How things work
How a seat belt knows to lock
A working model of the reel a seat belt winds onto, with its cover off. Nothing here is drawn by hand: the parts move the way the forces on them say they should.
The belt
The car
What is simplified
- Retractors differ from maker to maker. This one follows a common layout: a pawl on the spool, teeth in the frame, and a control wheel that both sensors act on. Others put the teeth on the spool and the pawl on the frame.
- Here the wheel's own weight makes it lag when the belt is yanked. Many retractors add a small separate weight on the wheel to do that job.
- The braking slider goes up to 1 g and the ball starts to roll at about a third of a g. Real sensors are set to trip somewhere below 0.7 g and to stay put when the car is tilted 15 degrees.
- Only part of the belt is shown on the spool, so the roll is thinner than in a car. The rewind spring is drawn evenly spaced; a real one bunches against the wall of its cup.
- Most modern belts also have a pretensioner and a load limiter. Neither is part of how the belt decides to lock, and neither is shown.
In short
The webbing of a seat belt winds onto a spool inside the retractor, and a spring keeps winding it back in. A steel pawl on the spool can swing out and catch a ring of teeth in the frame, which stops the spool. A light wheel riding on the spool controls the pawl. A sharp pull makes the spool leap ahead while the wheel lags behind, and that pushes the pawl out. In a hard stop, a steel ball rolls up its dish and lifts a lever that holds the wheel still, with the same result. The pull on the belt itself drives the pawl into the teeth.