Feathered vs Windmilling Prop

[chuckles1225]

Does anyone have a good physics-based description of why so much more drag is produced by a windmilling propeller than a feathered one? I have heard many illogical descriptions but never a fundamentally sound one that I can feel comfortable teaching my ME students.

For instance, I was taught that a windmilling propeller is comparable to a big circular piece of ply-wood out on the spinner with the diameter of the prop, but this seems incorrect because the surface area of the prop doesn't really change when it is spinning vs. stationary. This leads me to believe that there is something going on aerodynamically.

To take this one step farther, what about the free-spinning prop on a PT-6? Let's say there is a constant 20 knot wind blowing straight off the nose of a King Air while it is sitting shut down on the ground. If the blades were feathered and not spinning and I took an anemometer reading directly behind the prop it would obviously read 20 knots because it is not affected by the blades at all. But what if I gave the blades some pitch and they began to windmill and rotate at several hundred RPM. If I took another anemometer reading directly behind the spinning prop would it still read 20 knots? Would it read less because of the ply-wood hypothesis, or would it read more because the spinning prop is essentially a wing, accelerating air downward (backward in the case of a prop) as it spins?

 

[PFGiardino]

Forget the plywood.

Have you turned over a piston prop by hand before? You've got to put some muscle into it. The engine has resistance. When you let your propeller windmill, you're giving away some of your forward energy to turn the crank, pistons, accesories, etc. That forward energy could have been used to give you a farther glide. In the multi-world, it means a lower Vyse/better single engine climb. All of these are important for people trying to stave early death.

 

[tgrayson]

Does anyone have a good physics-based description of why so much more drag is produced by a windmilling propeller than a feathered one?

Windmilling props generate reverse thrust. The faster they windmill, the greater the reverse thrust. I haven't done any diagrams on this, but I think a thought experiment can show why this is true. Consider a propeller generating positive thrust and having the engine quit:

1. The drag on the propeller, which, by definition, acts opposite to the propeller rotation, will slow the propeller.
2. As the propeller slows, the AoA seen by the propeller get smaller, generating smaller thrust.
3. Eventually, the propeller slows to the point where the AoA is zero, but the propeller is still rotating.
4. Since there is still drag opposing the rotation of the propeller, it slows further, which moves the AoA into the negative range, since the relative wind is now in front of the propeller.
5. Since the aerodynamic force produced by an airfoil is perpendicular to the airflow causing the force, you suddenly have a component of this force acting in the same direction as the propeller rotation, and a component acting opposite the flight path of the aircraft, which makes it drag.
6. The component of the aerodynamic force of the propeller that acts in the direction of rotation is not sufficient to counteract the drag of the propeller opposite the direction of rotation, however, so it continues to slow.
7. As the propeller continues to slow, the negative AoA gets larger and larger and the force in the direction of propeller rotation gets stronger and stronger, and it eventually is strong enough to exactly neutralize the drag opposing the rotation of the propeller and the rotational speed reaches equilibrium.
8. At the same time, as the negative AoA gets bigger and bigger, and the propeller rotation slows, the component of reverse thrust opposite the flight path of the airplane gets smaller, meaning that the lower the RPM, the lower the drag.

Internal engine drag, since it makes the equilibrium RPM smaller, actually works to reduce drag, rather than increase it.

 

[PFGiardino]

TGrayson, thank you for your correction. It seems I've taken another unverified former CFI bit of "wisdom" for granted with out doing my dilligence. That's good info, but I'm still working though the mental picture of the force acting in the same direction of rotation.

OP... listen to him.

 

[tgrayson]

unverified former CFI bit of "wisdom" f

I probably taught the same thing until someone asked a question about it and I sat down and built an Excel spreadsheet to explore the issue. Like so many things, once you know the answer, lots of things click into place. For instance, when helicopters autorotate, they're using negative thrust to slow the descent rate. One thing they do to increase the negative thrust is disengage the rotor from the engine to allow it to windmill. Clearly, the drag is greater with the rotor not having to drag the engine along with it, which jibes with what my calculations showed.

As a reality check, one aerodynamics book I have throws out an empirical formula to estimate the drag:

D = (0.1/0.5(RPM_windmilling/RPM_rated)*(Power_rated*Velocity)
​

You can see the higher the windmilling RPM, the higher the drag. (source: Fluid Dynamic Drag, Hoerner, p. 13-22)

As for visualizing it, keep in mind that the AoA seen by the propeller is produced by the vector sum of the rotational velocity and the forward velocity of the airplane. As the rotational velocity decays, the forward velocity becomes more pronounced in comparison to the rotational velocity and the AoA begins to shrink.

 

[chuckles1225]

Very enlightening explanation tgrayson, thanks.

 

[shdw]

It seems this greatly overcomplicates a simple surface area versus drag issue. Sure AOA makes the amount of drag more from reverse thrust, but I question the helpfulness of such an in depth explanation to a pilot.

I would simplify it: Take just a plain old round stick and spin it, the faster the rotation the the more air is prevented from passing through the plane of rotation. The object, because of its rotation, presents more surface area to cause drag and slow us down. Now a prop, unlike this round stick, is a wing and thus factors such as AOA attack also play a roll. If you want to know more you can study how relative wind changes the AOA of a windmilling prop.



Tgray: I have some issues with your "negative thrust to slow the descent rate."

For a heli, positive thrust is thrust directed towards the ground, to lift them up. Negative thrust would be directed skyward, increasing descent rate.

Autorotation, the rotor is disengaged from the engine and rotor pitch is flattened to increase head speed. Then, just before slamming into the ground, the rotor pitch is increased. The result is positive thrust directed towards the ground, stopping the descent.

 

[tgrayson]

The object, because of its rotation, presents more surface area to cause drag and slow us down.

Except that this is horribly wrong. To flip it around, the space between the blades also moves rapidly around in a circle, so shouldn't the whole propeller become invisible the faster it moves?

Negative thrust would be directed skyward, increasing descent rate.

I'm defining negative thrust as thrust opposite the flight path.

Then, just before slamming into the ground, the rotor pitch is increased.

Relevance?

 

[chuckles1225]

The object, because of its rotation, presents more surface area to cause drag and slow us down.

This goes back to the old ply wood cut-out explanation. Why would rotation increase the surface area of an object? It would seem that if you took a snapshot of the rotating object at any point in time it would have the exact same surface area as when it is stationary. Is there a difference between actual surface area and effective surface area that you may be referring to?

 

[tgrayson]

Why would rotation increase the surface area of an object?

It doesn't. It has some intuitive appeal because, visually, we see a round disk appear before our eyes as the prop begins to spin. The logic being that something that looks solid to our eyes must be solid to the air.

The explanation doesn't survive casual inspection, as you noticed.

 

[shdw]

Except that this is horribly wrong. To flip it around, the space between the blades also moves rapidly around in a circle, so shouldn't the whole propeller become invisible the faster it moves?

Damn you why didn't you say something a couple months back when this was presented by someone else on a multiengine unfeathered prop drag question! You could have saved me from making a fool of myself here.

On a side note, do you think that explanation presents any negative transfer of learning to a pilot? I won't be using it anymore when I do get my MEI, but still curious your thoughts on that.


The relevance for the other comment was just to finish the explanation of the maneuver, explaining that positive thrust is what slows/stops descent, not negative thrust. I didn't realize you were using a different definition at the time though.

 

[tgrayson]

why didn't you say something a couple months back when this was presented by someone else on a multiengine unfeathered prop drag question!

I'm not sure if I saw it, but in discussion like that, enough problematic concepts pop up that it isn't feasible to challenge each one. Here's a thread similar to this one that's two years old and I pointed out the flaw in the flat plate explanation:

http://forums.jetcareers.com/cfi-corner/37467-windmilling-vs-stationary-prop-drag.html

On a side note, do you think that explanation presents any negative transfer of learning to a pilot? I won't be using it anymore when I do get my MEI, but still curious your thoughts on that.

I think that all false beliefs end up producing contradiction, if a person continues to be presented new information. What I suspect happens, though, is when a person encounters new information that requires giving up knowledge that he *thinks* he has, many people will choose not to accept the new information. This effectively shuts off that person's acquisition of knowledge. I've seen this process happen a lot, which is why a person's foundations often need to be restructured before you can build on them.

 

[tgrayson]

FYI:

I went looking through the archives I had of another site that was big in the 90s. A physicist, the author of the site Aviation Formulary, had done an analysis that showed that a windmilling prop wasn't anywhere close to the drag of a flat plate. I've got the analysis somewhere, but he provided this summary:

According to Skip Smith's "Illustrated Guide to Aerodynamics", a Cessna 182 has an "equivalent flat plate area" of 5.27 sq ft. In other words, the parasite drag of the entire airframe is the same as that of a 5.27 sq ft flat plate. The prop diameter of a 182 is 84", giving a swept area of 38.5 sq ft.

Were it true that the windmilling prop had the drag of a flat plate of its swept area, the drag of the 182 would increase by a factor of more than 8 when its prop is windmilling. This would be very noticeable if it were true!

This drag increases with the windmilling RPM, so you can improve glide ratio noticeably by pulling the prop control to low RPM. If you had a failure mode in which the prop was decoupled from the engine and the fine pitch stop on the governor failed to limit the prop RPM- on the way to prop failure, you'd get very high drag [emphasis mine], perhaps approaching that of a flat plate- but that isn't the scenario normally under consideration.

An autorotating helicopter blade likewise creates lots of drag - but it's critical the rotor RPM not be allowed to decay.

​

So his conclusion matches mine, for what that's worth.

 

[MikeD]

This drag increases with the windmilling RPM, so you can improve glide ratio noticeably by pulling the prop control to low RPM. If you had a failure mode in which the prop was decoupled from the engine and the fine pitch stop on the governor failed to limit the prop RPM- on the way to prop failure, you'd get very high drag [emphasis mine], perhaps approaching that of a flat plate- but that isn't the scenario normally under consideration.

Which is what I was always taught to do and have done with an engine out on an S/E that I knew I didn't have a chance of restarting.

An autorotating helicopter blade likewise creates lots of drag - but it's critical the rotor RPM not be allowed to decay.

.

In fact, your goal is to build and maintain rotor speed in the autorotation without overspeeding it.

 

[SHayes]

An autorotating helicopter blade likewise creates lots of drag - but it's critical the rotor RPM not be allowed to decay.

In fact, your goal is to build and maintain rotor speed in the autorotation without overspeeding it.

You are correct except that drag is not the only thing being produced by free spinning rotor blades. Lift and more importantly ENERGY is being produced also...unlike a windmilling prop on an airplane.

During vertical autorotation, the rotor disc is divided into three regions—the driven region, the driving region, and the stall region. The size of these regions vary with the blade pitch, rate of descent, and rotor rpm. When changing autorotative rpm, blade pitch, or rate of descent, the size of the regions change in relation to each other.

The driven region, also called the propeller region, is the region at the end of the blades. Normally, it consists of about 30 percent of the radius. It is the driven region that produces the most drag. The overall result is a deceleration in the rotation of the blade.

The driving region, or autorotative region, normally lies between 25 to 70 percent of the blade radius, which produces the forces needed to turn the blades during autorotation. Total aerodynamic force in the driving region is inclined slightly forward of the axis of rotation, producing a continual acceleration force. This inclination supplies thrust, which tends to accelerate the rotation of the blade. Driving region size varies with blade pitch setting, rate of descent, and rotor rpm.

The inner 25 percent of the rotor blade is referred to as the stall region and operates above its maximum angle of attack (stall angle) causing drag which tends to slow rotation of the blade. A constant rotor rpm is achieved by adjusting the collective pitch so blade acceleration forces from the driving region are balanced with the deceleration forces from the driven and stall regions.

By controlling the size of the driving region, the pilot can adjust autorotative rpm. For example, if the collective pitch is raised, the pitch angle increases in all regions. This causes the point of equilibrium to move inboard along the blade’s span, thus increasing the size of the driven region. The stall region also becomes larger while the driving region becomes smaller. Reducing the size of the driving region causes the acceleration force of the driving region and rpm to decrease.

 

[SHayes]

that I can feel comfortable teaching my ME students.

Remember the original question guys. He wants to explain this to his STUDENT. You are getting way too complicated for a student, hell I have a good amount of time as a MEI and I have trouble following you.

PFG was on track when it came to engine resistance. Add to that the fact of the surface area of the blades (NOT the propeller disc) that is hitting perpendicular to the wind. LOTS of drag.

Now look at a prop when it is feathered. It is not in the same position as when you shut the aircraft down on the ground. The feathered blades will "feather" skinny side to the relative wind. There is no little surface area it will not even try to turn the engine over...lots LESS drag.

 

[tgrayson]

During vertical autorotation, the rotor disc is divided into three regions—the driven region, the driving region, and the stall region....

Hmmm, verbatim cut and paste from the Rotorcraft Flying Handbook.

 

[tgrayson]

You are getting way too complicated for a student, hell I have a good amount of time as a MEI and I have trouble following you.

Not at all too complicated for a student, as long as the CFI understands it. Being a CFI is about the communication of knowledge. Knowledge means truth, and, if you don't know the truth, you should make it your business to learn it. If you don't care about the truth, then you should find another means of employment.

PFG was on track when it came to engine resistance.

No, he's not, for the reasons I explained. Note, too, that the rotational resistance is perpendicular to the flight path of the aircraft, not parallel, so it doesn't even operate in the proper direction.

Add to that the fact of the surface area of the blades (NOT the propeller disc) that is hitting perpendicular to the wind. LOTS of drag.

There is some drag there, but we know that when the blades stop spinning, the drag goes way down.

Now look at a prop when it is feathered. It is not in the same position as when you shut the aircraft down on the ground. The feathered blades will "feather" skinny side to the relative wind. There is no little surface area it will not even try to turn the engine over...lots LESS drag.

There is indeed less drag, but the primary source of drag reduction is that the blades will not be driven by the relative wind and hence will not develop reverse thrust.

 

[shdw]

There is some drag there, but we know that when the blades stop spinning, the drag goes way down, even if the propeller isn't in a feathered position.

I added the bold.

 

[N18040]

Forget the plywood.

Have you turned over a piston prop by hand before? You've got to put some muscle into it. The engine has resistance. When you let your propeller windmill, you're giving away some of your forward energy to turn the crank, pistons, accesories, etc. That forward energy could have been used to give you a farther glide. In the multi-world, it means a lower Vyse/better single engine climb. All of these are important for people trying to stave early death.

I could see engine resistance adding some drag in a recip, but how would you explain that for a turbine such as a PT-6. It has very little resistance to rotation, so if that was the only thing giving more drag, then there would be a very small change in drag between having the prop windmilling and feathered.