Germanwings A320 crash: Plane fell for 18 minutes (not 8)

The Scotsman ^ | March 25, 2015

[maggief]

That was quick.

http://www.nytimes.com/reuters/2015/03/25/world/middleeast/25reuters-france-crash-iran.html

Two Iranian Sports Journalists Among Victims of Germanwings Crash

...

The men were on their way to cover a match between the Iranian national soccer team and Chile scheduled to take place in Austria, the site reported.

[null and void]

compounded by the fire department running over and killing survivors.

A survivor. One. Covered in foam, laying down in a pile of foam.

[null and void]

I was going to post that if no one else did...

[Tzfat]

A flight expert on Fox just mentioned “angle of attack” sensors, which have failed before because they freeze up, citing an accident over the Mediterrean in 2008. It is a known problem and directives have been issued.

Partially true, that is actually conflating two separate issues. Air France 447 over the South Atlantic had frozen pitot probes, and the crew responded exactly opposite to what they should have done.

A recent Lufthansa had a Air Data Computer failure, which initiated an uncommanded 2,000 foot descent. The crew responded correctly and regained control.

This is a hot topic for Airbus pilots in the past few months. We have very specific and rapid procedures to handle both issues. None of which would prevent the crew from communicating for 10 minutes. Some emergency procedures are very involved and could preclude communication, but these are like that.

In the case of the first issue, there are no direct actions - just verification of attitude, and then a checklist. The second issue takes 2 seconds, pushing 2 buttons, and regains 100% control. That one is a memory-type response for all Airbus pilots world wide.

What is also missing is the descent into terrain. It simply cannot happen in an Airbus if the pilots are conscious. Similar to collision avoidance in modern autos, it sends off alerts well before hitting terrain. It measures rate of closure and alerts with sufficient time to completely avoid terrain, with maximum performance.

[Tzfat]

Just a hunch but the steady altitude loss described by all the sources of information for this accident, and the fact that they didn't appear to take action to avoid the terrain as they descended, and the fact that there was no distress call, strongly suggest incapacitation of the crew.

Yes, but not likely by hypoxia. All crew hypoxia accidents result in fuel starvation. This aircraft would have flown to the North Sea before crashing.

Hypoxia is extremely rare in commercial aircraft with two pilots. O2 masks are the first thing we do in any situation involving pressurization or smoke. We never would initiate a descent prior to putting on O2 masks.

[null and void]

Google says the altimeter on an A320 airbus is a good old fashioned dial.

A dial? What's driving the dial? Computer controlled, or a mechanical Bourdon tube to the outside world?

[Gritty]

Well, one thing we know isn't the cause of this crash is "terrorism". The US government ruled that out before the helicopters were on the scene.

[null and void]

Good points. But they didn’t crash onto level ground. Do you happen to know the altitude of the cliff they ended on, and the cloud ceiling?

[kjam22]

A computer controlled mechanical dial? Really?

[McGruff]

I wonder if a highly skilled programer would always have the same belief.

I was going to make a joke about the plane doing a Windows Update but decided not to, at least publicly.

In the interest of full disclosure, I are a somewhat skilled programer. This override check would be caught during certification testing. if not, someone's head would have rolled. Maybe I shouldn't use that phrase either.

[Tzfat]

As a matter of fact, I am a very experienced programmer too.

Consider that the A320 series has over 6,500 hulls flying. It may seem mysterious to many people how and why it does what it does, but tens of thousands of pilots know the A320 better than you know your car that you have driven for years.

[kjam22]

I just refuse to believe that the autopilot flew the plane from 38000 feet into mountains at 2000 feet while the pilots were sitting there, alert, doing their job, and somehow they just didn’t notice it until it was to late to do anything about it. You can argue that all day, but I won’t believe it.

[null and void]

Yeah, really. Dumber things have been done.

[Tzfat]

It’s not just me saying that, it’s Chesley Sullenberger, arguably the most famous A320 pilot in history. It’s not just me saying that, it’s Chesley Sullenberger, arguably the most famous A320 pilot in history.

I know Sully. If he said that, it would not matter. He has no credentials to prop up such a statement.

[Tzfat]

Is the altimeter a piece of software on an airbus?

2 digital and 1 analog. Not only that, there are at least 3 other systems that would start alerting long before impact with the terrain.

[null and void]

This override check would be caught during certification testing.

Unless it was an "after market" addition.

We've been known to add "upgrades" to Iranian SCADA controls on centrifuges.

No particular reason a sufficiently skilled and evil actor can't "upgrade" flight software.

But I do agree, that the factory software is as fail-safe as any human agency can make it.

[Tzfat]

I'd like to think that the radio communications don't get washed through the computer system. That would be silly, but engineers with a charter to build the first 100% computer controlled aircraft are apt to make everything computer controlled!

Yes, that would be silly. The primary radio will work if everything fails down to DC essential power. No computers.

BTW, the A320 is not 100% computer controlled. Not even close. Fly by wire simply means that it is electrical signals sent out to hydraulic actuators at the control surfaces.

[Toespi]

Tz, thank so much for your info. Funny thing you would mention the avoidance collision alert in cars. My husband and I are just settling a lemon law case with GM over our GMC Denali Terrain. This car is designed to have visual and audio alerts only of a potential crash which is fine. But, our car also does this crazy false alert thing, randomly out of nowhere, along with the loud audio warnings we get a sudden drag on the brakes. So you can be driving along and suddenly, ou of nowhere, the warnings go off and the brakes grab, which creates a sudden adrenaline rush and an “oh crap I’m going to wreck” moment. We instantly slam on our brakes, thinking emergency, once a car behind almost slammed into me. After multiple service appointments, GM has tried to say we are basically crazy, never had an issue Blah, blah, blah, only to find out many people have had problems. It is a false alert problem and the brake thing, which the car is not designed to do, is a computer glitch where it pre-charges the brakes in the event of am potential need to stop.

[kjam22]

Yep... I fly. Not those things. Look at the pics on my profile. Had the opportunity last summer to fly a WWII trainer. A tail dragger. Now that was cool! One pic of a friend dragging it out of the hanger. The other pic he took when he turned around and snapped one of me while I had the stick. Talk to you later!

[zipper]

Yes, but not likely by hypoxia. All crew hypoxia accidents result in fuel starvation. This aircraft would have flown to the North Sea before crashing.

Hypoxia is extremely rare in commercial aircraft with two pilots. O2 masks are the first thing we do in any situation involving pressurization or smoke. We never would initiate a descent prior to putting on O2 masks.

I said incapacitation, of which hypoxia is one possibility, or a contributing factor.

A rapid depressurization would put the crew to a test, of which they only have seconds to pass. There are no guarantees they will pass, especially if there are other factors in play (one is not in the cockpit, structural damage, insidious onset of symptoms, to name a few).