1. PF Contest - Win "Conquering the Physics GRE" book! Click Here to Enter
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

The real cause of refraction.

  1. Sep 12, 2010 #1
    Hello, while making my tea and watching the spoon inside the glass (with water), I start questioning myself what really is refraction. On class they just told me refraction is because light goes slower in a medium, so it bends according to Snell's Law.
    What is Snell's Law? Light bends to a certain angle when it goes change to a different refractive index medium, so it goes slower and therefore bends.
    This is pretty much a 'chicken or the egg paradox' for me.

    So why does light bends when it goes to a different refractive index. Why does the slower velocity cause it to bend?

    Thanks, Sakha.
  2. jcsd
  3. Sep 12, 2010 #2
    For any type of wave, even those at the surf at a beach, the wave will bend when it comes to a medium where it can propagate at a different speed. The name for this effect is refraction. The technical details are Snell's law. The reason is simple geometry, you easily look up a derivation of Snell's law if you want to understand better.

    The reason light refracts in a medium is because light is a wave that goes slower in a medium. The reason it goes slower in a medium is a consequence of the details of how any polarizable medium interacts with a changing EM field..
  4. Sep 12, 2010 #3
    So I guess my question can be rephrased as "Why will waves bends when they get to a medium where they propagate slower?".
  5. Sep 12, 2010 #4


    User Avatar
    Gold Member

    Don't quote me on the following.

    I do believe it is simple geometry, and is an inevitable consequence when you plot the movement of the wave point by point.

    I believe a wavefront can be viewed as the sum total of constructive interference of a wave radiating out from every point on the line that is the wave's crest.

    i.e. freeze an ocean wave in time. Treat every point on the line that is the crest of a wave as a point that emanates a new wave in a circular front moving out in all directions. The net results of all these infinite points radiating ends up cancelling everywhere except the leading edge of the wave moving forward.

    So far, all you've got is a transverse wave moving straight.

    Now, if you diagram this as it hits a new medium, the waves from every point as they encounter the boundary will propogate in slower, smaller circles. When you, as before, diagram these, then draw a line across the leading edges, the new line you draw (which is the new wavefront in the new medium) will be bent.

    (Note: there's only one wave crest in the diagram, moving over time. The "new" wave is just the one wave having moved forward.)

    Attached Files:

    Last edited: Sep 12, 2010
  6. Sep 13, 2010 #5
    Part of the concept is that frequency is constant but wavelength changes in proportion with speed. Think of cars going into a long tunnel. For concreteness, lets say the cars are going in at a rate of ten per minute and at a speed of 60kph, and lets say the cars are leaving at a speed of only 6kph (e.g., the different parts of the tunnel are in different speed zones). The cars still must be leaving at a rate of ten per minute. If they didn't leave at the same rate they entered, the total number of cars inside the tunnel would be changing, eventually there would be no more space to fit more cars in (or otherwise no more cars left to bring out). But the distance between the cars has changed. As they entered, the cars were each 100m apart. As they leave, there is only 10m from each car to the next. The cars are analogous to wave crests. For the same reasons, crests of a wave will bunch up when travelling to a medium where the speed of the wave is slower.

    Part of the concept is that the wave always travels in the direction perpendicular to the wave crest (with successive crests parallel if they are in the same medium). This is to do with Huygen's principle, which Dave is illustrating.

    And the rest is geometry. You can understand this by drawing an extended version of Dave's picture for yourself. If you fill the top half of a page with lots of parallel diagonal lines with 1cm separation, and want to fill the bottom of the page with lots of parallel diagonal lines with only 0.5cm separation, you'll see that the lines only match up in the middle if the lines at the bottom are at a different angle from the top. If you figure out the precise mathematics, you'll have Snell's law.
    Last edited: Sep 13, 2010
  7. Sep 13, 2010 #6
    1) Waves don't bend it's the wave train that does this. There is a big difference

    2) Refraction also happens when change to a medium where they they propagate faster.

    3) There's lots of good diagrams and explanation already in Physics Forums. I recommend a search.

  8. Sep 13, 2010 #7
    Snell's Law is empirical. You can experiment to find that it is true.

    The slowing of light waves in different medium is harder to see. You can't measure it directly, because the speed of light is hard enough to measure already.

    However, it's clear in slower moving mechanical waves that waves move at different speeds in different media. Sound moves faster through steel than through air. It is reasonable to assume light might be similar.

    Check out Fermat's Principle (and the more modern version, the Principle of Least Action). This theory provides a very succinct description of which path(s) light will take to get to its destination, solely based on the time (or action) it takes to get there. In other words, the path depends on the velocity of light through the media involved.
  9. Sep 13, 2010 #8


    User Avatar
    Gold Member

    He wasn't questioning thatit was true; I think he simply wanted to "zoom in" to see how it actually happens at the nuts & bolts level in any wave medium.
  10. Sep 13, 2010 #9


    User Avatar
    Gold Member

    Huh. So it is. I guess that was a piece of knowledge I'd picked up in the distant past. I was sort of making that explanation up as I went along. Even my picture atched the Wiki example. :smile:
  11. Sep 13, 2010 #10


    User Avatar
    Science Advisor

    Here's a very classical explanation: Imagine an army unit or marching band, with, say, 10 people across, any number of rows deep. They march from an asphalt lot onto soft, rough, ground where they have to march slower to be sure not to fall. And they come onto it at an angle to their front row- people on the left reaching it first. As the people on the left end march onto the field, they will march slower than the person to their left, so fall back a little. That causes the line to "break" (form an angle) at that point. As every one in the line marches onto the soft ground and slows, they will be in line again, but at the new angle.
  12. Sep 13, 2010 #11
    It's getting a bit clearer for me, after reading about Fermat's, Huygens, and the derivation of Snell's Law.

    A few more questions comes to mind:
    Sound waves then should follow this same laws and have all the characteristics of waves, but why don't we feel the sound 'bending'?

    As all this is about waves, a single photon will not experience this bending when it changes its medium of propagation right?

    Thanks for all responses.
  13. Sep 13, 2010 #12
    We don't necessarily have the sensitivity to overtly notice it when it happens, but it happens.

    A single photon will bend the same as a group of photons. It's a wave when it wants to be.
  14. Sep 13, 2010 #13
    Sound is indeed refracted.

    This phenomenon is very important in (anti) submarine warfare where sound is used instead of vision.
  15. Sep 13, 2010 #14
    Those indecisive annoying photons!
  16. Sep 13, 2010 #15


    User Avatar
    Gold Member

    What makes you think you don't?

    Something to note though: it is uncommon to be in a circumstance where sound passes through media of differing refractive indices. Almost exclusively, our day-to-day experience of sounds is through air, which is, for the most part, of a uniform density. So, observing refractive sound will be challenging at best.

    Now it does happen (sounds will refract when entering water, or steel, etc.) but the observation of it has to be in fairly contrived and controlled circumstances.
  17. Sep 13, 2010 #16


    User Avatar
    Gold Member


    (as the right wheel hits the grass, it slows down, but the other wheel is still going fast, resulting in rotation.
  18. Sep 13, 2010 #17
    I really like this example, it's as if there was some kind of torque working there.
    What is messing with me probably is my intuition of thinking of light as a straight ray instead of a wave.
  19. Sep 13, 2010 #18
    I like that example. Do you remember where it came from?
    Last edited: Sep 13, 2010
  20. Sep 13, 2010 #19


    User Avatar
    Gold Member

    Ubiquitous. It was what I was taught in school.

    And that waren't jes yesterday...
  21. Aug 8, 2011 #20
    light has a dual property, i.e, wave property and particle property. if we consider the wave property then light travels in a straight line as it is the shortest path and uses less energy. unelike any other substance or any form of energy in the universe, light also has the basic instinct, that is to save its energy and to have high energy levels like sphere. light travels through different speed in different media. the medium in which light travels faster is called a rarer medium. e.g, air. the medium in which light travels slower is called a denser medium. e.g, water. so, as light travels from a rarer medium to a denser medium, then the phenomena of refraction takes place. if we consider, the rarer medium to be air and denser be water, then a ray of light, travelling obliquely from air, enters water. in the boundary of separation of the two media, it bends.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook