Mathematics, Sciences, and Technologies

“Every secret of a writer’s soul, every experience of his life, every quality of his mind is written large in his works.”
– Virginia Woolf

ODEs 7-10 (PDEs): Wave Equation (Part 2)

,

The Vibrating Circular Membrane Problem

The wave equation can be easily generalized to a two-dimensional space. Particularly, we are often interested in the so-called vibrating circular membrane problem, e.g. a drumhead, so it is natural to expand the Laplacian operator in polar coordinates. The Laplace equation in this case is then expressed as

\begin{align}
\frac{\partial^2 z}{\partial t^2} -b^2\nabla^2 z = \frac{\partial^2 z}{\partial t^2} -b^2 (\frac{\partial^2 z}{\partial r^2} + \frac{1}{r}\frac{\partial z}{\partial r} + \frac{1}{r^2} \frac{\partial^2 z}{\partial \theta^2}) = 0 \tag{1}
\end{align}

where \( z \) is the vertical displacement of the membrane and \( b^2 = T/m \) is the ratio of the tension to mass per unit area. Again, we will need the method of Separation of Variables, and suppose that

\begin{align}
z(r, \theta, t) = R(r)\Theta(\theta)T(t) \tag{2}
\end{align}

Substituting (2) into (1), we have

\begin{align}
\begin{aligned} R(r)\Theta(\theta)T^{\prime\prime}(t) -b^2 (R^{\prime\prime}(r)\Theta(\theta)T(t) \\
+ \frac{1}{r}R'(r)\Theta(\theta)T(t) + \frac{1}{r^2} R(r)\Theta^{\prime\prime}(\theta)T(t)) \end{aligned} &= 0 \\
\frac{1}{b^2}\frac{T^{\prime\prime}(t)}{T(t)} -(\frac{R^{\prime\prime}(r)}{R(r)} + \frac{1}{r}\frac{R'(r)}{R(r)} + \frac{1}{r^2} \frac{\Theta^{\prime\prime}(\theta)}{\Theta(\theta)}) &= 0 \tag{3}
\end{align}

divided by \(R\Theta T\) as before. We now first separate the first term that depends on \(t\) only and the other terms that depend on \(r, \theta\):

\begin{align}
\frac{T^{\prime\prime}(t)}{b^2T(t)} = -k^2 = \frac{R^{\prime\prime}(r)}{R(r)} + \frac{1}{r}\frac{R'(r)}{R(r)} + \frac{1}{r^2} \frac{\Theta^{\prime\prime}(\theta)}{\Theta(\theta)} \tag{4}
\end{align}

The time part is simply solved via

\begin{align}
T^{\prime\prime} + k^2b^2 T = 0 \tag{5}
\end{align}

hence

\begin{align}
T = E\cos(kbt) + F\sin(kbt) \tag{6}
\end{align}

The second part in (4) can be further separated into \(r\) and \(\theta\) components respectively:

\begin{align}
r^2\frac{R^{\prime\prime}(r)}{R(r)} + r\frac{R'(r)}{R(r)} + k^2r^2 + \frac{\Theta^{\prime\prime}(\theta)}{\Theta(\theta)} &= 0 \tag{7}
\end{align}

\begin{align}
r^2\frac{R^{\prime\prime}(r)}{R(r)} + r\frac{R'(r)}{R(r)} + k^2r^2 = m^2 = -\frac{\Theta^{\prime\prime}(\theta)}{\Theta(\theta)} \tag{8}
\end{align}

The azimuthal part is also easily solved as

\begin{align}
\Theta^{\prime\prime} + m^2\Theta = 0 \tag{9}
\end{align}

\begin{align}
\Theta = C \cos(m\theta) + D \sin(m\theta) \tag{10}
\end{align}

where \( m = 0,1,2,\ldots \) by single-valuedness. Finally, the radial part is processed in a way very similar to what we have seen in this tutorial ((8)-(13)). We have

\begin{align}
r^2R^{\prime\prime} + rR’ + (k^2r^2-m^2)R = 0 \tag{11}
\end{align}

that is the same as the Bessel equation ((1) of this tutorial), again with \( r \rightarrow kr \) and \( \nu = m\). Therefore

\begin{align}
R = AJ_m(kr) + BY_m(kr) \tag{12}
\end{align}

The physical condition requires that the solution is bounded at \(r = 0\), so \(B=0\) in (12). Meanwhile, the boundary condition of a membrane is clearly a Dirichlet one, fixed to the rim: \( z(a,\theta,t) = 0, R(a) = 0 \) where \(a\) is the radius of the membrane. This means that

\begin{align}
R = J_m(\frac{k_{mn}}{a}r) \tag{13}
\end{align}

where \( k_{mn} = ka \) will be the \(n\)-th zero of the Bessel function of the first kind of order \(m\), \(J_m\), just as (17) of this tutorial. The full solution is then a superposition of

\begin{align}
z(r,\theta,t) = \sum_{m=0}^{\infty}\sum_{n=1}^{\infty} J_m(\frac{k_{mn}}{a}r) (C_{mn} \cos(m\theta) + D_{mn} \sin(m\theta))(E_{mn}\cos(\frac{k_{mn}b}{a}t) + F_{mn}\sin(\frac{k_{mn}b}{a}t)) \tag{14}
\end{align}

Now, if we are given the I.C.s that correspond to releasing the membrane from some position to rest, i.e.

\begin{align}
\left\{\begin{aligned}
z(r,\theta,0) &= f(r,\theta) \\
z_t(r,\theta,0) &= 0
\end{aligned}\right. \tag{15}
\end{align}

where \( f(r,\theta) \) can be a function of both \(r\) and \(\theta\), then we can match them by using the technique in the same tutorial, (19)-(22). First, notice that the second I.C. means that \( F = 0 \). Then, we can assume that the first I.C. can be expressed as Fourier modes with certain radial structures:

\begin{align}
z(r,\theta,0) = f(r,\theta) = \sum_{m=0}^{\infty} (C_m(r) \cos(m\theta) + D_m(r)\sin(m\theta)) \tag{16}
\end{align}

Comparing the Fourier modes in (16) with (14) at \( t = 0 \) shows that

\begin{align}
C_m(r) = \sum_{n=1}^{\infty}C_{mn}J_m(\frac{k_{mn}}{a}r) \tag{17}
\end{align}

where \( E_{mn} \) is absorbed. It is similar for \(D_m\) and \( D_{mn} \). \( C_{mn} \) is then given as a Bessel series following the orthonormality of the Bessel function ((30) of this tutorial):

\begin{align}
C_{mn} &= \frac{\int_{0}^{a} \frac{r}{a} C_m(r) J_m(k_{mn}\frac{r}{a}) d\frac{r}{a}}{(\int_{0}^{a} \frac{r}{a} [J_m(k_{mn}\frac{r}{a})]^2 d\frac{r}{a})} \\
&= \frac{2\int_{0}^{a} r C_m(r) J_m(k_{mn}\frac{r}{a}) dr}{a^2[J_{m+1}(k_{mn})]^2} \tag{18}
\end{align}

And \(D_{mn}\) will be computed in the exact same way. The full solution will therefore be

\begin{align}
z(r,\theta,t) &= \sum_{m=0}^{\infty}\sum_{n=1}^{\infty} J_m(\frac{k_{mn}}{a}r) ( \frac{2\int_{0}^{a} r C_m(r) J_m(k_{mn}\frac{r}{a}) dr}{a^2[J_{m+1}(k_{mn})]^2} \cos(m\theta) \\
&\quad + \frac{2\int_{0}^{a} r D_m(r) J_m(k_{mn}\frac{r}{a}) dr}{a^2[J_{m+1}(k_{mn})]^2} \sin(m\theta))\cos(\frac{k_{mn}b}{a}t) \tag{19}
\end{align}

For example, if the initial displacement is axisymmetric and has the form \( z(r,\theta,0) = 1 -\frac{r}{a} \), then \( m = 0\) and simply \(\Theta = C_{0n}, C_0(r) = 1 -\frac{r}{a} \). The evaluated expression will be

\begin{align}
C_{0n} &= \frac{2\int_{0}^{a} r (1-\frac{r}{a}) J_0(k_{0n}\frac{r}{a}) dr}{a^2[J_{1}(k_{0n})]^2} \tag{20}
\end{align}

and the complete solution will be

\begin{align}
z(r,\theta,t) &= \sum_{n=1}^{\infty} J_0(\frac{k_{0n}}{a}r) \frac{2\int_{0}^{a} r (1-\frac{r}{a}) J_0(k_{0n}\frac{r}{a}) dr}{a^2[J_{1}(k_{0n})]^2} \cos(\frac{k_{0n}b}{a}t) \tag{21}
\end{align}

Exercise

If the I.C.s are now

\begin{align}
\left\{\begin{aligned}
z(r,\theta,0) &= 0\\
z_t(r,\theta,0) &= g(r,\theta)
\end{aligned}\right. \tag{22}
\end{align}

so that the instantaneous displacement of the membrane is zero but there is some non-zero velocity, find the new solution form. Hence, what will be the full solution if the I.C.s are in the general form of

\begin{align}
\left\{\begin{aligned}
z(r,\theta,0) &= f(r,\theta) \\
z_t(r,\theta,0) &= g(r,\theta)
\end{aligned}\right. \tag{23}
\end{align}

Answer

With the new I.C.s, this time \( E = 0 \), and the velocity is given by

\begin{align}
z_t(r,\theta,t) = \sum_{m=0}^{\infty}\sum_{n=1}^{\infty} J_m(\frac{k_{mn}}{a}r) (C_{mn} \cos(m\theta) + D_{mn} \sin(m\theta)) \frac{k_{mn}b}{a} \cos(\frac{k_{mn}b}{a}t)
\end{align}

where we have absorbed the \(F_{mn}\) coefficient. To match the second I.C., we again assume a decomposition in the form of (16)

\begin{align}
z_t(r,\theta,0) &= g(r,\theta) = \sum_{m=0}^{\infty} (C_m(r) \cos(m\theta) + D_m(r)\sin(m\theta))
\end{align}

Comparing these two expressions then yields

\begin{align}
C_m(r) = \sum_{n=1}^{\infty}C_{mn}\frac{k_{mn}b}{a}J_m(\frac{k_{mn}}{a}r)
\end{align}

and by the same argument as before

\begin{align}
C_{mn} &= \frac{a}{k_{mn}b}\frac{2\int_{0}^{a} r C_m(r) J_m(k_{mn}\frac{r}{a}) dr}{a^2[J_{m+1}(k_{mn})]^2} \\
&= \frac{1}{k_{mn}ab}\frac{2\int_{0}^{a} r C_m(r) J_m(k_{mn}\frac{r}{a}) dr}{[J_{m+1}(k_{mn})]^2}
\end{align}

It is similar for \(D_m\) and \(D_{mn}\). Therefore, the corresponding solution will be

\begin{align}
z(r,\theta,t) &= \sum_{m=0}^{\infty}\sum_{n=1}^{\infty} J_m(\frac{k_{mn}}{a}r) (\frac{2\int_{0}^{a} r C_m(r) J_m(k_{mn}\frac{r}{a}) dr}{[J_{m+1}(k_{mn})]^2} \cos(m\theta) \\
&\quad + \frac{2\int_{0}^{a} r D_m(r) J_m(k_{mn}\frac{r}{a}) dr}{[J_{m+1}(k_{mn})]^2} \sin(m\theta)) \frac{1}{k_{mn}ab}\sin(\frac{k_{mn}b}{a}t)
\end{align}

The full solution for the general I.C.s will then be the sum of this and (19) due to linearity.

Leave a Reply

I’m Benjamin

Welcome to my Mathematical World! Here you can find posts and tutorials related to applied topics like Linear Algebra, Calculus, Differential Equations, as well as Programming. Feel free to leave comments and suggestions!

Let’s connect

Discover more from Benjamin's Maths World

Subscribe now to keep reading and get access to the full archive.

Continue reading