# Frozen Waves

1. Jan 15, 2007

### Gaijinja

Ok, first I'll mention that I'm completely new to these forms (first post) and all that jazz. Anyway, on to the subject of the post.

In my physics class (high school level only, mind you) we were doing a lab with long slinkies to show the properties of waves and such. Our teacher wanted us to try and measure the amplitude of a moving wave, both transversal and longitudinal, mainly to prove how hard it is. My lab partners and I discovered that if one person bunched up a lot of the slinky at one end (effectively making the slinky more taught) while the person on the other end of the slinky generated a wave of either type and then the person with the bunched up slinky let go, the wave effectively froze in place while the slinky sprung back to a less deformed position, allowing us a couple of seconds to quickly measure amplitude. When we showed our little trick to the teacher, he was quite impressed, and said that he was completely unfamiliar with that phenomena in waves. He claimed that we might be able to do some research on the subject, and get it published in a small science journal or something similar. Now, as much as I would like to get something published, I realize that this probably isn't a new discovery, however, I've been searching on google for awhile, and have not run across anything explaining this phenomena. I have read a bunch about Standing Waves, where two waves interfere in such a way as to create specific nodes that don't move. This, however, is not the phenomena we saw.

So, after that long introduction, my question is what is this phenomena and what is the math associated with it?

2. Jan 15, 2007

### planish

If the "wave generator" person was making a single longitudinal wave, that would be a compression followed immediately by an expansion (or vice-versa). The person releasing the bunched-up end would be creating a (I think) compression pulse with probably a different amplititude. If they meet in the middle, the two should pass each other without stopping, only momentarily creating an interference pattern, but only for a fraction of a second.
If the "wave generator" person made a transverse wave, the I would have expected similar results, but with wierder motion as they passed each other.

How did you measure the amplitude of the longitudinal wave? Wasn't it hard to choose the start and end points with any degree of accuracy?

I wonder if you couldn't call it a "stationary soliton" or something like that, as impossible as it sounds. See http://en.wikipedia.org/wiki/Soliton

Magic. There is no math, only a random concentration of mana in the vicinity - http://en.wikipedia.org/wiki/Mana :rofl:

Seriously, I got nothing.
If you do investigate it further, I'd suggest videotaping it in front of a wall with a grid pattern so you can freeze-frame it at various points, measure points of interest, and study the timing (eg. by counting video frame numbers).
Also come up with some mechanism for creating the pulses, which would then be adjustable and repeatable as far as timing and amplitude go, between one take and the next.
Maybe even paint contrasting dots on the slinky every 10th turn or so, in case the camera cannot resolve individual turns when they in motion.

3. Jan 15, 2007

### cesiumfrog

It sounds like you're trying to suddenly decrease the tension in the slinky, ie., to suddenly decrease the wave velocity in the medium.

4. Jan 15, 2007

### HallsofIvy

By the way, the slinky was "taut". You just can't teach a slinky anything!

5. Jan 15, 2007

### Gaijinja

Yes, that's one of our hypotheses. The interesting thing to note was that regardless of how fast the wave was moving, it always froze when the tension was released.

That's what you think.

Invariably so, but that was the point of the lab (showing how hard it is to measure dynamic waves), and is rather independent of the phenomena I've described.

6. Jan 15, 2007

### Claude Bile

Here's my theory.

When you release the taut end of the slinky, rings of the slinky must propagate in the opposite direction to the wave (to equalise the tension). In a manner of speaking, the 'velocity' of the medium is countering the velocity of the wave, which makes sense because both 'motions' are governed by the same parameters. The end result is a stationary wave during the period of tension equalisation.

I have a sudden urge to buy a slinky to try this out for some reason.

Claude.

7. Jan 15, 2007

### cesiumfrog

I would expect to find a relation between tension and wave propagation speed written in one of your physics teacher's general textbooks, being a reasonably basic concept (eg. tuning of guitars). I do presume you can imagine a way to test this hypothesis?

Last edited: Jan 15, 2007
8. Jan 15, 2007

### Gaijinja

Yes I can, I was not asking for ideas, I was merely asking if it would be worth investing my time in, or if it was just a really basic phenomena my teacher just wasn't aware of.

Anyway, thanks for all the responses, they've all been very helpful.

Last edited: Jan 15, 2007
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook