Why do bendy glass threads carry far more of your video call across the world without it fading?
After you watchWhy do bendy glass threads carry far more of your video call across the world without it fading?
The short answer
Fiber carries your call as flashes of light trapped inside a hair-thin glass thread. When the light hits the inside wall at a shallow, grazing angle, the wall acts like a perfect mirror and bounces it straight back in, so the beam zigzags down the thread — even around bends — without leaking out. That trick is called total internal reflection.
Try this next
- What if you bend the fiber sharper and sharper? Tighten the bend in the explainer and predict first: at which curve does the trapped ray finally hit the wall too steeply and leak out the side?
- What if the ray starts out closer to head-on instead of grazing? Aim the ray steeper before sending the pulse and guess whether it stays trapped or escapes at the very first wall it hits.
Now you — bend it
- What if In the bendy-fiber lab, drag the aim slider to sit just barely shallower than the violet critical line, then send a pulse.A ray skimming right at the edge only just survives each bounce. Predict what one slightly-too-tight bend would do to a ray that close to the limit, versus a ray aimed very shallow with margin to spare.
- What if Imagine swapping the air outside the glass for a cladding whose index is much closer to the core's index, so the critical angle slides from about 42 degrees up toward 80-plus degrees from straight-on.A bigger critical angle means only the shallowest, most grazing rays stay trapped. Predict whether that narrower 'acceptance cone' lets in more rays or fewer, and what that does to how much light you can pour in at the start.
- What if The lab fiber is short. Imagine stretching it to 100 km and sending one trapped pulse versus a steep ray that leaks at the first bend.A trapped ray still loses a tiny fraction of its power per kilometer to scattering and absorption (real fiber is roughly 0.2 dB/km, about 5% lost each km). Predict whether 'trapped' means 'arrives at full strength' over that distance, or whether it still needs a boost along the way.
Can you prove it?Whether a ray stays trapped depends only on the angle it hits the wall, not on how bright the pulse is. — Fix the aim slider at one shallow angle and send pulses; then keep that exact angle and only change the brightness in your head. The critical angle comes from the two glasses' indices and Snell's law (the sine of the critical angle equals n_cladding divided by n_core), which has no brightness term in it. Now slide steeper at the same brightness and watch it leak — showing the angle is the switch, not the power.
Design your own test:Before you send each pulse, predict the exact slider spot where 'arrives at the catcher' flips to 'leaks at the first bend' — then test whether that flip is sharp and sudden (a true tipping point) or a gradual fade.
Explain it to a 6-year-old: If the light slides along the glass wall almost flat, the wall acts like a perfect mirror and the light skips all the way around the bend without sneaking out.
The whole story
How it works
Light usually travels in a straight line. Inside a glass fiber, the beam keeps hitting the inside wall. If it hits at a shallow grazing angle, all of the light reflects back inside instead of passing out into the air, so it bounces again and again and travels down the thread. There is a special tipping-point angle, the critical angle (about 42 degrees from straight-on for glass and air). Hit the wall shallower than that and the light stays trapped; hit it steeper, closer to head-on, and the light leaks out the side and the signal is lost. Because the light always reflects, it can follow the cable around curves all the way to the far end.
What people get wrong
People often think light only goes in straight lines, so it could never follow the bends in a cable and would just shoot out at the first curve. In fact, the glass wall bounces the light back in every time it hits at a shallow enough angle, so the beam zigzags around bends and stays inside.
The catch
Fiber (light in glass) carries a huge amount of information at once and barely fades over long distances, but the glass must be very pure and you cannot bend it too sharply or even a grazing ray finally hits too steeply and leaks out. Copper (electricity in metal) is cheap, tough, and easy to bend and connect, but the signal fades and picks up noise much faster and cannot carry nearly as much at once.
Questions kids ask
How does light go around the bends in a fiber cable?
Every time the light meets the inside wall at a shallow grazing angle, the wall reflects all of it back inside. So the beam zigzags down the thread and follows gentle curves without leaking out. Bend the cable too sharply, though, and the light finally hits the wall too steeply and escapes.
What is the critical angle?
It is the tipping-point angle for light trying to leave glass for air, about 42 degrees from straight-on (the normal). Hit the wall shallower than that and 100 percent of the light reflects back inside (total internal reflection); hit it steeper, closer to head-on, and the light passes out through the side.
Why does fiber beat copper wire for internet?
It is not that the light itself travels faster — light in glass actually moves a bit slower than an electric signal races through copper. Fiber wins because it can carry far more information at once and the signal barely fades, so one thread can send a huge amount of data all the way across cities and oceans without it getting weak. Copper carries less at a time and its signal fades and picks up noise much faster over distance.
Does the light bounce off a mirror inside the fiber?
There is no shiny metal mirror. The glass wall itself acts like a perfect mirror whenever the light hits it at a shallow enough angle. That is total internal reflection: the boundary between glass and air bounces all the light back in, no coating needed.
Talk about it
- Light usually goes in a straight line — so how do you think it follows a bendy cable around corners? Guess before we test it.
- If there's no shiny mirror inside the glass, what do you think is bouncing the light back in?
- Why might the angle the light hits the wall matter more than how bright it is?
For grown-ups
When light passes from a dense medium (glass) toward a less dense one (air) it bends away from the normal; beyond a critical angle (about 42 degrees from the normal for a glass-air boundary, set by Snell's law and the two refractive indices) it cannot exit at all and undergoes total internal reflection. A fiber's core is wrapped in a slightly lower-index cladding, so rays inside the acceptance cone keep reflecting and stay guided for many kilometers. Light in glass travels at roughly two-thirds the speed of light in vacuum, yet carries enormous bandwidth with very low loss, which is why fiber beats copper for long, fast data.
Keep going
What else makes you wonder?
- If light keeps bouncing instead of going straight, does it travel a longer path than a straight beam would?
- Could you trap a beam inside something other than glass, like water or clear jelly?
- How do they get a flash of light to start out at just the right shallow angle to stay trapped?