# Light travelling on curved surface

1. Nov 7, 2015

### SataSata

Can anybody explain under what conditions does this happens?
Why does the light beam travel on the curved surface? How does this happen? Isn't light suppose to travel in straight line?

#### Attached Files:

• ###### curvedlight.png
File size:
110.4 KB
Views:
160
2. Nov 7, 2015

Staff Emeritus
Why do you think the light beam is traveling on the curved surface? If you drilled a tiny hole in the top of the plastic, what would you see?

3. Nov 7, 2015

### sophiecentaur

Light travels in a straight line in space. In your picture, the light is being ducted along the curved channel in a series of short straight lines between reflections against the walls. The phenomenon is probably Total Internal Reflection, if it is travelling through a plastic rod.

4. Nov 7, 2015

### SataSata

I don't understand what I would see. Can you elaborate?
I'm not sure how there's a curved channel in this. I got the picture from this video:
It happens around 2.55 mins and only for a short window of angle of incidence.

5. Nov 7, 2015

### sophiecentaur

That's a better picture. It was not clear what the original picture was showing. It isn't a curved channel - it's just the inside of the D shape.
Do you know about Total Internal Reflection? You can see many pictures of it happening just once as it strikes the internal surface of glass / water / plastic etc. For it to happen, the angle of incidence must be greater than the Critical angle. In your first picture. the light is being reflected internally many times by the internal curved surface. The picture is too blurry to see that happening but you could draw the effect if you draw a circular arc and draw a line, striking it very obliquely (inside the curve) the ray will be reflected and hit the circle again and again (following the i = r rule at each contact) If the ray hits the curve too steeply, it will just escape and if it enters the D too shallow, it will not make it to the curved surface.

6. Nov 7, 2015

### SataSata

I still don't understand how total internal reflection on the curve could have happen at that angle. Look at my attached drawings. I can understand total internal reflection happening for the left drawing. But in the video, the beam came in from a different side which is shown in my drawing on the right. I couldn't come up with any way total internal reflection can happen on the D shape side.

File size:
11.2 KB
Views:
105
7. Nov 7, 2015

### Staff: Mentor

I don't think it's total internal reflection. That requires a duct and small angles. To me it just looks like some minor scattering inside a not perfectly clear prism.

8. Nov 7, 2015

### sophiecentaur

It is possible to launch a ray that will pass along the circular block as long as the angle is right. (As a series of cords to the curve). I think there must be a curved section at the transition between flat and semicircular sides (the corner`) that causes light to be refracted at the correct angle. This will be restricted to only a small range of incident angles. It would be easier to explain with a different shaped block - one with a flat side which is almost at right angles to the present flat edge.
If you can launch the ray at the correct angle, you can do it with reflections on just one concave internal surface. You can draw a diagram yourself to show it happening. Imo, the D block has some pretty subtle shaping of the corners that allow it to happen.

9. Nov 7, 2015

### SataSata

Yes I think it's possible depending on how the corner of the D block is shaped. It won't work if the corner is a perfect right angle which is unlikely. It probably has a little curvature at the corner.

#### Attached Files:

• ###### dshapecorner.png
File size:
1 KB
Views:
79
10. Nov 7, 2015

### Staff: Mentor

Isn't "the correct angle" small? The angle just in from the corner is almost 90 degrees!

11. Nov 8, 2015

### sophiecentaur

Yes, that would produce the effect but the incidence would have to be almost normal to the flat side. The picture in the OP shows the light coming in, apparently nearly parallel to the flat side. It's a bit confusing.