2

My approach is the following:

we take Fourier transform with respect to $x$, where $k$ is the variable the resulting fourier transform is in.

$\hat{u}_{tt} + k^2 \hat{u} = 0$

Solving this gives me (I think is where I am wrong)

$A(x)e^{ikt}+B(x)e^{-ikt}$

Using the shift rule, I can recover the $F(x-t)$ and $F(x+t)$ but how am I going to get the integral of $G$?enter image description here

Amzoti
  • 56,093
Lost1
  • 7,895

1 Answers1

2

Solving should give you, rather

$$\hat{u}(k,t) = A(k) e^{i k t} + B(k) e^{-i k t}$$

$A$ and $B$ come from initial conditions. You then inverse FT to get $u(x,t)$.

In your case, $u(x,0) = F(x)$ means that, taking the FT

$$\hat{u}(k,0) = A(k) + B(k) = \hat{F}(k)$$

where

$$\hat{F}(k) = \int_{-\infty}^{\infty} dx \: F(x) e^{i k x}$$

The other initial condition is

$$\frac{\partial}{\partial t} \hat{u}(k,0) = i k [A(k) - B(k)] = \hat{G}(k)$$

where

$$\hat{G}(k) = \int_{-\infty}^{\infty} dx \: G(x) e^{i k x}$$

Two equations, two unknowns...

Ron Gordon
  • 138,521