6
$\begingroup$

What is meant by a 'pure' wave?

I know it might sound like a basic question, but I've never been taught this.

I saw that a sine wave is a pure wave. I tried Googling what a pure wave is, but all I get is links regarding Pure Wave Inverters for sale...which is not what I'm looking for.

$\endgroup$
  • 1
    $\begingroup$ I think symmetric as its periodic over $2π$ and has equal crest and trough $\endgroup$ – Archis Welankar Feb 4 '16 at 4:58
13
$\begingroup$

A sine or cosine wave has just one frequency and it called a pure wave for that reason. If you have a periodic function with a different shape you can Fourier analyze it and get some number of different frequencies. These are not pure waves because of the multiple frequencies.

| cite | improve this answer | |
$\endgroup$
  • 2
    $\begingroup$ THANK YOU :D ...this helped more than you think :) $\endgroup$ – Max Echendu Feb 4 '16 at 5:11
9
$\begingroup$

A "pure" sine (or cosine) wave is a wave of the form $y(t)=c_1\sin(\omega t+\phi)$ or $y(t)=c_2\cos(\omega t+\phi)$. Essentially, it's a wave that may have been translated, scaled, or have its period modified, but at it's core it's still a sine or cosine wave.

This is contrary to sums of various waves. A core result of a field known as Fourier Analysis is that any periodic function can be approximated as a sum of the "pure" sine and cosine waves. Here, the sines and cosines are of the form $y_n(t)=b_n\sin(nt)$ or $y_n(t)=a_n\cos(nt)$ for certain series $a_n$, $b_n$ known as the fourier series of your function.

So, given a periodic function $f(t)$, you can find that $f(t)=\frac{1}{2}a_0+\sum_{n=1}^\infty a_n\cos(nt)+\sum_{n=1}^\infty b_n\sin(nt)$. So, you can use an infinite number of "pure" sine/cosine waves to represent any periodic function.

Note that a pure wave isn't just having certain properties sine/cosine have - even waves that look fairly similar to sine to the untrained eye (like the triangle wave - while it has sharp boundaries, it (can) have similar period and amplitude) can be very complex when written in terms of sines/cosines. While you (usually) need the full infinite number to perfectly describe a wave, you can get surprisingly good approximations relatively quickly (here $n$ is the number of sine terms, and there are no cosine terms).

| cite | improve this answer | |
$\endgroup$
3
$\begingroup$

In electrical technology any periodic motion can be expressed as a sum of the fundamental wave and several harmonics whose frequencies are multiple of the fundamental, expressible in Fourier series. If the wave is pure it means that amplitudes other than the fundamental are zero. The other harmonics are seen as "contaminating" the pure sine wave.

It is pure or aka simple harmonic, as is represented by the harmonic dynamic time differential function which strictly has a time period $T$ where $ \omega T = 2 \pi $ in:

$$ \ddot x + \omega^2 x =0. $$

A square wave for example has several non-fundamental additives to the pure signal. It consistes of third, fifth,.. order harmonics:

Other Harmonics in Square Wave

| cite | improve this answer | |
$\endgroup$
1
$\begingroup$

It's the simplest form and satisfies a linear wave equation $\frac{\partial^{2} \psi}{\partial t^{2}}=c^{2}\frac{\partial^{2} \psi}{\partial x^{2}}$ and hence can superimpose one another. Sinusoidal wave consists of single frequency (or wavelength) whereas wave packet consists spectrum of frequencies. This is analogous to uncertainty principle in quantum mechanics $\Delta p \Delta x \geq \hbar \leftrightarrow \Delta k \Delta x \approx 1$.

There're other kinds of waves:

A soliton $\psi(x,t)=\frac{c}{2}\, \text{sech}^{2} \left[ \frac{\sqrt{c}}{2}\, (x-ct) \right]$ is a single wave packet travelling at single speed without dispersion. It satisfies the Korteweg-deVries (KdV) equation $\psi_{t}+\psi_{xxx}+6\psi \psi_{x}=0$ which is non-linear.

Water wave has a dispersion relation $$c=\sqrt{\left( \frac{g\lambda}{2\pi}+\frac{2\pi \gamma}{\rho \lambda} \right) \tanh \frac{2\pi h}{\lambda}}$$ with displacement profile $$\xi+\eta i=\frac{ga}{kc^{2}} \frac{\cosh[k(y+h)+i(kx-\omega t)]}{\cosh kh}$$

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ $f(x) = \displaystyle e^{i \sum \omega_k x_k}$ (for certain values of $\omega$) is the basis of solutions to any homogenous linear differential equation (this is why the Fourier transform/series was created at first) $\endgroup$ – reuns Feb 4 '16 at 12:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.