On a light twin-engine airplane, the propeller systems are far more sophisticated than the fixed-pitch props found on many single-engine trainers. Each engine drives a constant-speed propeller whose blade pitch is continuously adjusted by a governor to maintain a pilot-selected RPM regardless of airspeed or power changes. More importantly for engine-out operations, these propellers can be feathered — rotating the blades edge-on into the wind to eliminate drag and stop rotation. Understanding how these systems work, and why feathering is so operationally critical, is fundamental knowledge for any pilot seeking a multi-engine rating.
The FAA Airplane Flying Handbook (FAA-H-8083-3C, Chapter 13) treats propeller management as central to safe multi-engine operations, and the subject appears consistently on both the knowledge test and the practical exam. What follows is a thorough breakdown of how constant-speed and feathering systems function, why they matter to single-engine performance, and the key operational facts every student must know.
How Constant-Speed Propellers Work
A constant-speed propeller uses a governor — an engine-driven, oil-pressure-operated control unit — to maintain a chosen RPM by automatically varying blade pitch. The pilot sets a desired RPM with the propeller control lever in the cockpit; the governor then acts as a closed-loop feedback system.
When the airplane accelerates or the pilot advances the throttle, engine torque tends to increase RPM above the set value. The governor senses this, increases oil pressure directed to the propeller hub, and drives the blades toward a coarser (higher) pitch — increasing the load on the engine and returning RPM to the set value. Conversely, if airspeed decreases or power is reduced, RPM tends to drop; the governor reduces oil pressure, a mechanical spring (and in most designs, counterweights on the blade shanks) drives the blades toward a finer (lower) pitch, reducing the load and restoring RPM.
This constant-pitch adjustment allows the engine to operate at its most efficient RPM across a wide range of flight conditions. On a twin, having independent propeller controls for each engine is essential — the pilot can fine-tune each propeller separately to synchronize RPM, which reduces the annoying beat frequency (audible as a pulsing drone) that results when two engines run at slightly different RPM.
Oil Pressure Logic and Failure Modes
Most light-twin propeller systems are designed so that high oil pressure drives the blade toward fine pitch (low pitch), and loss of oil pressure allows the blade to move toward coarse pitch due to the springs and counterweights. This is the so-called single-acting design. An important implication: if the governor fails or oil pressure is lost, the propeller will move toward the high-pitch (coarse) end of its travel — ideally toward feather. In practice, many single-acting systems will feather automatically on complete oil pressure loss, which is a safety feature during engine failure.
Some light twins use a double-acting system where oil pressure can drive blades in either direction, providing more precise control but requiring additional consideration during failures.
Feathering: What It Is and Why It Is Critical
Feathering means rotating the propeller blades so their chord line is nearly parallel to the relative wind — approximately 90° of pitch change from the fine-pitch position. A feathered propeller presents its leading edge to the airflow, generating almost no aerodynamic drag and stopping rotation entirely.
Why does this matter so much? When an engine fails on a twin, a windmilling propeller — one that continues to spin freely in the airflow — generates an enormous amount of parasite drag. FAA data cited in Chapter 13 of the Airplane Flying Handbook makes the point dramatically: the drag of a windmilling propeller on a failed engine can equal or exceed the drag of the entire airframe. This drag not only degrades climb performance but also worsens the yawing moment toward the dead engine, increasing the rudder force required to maintain directional control.
By feathering the failed engine's propeller, the pilot accomplishes two critical goals simultaneously: drag is reduced dramatically, and the asymmetric yawing moment is lessened, reducing the rudder force needed and improving the airplane's ability to climb (or at least reduce descent rate) on the remaining engine. On most light twins, the difference between a windmilling and a feathered propeller can mean the difference between a controlled single-engine climb and an uncontrollable descent.
The Feathering Mechanism
To feather the propeller, the pilot moves the propeller control for the affected engine all the way aft to the feather detent (or position). This cuts off the governor oil supply and allows residual oil to drain from the propeller hub, permitting the blade-pitch-change mechanism — powered by nitrogen gas charge, high-pitch springs, and/or counterweights — to drive the blades fully toward the feather angle. On most light twins, feathering takes only a few seconds once the control is selected.
A feathering lock or centrifugal latch is incorporated in many designs to prevent the blades from feathering at low RPM while the engine is still running on the ground (where a feathered prop on a running engine would create a dangerous high-power, zero-thrust condition). The latch disengages above a certain RPM (typically around 800–1,000 RPM), so feathering is only mechanically possible when the engine has already slowed significantly or stopped.
Why It Matters: Single-Engine Performance and Survival
The entire architecture of multi-engine single-engine-out (OEI) performance depends on effective propeller management. The FAA specifies Vyse (best single-engine rate-of-climb speed, marked by the blue line on the airspeed indicator) and Vxse (best single-engine angle of climb speed) as the target speeds for OEI operations. These performance figures are predicated on the failed engine's propeller being feathered.
With a windmilling prop, actual single-engine climb performance will be significantly worse than the published figures — often to the point that the airplane cannot maintain altitude even at maximum power on the good engine. The difference is not subtle; pilots who have flown twins with a simulated windmilling prop (propeller at fine pitch, engine at idle) versus a feathered prop consistently report a dramatic improvement in aircraft performance after feathering.
Feathering also reduces rudder workload. Less drag on the dead side means a smaller yawing moment toward that side, so the required rudder deflection — and the associated aileron input to maintain bank — decreases. This directly affects the airplane's ability to stay above Vmc (minimum control speed), since maintaining control is easier with less asymmetric thrust and drag.
The optimal single-engine climb configuration identified in FAA-H-8083-3C is zero sideslip, achieved with approximately a 2° bank toward the operating engine combined with appropriate rudder. This configuration minimizes total drag and yields the best possible OEI climb gradient — better than wings-level and better than any excessive bank angle.
Key Numbers and Rules
- Blue line (Vyse): The airspeed for best single-engine rate of climb, marked by a blue radial line on the airspeed indicator. Always target this speed during OEI climb.
- Vxse: Best single-engine angle of climb — used when obstacle clearance is needed; lower than Vyse and requires precise pitch control.
- Vsse: Safe single-engine demonstration speed — the minimum speed at which intentional engine cuts should be performed during training. Set to provide adequate margin above Vmc.
- Feathering detent: Located at the full-aft position of the propeller control; pulling through a friction gate or detent prevents inadvertent feathering.
- Windmilling vs. feathered drag: FAA guidance notes that drag from a windmilling propeller on the failed engine can approximate the parasite drag of the entire rest of the airplane — a dramatic illustration of why prompt feathering is non-negotiable.
- Bank angle for OEI control: Maintain approximately 2° bank into the operating engine (zero sideslip) for best climb; never exceed 5° bank, which is the limit used to certify Vmc under 14 CFR Part 23.
- Feathering RPM lock: Typically disengages above ~800–1,000 RPM (varies by make/model); below this threshold, a centrifugal latch prevents inadvertent feathering on the ground.
Memory Aid
"Dead Foot, Dead Engine — then Feather" is the classic twin-engine memory aid: identify the failed engine by the rudder foot that is pushing (the foot pressing to maintain directional control is on the side of the operating engine, because you push away from the dead side), confirm with throttle/instrument crosscheck, then move that propeller control to feather. Some instructors phrase it as "Identify, Verify, Feather" — a three-step checklist that prevents feathering the wrong engine (a catastrophic mistake).
Common Test Traps
- Confusing Vmc with a climb speed: Vmc is purely a directional control speed — it defines the minimum speed at which you can maintain straight flight with the critical engine out. It has no climb requirement built in. You can be above Vmc and still descend.
- Assuming feathering is optional: Many students underestimate how severe windmilling-prop drag is. The published OEI performance data assumes a feathered prop; without feathering, real-world performance is far worse.
- Feathering the wrong engine: Always confirm which engine failed before feathering. A common mnemonic: reduce throttle on the suspected failed engine — if performance does not deteriorate further, that's the dead one. Feathering the good engine is fatal.
- Thinking Vmc is determined at max gross weight: Vmc is actually determined at the most unfavorable weight — which is the lightest weight. A lighter airplane generates less lift, so the rudder gets less help from the banked-lift vector, and Vmc is higher. Never say Vmc is set at max gross weight.
- Forgetting the centrifugal feathering lock: On the ground with a running engine at low RPM, the prop cannot feather due to the latch. Students sometimes expect feathering to work in all conditions — it will not below the latch-release RPM threshold.