of {grad}(\frac{1}{r}) across the \partial D. Since the gradient gives the direction of the greatest ascent, it follows that the gradient of \frac{1}{r} at each point points toward the origin. From the previous page, we know that U \propto \frac{1}{r}, where U is the potential. Since F = \nabla U, it follows that F \propto \nabla \frac{1}{r} = grad(\frac{1}{r}), where F is the corresponding force. Since F \propto \frac{1}{r^2}, it follows that grad (\frac{1}{r}) \propto \frac{1}{r^2}. Hence, we have:

\frac{\text{flux across sphere of radius R}}{\text{flux across sphere of radius r}} = \frac{R^2}{r^2}.

We also know that:

\frac{\text{surface area of circle of radius R}}{\text{surface area of circle of radius r}} = \frac{r^2}{R^2}.

Thus, the flux of f across the surface of the two spheres are equal, and so the total flux across the region is 0.

It follows by the divergence theorem that:

\iiint_{D} \nabla \cdot grad(\frac{1}{r}) dV = \iint_{\partial D} grad(\frac{1}{r}) \cdot \overrightarrow{n} dS = 0.

Since r, R had been arbitrary, it follows that \nabla \cdot grad(\frac{1}{r}) = 0. Hence, \frac{1}{r} is harmonic.

Important: The function f(x, y) = \frac{1}{sqrt(x^2 + y^2)} is not harmonic. What is difference?

The difference is that

\frac{\text{flux across circle of radius R}}{\text{flux across circle of radius r}} = \frac{r^2}{R^2}

but

\frac{\text{circumference of circle of radius R}}{\text{circumference of circle of radius r}} = \frac{R}{r}

and so the flux across the two boundaries are no longer the same. Hence, the total flux across the region is non-zero.

Goal 2: Suppose u(x, y) is a harmonic function on \text{D} \subset \mathbb{R}^2. Why do we require the condition that it be defined on a simply connected domain in order for it to be the real part of a holomorphic function f(z) = u(z) + iv(z)?

By the Cauchy-Riemann equations, it is not difficult to see that the real part of any holomorphic function is harmonic. (Note: proof of the converse will follow in a later post.)

We use the example of f(x, y) = log(\sqrt{x^2+y^2}) defined on the punctured plane mathbb{R}^2 \setminus {(0,0)}. Note that log(\sqrt{x^2 + y^2}) = Re(log(z)), but log(z) is multi-valued, hence it is not holomorphic.

Claim: S is a basis for the space of harmonic polynomials.

It suffices to show that the dimension of the space of homogeneous harmonic polynomial of fixed degree r \geq 1 is 2.

Given n \in \mathbb{N}. Given a homogeneous harmonic polynomial of degree n, suppose that it contains x^ry^{n-r}, where latex 0 \leq r \leq n$. Notice that it would then have to contain x^{r-2}y^{n-r+2}, x^{r-4}y^{n-r+4}, ... and x^{r+2}y^{n-r-2}, x^{r+4}y^{n-r-4}, ..., i.e. it will contain all terms of the form x^{r+2k} where k \in \mathbb{Z} and 0 \leq r+2k \leq n. And if the polynomial contains x^{r+1}y^{n-r-1} it follows that it will contain all terms of the form x^{r+2k-1} where k \in \mathbb{Z} and 0 \leq r+2k-1 \leq n. Hence, the space of homogeneous harmonic polynomial of fixed degree r \geq 1 is at least 2. This concludes the proof.

One response

  1. May 9th, Thoughts: A Comparison and a Trick – Menagerie of Mathematics Avatar

    […] is a solution. The method is very much the same as the one I used to prove the 3-dim. case in: https://menagerieofmathematics.wordpress.com/2025/05/07/may-6th-reading-thoughts-1-r-harmonic-functi…. I’m currently feeling much too lazy to repeat the […]

    Like

Leave a Reply

Is this your new site? Log in to activate admin features and dismiss this message
Log In