Sunday, March 06, 2011

No Stalling

No, I'm not slacking off or stalling by blogging. Telling other people how a system works is one of my favourite ways of integrating knowledge. And stall protection is an important system on this airplane. Aviation people know what a "stall" is and non-aviation people can probably follow this if I say merely that a stall is a loss of aerodynamic lift experienced when an airplane is not getting the right airflow over its wings, typically a result of going too slowly. Its only relation to the locomotion problem of the same name experienced in an automobile (say if there's a problem with the engine or if the driver is out of practice operating a manual transmission) is that the vehicle in question stops going forward. This is especially bad in an airplane because then it goes down. This is a gross simplification to get to the topic. For more information on the physics of the stall, see your local aerodynamics textbook or the comment pedants below. If you are bored or baffled by the technical aircraft information in this and the following posts, recall that the other option would be no posts at all while I am busy.

The Screaming Whippet has a multistage stall prevention system. Indications to the pilot of the possibility of a stall include an angle of attack indicator with coloured sectors (green for cruise, white for approach, yellow for slow, black and red striped for the warning area and red for stall), an audible stall horn, a stick pusher, and a plethora of annunciator lights for the various states of the system.

The system gets its inputs from an angle of attack vane (which may be either a paddle that is pivoted so that it is free to align with the airflow against the side of the left forward fuselage or a motoring pitot tube-like arrangement on the outboard right wing leading edge), a flap position transmitter (the same speed may be safe or not, depending on flap position), an airspeed limit switch in the copilot's pitot system, a squat switch inboard on the left main gear (it is not a danger to be going slowly while on the ground), and a test circuit. The first three inputs go to a computer which uses them to calculate the ratio of V/Vs (current speed to stall speed) and output it to the pointer on the AoA indicator. At approximately 1.1 Vs, or 5-10 kts above stall, a warning horn will sound. At stall speed (from 1 knot below to 4 knots above) a stick pusher will apply 60 lbs of forward force to the control column, commanding the airplane into a nose-down attitude, the first step in stall recovery.

The motive force for the stick pusher comes from a loud servo motor (everything on the Whippet screams) for which there is a slip clutch, so that it can be overpowered by the pilot if need be, and a magnetic clutch to enable it to give a varying amount of pulling force. It's attached to the elevator actuator beam, which is linked directly to the yoke my cables. Its engagement serves to both jolt the nose down and jolt the pilot out of whatever stupidity caused her to almost stall the aircraft. The servo is not powered unless the squat switch indicates that there is no weight on the wheels, the airspeed limit switch indicates that the airspeed is below 140 (+/-5) kts, and the emergency override switch has not been set to disengage the clutch. So if the airplane is flying, but below 140 kts, that switch should be on and an annunciator light should indicate that the system is armed.

If something is not working, the system tells me about that too. The warning light flashes if the servo has stopped or the clutch is disengaged, and comes on steady if the computer fails. (It's steady if the servo and/or clutch has failed as well as the computer).

The system must be tested before flight, so there is a test function. It consists of a three position switch, spring-loaded to the centre off position. The other tow positions connects the servo to power and respectively give the computer stall and cruise inputs. The indicator, warning horn and stick pusher should all respond as at a stall to the first position, and the indicator should show 1.3 Vs, with no horn or pusher at the cruise position. It's important to ensure that the gust lock system is disengaged before testing the stick pusher, and also to hold the control column firmly to prevent the stick pusher from smashing any instruments.

We were told in class that if the test mode was unserviceable, we could still test it on the ground by manually moving the external AoA vane, from the stalled position to the cruise position, while having someone manually depress the squat switch. (I know it sounds odd that a squat switch could be manually bypassed without lifting the airplane, but this is an odd design where lifting the weight off the wheels depresses the switch and placing weight on the wheels unpresses it).

Also, this hotel room is freezing. What is this? Winter in Canada?

And in isn't it supposed to only be in movies that you can escape the passenger cabin through an aircraft toilet?


Ihab Awad said...

Hm. I'm surprised that this system would abruptly apply down elevator based on indirect evidence of stall (speed and other inputs), rather than directly detect flow separation by aerodynamic means. I guess I'm wondering if there are situations which would be survivable but for the "safety" system tucking the nose into the ground.

amulbunny's random thoughts said...

I was watching Flying Wild Alaska last night and part of the show was talking about stalling. I didn't realize that it meant the engine was working but the airplane wasn't in the right attitude.

I have a lot of respect for you flying into those little gravel runways now after watching this show.

zb said...

Nice. A technical post again! And the whippet obviously has the best of both worlds: Cables (i.e. mechanical cables) and computers. Also, I like that you are not only giving typical numbers, but ranges (-1 ... 4 kts). Coming to think of it, the computers might even be analogue. It is amazing how long electronic stuff lives when it's not for the consumer market, but part of a larger, very expensive system like a plane.

Sarah said...

I love technical posts. Go deeper!

Ihab Awad: I'm surprised that this system would abruptly apply down elevator based on indirect evidence of stall (speed and other inputs), rather than directly detect flow separation by aerodynamic means.

But AoA is a direct aerodynamic measure. A wing will stall beyond the critical angle at any weight or airspeed. That's the beauty of it, and why a rudimentary stall warning vane is found on the simplest power airplanes.

There is a clutch for override of the stick-pusher by the pilot. Unfortunately, override is sometimes incorrect ( c.f. Colgan 3407 and many others. )


D.B. said...

That's a clue for those of use trying to work out what a "Screaming Whippet" might be. I'm guessing it's a "T-tail" aircraft, one where the horizontal tail surfaces are mounted at the top of the tail. The clue is based how very much the designers want the aircraft to avoid entering a stall - aircraft with T-tails can get into an unrecoverable "deep-stall" where the turbulent airflow coming from the wings blankets the elevator, so that the normal stall recovery technique of pushing the nose down to regain airspeed won't work. The prototype BAC 1-11 was lost for that reason, as was the prototype TU-134 and later a HS Trident taking off from Heathrow. As a result, T-Tail jet aircraft with swept wings all have either a stick pusher or a stick shaker, or both, which alert the pilot to an impending stall, and in some cases take the corrective action automatically.

Is there an override? At very low altitudes, it could be preferable to let the deep stall develop, rather than nose over.

I tried to avoid using my aeronautical engineering degree, so as not to be a "pedant" on exactly why a wing stalls.....

Aviatrix said...

D.B.: There is an override in two senses. The slip clutch allows the pilot to overpower the system if it activates in error and the entire system can be turned off with the flick of a switch.

Plus you are all WELCOME to expand on the topics I skip. If you'd rather do a guest post on the aerodynamics of the stall, e-mail it to me and then I can link to it every time I need to do a rant on the news media not knowing what a stall is.

Ihab Awad said...

Sarah: Got it, thanks!

Anonymous said...

Rudder Pedal Potentiometer and Nose Gear Follow-up Potentiometer Position." (Why yes, this training manual is apparently so old that it predates the late eighteenth century English move to distance itself from its Germanic roots and not capitalize all nouns).

Why not at all. It is obvious that 30 pages later you will be confronted with the term RPPNGFPP which you will immediately recognize as Rudder Pedal Potentiometer and Nose Gear Follow-up Potentiometer Position.

They have provided you with the acronym definition using the capitalization method.