The Variable-Pitch Propeller

Quantify the loss before engineering it away.

1. Take two fixed propellers of the same diameter but different pitch — one fine, one coarse.
2. On the thrust rig, measure static thrust and rpm for each at full power.
3. Now measure both in the wind tunnel at a simulated cruise airspeed, recording thrust and rpm.
4. Tabulate all four results.

The fine propeller wins at zero airspeed and overspeeds the motor at cruise; the coarse one is feeble static and efficient at speed. Neither is good at both, and the gap between them is the performance a fixed propeller throws away.

The whole mechanism is a blade root that turns in a bearing while transmitting enormous loads.

1. Turn a cylindrical hub with two radial bores, one per blade.
2. Machine each blade root as a matching shank with a shoulder that seats against a thrust bearing.
3. Fit each root with a needle or ball thrust bearing so it can rotate under load.
4. Attach a short crank pin to the inboard end of each blade root, offset from the rotation axis.
5. Confirm each blade turns smoothly through at least 30° with no radial play.

The thrust bearing is doing the difficult work: it must let the blade twist freely while carrying the full centrifugal load, which on a real propeller is measured in tonnes. Radial play is unacceptable — any slop becomes a blade angle that varies with load, which is exactly the imprecision the mechanism exists to remove.

Two blades that disagree about their pitch are worse than two blades that are both wrong.

1. Make a sliding yoke inside the hub, free to move along the shaft axis.
2. Cut a slot in the yoke to engage each blade’s crank pin.
3. Confirm that moving the yoke axially rotates both blades by the same amount, in the same direction.
4. Measure both blade angles with an angle gauge at three yoke positions and check they agree within a fraction of a degree.

Converting axial motion into synchronised rotation is the mechanical heart of it, and the same slot-and-crankpin arrangement appears in almost every real design. One actuator, one yoke, all blades identical by construction rather than by adjustment.

Now make it adjust itself, which is where it stops being a mechanism and becomes a control system.

1. Fit a pair of flyweights to the hub, hinged so that spinning throws them outward.
2. Link the flyweights to the sliding yoke so that outward movement drives the blades toward COARSE pitch.
3. Oppose them with a compression spring whose preload you can adjust.
4. Spin the assembly and observe blade angle against rpm.
5. Change the spring preload and repeat.

The propeller now holds a roughly constant speed on its own. If rpm rises, the flyweights swing out, pitch coarsens, the blades take a bigger bite, the load on the engine increases and rpm falls back. If rpm drops, the spring wins, pitch fines off, load decreases and rpm recovers. The spring preload sets the target speed — which is exactly what the blue-knobbed propeller lever in the cockpit adjusts.

It is a negative-feedback governor, mechanically identical in principle to the flyball governor on a steam engine. The error signal is rotational speed, the correction is blade angle, and no electronics are involved anywhere.

The most valuable blade angle is the one that produces no thrust at all.

1. Extend the pitch range so the blades can rotate to roughly 90° — edge-on to the airflow.
2. With the rig in the tunnel and the motor unpowered, measure the drag at normal pitch.
3. Now rotate the blades to the feathered position and measure again.

A stopped propeller at normal pitch is a large flat disc of drag; feathered, it is a few knife-edges. The difference is dramatic, and on a twin-engined aircraft it is the difference between flying on one engine and descending on one engine.

Worse than the drag is what an unfeathered propeller does while windmilling: the airflow spins it, it drags the dead engine round with it, and it produces a large asymmetric drag on one wing that must be held off with rudder. Feathering removes the drag and stops the rotation together.