How? When your engine quits, you immediately change the pitch angle of the main rotor blades from positive pitch to negative pitch. By doing this the rotor blades will continue to keep spinning in a forward direction as the air rushes by them while the helicopter is descending.

The whole idea behind this is to build up rotor speed as the heli descends and then as the helicopter nears the ground you change the pitch angle of the rotor blades back to positive. The kinetic energy built up in the rotor allows you enough power to stop your descent and make a soft landing.

This is a bit tricky. Having the right amount of negative pitch and positive pitch is very important (generally -3 degrees to +12 degrees but it depends on your helicopter). Auto rotations work much better if you have some forward speed providing "clean" air (also referred to as translational lift ) preventing the heli from descending too quickly while still generating lots of energy in the rotor. This is generally referred to as "glide path" and a forward glide path of about 45 degrees provides a good slow descent rate while building lots of rotor energy.

The key to a successful autorotation is all in the timing and knowing when to slow your descent and forward speed by flaring with a rear cyclic command, applying positive collective pitch, and then leveling out the heli just before touch down with forward cyclic to land gently. All that without dissipating all the energy in the rotor before you land - remember, you only get one shot at this.

Practice makes perfect and this is where your trusty Flight Simulator can help with the dynamics of what all is going on.

If you had fixed pitch or kept the rotor blades at a positive pitch angle, they would soon stall and stop rotating. Yup - the helicopter has now become an expensive rock. This is why autoroations rely on collective pitch.

We don’t actually physically disconnect the rotor shaft from the rest of the drive components - but we do allow it to spin freely, by using one way bearings on the rotor shaft. This is the same idea as a bicycle’s rear wheel. While you are pedaling, there is power going to the rear wheel - but when you stop pedaling, the rear wheel is free to spin. If this wasn’t the case, the pedals would keep rotating and you couldn’t coast.

This is exactly the same principle that is used in a RC helicopter with autorotation capability. As long as the engine is supplying power, the energy is transmitted to the main rotor shaft through the locked one-way bearings. If the engine quits, the bearings now rotate and the rotor shaft is allowed to spin freely – just like coasting on your bicycle.

With no engine noise, the only thing you can hear is the rotor slicing through the air. Hearing how the noise changes as the blades are changed from negative to positive pitch just before touch down is very cool to experience and is why the auto has become a crowd favorite and is so popular in RC Heli competition events.

Here's a pretty neat video of a full size S-92 heli performing an auto rotation. Note the huge amount of flare required to dissipate forward speed.