Food for throught: Airbus A3XX crash

General conversations about BMW E28s and the people who own them.
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

I'd LOVE to talk to you about the NEO project...we have about 40 NEO's ordered.
You got a cool job man...
Justin_FL
MyE28 IT Guru
MyE28 IT Guru
Posts: 2822
Joined: Feb 12, 2006 12:00 PM
Location: Palm Beach
Contact:

Post by Justin_FL »

Unfortunately, I'm just the slave labor, doing 3D design/modeling for ground testing support. The engineers up north in Connecticut design the important spinning bits.

But since there are so many test stands at the WPB facility, it is neat that I can go out and put hands on the state of the art engine hardware and hear/see it run up close. Nothing like having your heavily reinforced office building shake all day from military endurance test runs ;)

BTW, I do enjoy reading your experience and perspectives as a pilot in this thread. Thanks for sharing!
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

Justin_FL wrote:Unfortunately, I'm just the slave labor
:laugh:
Me too...
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

Apparently, the event was 4 minutes...not 15 as previously thought. Naturally, the reports from news agencies who don't understand the plane are a little off.

They say the "senior pilot" the captain was out of the flight deck when it all started. Thats normal. Long crossing require 3 type rated pilots anyway...and they rotate in shifts taking a break in the cabin....or a snooze.
They say the crew got a stall warning...the airbus doesn't give a traditional stall warning, because it can't stall in "normal law" (all computers up and working with accurate data)...so these guys never hear it under normal conditions. Probably confused them.

Here's the flight characteristics of the Airbus in "alternate law"
(some or all flight control computers failed or off)

All protections except for load factor maneuvering protection are lost.
• The load factor limitation is similar to to that under Normal Law.
• Amber XX's replace the green = attitude limits on the PFD.
• A low speed stability function replaces the normal angle-of-attack protection
o System introduces a progressive nose down command which attempts to prevent the speed from decaying further.
o This command CAN be overridden by sidestick input.
o The airplane CAN be stalled in Alternate Law.
o An audio stall warning consisting of "crickets" and a "STALL" aural message is activated.
o The Alpha Floor function is inoperative.
• The PFD airspeed scale is modified:
o VLS remains displayed
o VALPHA PROT and VALPHA MAX are removed
o They are replaced by a red and black barber pole, the top indicating the stall warning speed VSW
• A nose up command is introduced any time the airplane exceeds VMO/MMO to keep the speed from increasing further, which CAN be overridden by the sidestick.
Bank angle protection is lost.
• Certain failures cause the system to revert to Alternate Law without speed stability.
Yaw damping is lost if the fault is a triple ADR failure.

The bold items are the critical ones for AF447.
It sounds like they were trying to manually fly the plane by hand without any instuments, and without any of the protections that an airbus typically has...without being able to see outside.
Reports have them with a pitch angle of 15 degrees up, descending at 10,000fpm. Thats a fully stalled swept-wing aircraft. They talk about heavy banking left and right..also a sign of a stalled swept wing aircraft.
Alpha Floor lost- thats the thrust mode that firewalls the engines at critically slow speeds.
Alpha Prot lost- slow speed pitch protection
Alpha Max lost-- high speed protection.
cgraff
Posts: 1387
Joined: Feb 12, 2006 12:00 PM
Location: New Jersey
Contact:

Post by cgraff »

Link to published report?
JohnH
Posts: 693
Joined: Jan 24, 2007 9:55 AM
Location: Surrey,UK

Post by JohnH »

TSMacNeil wrote:It sounds like they were trying to manually fly the plane by hand without any instuments, and without any of the protections that an airbus typically has...without being able to see outside.
Thats progress for you. Even Alcock and Brown had a couple of spirit levels for pitch and bank.
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

cgraff wrote:Link to published report?
There isn't one yet. This is just gathered statements from the French investigators at this point.
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

TSMacNeil wrote:
cgraff wrote:Link to published report?
There isn't one yet. This is just gathered statements from the French investigators at this point.
Found this:
http://www.reuters.com/article/2011/05/ ... 8420110527
cgraff
Posts: 1387
Joined: Feb 12, 2006 12:00 PM
Location: New Jersey
Contact:

Post by cgraff »

TSMacNeil wrote:
TSMacNeil wrote:
cgraff wrote:Link to published report?
There isn't one yet. This is just gathered statements from the French investigators at this point.
Found this:
http://www.reuters.com/article/2011/05/ ... 8420110527
Care to comment on this?

http://www.businessinsider.com/air-fran ... ror-2011-5

-Chris
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

The guy who wrote that is a fucking moron. Doesn't know what he's talking about....but he is a private pilot who has flown small airplanes...and he stayed at a Holiday Inn. Did you read it?

"Based on this preliminary report, and speaking as someone who has a pilot's license and used to fly small propeller planes, it appears the co-pilot screwed up. He was flying partially blind--one of the plane's airspeed indicators blipped out--but he then moved the controls in a way that slowed the plane down and caused it to stall."

Maybe he should speak to an Airbus pilot....he'd know a little more before he opines...
:roll:

This is what I'm reading...
http://www.scribd.com/doc/56453341/Poin ... mai2011-En

Pick it up at 2'10'05...thats where the pitot/static system froze over, and the shit hit the fan. BTW-60kts indicated is the bottom peg...meaning zero. Its important to note that the window between overspeeding and stalling at mach .80 and weighing 270tons is about 10-15 kts...not much. AKA "coffin corner" speed up and you overspeed, slow down and you stall...
Finally, I see some disturbing AOA (angle of attack) numbers relative to actual pitch attitudes. This may be the AOA vanes partially frozen and moving erratically. The AOA being up to +40 degrees but the pitch being +15 or so is disturbing...
Last edited by TSMacNeil on May 27, 2011 11:09 AM, edited 2 times in total.
cgraff
Posts: 1387
Joined: Feb 12, 2006 12:00 PM
Location: New Jersey
Contact:

Post by cgraff »

TSMacNeil wrote:The guy who wrote that is a fucking moron. Doesn't know what he's talking about....but he is a private pilot who has flown small airplanes...and he stayed at a Holiday Inn. Did you read it?

"Based on this preliminary report, and speaking as someone who has a pilot's license and used to fly small propeller planes, it appears the co-pilot screwed up. He was flying partially blind--one of the plane's airspeed indicators blipped out--but he then moved the controls in a way that slowed the plane down and caused it to stall."

Maybe he should speak to an Airbus pilot....he'd know a little more before he opines...
:roll:

This is what I'm reading...
http://www.scribd.com/doc/56453341/Poin ... mai2011-En

Pick it up at 2'10'05...thats where the pitot/static system froze over, and the shit hit the fan. BTW-60kts indicated is the bottom peg...meaning zero. Its important to note that the window between overspeeding and stalling at mach .80 and weighing 270tons is about 10-15 kts...not much.
I read that too and rolled my eyes (re: your bold type)

http://www.bea.aero/fr/enquetes/vol.af. ... 011.en.pdf

That's link to PDF.

-Chris
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

Ah, much easier to read-thank you. I see that the PF (pilot flying) maintained a nose-high attitude throughout this event. Depending on what instruments he had available to him (none, from the ECAM info) this is understandable.

The sidestick and FBW system is kinda like power steering in a '53 Buick. Its like power steering on steroids....and there's only simulated "feel"...which is unavailable when all the flight control computers fail like they did. It holds the last position inputted.
Example, pitch up for a 500fpm climb, and let go....it stays in the 500fpm climb. Same with turns, bank 15 degrees left and let go...it stays in the 15 degree turn forever. It would be easy to be pitched up and not know it without any reliable gauges.

I'm not making excuses, but it's important to really know what was going on up there, considering the weather, alarms, checklists, etc...I think my conclusion is the correct one.
cgraff
Posts: 1387
Joined: Feb 12, 2006 12:00 PM
Location: New Jersey
Contact:

Post by cgraff »

TSMacNeil wrote:Ah, much easier to read-thank you. I see that the PF (pilot flying) maintained a nose-high attitude throughout this event. Depending on what instruments he had available to him (none, from the ECAM info) this is understandable.

The sidestick and FBW system is kinda like power steering in a '53 Buick. Its like power steering on steroids....and there's only simulated "feel"...which is unavailable when all the flight control computers fail like they did. It holds the last position inputted.
Example, pitch up for a 500fpm climb, and let go....it stays in the 500fpm climb. Same with turns, bank 15 degrees left and let go...it stays in the 15 degree turn forever. It would be easy to be pitched up and not know it without any reliable gauges.

I'm not making excuses, but it's important to really know what was going on up there, considering the weather, alarms, checklists, etc...I think my conclusion is the correct one.
Maybe you can elaborate, and also I haven't had time to really dig in through all the details, but my observations and questions are the following (and again, forgive as I haven't had a change to go thru all the details):

1. The plane was straight and level, and then they changed heading. Did they also pitch up to change altitude? Was this before all the failures?

2. Engines were at 100%, then at 55%...why was this not (apparently) observed or corrected for?

3. They kept on losing altitude, and there were callouts for this. Why was it not a foremost priority to keep altitude? Especially given the 10k ft/min descent rate?

Again, would love for you to go through in detail a bit more to give us a better understanding of what's happening, what all the error and alerts and warnings were doing, what the plane is doing, and what the pilot may be thinking about and dealing with, now that we have the report.

-Chris
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

Well, I'll start with a video on the fly by wire system by airbus...this will show how it works when everything is working (in normal law)
http://www.youtube.com/watch?v=WCc-R4xXZPU

Then, an example of the cascading ECAM messages and how they are presented...(gives an idea of how fast you can get task saturated) They dont have one for pitot system failure, so this is a dual engine failure...you'll get the idea.
http://www.youtube.com/watch?v=Uz3XAyyWNcY

Then, it's important to recall that the only thing these guys had was the engine instruments...all the other flight and navigation systems were failed...just red flags, no useful information to fly with. They were immediately in "alternate law"...where they had no high or low speed protection, no bank angle protection, and they could (and did) stall the airplane. They slowed from mach .82 to .80...and were at 38,000' and heavy. (about 500,000lbs)
This means they were right up into "coffin corner"...a dangerous place to be. The least little increase or decrease in airspeed either stalls the plane, or overspeeds the plane.
(If he had descended a few thousand feet he would have had a bigger margin of airspeed to deal with)
As soon as the airspeed went from 235kts to 60kts ( I say thats icing) the autopilot kicked off, the flight computers all faulted, and he stalled...almost simultaneously. No flight instruments, no heading or airspeed readings, no vertical speed readings...nothing. The data in this report is logged in the Flight Data Recorder, so we can see in hindsight what was happening- they didn't have any of that information presented to them in the cockpit. He didn't know he was descending at 10,000fpm...Im sure he felt something, but vertigo is a funny thing.
Thats important- he hears the overspeed warning (clacker), but doesn't know by how much, or how fast they are. He has a stall cricket and aural...but, again, how slow? How fast? He has no vertical speed gauge, no altimeter...he knows he's wallowing around, but doesnt know what wings level is (no instruments) doesn't know his sink rate, airspeed...nothing.
Can't see out the windows-nightime in the Atlantic.
His pulling power to idle was probably a reflex instinctive action to silence the overspeed clack and not pull a wing off.
Likewise for the full power use, he's being told he's stalling...natural reaction is to firewall the power...don't really have to pitch forward in a jet, it has enough thrust to fly out of a stall, if you can keep the wings level...which he had no way of doing positively.

Gotta run just now, but Ill pop back in later-
Mark 88/M5 Houston
Posts: 8548
Joined: Feb 12, 2006 12:00 PM
Location: Far North Houston

Post by Mark 88/M5 Houston »

What a nightmare! Even with the few hours I have I can't imagine anyone being able to save that aircraft under those conditions. Definitely a comprehensive and cascading design theory/philosophy problem. I guess we will see the French legal system playing "Hot Potato" with this one.

Note to self: Ask to see the paperwork for the updated pitot assemblies before I get on any Airbus.
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

Well, many aircraft have an inclinometer on board for manual fueling. It is used to determine when the plane is level on the ramp so they can use dripsticks properly. I've seen them in cockpits for this purpose.

That would be useful for determining level flight...and setting a power of about 65-70% N1 would sustain fairly unaccelerated flight. Of course, turbulence plays a big role in even being able to read it too.

No matter how you slice it, this was a bad dealio.
Matt
Posts: 2345
Joined: Feb 12, 2006 12:00 PM
Location: Fargo

Post by Matt »

Since I did't know what a coffin corner was, and was having a hard time understanding how 15 kts could be the entire usable range of speeds in an aircraft, i looked up "Coffin Corner" on wikipedia and found this:

http://en.wikipedia.org/wiki/Coffin_corner_(aviation)

The idea that a commercial aircraft would regularly operate anywhere near this range, or that the margin of error could be as low as 10 kts, is really surprising.
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

Yep. Any prudent pilot recognizes when he's close to coffin corner. Especially in newer glass cockpits, they show you on your airspeed tape your maximum and minimum speeds.

This image will help illustrate the airspeed gauge.
Image

On the left of the image is the vertical airspeed "tape"...(the speedometer, if you will) At the top of that scale is a red and black barberpole...that is Vmax...."redline". That particular number is ever-changing based on current weight, altitude, etc...
You can watch the barberpole move up and down that tape as the dynamics change.
At the bottom of that tape is an amber barberpole looking indication....thats Vmin...minimum flight speed.
The higher you fly+the heavier you are= how close together those two speeds are...and the two barberpoles move closer together giving you a visual representation of your margin between the two.
In my experience, as long as the air is smooth, 15-20 kts is common....as soon as it gets bumpy, those "rattlesnakes" start moving...the red one moves down (slower) and the amber one moves up (faster)....if they meet....you have big-big problems.
In this picture, the speed margin is about 45 kts...the higher you go, the less fuel you burn, but the closer these two figures get to each other...
So you can see, as soon as the autopilot kicked off on these guys...the least little pitch up by a surprised pilot could have immediately put them into a stall...sounds like thats exactly what happened.

Older aircraft couldn't show you visually this dynamic range...you had to refer to charts...I did it for years.

I hope this helps a little, prolly boring the fuck outta you guys...sorry.
Last edited by TSMacNeil on May 27, 2011 9:25 PM, edited 1 time in total.
Matt
Posts: 2345
Joined: Feb 12, 2006 12:00 PM
Location: Fargo

Post by Matt »

In this picture, am i understanding it correctly:

current speed: Mach .780 (259kts)
vmax: 275kts
vmin: 230kts
alt: 35960
heading: 300
~2 degree climb?

and the right display is nav?

Also, isn't 259kts about 300mph? Isn't that slow for someone at 35,000 ft?
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

You got it...exactly right.

At about 31,000' indicated "airspeed" converts to mach for the rest of the climb. This is to account for aerodynamics at higher altitudes. So, yes, 250kts indicated seems slow, but when you calculate that .78 mach...their cooking along A-OK.

Lets say its 768mph X.78= 599mph- no corrections applied....
but you get the idea.
Last edited by TSMacNeil on May 27, 2011 9:32 PM, edited 2 times in total.
mitch
Posts: 961
Joined: Feb 12, 2006 12:00 PM
Location: Wheaton, IL

Post by mitch »

TSMacNeil wrote:I hope this helps a little, prolly boring the fuck outta you guys...sorry.
Nope. Highly interesting and informative.
Matt
Posts: 2345
Joined: Feb 12, 2006 12:00 PM
Location: Fargo

Post by Matt »

TSMacNeil wrote:You got it...exactly right.

At about 31,000' indicted "airspeed" converts to mach for the rest of the climb. this is to account for aerodynamics at higher altitudes. So, yes, 250kts indicated seems slow, but when you calculate that .78 mach...their cooking along A-OK.

Lets say its 768mph X.78= 599mph
Oh! So is the ground speed more like the mach number corrected to sea level -- the 599 figure you mention ? If so, that makes much more sense to me.
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

WINNING!
:laugh:
You just passed aerodynamics 301 with an A!
Matt
Posts: 2345
Joined: Feb 12, 2006 12:00 PM
Location: Fargo

Post by Matt »

Ok staying in a 10-15 kt range around a 260kt "center" is less "holy shit crazy" than what i was thinking of, which is more like staying 600 mph +- 15mph :)

If the conditions are calm, are there a lot of throttle (or other?) adjustments required to stay in the safe speed range? And is that engine control something the equipment does somewhat autonomously?
TSMacNeil
Posts: 7014
Joined: Jan 22, 2009 5:37 PM
Location: Phoenix, Arizona

Post by TSMacNeil »

You're asking great questions...smart.

Yes, the autothrust (autothrottles) will maintain a commanded or programmed speed pretty well. Of course, there is a small lag time between the computers detecting the "energy" (speed) slowing, and adding corrective power...but it's very good.

If you look carefully at that gauge, you may see right at the 260kt mark (current indicated speed) a very short amber arrowhead beginning to form pointing down (slower)...thats a predictive airspeed trend indication. That sucker is always jumping all over the place...it's looking about 7 seconds into the future to show you a speed trend.

Turbulence is the single biggest "affector" of speed control and "coffin corner" at any given time. When you fly, if it gets sufficiently bumpy, you'll notice the engines spool down. That is to keep from getting into the high speed barberpole. Turbulence makes these two parameters jump all over the place...AF447 was in the heart of a thunderstorm when this all went down...a fart could have made them stall.
Descending is the only action that spreads out the two speeds and provides a wider margin to work with.

A few more illustrations:
Image
this airplane is at 34,000' but its' margin is narrower than the other (30kts)...this plane is heavier....his "coffin corner" is at a lower altitude.

A depiction:
Image

I'm only elaborating so that all the Monday morning quarterbacking doesn't throw this guy down the shitter for being in a stall...it doesn't take much to stall a jet at these altitudes...and he didn't have the proper gauges to affect a recovery like he's undoubtedly performed 1000 times in his career....just like I have.
Last edited by TSMacNeil on May 27, 2011 11:44 PM, edited 2 times in total.
Post Reply