4.1: One-Dimensional Wave Equation
( \newcommand{\kernel}{\mathrm{null}\,}\)
The one-dimensional wave equation is given by
\begin{equation}
\label{waveone}
\dfrac{1}{c^2}u_{tt}-u_{xx}=0,
\end{equation}
where u=u(x,t) is a scalar function of two variables and c is a positive constant. According to previous considerations, all C^2-solutions of the wave equation are
\begin{equation} \label{wavegen} u(x,t)=f(x+ct)+g(x-ct), \end{equation}
with arbitrary C^2-functions f and g
The Cauchy initial value problem for the wave equation is to find a C^2-solution of
\begin{eqnarray*} \dfrac{1}{c^2}u_{tt}-u_{xx}&=&0\\ u(x,0)&=&\alpha(x)\\ u_t(x,0)&=&\beta(x), \end{eqnarray*}
where \alpha,\ \beta\in C^2(-\infty,\infty) are given.
Theorem 4.1. There exists a unique C^2(\mathbb{R}^1\times\mathbb{R}^1)-solution of the Cauchy initial value problem, and this solution is given by d'Alembert's1 formula
\begin{equation} \label{waveform} u(x,t)=\dfrac{\alpha(x+ct)+\alpha(x-ct)}{2}+\dfrac{1}{2c}\int_{x-ct}^{x+ct}\ \beta(s)\ ds. \end{equation}
Proof. Assume there is a solution u(x,t) of the Cauchy initial value problem, then it follows from (\ref{wavegen}) that
\begin{eqnarray} \label{ini1} u(x,0)&=&f(x)+g(x)=\alpha(x)\\ \label{ini2} u_t(x,0)&=&cf'(x)-cg'(x)=\beta(x). \end{eqnarray}
From (\ref{ini1}) we obtain
f'(x)+g'(x)=\alpha'(x),
which implies, together with (\ref{ini2}), that
\begin{eqnarray*} \label{12a} f'(x)&=&\dfrac{\alpha'(x)+\beta(x)/c}{2}\\ \label{12b}g'(x)&=&\dfrac{\alpha'(x)-\beta(x)/c}{2}. \end{eqnarray*}
Then
\begin{eqnarray*} f(x)&=&\dfrac{\alpha(x)}{2}+\dfrac{1}{2c}\int_0^x\ \beta(s)\ ds +C_1\\ g(x)&=&\dfrac{\alpha(x)}{2}-\dfrac{1}{2c}\int_0^x\ \beta(s)\ ds +C_2. \end{eqnarray*}
The constants C_1, C_2 satisfy
C_1+C_2=f(x)+g(x)-\alpha(x)=0,
see (\ref{ini1}). Thus each C^2-solution of the Cauchy initial value problem is given by d'Alembert's formula. On the other hand, the function u(x,t) defined by the right hand side of (\ref{waveform}) is a solution of the initial value problem.
\Box
Corollaries. 1. The solution u(x,t) of the initial value problem depends on the values of \alpha at the endpoints of the interval [x-ct,x+ct] and on the values of \beta on this interval only, see Figure 4.1.1. The interval [x-ct,x+ct] is called {\it domain of dependence}.
Figure 4.1.1: Interval of dependence
2. Let P be a point on the x-axis. Then we ask which points (x,t) need values of \alpha or \beta at P in order to calculate u(x,t)? From the d'Alembert formula it follows that this domain is a cone, see Figure 4.2.1. This set is called domain of influence.
Figure 4.2.1: Domain of influence
1 d'Alembert, Jean Babtiste le Rond, 1717-1783
Contributors and Attributions
Integrated by Justin Marshall.