When one engine fails on a light twin, the pilot faces two immediate problems: the loss of roughly half the total thrust and a dramatic increase in asymmetric drag. What happens to the failed engine's propeller determines how severe both of those problems become. A propeller spinning freely in the slipstream — called a windmilling propeller — generates enormous drag and can make the difference between a controlled climb and an uncontrolled descent. A feathered propeller, with its blades rotated edge-on into the relative wind, reduces that drag to a fraction of the windmilling value and is central to any realistic one-engine-inoperative (OEI) performance discussion.
This article explains the aerodynamic mechanics of both states, quantifies the performance difference, and connects the concept to the practical skills and examiner questions that appear on multi-engine checkrides and written tests.
How a Windmilling Propeller Creates Drag
Under normal powered flight, a propeller blade operates at a positive angle of attack relative to the local airflow, producing thrust. When an engine fails and the throttle is not cut, the propeller continues to rotate — but now the airflow drives the rotation rather than the engine driving it. This is the windmilling condition. The blades are at a large negative angle of attack relative to the oncoming air, and instead of producing thrust they produce a large rearward aerodynamic force: drag.
The amount of drag is substantial. A windmilling propeller on a light twin can generate as much parasite drag as the entire airframe of the aircraft. FAA-H-8083-3C Chapter 13 emphasizes that a windmilling propeller creates significantly more drag than a stopped, unfeathered propeller, and far more than a feathered propeller. The exact drag depends on the blade area, pitch, and airspeed, but the conceptual ranking is consistent: feathered propeller produces the least drag; a stopped flat-pitch propeller produces more; a windmilling propeller produces the most.
What Feathering Accomplishes
A feathering propeller system allows the blade pitch to be rotated to approximately 90° of pitch (blade chord nearly parallel to the direction of flight). In this position the blade presents its thin leading edge to the relative wind rather than its broad face. The result is a dramatic reduction in aerodynamic drag — roughly equivalent to a streamlined fixed object of the same frontal area. The propeller also stops rotating, eliminating the gyroscopic and P-factor moments that a spinning propeller contributes.
The aerodynamic benefit is not subtle. Published OEI climb performance data in a Pilot's Operating Handbook (POH) almost always assumes the failed-engine propeller is feathered. If you attempt to use that data with a windmilling propeller — because the engine failure caught you below the altitude where you could feather, or because the feathering system malfunctioned — you will find actual climb performance significantly worse, potentially zero or negative even when the POH predicts a positive climb rate.
Performance Numbers: The Practical Impact
Consider a typical light twin with a published Vyse (blue line) OEI climb rate of 200–300 feet per minute at sea level, maximum gross weight, and standard day. That figure assumes the failed engine's propeller is feathered. With the propeller windmilling, the same aircraft may climb at less than 50 feet per minute, or may not climb at all. At higher density altitudes the situation becomes more extreme: an aircraft that can barely maintain altitude with a feathered propeller may descend at several hundred feet per minute with a windmilling propeller.
This is why the first items in any multi-engine emergency procedure after identifying the failed engine are the memory items leading to feathering: mixture to rich (or auto-rich), propeller to high RPM (to increase oil pressure and aid feathering), and then, after confirming the failed engine, condition lever or propeller control to feather. Speed must be at or above Vyse — the best single-engine rate-of-climb speed shown by the blue line on the airspeed indicator — to give the remaining engine the most efficient operating condition and to maximize OEI climb rate.
The Drag Comparison in Context: Vmc and Control
The drag difference between a windmilling and a feathered propeller also affects directional control. Vmc — the minimum control speed with the critical engine inoperative — is published by the manufacturer and must be demonstrated to not exceed a regulatory maximum under 14 CFR Part 23. The published Vmc is determined with the critical engine's propeller windmilling, not feathered. This is deliberately the worst-case condition, because the windmilling propeller adds asymmetric drag on the failed-engine side in addition to the thrust asymmetry, which together create a larger yawing moment that the rudder must overcome.
On a conventional light twin where both propellers rotate clockwise as viewed from the pilot's seat, the left engine is the critical engine. Its failure leaves the right engine operating with its descending blade — which generates more thrust due to P-factor — positioned farthest from the aircraft centerline, creating the greatest possible yawing moment. Because the windmilling propeller on the left engine adds asymmetric drag, Vmc in the windmilling condition is higher than it would be with a feathered propeller on the failed engine side. Once the left engine propeller is feathered, the effective Vmc for continued flight is somewhat lower, meaning the pilot has a slightly larger safety margin above the actual control limit.
Why It Matters Operationally
The feathered-versus-windmilling distinction has direct consequences during takeoff, initial climb, and en-route cruise after an engine failure.
- Takeoff and initial climb: If an engine fails at or near liftoff, airspeed may be close to Vmc and the altitude is insufficient to feather and recover. The immediate priority is directional control and preventing a departure below Vmc by reducing power on the operating engine if necessary, then establishing Vyse and climbing before attempting to secure the failed engine.
- Accelerate-stop versus accelerate-go decisions: The increased drag of a windmilling propeller makes it harder to climb. Pre-planned decision speeds (commonly called accelerate-stop and accelerate-go speeds) must account for whether OEI climb data assumes a feathered propeller. If you cannot feather before those performance calculations matter, the actual terrain-clearance margin is smaller than the POH implies.
- En-route single-engine service ceiling: The single-engine service ceiling listed in the POH (the altitude at which OEI climb rate falls to 50 feet per minute) is computed with the failed propeller feathered. At the same altitude with a windmilling propeller the aircraft may already be at or below its performance ceiling.
- Zero sideslip technique: For best OEI climb performance, the pilot should bank approximately 2° toward the operating engine and coordinate with rudder to eliminate sideslip. This technique — sometimes called zero sideslip — gives slightly better performance than a wings-level, ball-centered approach because it reduces fuselage drag while still generating a small lateral component of lift that unloads the rudder. This benefit, though real, is secondary to having the failed engine's propeller feathered.
Key Numbers and Rules
- A windmilling propeller can produce drag approximately equal to or greater than the rest of the airframe combined on many light twins.
- Published OEI climb performance in the POH assumes the failed engine propeller is feathered.
- Vmc is established with the critical engine propeller windmilling — the worst-case drag condition.
- Vyse (blue line) is the speed that maximizes OEI rate of climb; it must be maintained to achieve POH-published climb data.
- The critical engine on a conventional light twin (both props turning clockwise from the pilot's view) is the left engine.
- Banking up to 5° into the operating engine reduces Vmc by as much as approximately 3 knots per degree of bank; zero bank (wings level) raises Vmc; greater than 5° bank is not permitted in Vmc determination.
- Below Vmc, the immediate corrective action is to reduce power on the operating engine and lower the nose to accelerate — not to use more rudder.
Common Test Traps
- Assuming POH climb data applies with a windmilling propeller. It does not. Published OEI figures assume a feathered prop; a windmilling prop will dramatically degrade actual performance.
- Confusing Vmc conditions. Vmc is published with the propeller windmilling, not feathered. Feathering after engine failure actually reduces the effective asymmetric yawing moment, but Vmc is a certification number tied to the worst-case (windmilling) scenario.
- Thinking the right engine is critical on a conventional twin. The left engine is critical on a standard light twin with both propellers turning clockwise. The right engine's descending blade has the longer moment arm; losing the left engine leaves the right engine to create maximum yaw.
- Believing heavier weight makes Vmc worse. Vmc is most adverse (highest) at light weight. A heavier aircraft's larger lift vector, combined with up to 5° of bank toward the operating engine, provides more lateral force to assist the rudder, effectively lowering Vmc. The Vmc certification standard uses the most unfavorable (lightest) weight.
- Treating power reduction below Vmc as a last resort. Reducing power on the operating engine when below Vmc is the primary, immediate corrective action — it removes the asymmetric thrust that is causing loss of directional control. Waiting too long to do this is a fatal mistake.