Thanks for your reply. To be honest, I hoped my question would generate something similar to what you wrote. Maybe I can learn something here.
It appears that the DFDR did not have the facility to record control column forces as opposed to positions (thats a relatively recent requirement for FDRs). So no-one is ever going to meet your apparent standards of proof that the control column was "pulled". Maybe it moved full aft of it's own accord? But the only logical conclusion is that the pilot flying the aircraft was moving the control column. Especially since it did the opposite of what the stick pusher would have had it do if there were no pilot intervention.
First - let me state clearly that I am not an expert on interpreting DFDR traces. Keep that in mind. Additionally, I am not rated in the Q-400; keep that in mind as well.
As you point out we are not recording 'forces' we are recording positions. However, I see readouts in lbs on the side of many relevant traces. Here is my thinking - please shoot it down where it goes wrong.
A) I am not interested in what happened after the stick pusher activated. I accept the premise that the PF may have been fighting the pusher but, from my perspective, the upset had already occurred at that point. By the time the pusher triggered, for all practical purposes this accident had already occurred. My interest is what happened before pusher activation. Whatever that was it is what, in my opinion, caused the upset and stall that activated the pusher.
Some things we believe that I do not dispute.
a) the REF SPEED switch (whatever its real name may be) was turned on. It reprograms the shaker and the pusher to activate at a higher airspeed.
b) the aircraft was being flown by the autopilot. It descended and leveled on the autopilot. Additional power was not added by the flight crew when the autopilot leveled the aircraft. As a result the airspeed continued to dcay, the pitch and AOA to increase, until the stick shaker activated and the auto-pilot disconnected.
c) at that point, one of the pilots added maximum power to the engines - within one second.
d) the addition of max power following shaker activation is a standard response and does not indicate suspicion of a tail stall.
Something I don’t believe and do dispute: The pilot intentionally yanked back on the control column as soon as the shaker went off and the autopilot disconnected.
Now let's go back to the NTSB report, figure A2.1 on pg 18 and look at the various traces – in all of the figures.
[Note: We can tell when a force is moving the elevators and the direction of movement. We can also tell when the control column moves and the direction of movement. But, we cannot tell which pilot is actually applying that force/movement (the reading comes form paired transducers). We also cannot tell what moved the control column - we only know that it moved and the direction of movement.
It is reasonable to assume that it is the pilot in the left seat since that was the PF. However it could have been either pilot or both pilots. There is an F/O in the cockpit that does things independent of coordination with the captain, let alone command (“I put the flaps up”).]
My question is this: It may be reasonable to assume that the PF flying moved the control column aft. But, is it logical to assume that the PF intentionally pulled back on the yoke when the shaker first went off and never relaxed that input? I do not think it is logical. Is it possible that something else could have produced the aft movement recorded by the transducers on the control columns? If so, what would that be?
I note that the power levers are reduced to an angle of 42 deg (I’m told by the report that 35deg is the equivalent of flight idle). I can see the auto-pilot engage and the a/c level at 2400 pa in alt hold mode. I note the airspeed, which continues to decline progressively and I also note that the power lever position does not change – until the shaker activates. [During this phase the crew is not fully aware of what the aircraft is doing.]
I also note that the elevator trim is moving nose up progressively as the airspeed continues to decrease. I see the flaps selected and extending to 10 deg. Further airspeed decay and further nose up elevator trim (which appears to reach the maximum). I see the AOA increasing – pitch attitude increasing to about +7 deg. then +9 deg. I see the shaker activate and the autopilot disconnect (at the same time). I then see power lever angle increase to about 70 deg. and pitch increase to about +20 deg and continue to about +30 deg. I do not see any change in elevator trim. As the pitch increases the control column moves aft. Very shortly thereafter the aircraft stalls, rolls left and the pusher activates. The upset is in progress. I then see the pilot(s) fighting the controls with extreme aileron and rudder inputs. The pitch goes to 0 at some point but very briefly. Shortly thereafter the pusher activates for the second time and the rest is history.
Observation/conjecture: Until the shaker activated and the autopilot disconnected – the wing was flying and the aircraft was NOT stalled. The pitch up following power increase resulted in the stall and pusher activation.
This is a high-wing aircraft with a great deal of available power and large propellers. According to the manufacturer it is not susceptible to tail stalls. I also do not believe that either of the pilots thought they were in a tail stall. Had they known enough to think that and react by applying back pressure – they would also have known enough not to add power. I rule out the tail stall theory as a source of the apparent initial control movement.
The elevator is hydraulically actuated. There are no trim tabs on the elevator. Therefore when it is “trimmed” (which is electrically controlled) either by the trim switches on the yoke or through the auto pilot, the actuators displace the entire elevator [similar to a trimable stabilizer]. The elevator will stay in that position until the trim is changed [it never was]. A large increase in power – which took place immediately after shaker activation – will raise the nose. If the elevator is already trimmed at or close to the maximum nose up [see the FDR trace], it seems that the addition of near maximum power will help to raise the nose quite rapidly and much farther. A lot of forward pressure (not anticipated) would be required to stop it. If the pilot never touched the control column the resulting pitch up moment is highly likely to exceed the critical AOA and stall the wing [with no applied back pressure].
It is pretty obvious that neither pilot realized that the airspeed was decaying rapidly, due to insufficient power to sustain level flight. The autopilot was trimming the nose up rapidly (there is no trim wheel to observe) and the aircraft was very close to the artificially augmented stall speed increased by the REF SPEED switch in the INCR position. When the shaker activated and the auto-pilot disconnected they were most likely totally surprised. They added near max power immediately and the nose pitched up very rapidly, almost certainly another ‘surprise”. If they did not observe the low airspeed it is highly unlikely they would have expected the nose to be trimmed full up. The autopilot did that to maintain the selected altitude – as it is supposed to. Sudden and unexpected control pressures are rarely corrected instantly.
The FDR shows the trim increasing [to near max] and the elevator moving, just before the upset. It does not show movement of the control column because it doesn’t need to move [the autopilot is just balancing the pressures to hold the altitude – ortherwise the pitch would decrease]. The FDR shows the pitch and AOA increasing. When the autopilot disconnects and near max power is applied the trim is still there and the nose pitches up to +20. In the process the control column transducers show an aft movement of the column. Of course they do, it is now responding to the fixed up-elevator position induced by trim.
Pilot response/reaction time to anomalies is normally calculated to be 3 seconds. Look at the time-trace following autopilot disconnect:
22:16:26.9 – CAS-130, Pitch +9, AOA +12 – Autopilot disengages/shaker activates
22:16:27.9 – Power levers move to 70 deg position (max is 80)
22.16.29.9 – CAS 122, Pitch +20, AOA +20 – Aircraft begins roll to left – right aileron and rudder applied – roll continues which means the left wing has lost lift. Time elapsed = 3 seconds.
22:16:31.9 – left roll angle reaches 45 deg and begins to recover.
22:16:33.9 – CAS -90/110, Pitch +31, AOA +31 – Aircraft rolls thru wings level and continues roll to right, Flaps retracted (oops), Stick pusher fires. Total time elapsed = 7 seconds.
I find it hard to believe that the PF intentionally applied back pressure when the shaker went off. I think the pitch up was caused by the elevator trim setting and the power application. After the aircraft upset and was snapping left and right in a complete stall it is not too surprising that they didn’t know what to do and applied incorrect control movements. When the bank angle is 105 deg and you’re on instruments which way is up/down isn’t readily figured out [the AI screen was most likely solid Blue]. When you’re rapidly applying full aileron and rudder from one side to the other – the airplane is way ahead of the crew.
I have no idea what the stall characteristics of a Q-400 might be, but given the left roll experienced before the pusher it does not appear that they are benign. Usually, though not always, an airplane needs a shaker and pusher because its stall characteristics are abrupt and with little or no warning – so both devices are installed to prevent stalls. I don’t know what AOA is required to make this wing stall but it appears to be in the neighborhood of +20 degrees. The energy/momentum of the pitch up kept it going to +30 degrees. Much sustained forward pressure would have been required to lower the nose – it was never applied. But, that doesn’t necessarily mean that back pressure was applied immediately following shaker activation. Something did move the column aft – was it the pilot or was it the elevator that was already displaced to near max nose-up trim. I don’t know the answer but I’m just thinking out of the box.
Situational awareness was lost before the upset. It is highly improbable that it would be regained while the upset is in progress. Even if it were, it is rare indeed that a transport that upsets (for whatever reason) at 1500 ft +- AGL can be recovered successfully. Most likely the loss of situational awareness was the result of complacency, inadequate training, and inexperience [all working together at the same time]. IMO, this crew was not qualified to operate this aircraft [yes, I know they were “properly rated”]. I further think that the PF flying was not ready to ‘command’ anything and the F/O was so inexperienced as to be of virtually no help. Additionally, the number of flight hours is not an accurate measure of experience or command ability. I argue that there is a substantial difference between being “PIC” in a light aircraft and Captain of a transport category aircraft of the size and relative sophistication of a Q-400. I realize it’s not an Airbus, but it’s not a Saab 340 either. CRM is a good thing; a complete lack of cockpit discipline is not.
Being paid more money or less money has nothing to do with one’s piloting skills. It does however, result in management’s need to accept lower levels of experience and minimal qualifications. Don’t blame pilots for accepting jobs that pay less than others – that’s a system created by management and sanctioned by government. In many instances in major carriers, management has reduced pilot pay by as much as fifty percent, eliminated pensions, greatly reduced medical benefits and devastated work rules returning them to FAR standars, which are rediculous. Pilots in legacy carriers have accepted these changes because they have no choice – but their aircraft haven’t started falling out of the sky because their pay is now lower than it once was. Pilots in regional/feeder carriers have done the same. Lower pay does not decrease individual qualifications and pilots have no real control over their compensation. When you need to feed your family or yourself you take the job that’s available. We all do this and it is not limited to the airline industry.
I'm jurasic and date from a time when being an airline pilot was a respectable and desirable career profession. That has changed and today it has regressed to the point where, candidly, it is an insecure profession filled with undue stress - an undesirable job. That's truly sad. I never wanted to be a bus driver but that's what I felt like by the time I reached retirement. I watched my friends and myself deprived of our pensions while incompetent managers walked away with millions leaving our companies in bankruptcy. You could say I'm bitter about that but somehow I managed to avoid letting it degrade my on-the-job performance. I'm not a member of the ME generation. Sorry, I digress.
An experienced captain can ‘carry’ an inexperienced first officer so long as things don’t get too critical. An inexperienced captain can’t, and pairing the two together is high risk. As the complexity of the aircraft increases, the need for maturity, and training increases with it. Due to economic considerations, regulations have not kept pace with advances in aircraft technology and design. FAA ‘minimum standards’ are outdated and inadequate. That’s not the pilot’s fault and it was not the fault of the Colgan crew. If they were products of the pay-for-training cadre they are not guilty of wrong doing, they are victims of wrong doing.
I have no desire to crucify this crew because mistakes were made. They are just victims of a system that is essentially sick at too many critical levels. Expedience led to them being where they were when they were and as far as I’m concerned that was not of their making. When the dollar or the pound or the euro is allowed to replace the sound judgment of managers and regulators, the result will be more accidents of this nature.
Many accidents and incidents in regional carriers in the last few years are indicative of this. The upset and dual flameout of the CRJ-200 at FL410 is a good example. The PIC of that a/c had more than 6000 flight hours [in what?] but neither member of that crew had a clue related to high altitude operations. They demonstrated a total lack of judgment and it cost them their lives. Their lack of judgment was first preceded by management’s lack of judgment – for not ensuring that they were properly trained. They died with a permanent black mark on their records. Management collected the insurance and walked away with a smile to repeat the error of their ways. The same is true of the CRJ-200 departure accident in KY. That crew didn’t know what runway they were on, but they did know that the runway was unlighted. Nevertheless, the “captain” elected to take off without so much as a second thought and the F/O said not a word. Is it ‘normal’ for airliners to depart on unlighted runways? Somehow I don’t think so. Fifty people died in the resulting wreck. There was nothing wrong with the manipulation of controls in either case. There was everything wrong with their training and their judgment. No one in their company ever told them that wasn't a good idea. There was no SOP. They are balmed. Management has since been promoted and collected their bonuses. Same story, different day.
In contrast, the USAir flight crew that landed in the Hudson River was very seasoned – even though they had lost 50% of their former pay, all of their pension benefits, were embroiled in a nasty merger and likely to lose their seniority, and the commander commuted from the West coast to the NY area. They experienced a double flame out shortly after take off in the middle of NYC. The outcome was successful. It wasn’t because the captain demonstrated any unusual stick and rudder technique. However, both crewmembers did demonstrate excellent judgment, both dated from a era when training was adequate and, as a consequence, made the only logical decision available to them. Sure they were lucky that the river was there and that the aircraft remained essentially intact on touchdown but luck didn’t save the day, sound judgment and adequate training did. Will the younger folks coming along in the same airline have the same experience? I doubt it – times have changed and now the bare minimum has become the norm. If that happens again the outcome is quite likely to be different.
Very few accidents, if any, result from moving the controls the wrong way. A great many accidents result from poor judgment. Good judgment is acquired over time and with maturity - sometimes never. You can’t teach judgment in a simulator. The chain of most accidents results from errors in judgment – not stick and rudder manipulation. This accident is a sad demonstration of just that. However in defense of the crew, the bad judgment that ultimately led to this accident did not begin in the cockpit; it began in the board room.
Apologies for the rant; I’ll get off the soapbox now.