Gurudriver: Air France

The Coffee Shop
1

I was just reading a brief synopsis of the cockpit chatter and data recorder details. They know stall alarms were going off, 6 times in fact. They know the FO was trying to climb. What possible scenario would a pilot try and climb @ 35K-ish feet if he felt his airspeed was compromised this badly? Let’s say they don’t trust their indicated airspeed or the pitot’s are fouled. Why would you not look at your GPS for airspeed there? No way an Airbus doesn’t have GPS, right? Heck we talked about stalls b/f my discovery flight and were practicing stall recovery at maybe 20 hours.

I know high speed stalls and jets are a completely different ball of wax and would love to hear a commercial jockey comment a bit. Do airliners flirt with coffins corner?

2

Disclaimer: I fly Lockheed, so this comes from co-workers who fly the A319 series.

Airbus builds their planes so even a caveman can fly it. As such, the computer will literally not let you pull to a stall unless there is a significant failure, so their stall recovery is to hold the stick back and let the plane max perform itself. IIRC, they encountered a failure, but Co kept pulling through the buffet and into the full stall. CW says do some of that pilot shit and recognize it, but with a rough ride and things going crazy in the cockpit, they may not have recognized the usual signs until it was too late. BL: “Max, relax, roll” isn’t how they train for stalls.

3

What happened?

4

I’m guessing they’re talking about the AF flight that crashed in a Brazil-France flight 2 years ago.
AF447.

5

Oh, yes.

6

Sometimes, the speed indicator can fail and go to zero, stick in its current position (hardly noticeable), show higher indicated, or something in between. The airpseed indicator may even alternate high and low. If the airspeed reads too high, my first inclination is to pull the power back or to raise my nose to bleed off the speed. In the case of AF, they actually pulled the power off and raised the nose, inducing a stall on a perfectly flyable airplane. The AF jet’s speed sensing probes, on the outside of the airplane, were frozen. This causes erroneous indications. But those probes are electrically heated and should not ice up.

These erroneous indications may not be readily recognized if you are flying in a turbulent area, too, as the airspeed may shift around anyway. Convective activity can cause shears of wind with wild airspeed swings on the indicator.

In the USAF, they basically teach “Recognize, Confirm, Recover”: Recognize there is a problem (duh, but sometimes the “unusual attitude” can be insidious), Confirm using the other pilot’s instruments + the standby instruments (the gps computer works well but I think angle of attack is quicker and easier to read), and Recover the airplane to a normal pitch, power, and speed. Recognize and Recover are usually straight forward but that Confirm part often takes a little time. The Confirm may have to be done at night, in the weather, in turbulence, the airplane’s autopilot bells going off, while the airplane is rolling over on its belly, pitching down, zooming toward a stall, etc. This is tough when the airspeed indicator shows plenty of airspeed while the stall warning indicators are all going off. Cockpit warnings, thunderstorms, bad instruments, night, jet lag, not getting along with a crewmember, poor sleep on the layover, fight with the wife before the trip - utter mayhem!

My job is pretty easy, usually, but when the shit hits the fan you have to take in a ton of data, evaluate the veracity of that data, then choose a path fairly quickly. If that path is the wrong one, you have to know how to back out of that and start over. The guys that flunked out of USAF pilot training were the ones that would give up and shut down.

Oh, and my pay has never reached back up to year 2001 levels and we pilots are operating under 1993 pay rates at my airline! But the airplane doesn’t care how much I get paid, ie. I strive for the same skill level on every flight. There’s no other choice. Always fly as though what happened to the “other guy” could actually get you, too. Hope this helps.

(Btw, I was a USAF accident investigator for a few years at Ramstein Air Base and had the unfortunate duty of investigating Commerce Secretary Ron Brown’s crash on April 3, 1996)

7

Wow that was great man thanks for the insight. I cannot imagine with all the years and hours in both seats what must have been going through their collective heads. Ugh.

I know all airframes are different, but what kind of altitudes can one recover from in a heavy for a high speed stall? Approach stall? If you screw the pooch at 3K can you recover? I mean a total departure…can that be worked on or are you a smoking hole in the ground?

I was always under the impression that one of the great attributes of the Airbus line was their elaborate flight envelope protection?

Thanks for the great stuff Guru.

8

I thought (given our obvious different views on science, God and all) that I would never take the time
to read one of your posts. I guess I was way wrong. Very instructive stuff there. By the way, I absolutely
hate flying since a really nasty flight in 2003 in the Houston area.

9

What a great post.

I’ll never cease to be amazed at the variety of skills here on ST.

10

Wow that was great man thanks for the insight. I cannot imagine with all the years and hours in both seats what must have been going through their collective heads. Ugh.

I know all airframes are different, but what kind of altitudes can one recover from in a heavy for a high speed stall? Approach stall? If you screw the pooch at 3K can you recover? I mean a total departure…can that be worked on or are you a smoking hole in the ground?

I was always under the impression that one of the great attributes of the Airbus line was their elaborate flight envelope protection?

Thanks for the great stuff Guru.

That sort of thing works great until erroneous data gets into the system: http://digitaljournal.com/article/261141

There have been issues with bad data getting past all of the safeguards and into the digital flight control systems, resulting in what is known as an “inflight upset.” And lest anyone accuse me of bias, there has been at least one incident with a bad Air Data Inertial Reference Unit (ADIRU) with a Boeing 777, IIRC. There has been talk in the industry about the pitfalls of relying so much on technology instead of basic piloting skills. In this case, more technology may not actually be better.

Spot

11

Wow that was great man thanks for the insight. I cannot imagine with all the years and hours in both seats what must have been going through their collective heads. Ugh.

I know all airframes are different, but what kind of altitudes can one recover from in a heavy for a high speed stall? Approach stall? If you screw the pooch at 3K can you recover? I mean a total departure…can that be worked on or are you a smoking hole in the ground?

I was always under the impression that one of the great attributes of the Airbus line was their elaborate flight envelope protection?

Thanks for the great stuff Guru.

I’m not judging, but those guys were pretty confused all the way down to the crash site.

Man, you would be very surprised at how a jet airplane can perform out of a stall if you firewall the throttles and fly the airplane away from the ground just above stall speed! A stall at 3,000’ probably means you have flaps down so the recovery would be quick, a few hundred feet up to maybe a thousand feet of altitude loss.

We practice this stuff in the simulator every 9 months. I’m 44 years old and I still look forward to “dial-a-death” scenarios in the sim! They give us a decreasing performance windshear on short final, meaning your airspeed just drops off around 2-500’ above ground. Depending on the severity of the shear, it’s recoverable. You see about 90-100 knots in some of those scenarios (yikes!) when the jet should be around 140-150 knots for landing. In those cases, you just can’t turn at all, which puts a load on the wings, which in turn raises the stall speed. So, it’s straight ahead and recover, if terrain isn’t a factor.

I don’t know much about Airbus envelope protection personally. I flew B737-200s, C-5 Galaxys, Gulfstream III (a very sexy jet: http://www.militaryaircraft.de/pictures/military/aircraft/C-20/C-20H_RIAT2007_0847_800.jpg), B727, MD-80s, and now the B737-800. All the airplanes have a low speed warning, then a stall warning, and then the stick shaker, typically. Departure from controlled flight is pretty rare. These transport planes just don’t spin or get sideways. The fighter and acrobatic dudes can get airplanes in some crazy situations.

Hey, we did refuse an airplane on Thursday after flying from DCA-ORD (Washington to Chicago). When the flaps were selected to 30 (30 and 40 are the full flap landing configurations), the airplane would roll to the right. The mechanics said the landing gear doors airflow were causing the flaps on the right side to retract a little, causing a rolling moment. Well, that’s wrong and we refused to fly it further. The flap tracks were worn down? They towed it to the hangar, we got another plane, and everyone was 2 hours late : ( Better safe than sorry.

12

I thought (given our obvious different views on science, God and all) that I would never take the time
to read one of your posts. I guess I was way wrong. Very instructive stuff there. By the way, I absolutely
hate flying since a really nasty flight in 2003 in the Houston area.

What happened in Houston? Turbulence?

13

What a great post.

I’ll never cease to be amazed at the variety of skills here on ST.

There are some seriously qualified and smart folks on ST. The triathlete world is just full of physically and mentally fit people. Kinda’ goes hand in hand.

14

That sort of thing works great until erroneous data gets into the system: http://digitaljournal.com/article/261141

There have been issues with bad data getting past all of the safeguards and into the digital flight control systems, resulting in what is known as an “inflight upset.” And lest anyone accuse me of bias, there has been at least one incident with a bad Air Data Inertial Reference Unit (ADIRU) with a Boeing 777, IIRC. There has been talk in the industry about the pitfalls of relying so much on technology instead of basic piloting skills. In this case, more technology may not actually be better.

Spot

So true and I think you just have to revert to or start with the basics in every situation. What does the data tell me and can I trust it? It’s very rare to have double ADI failure but it has been documented. Usually, one can notice a bad horizon indicator on one pilot’s side, crosscheck with the other pilot, then crosscheck those with the standby attitude in about 1.5 seconds. Best 2 of 3 wins! Same with the airspeed indicator and altitude. If you can’t get 2 to agree, the standby system is probably correct.

15

Pretty heavy ones, two drops, one flight attendant looking dead scared and crossing herself, total fun :slight_smile:
.

16

Pretty heavy ones, two drops, one flight attendant looking dead scared and crossing herself, total fun :slight_smile:

It’s only happened once and it was years ago, but I was in turbulence and got real queasy while flying. I turned to the other pilot and said, “We really need to find smooth air…” It’s a bad feeling to be in charge of the airplane but on the verge of losing your stomach!

17

Interesting stuff.

I had a friend whose dad was a pilot for UPS. He used to say that the flight crew of an Airbus aircraft consisted of a pilot and a dog. The pilot’s job was to feed the dog, and the dog’s job was to bite the pilot if he tried to touch anything.

I’ve actually been getting some mileage out of that quote at my own job (x-ray and CT), with the increasing tendency to automate everything about image acquisition. I believe strongly that automation makes the easy stuff easier and the hard stuff harder. You have to know a dozen ways of getting around it in order to make your equipment perform to its full potential.

18

Interesting stuff.

I had a friend whose dad was a pilot for UPS. He used to say that the flight crew of an Airbus aircraft consisted of a pilot and a dog. The pilot’s job was to feed the dog, and the dog’s job was to bite the pilot if he tried to touch anything.

I’ve actually been getting some mileage out of that quote at my own job (x-ray and CT), with the increasing tendency to automate everything about image acquisition. I believe strongly that automation makes the easy stuff easier and the hard stuff harder. You have to know a dozen ways of getting around it in order to make your equipment perform to its full potential.

I’ve always liked that joke! The new phrase amongst pilots is, “What’s it doing?” while the airplane is flipping on its back or making some action the pilots never intended. More and more, the data from various parts of the airplane is merely a presentation of what the airplane “thinks” is happening. Used to be that an indicator had some direct linkage to a real tachometer, airspeed indicator, or heat sensor. Increasingly, some black box collates data and then tells you, according to a program, what should be presented. So, you may not be looking at the truth.

19

That sort of thing works great until erroneous data gets into the system: http://digitaljournal.com/article/261141

There have been issues with bad data getting past all of the safeguards and into the digital flight control systems, resulting in what is known as an “inflight upset.” And lest anyone accuse me of bias, there has been at least one incident with a bad Air Data Inertial Reference Unit (ADIRU) with a Boeing 777, IIRC. There has been talk in the industry about the pitfalls of relying so much on technology instead of basic piloting skills. In this case, more technology may not actually be better.

Spot

So true and I think you just have to revert to or start with the basics in every situation. What does the data tell me and can I trust it? It’s very rare to have double ADI failure but it has been documented. Usually, one can notice a bad horizon indicator on one pilot’s side, crosscheck with the other pilot, then crosscheck those with the standby attitude in about 1.5 seconds. Best 2 of 3 wins! Same with the airspeed indicator and altitude. If you can’t get 2 to agree, the standby system is probably correct.

Yep, double ADI failure is very rare, but it’s happened: http://usaf.aib.law.af.mil/B-1_DiegoGarcia_12Dec01.pdf The Bone that was lost out of Diego Garcia had that happen at night with no discernible horizon, and the crew ended up ejecting. I got to fly a sortie where we tried to replicate what happened (in the daytime with clear blue skies, of course!) when I was a WSO with the B-1 OT guys.

Spot

20

How does using engine RPM’s work in the jet world as a possible indicator of what’s going on? I mean if you look down and see a power setting of 90% on the throttles and RPM’s indicate roughly the same thing…would that not help? Did I understand that these guys didn’t just initiate a climb BUT also pulled the throttles back to idle!!! I was not aware of the throttle part, yikes.