I recently had a conversation with a representative of one of the propeller manufacturers and it became apparent that he had a couple of misconceptions about the behavior of the King Air propeller both during in-flight shutdowns and airstarts. That prompted me to write this article in an attempt to provide clarification.
As we have discussed in some past articles, the propeller on a PT6 will feather itself after shutdown on the ramp due to lack of oil pressure. Safety is compromised by doing so, however, since the King Air propellers spin for a much longer time than when we feather them manually … that long time period of lethally spinning propellers makes it more likely for a passenger or line person to get injured or killed.
The more headwind there is on the ramp, the longer the propellers will spin, being driven by the wind. Well, when we shut an engine down in flight, there is a lot of headwind, right? This does two things of interest involving the propeller. First, the windmilling blast will keep the propeller turning. In fact, hard to believe at first, but at any airspeed above about 140 KIAS, the propeller can windmill as fast as takeoff RPM even with no fuel flow! Second, the cowling design provides ram air to the compressor inlet and the rotating power turbine (the one connected to the propeller) helps suck air through the engine from the compressor out through the exhaust stacks. This combination of push (from the ram air in the cowling) and pull (from the power turbine) keeps the compressor from ever stopping … yielding an N1 or Ng speed of 12% or more, depending upon altitude and airspeed.
So not only does the oil pump inside the propeller governor keep turning at high speed and hence providing normal propeller oil pressure, but also this pump keeps getting fed new oil supplied by the oil pressure pump in the accessory gearbox. Granted engine oil pressure will have dropped well out of the normal range since the pump is only turning at 12% speed or so, but that is still enough speed to get the oil from the tank to the propeller governor and also to scavenge the oil from the nose case back to the tank.
Unless we manually feather the propeller (dumping propeller oil into the nose case through the primary propeller governor) or have autofeather activate (dumping oil through the overspeed governor) the propeller will continue to windmill indefinitely following a fuel shut-off in flight.
You probably know how much I like autofeather, but if you are flying a model without it, expect significant windmilling propeller drag until you manually feather.
So what happens to propeller rotation when we feather? Well, duh, it stops, doesn’t it? Not always.
You see, unlike a piston engine, the free-turbine PT6 has almost no resistance to propeller rotation. So unless the feather blade angles are set exactly correct and we are flying without sideslip, it is not uncommon to see slight propeller rotation after feathering a shutdown engine in flight. Sometimes the propeller rotates in the forward direction, sometimes it goes backwards and, yes, sometimes it does completely stop. The Raisbeck-Hartzell “Quiet Turbofan” propellers almost always rotate quite aggressively in the normal direction when feathered … about 20 RPM. My friend James Raisbeck taught me years ago that his propellers have such a large amount of twist in them that airframe drag would actually be higher if he designed the feather angle such that they would truly stop.
Once the feathered propeller stops or at least almost stops, the sucking or pulling action of the power turbine that was helping move air through the engine then goes away. Without it, the only factor making the compressor rotate is the ram air supplied by the cowling. It is my experience that the N1 we see now will vary from roughly 3% to 10%. The lower numbers will be seen in level flight at typical single-engine airspeeds. The higher numbers will be seen only in a dive, with IAS above 200 knots. Also, the thicker air at lower altitudes provides more ram effect and tends to rotate the compressor faster. If the propeller had never been feathered, the windmilling compressor speeds will tend to be about 10% faster due to the sucking action of the power turbine. In other words, instead of a range of 3% to 10%, now we will see probably 13% to 20%.
Pop quiz question: What is the minimum N1 required before introducing fuel during an airstart?
Did you answer 12%? Most people do, but they’re wrong. Yes, 12% is stated in the POH as the minimum compressor speed needed before introducing fuel during a ground start. If we can get that much or more for an airstart, great. However, there is no minimum speed limit listed for an airstart. I believe the King Air 300-series’ POHs “encourage” obtaining at least 10%, but it is not an actual requirement.
Can you guess where I am going with this? Here it is: Can we do a windmilling (no starter assist) airstart with the propeller feathered? The answer is a qualified “Yes.”
Only in the 300-series – with their splendid performance and their pitot-cowls – is the title of the appropriate procedure “No Starter-Assisted Airstart (propeller feathered or windmilling).” For the other models, the title is “No Starter-Assisted Airstart (windmilling engine and propeller).” Getting the “encouraged” 10% N1 or more before introducing fuel is easy with a windmilling propeller but much more difficult with a feathered propeller. The windmilling airstart envelop for all King Air models is a minimum of 140 KIAS and below 20,000 feet. Of course, the higher the speed and the lower the altitude (within reason!), the more N1 speed we will obtain. Hence, diving at high speed down below 10,000 feet can yield just as much compressor speed with a feathered propeller than having, say, 150 KIAS at 17,000 feet with a windmilling propeller.
Folks, let’s be realistic here. Are you going to actually do this? Do an airstart with a feathered propeller? I bet not. First, how many times would you choose to restart an engine that you decided to shut down in flight, other than during a training session? Second, if the propeller had been feathered – as it almost certainly would have been when completing the shutdown procedure – why wouldn’t you just go ahead and use the starter? Third, only if the starter chose this particular time to become inoperative would there be any need for a windmilling airstart with a feathered propeller!
Even during flight training this is not a very desirable procedure to practice or demonstrate. Why? Because although the peak ITT will likely be well below the starting limit, it will be quite a bit higher than you are used to seeing. Why subject your hot section to that higher temperature?
So, for me, I prefer to always have a propeller windmilling at takeoff RPM while conducting a windmilling airstart, since it leads to very comfortable starting ITTs.
Now I’ll let you in on a little secret. Remember those Raisbeck-Hartzell propellers that rotate about 20 RPM when feathered? Or, even another propeller that happens to have its feather blade angle set such that it rotates in the forward direction? Well, move the propeller lever from feather to the full forward position and lower the nose to pick up some extra airspeed. See what is happening? The propeller is unfeathering itself! (You remembered to turn off the Autofeather switch as part of the inoperative engine cleanup procedure, didn’t you?)
No King Air has ever been built with unfeathering accumulators, a popular option on some piston twins, especially ones used for multi-engine training. Without that type of device, the only way in which oil pressure can be created and sent into the propeller dome to drive the propeller blades to lesser angles is to use the pump inside the primary propeller governor. Since that governor is driven by the propeller shaft, no propeller rotation means no oil pressure. Only with propeller rotation can we create the oil pressure necessary for unfeathering.
I have seen situations in which the forward rotation of the propeller was just too darn slow to accomplish the unfeathering. But if you have about 10 RPM or more, it will work. Just place the propeller lever forward, be patient, and soon you will observe the propeller is starting to rotate faster. And the faster it goes, the more oil pressure is created so it speeds up even more rapidly.
If you leave the propeller lever fully forward as the propeller increases its speed, it will eventually hit the speed at which the governor is operating. With the lever fully forward that is takeoff RPM, of course. There will be quite a bit of surging as the governor takes over and a lot of drag since, with no fuel flow yet, the blade angle will be quite small in order to reach takeoff RPM.
A technique I use is to pull the propeller lever back to where it is just touching, but not into, the feather detent once I see the propeller starting to increase speed. This sets the governor at the minimum possible speed. Now when the unfeathering propeller hits the governor, it will be at a lower RPM and with a bigger blade angle … yielding less drag and with the propeller speed never surging past redline.
With the propeller stabilized at minimum governing speed, slowly move the propeller lever fully forward to get as much airflow through the engine as possible before advancing the condition lever – the next step in the windmilling airstart checklist. As soon as N1 starts increasing and ITT starts to rise, we have verification that lightoff has occurred and it is a good time to start leveling off from the dive we have been in to get maximum airflow.
Speaking of the checklist, I believe that the airstart procedures – both no starter-assisted and starter-assisted – contain more steps than any other abnormal procedure. Since this is something done rarely, take your time and go through the checklist slowly and methodically. A benefit of doing the windmilling (not the starter-assisted) procedure is that no large electrical demands and voltage transients are experienced so there is no need to turn off things like EFIS tubes, windshield heat, Cabin Temp Mode selector, etc. But make sure you remember to move the auto-ignition switch to “Arm” when it is specified, or the fuel and air mixture has no source of ignition!
So, Mr. Propeller-Manufacturer-Service-Tech, please realize that (1) a PT6 propeller will never feather itself in flight, it must have a pilot or the autofeather system to do so, and (2) it will never unfeather itself in flight until propeller rotation occurs.
This article was republished from the August 2013 issue.