Not so good news about flight 3407

[Crossky]

In reading the Q drivers explanation of how the AOA/Stall system is rescheduled in icing, the wing may not have actually aerodynamically stalled until it reached it's critical AOA during the abrupt pull-up maneuver.

There may not have actually been an aerodynamic stall when the pusher activated, since the pusher is biased by 20 knots as a function of being in icing conditions.

Put another way, with the 'icing switch' off, stall speed was shown to be computed at 90 kts or so.

. . .
Likely, as is typically the case, a combination. But it is clear the speed was allowed to decay to a very low value. After this, the abrupt pull-up sealed their fate at such a low altitude.

I agree with this scenario.

We'll know perhaps whether the crew was concerned about a tail stall from the CVR. They seemingly reacted to a perceived tail stall: extend flaps, the AP disconnects and they get shaker, then pusher. In a tail stall the nose drops, the remedy is to retract flaps and increase back pressure. Only thing is, they could've misinterpreted it. Stick shaker and pusher means WING STALL.

Fly the plane first, if the plane is still flying, don't mess with it. I got AP disconnect, shaker and pusher once with the boards out when new in the RJ on a visual. I had focused on descending after ALTS CAP to catch the GS, only a lower ALT hadn't been selected. I had increased thrust, but not enough to compensate for the drag from the flight spoilers. It took the jumpseater behind us to say 'spoilers' (yes this was humbling for me) before I clued in. But in that momemt before I got SA back I had increased thrust some, and done nothing to the pitch attitude.

Pilot error is sad, tragic, and unfortunately often a fact of life in aviation accidents. NBC didn't do regional pilots any favors tonight, explaining how long (short really) each had been employed by Colgan.

 

[joe_pilot]

There are flight directors that will drive to a 30 degree pitch up?!?!?

That was my first thought.

Also, didn't 3407 intercept from the South, ie. left traffic? I don't remember from the ATC tapes, maybe someone out there can refresh my memory? Thx.

Yup, from the south. Last assigned hdg was 260 to join

 

[joe_pilot]

I agree with this scenario.

We'll know perhaps whether the crew was concerned about a tail stall from the CVR. They seemingly reacted to a perceived tail stall: extend flaps, the AP disconnects and they get shaker, then pusher. In a tail stall the nose drops, the remedy is to retract flaps and increase back pressure. Only thing is, they could've misinterpreted it. Stick shaker and pusher means WING STALL.

Fly the plane first, if the plane is still flying, don't mess with it. I got AP disconnect, shaker and pusher once with the boards out when new in the RJ on a visual. I had focused on descending after ALTS CAP to catch the GS, only a lower ALT hadn't been selected. I had increased thrust, but not enough to compensate for the drag from the flight spoilers. It took the jumpseater behind us to say 'spoilers' (yes this was humbling for me) before I clued in. But in that momemt before I got SA back I had increased thrust some, and done nothing to the pitch attitude.

Pilot error is sad, tragic, and unfortunately often a fact of life in aviation accidents. NBC didn't do regional pilots any favors tonight, explaining how long (short really) each had been employed by Colgan.

Your candor about the brain-cramp is refreshing. We all need to work on this.

 

[skiandsurf]

It is sad that pilot error is even still allowed to be labeled as the direct cause of an accident. It is well established that humans will always err, but our aviation system has evolved to a state of being able to trap most errors. The only way to prevent further accidents is to explore why the system failed to trap the error. This cannot be meaningfully done by simply closing the case with "pilot error" yet again.

We, as pilots, will make mistakes. But its the severity of the mistakes that we can improve on. I always tell my flying partners that there are 2 types of mistakes: 1) one that we talk and laugh about and 2) one that others talk and laugh about. Lets keep them between the two of us.

 

[livin'thesim]

An operator of mid-sized turboprops had an interesting autopilot anomaly at a Michigan airport about a year or so ago.

Turned out to be the electronics playing tricks on the flight crew. I believe the actual flight control computer had a fault in it.

Maybe I'm just a little too crusty, but I feel that the automation has built a certain set of hazards into the system that we have yet to fully appreciate. And the automation seems to be expanding in new ways.

The biggest problem I have is that the box is often making many decisions that are not even documented in the manuals that pilots are issued.

There is often too much emphasis in 121 training on full reliance on automation. But think about TAWS. If you hear a terrain warning, you click off the AP and fly the escape maneuver. If you get a TCAS alert, you click off the AP and avoid the conflicting aircraft.

Why then, is it not safe to click off the AP in other instances where the pilot is behind the automation, or where the automation is behind the airplane?

 

[belchfire]

If the stall speed was 90 and they saw 2 g's in a pull up, the stall speed jumps to 127 on a clean wing. SInce the flaps were in transit and the airspeed was slowing rapidly with the gear and flap extension, that was surely met.

I think that we need the real timeline for events-my impression was the the stick shaker was before the pitch excursion.

 

[Oakum_Boy]

There are flight directors that will drive to a 30 degree pitch up?!?!?

I've witnessed something similar in atlanta when they're taxiing airplanes through the ILS critial area.

The glideslope will shoot down from above just enough to trigger at GS capture and then it heads right back up just as quickly. If you're not on top of it, the airplane will pitch up quickly with resulting airspeed decay. Even disconnecting in a timely manner you can lose 15 knots in a matter of a couple seconds.

Remember on these cheaper FCC's without a SPD REF (or autothrottle mode) when you're in APPR mode it is a purely reactive mode, although I don't think it's programmed to command more that 27 degrees of bank, one might assume that there is a pitch limit at some point. Who knows.

 

[deadstick]

FWIW, it could be any of us, and if you think you are above it, you have not been aviation long enough.

I understand what you meant, but your statement should read:

......and if you think you are above it, you have been in aviation too long already.

Good point, but the problem gets people at both ends and in the middle too.

Joe, sums it up well. As professionals, we should strive to always improve our skills. That requires an open mind and a willingness to learn. When you want to quit learning, it's time to retire.

I found this WSJ article interesting, particularly this paragraph:

"The safety board, among other issues, is looking into why Colgan’s training programs apparently stop short of allowing pilots in simulators to feel the stick-pusher activate, according to people familiar with the issue."

Considering the source (the media), I have to question the validity of the claim.

http://blogs.wsj.com/middleseat/200...-focus-on-stick-pusher-training/?mod=yahoo_hs

 

[WSurf]

Just wanna add a problem I had on a flight in a Dash 8 300 series. Note, this has never ever in my 8 years on the plane happened before.
CLT had me slowed to 150kts about 15 miles out. Thats pretty slow clean in a 300 series and usually has about a 5 Degree pitch up with zero flaps put in with a heavy load in straight and level flight.
We were cleared for the approach and the autopilot was on with the weather near mins. I would much rather 'george' flying it and I watch. Approach mode was selected.

As we got closer to the Airport I looked down at my approach plate to verify the mins. for the approach.
When I looked up the Airspeed started bleeding off fast and I added power, and more more power. Finally I disengaged the autopilot and at that time and noticed the plane had started a climb.
FO and I were dumbfounded on what happened. As we continued the approach with no problems I figured it out.
Once the Glideslope was started to come down, the damn plane for some reason or another flew up to Capture it. Being in smooth IMC and if we had not been playing attention this coulda lead to a worse problem.
Like I said just something to consider and it has never ever happen to me again or before.

Computers, u just never know. So be careful out there.

 

[DiverDriver]

Which is why I'm adiment about having someone shadow the controls getting low to the ground and completely dumbfounded why I even have to say "please shadow the controls we're close to the ground"....several times to the same pilots. The AP may be on but we don't let the airplane fly us. WE fly the plane. Right?

Birds, transient aircraft, wake turbulence encounters, mechanical problems, navigation annomilies, ice, unknowns all can happen fast and require immediate action by taking over from the auto pilot. So why do so few pilots shadow the controls when low to the ground even if AP is on? You think the AP is that good? I don't. And sh!t can happen too fast.

 

[WSurf]

Which is why I'm adiment about having someone shadow the controls getting low to the ground and completely dumbfounded why I even have to say "please shadow the controls we're close to the ground"....several times to the same pilots. The AP may be on but we don't let the airplane fly us. WE fly the plane. Right?

Birds, transient aircraft, wake turbulence encounters, mechanical problems, navigation annomilies, ice, unknowns all can happen fast and require immediate action by taking over from the auto pilot. So why do so few pilots shadow the controls when low to the ground even if AP is on? You think the AP is that good? I don't. And sh!t can happen too fast.

The autopilot is there to decrease your workload, and thats it. It gives you the opportunity to watch and scan things much better. Like I said, this happened in a spilt second of me just glancing down to verify the Mins for the approach again.

I couldn't imagine this happening while both pilots were outside looking at the ice on the wings. But if it did, it could add up getting into something serious.

 

[WSurf]

You think the AP is that good? I don't. And sh!t can happen too fast.

Of course I think the autopilot is that good. You can't come close to flying an approach or in that case the plane at all as good as it. I laugh when I hear people say they can.
But just like your partner flying it, you have to monitor it at all times.

 

[surplus1]

I gotta agree with your last WSurf. Just try hand flying a CATIII B. It ain't gonna happen.

Caveat: It depends on what kind of autopilot you have. They're not all the same.

 

[DiverDriver]

The autopilot is there to decrease your workload, and thats it. It gives you the opportunity to watch and scan things much better. Like I said, this happened in a spilt second of me just glancing down to verify the Mins for the approach again.

I couldn't imagine this happening while both pilots were outside looking at the ice on the wings. But if it did, it could add up getting into something serious.

Sorry Surf, in no way was I directing my comments at you. Your post was good for all to read and see how fast it can happen. Which goes along with why I wish more pilots I see would shadow the controls and be ready.

 

[ReverseSensing]

The Q400 autopilot/FD is weird, especially compared to the 200. It has trouble capturing LOC well, it will often initially turn the wrong way, for example, when proceeding DTO a fix in LNAV, especially if it is a fairly large angle turn. It corrects itself after a second or two. We call it the AP clearing turn.

On an ILS once, fully captured, passing through a windshear, I witnessed the FD command a 15 degree nosedown pitch to maintain the GS. The autopilot attempted to dutifully comply, until I intervened.

And don't even get me started on capturing a VNAV. Talk about having to guard against putting the FAs on the ceiling of the cabin.