There is more than one meaning or situation that can be addressed by the term “Windmilling in Reverse” and I hope to cover all of them in this article.
First, why would the feathered propeller of a PT6 engine that has been shut down in flight rotate backward – turn counterclockwise (CCW) as viewed from the pilot’s seat – instead of being stopped or rotating in the clockwise (CW), normal, direction? It was typical in the original three-bladed PT6s used on earlier King Airs that the propeller would indeed be stationary on a shutdown engine in flight. Since there is extremely low resistance to rotation – due to the fact that the input shaft to the engine’s Reduction Gearbox (RGB) is the free turbine that is not connected to the compressor and its accessories – even the smallest error in the feathered blade angle can cause rotation. Sometimes the misadjustment leads to normal, CW propeller rotation and sometimes it goes in the opposite direction, CCW. If the propeller blade angle doesn’t quite go far enough to streamline into the relative wind – close to a 90° blade angle, but with the actual number depending upon both the distance from the center and the twist with which the blade is designed – then the relative wind creates a force that tries to rotate the propeller in its normal direction.
On the other hand, if the propeller blade angle goes a bit past 90° now the relative wind leads to a CCW, backward rotation.
Although it is pleasing to have the propeller come to a dead stop when shutdown and feathered in flight – like it was on the piston twin trainers in which we learned – it is not at all uncommon to experience slight rotation in either the normal or the backward direction. Don’t worry about this being harmful to the engine … it’s not. Why? Because the oil supply and scavenge pumps are driven by the compressor’s rotation: N1 or Ng. Unless something jams the compressor, the ram air through the cowling and engine causes sufficient compressor rotation to supply a continuous supply of oil to the rotating propeller, its RGB and the power turbine’s shaft: N2 or Nf.
When Raisbeck Engineering developed the first four-blade propeller for the King Air 200 model – the excellent “Quiet Turbofan” propeller – it had a very pronounced twist designed into the propeller blades. As I have mentioned before, this prop doesn’t even come close to stopping rotation when feathered on a shutdown engine. Expect to see 10 to 20 RPM in the normal, CW direction … which is enough to create sufficient propeller oil pressure – from the pump inside the propeller governor – to allow the prop to unfeather itself if the propeller lever is not kept in the feathered position, keeping the path open for propeller oil to exit back into the engine. Even if your King Air has autofeather, be certain to complete the full shutdown procedure and move the prop lever manually into the fully aft, feather position!
The second situation of windmilling in reverse that I wish to present is more obtuse, by far. We all expect that a fixed pitch propeller will turn faster when subjected to either more power or more airspeed. Of course, that is correct and it is what a constant speed propeller governor is meant to overcome. Namely, whenever the propeller momentarily goes faster (i.e., it overspeeds), the governor sends the blades to a larger angle, a bigger bite. The bigger bite creates more rotational resistance, slows down the propeller, and restores the on-speed condition. Vice versa, when the propeller underspeeds, the governor decreases the blade angle to restore the on-speed condition.
As a mental imaging exercise only (Don’t try this at home, folks!), consider what happens to the propeller when we cut off the fuel to a PT6 in cruise flight. Momentarily, due to the total reduction of power, the propeller underspeeds and the governor flattens the blade angle. Because the propeller of a PT6 has so little resistance to rotation – since it is only connected to the RGB and the power turbine, as well as to three accessory drive pads – at any airspeed above about 140 KIAS, it can return to constant speed operation at any speed selected by the prop lever … even takeoff RPM! Only when the airspeed drops below 140 do the flattening propeller blades finally reach the Low Pitch Stop (LPS). Now the blade angle becomes fixed, cannot flatten any more, and indeed the propeller speed starts to slow down.
As I hope you realize, when the pilot moves the power lever into the Beta and Reverse ranges, he or she is moving the LPS to a progressively flatter angle. Ignore the tremendous drag that is being produced as we continue our mental imaging exercise and theorize what would happen if the blade angle went totally “flat.” Does it make sense that our windmilling tendency would be lost? To make it simpler, imagine blades with no twist … just flat boards replacing the actual propeller blades. It those boards were at a zero-degree blade angle, the relative wind would be producing lots of drag but no rotational tendency. It’d be like holding your hand out of the car window on the freeway with your palm facing forward. There is lots of force pushing your hand back but no tendency to make your arm move up or down. Now rotate your hand such that your thumb moves forward. Hand wants to rise, right? Vice versa, rotate your hand to move the thumb back a little and now the arm wants to move down. (Like me, you did do this in your parents’ car as a kid, didn’t you?)
I hope that the conclusion being reached is that a totally flat propeller blade has no windmilling tendency, same as a feathered blade. Given enough time, all prop rotation would stop.
To “prove” that this in merely a mental imaging exercise, keep in mind that oil pressure from the pump inside the propeller governor provides the force that overcomes the feathering springs’ efforts to send the blade angle to feather. What is the pump doing when rotation stops? Nothing! It’s not turning and hence is creating no additional oil pressure. Therefore, we could never really reach that totally flat blade angle position – the one that gives no rotational tendency – because there’d be no way to keep the blade at that position. It would always start leaking toward feather, and then rotation in the normal CW direction would begin again.
Why have I taken your time to read these past few paragraphs if it is an impossible situation to achieve? Because I want you to consider the effect of windmilling when (1) using reverse thrust after landing, and (2) when conducting LPS run-up tests.
Here we are on a short runway using the Maximum Reverse procedure: High Idle, prop levers fully forward on final, and lifting and pulling back as far as the power levers will go right at touchdown. We are now moving the LPS to its most negative blade angle position and simultaneously asking for about 85% compressor speed. As the blade angle becomes negative, the relative wind is now serving to resist, not aid, propeller rotation. The faster we are moving, the more resistance there is. This explains why it is common to see the propeller speed slightly increase while in maximum reverse as the airspeed slows and approaches 60 knots, the point at which we should be easing out of the reverse range so as to be at Ground Fine by 40 knots … to decrease the potential for propeller erosion and FOD (Foreign Object Damage). Proper power lever/engine rigging should be done on a run-up pad that has been swept clean of any debris. Now maximum reverse is obtained with zero airspeed, the only relative wind being what the prop wash itself provides. Do you see that the results observed during this check will not be identical to those observed at 80 knots after touchdown? Namely, the airspeed causes more resistance to propeller rotation and, hence, slightly less propeller speed. Observing different results on the maintenance run-up pad versus on the runway with significant airspeed is to be expected. It’s not cause for concern.
The “Flight Idle Torque Test” is how the pilots and mechanics can determine if the low pitch stop is set at the correct blade angle. It should be labeled the “Low Pitch Stop Test” and the Raisbeck Engineering Maintenance Manual does indeed use that name. For each different propeller – manufacturer, number of blades, designer – a chart exists that presents target torque to be achieved at a specified RPM. The existing Pressure Altitude and OAT are the variable parameters that the chart uses (see example below).
For an example, let’s consider the four-blade “Quiet Turbofan” Raisbeck-Hartzell propeller that is widely used on many 90-series airplanes. With the propeller levers fully forward – setting the primary governor to its maximum, takeoff, propeller speed – power is added to bring the propeller speed up to 1,800 RPM. Since we are not yet on the governor, still in an underspeed condition, the propeller blade angle is at the LPS setting. For standard Sea Level conditions, 15°C, the torque should now be 500 ft-lb with a tolerance of +20/-0 ft-lbs and within 20 ft-lbs of the other side.
Seeing torques of, say, 560 on the left side and 480 on the right side, tells us three things: First, the LPS is incorrect on both sides and needs adjustment. Second, the higher torque on the left side indicates that there is more resistance to rotation on that side. Why? Because the blades are taking too big of a bite of air … the LPS is at a larger blade angle, coarser than it should be. Likewise, the right side’s LPS is too fine, at too small of an angle. (To make this easier to understand, just consider how difficult it would be to spin a feathered propeller at 1,800 RPM. Why, we would likely hit the torque limit first! Vice versa, if the blade angle were totally flat, near zero degrees, it would spin quite rapidly very easily, with very low torque applied.) The third thing we can learn from this mis-match in “Flight Idle” torques is that we can anticipate a yaw to the right in the flare … more drag on that side when the propellers finally come off of the governors and hit their respective LPSs.
Quiz time: In addition to Pressure Altitude and OAT, what other variable factor will affect the LPS torque test results?
That’s right … wind! Based on the whole premise of this article, headwind in the run-up area helps the propeller to rotate. With that help, the flow of exhaust gases across the power turbine blades don’t have to do as much work, don’t have to deliver as much torque. Vice versa – a tailwind in the run-up area leads to that “Reverse Windmilling” tendency and acts to slow the propeller down. Thus, more torque will be required to reach the specified 1,800, RPM value.
And now you know why, when the wind is not calm that the Low Pitch Stop torque value must be the average of one reading taken while facing into the wind and one taken while facing downwind.
OK, OK, I hear you. Next month I will pick a topic that is not so technical! See you then.
I am very happy to report that “The King Air Book – Volume II” is now available! This new book, which I finished creating a couple of months ago, is a compilation of all of the articles that I have written for this magazine. They are indexed according to topic, making it easy to find your areas of interest. In addition, a digital copy that is searchable and in color is also available for ordering. Go to www.kingairacademy.com to find the links for both. I appreciate your support!