Graphical understanding of the Laplace transform of sin(x).

76 Views Asked by At

The Laplace transform transforms a function from the time domain to the frequency domain. But I am having trouble understanding what this really means. I assume this picture is correct

Then, a frequency domain - representation of the function $\sin(t)$, should be a line of height = amplitude of $\sin(x) = 1$, and it should have no width. On the $x-$axis, the line should be located at $x=\frac{1}{2\pi}$, since that is the frequency of $\sin(x)$.

But the Laplace transform of $\sin(t)$ is $\frac{1}{s^2+1}$. This is not a straight line. Why not? I dont understand how $\sin(t)$ can be equal to $\frac{1}{s^2+1}$ in the frequency domain?

Any help would be much apreciated!!

$\sin(x)$ and $\frac{1}{x^2+1}$ graphed

1

There are 1 best solutions below

0
On

I believe your confusing the unilateral Laplace transform

$$\mathcal{L}_t[f(t)](s)=\int\limits_0^{\infty} f(t)\, e^{-s t}\,dt\tag{1}$$

with the Fourier transform

$$F(\omega)=\mathcal{F}_t[f(t)](\omega)=\int\limits_{-\infty}^{\infty} f(t)\, e^{-i \omega t}\,dt\tag{2}$$

which is equivalent to the bilateral Laplace transform

$$\mathcal{BL}_t[f(t)](s)=\int\limits_{-\infty}^{\infty} f(t)\, e^{-s t}\,dt\tag{3}$$

evaluated at $s=i \omega$.


The unilateral Laplace transform of $\sin(t)$ is $$\mathcal{L}_t[\sin(t)](s)=\int\limits_0^{\infty} \sin(t)\, e^{-s t}\,dt=\frac{1}{s^2+1}\,,\quad\Re(s)>0\tag{4}$$

whereas the Fourier transform of $\sin(t)$ is

$$\mathcal{F}_t[\sin(t)](\omega)=\int\limits_{-\infty}^{\infty} \sin(t)\, e^{-i \omega t}\,dt=i \pi\, \delta(\omega+1)-i \pi\, \delta(\omega-1)\tag{5}$$

which evaluates to $0$ for $\omega\in\mathbb{R}\land\omega\ne\pm1$ and is undefined at $\omega=\pm1$.


The inverse Fourier transform associated with formula (2) above is

$$f(t)=\mathcal{F}_{\omega}^{-1}[F(\omega)](t)=\frac{1}{2 \pi} \int\limits_{-\infty}^{\infty} F(\omega)\, e^{i t \omega}\,d\omega\tag{6}.$$


Formula (5) above can be validated via the inverse Fourier transform

$$\mathcal{F}_{\omega}^{-1}[i \pi\, \delta(\omega+1)-i \pi\, \delta(\omega-1)](t)=$$ $$\frac{1}{2 \pi} \int\limits_{-\infty}^{\infty} (i \pi\, \delta(\omega+1)-i \pi\, \delta(\omega-1))\, e^{i t \omega}\,d\omega=\frac{1}{2} i\, e^{-i t}-\frac{1}{2} i\, e^{i t}=\sin (t)\tag{7}$$

which can be derived from the following property of the Dirac delta function $\delta(\omega)$:

$$\int\limits_{-\infty}^{\infty} \delta(\omega-\alpha)\, F(\omega)\,d\omega=F(\alpha)\,,\quad\alpha\in\mathbb{R}\tag{8}.$$


It is not true that $\delta(0)=1$ (actually $\delta(0)$ is undefined), rather it is true that

$$\int\limits_{-\beta}^{\beta} \delta(\omega)\,d\omega=1\tag{9}$$

for all $\beta>0$.


The Fourier and inverse Fourier transforms can also be defined as

$$F(\omega)=\mathcal{F}_t[f(t)](\omega)=\int\limits_{-\infty}^{\infty} f(t)\, e^{-i 2 \pi \omega t}\,dt\tag{10}$$

and

$$f(t)=\mathcal{F}_{\omega}^{-1}[F(\omega)](t)=\int\limits_{-\infty}^{\infty} F(\omega)\, e^{i 2 \pi t \omega}\ d\omega\tag{11}$$

which for $f(t)=\sin(t)$ leads to

$$\mathcal{F}_t[\sin(t)](\omega)=\int\limits_{-\infty}^{\infty} \sin(t)\, e^{-i 2 \pi \omega t}\,dt=\frac{1}{2} i\, \delta\left(\omega+\frac{1}{2 \pi}\right)-\frac{1}{2} i\, \delta\left(\omega-\frac{1}{2 \pi}\right)\tag{12}$$

and

$$\mathcal{F}_{\omega}^{-1}\left[\frac{1}{2} i\, \delta\left(\omega+\frac{1}{2 \pi}\right)-\frac{1}{2} i\, \delta\left(\omega-\frac{1}{2 \pi}\right)\right](t)=$$ $$\int\limits_{-\infty}^{\infty} \left(\frac{1}{2} i\, \delta\left(\omega+\frac{1}{2 \pi}\right)-\frac{1}{2} i\, \delta\left(\omega-\frac{1}{2 \pi}\right)\right)\, e^{i 2 \pi t \omega}\,d\omega=\frac{1}{2} i\, e^{-i t}-\frac{1}{2} i\, e^{i t}=\sin(t)\tag{13}$$

which seem more consistent with some of the expected results outlined in the question, and for $f(t)=\cos(t)$ leads to

$$\mathcal{F}_t[\cos(t)](\omega)=\int_{-\infty}^{\infty} \cos(t)\, e^{-i 2 \pi \omega t}\,dt=\frac{1}{2}\, \delta\left(\omega+\frac{1}{2 \pi}\right)+\frac{1}{2}\, \delta\left(\omega-\frac{1}{2 \pi}\right)\tag{14}$$

and

$$\mathcal{F}_{\omega}^{-1}\left[\frac{1}{2}\, \delta\left(\omega+\frac{1}{2 \pi}\right)+\frac{1}{2}\, \delta\left(\omega-\frac{1}{2 \pi}\right)\right](t)=$$ $$\int\limits_{-\infty}^{\infty} \left(\frac{1}{2}\, \delta\left(\omega+\frac{1}{2 \pi}\right)+\frac{1}{2}\, \delta\left(\omega-\frac{1}{2 \pi}\right)\right)\, e^{i 2 \pi t \omega } \, d\omega=\frac{1}{2}\, e^{-i t}+\frac{1}{2}\, e^{i t}=\cos(t)\tag{15}.$$