The spinning ATR plane crash in Brazil | Hacker News Reader

The spinning ATR plane crash in Brazil

philip.greenspun.com

58 pointsmhb28 comments

> The aspiration seems to be an LCD version of the gauges and dials that were in a B-17 bomber over Germany in World War II. The computers on board the aircraft had all of the information that they needed to warn the crew “you can’t hold altitude at this power setting” long before they came anywhere near stalling.

I feel this way in cars as well - with a full LCD infotainment center why am I still forced to try and interpret inscrutable indicator lights?

My 2019 Subaru Outback will at least display things like "low fuel" and "low tire pressure" on the center display as well as on the drivers mini display. I haven't gotten into any check engine situations, or anything really scary like overheat or low oil pressure, but IMO even the data from a scan tool is somewhat inscrutable if you are not an expert.

Because they don’t distract.

Nothing I hate more than my Audi loudly beeping for non-critical infos/warnings. The only appropriate time for a car to loudly beep unprompted is if I need to stop immediately.

This reminds me of the BMW 3 degrees C bling. Which is the same bling as parking sensors dirty and thus no longer reliable.

I agree. If I want to drive at 15km/h without a seat belt in an empty car park the car shouldn't be beeping at me at 50 decibels warning me of immediate death.

Because if the automakers were left to their own devices, the infotainment display would need to show a loading bar before an ABS warning, and the speedometer would be buried there menus deep.

Check engine light can mean 100s of things. It would be cool to see the actual codes on the infotainment. Why do I need to hook up an OBD scan tool?

https://www.youtube.com/watch?v=5zaRCbnMhFo ATR-72 EN VRILLE by ATE CHUET

In French, a bit more information about ATR and icing as potential cause for this crash.

> Captain Sully had the stick full back during his heroic single-pilot landing on the Hudson, just like a panicked student pilot

Ouch!

I think the original author is drawing the wrong conclusion here. Sully may have had the stick all the way back, but that doesn't mean it was wrong. The lowest energy way to land is to try to prevent the aircraft from landing. So when you are in the flare, the technique is to pull back on the stick to keep the aircraft in the air a few feet above the runway, until you have no more elevator authority, and the aircraft will settle onto the runway with the least amount of energy. Exactly what you want in a off-field landing too.

As my glider instructor put it: those last few seconds before touchdown are the prime rib of flying. Keep pulling back on the stick and make it last.

In fact if you look at the Wikipedia article that the OP linked to it suggests that Sully was unhappy that the aircraft was not responding to his full back stick:

> However, Sullenberger said that these computer-imposed limits also prevented him from achieving the optimal landing flare for the ditching, which would have softened the impact.

My understanding is that transport-class aircraft (i.e. large passenger jets) are typically landed differently than smaller planes like a Cessna Skyhawk or a glider. In my brief flight training experience in a small plane, I was taught to stall the A/C just above the runway as you describe. However, I believe ATPs are taught to land at about 1.3x stall speed, then use the spoilers to dump the lift entirely and get full weight on the wheels in a very controlled way.

I'm of course discussing typical on-field operations, not emergency dead stick ditchings. I don't know how those are supposed to go, but one might imagine that going to full stall just before impact would indeed result in the minimum energy state. The trick is to not accidentally do that too high!

From what I read "the stick all the way back" is correct, because it's telling the computer your intent, "raise the nose of the plane". Airbus aircraft use flight laws to control the actual control surfaces, so there's no reason to make your intent less strong than it actually is.

Airbus airplanes have several different flight law modes (including direct law, which has no protections). So the biggest risk is probably that it switches laws and actually tries to carry out what you're doing.

I know, for all of Boeing's problems not doing that is a big plus. At least on the 777. I hope they haven't changed that since.

If your car suddenly reduced your turning authority, and you wanted to turn urgently, you'd crank the wheel harder.

The French designed a plane which actively interferes with your authority _while_ you're trying to use it.

And if we're going to give them credit for saving USAir 1549, then we have to give them blame for Air France 447, for making a plane that _averages_ the inputs of the two pilots to arrive at an overall control output, without any real warning, allowing two pilots to split controls and perform controlled flight into the ocean.

> for making a plane that _averages_ the inputs of the two pilots to arrive at an overall control output, without any real warning,

For what it's worth, the 777 does the same when there is too much opposing force on the yokes. At least on Airbuses, there is an aural warning.

You'd have to physically break through the jammed control mechanism to get to that point, otherwise, they're mechanically linked up to a certain amount of pressure.

Anyways, the point is, assigning blame in this way is petty and weird on the authors part.

> You'd have to physically break through the jammed control mechanism to get to that point, otherwise, they're mechanically linked up to a certain amount of pressure.

That's what I said.

Interestingly, the BEA pointed out that the dual input rate is the same on Airbuses and Boeings (cf. https://bea.aero/fileadmin/user_upload/F-GSQJ_finalreport_EN..., page 47). Lack of communication and stressful environment here too, maybe coupled commands would not have saved AF447 after all…

> Anyways, the point is, assigning blame in this way is petty and weird on the authors part.

I agree, but you're doing the same thing as the author of the article: taking the fact that the captain pulled on the stick, and try to find something to blame the plane.

Unlike in your car example where the steering authority is “suddenly” reduced, the flight envelope protection system is always here, and pilots are trained accordingly. Full stick back is not uncommon on Airbuses, e.g. in a windshear escape manoeuvre. And as pointed out by others in this thread, it was actually the correct action. Oh well, if alpha max was 1° higher on the 320, maybe the landing would have been softer.

See "Fly by Wire: The Geese, the Glide, the Miracle on the Hudson", by William Langewiesche. He points out that pulling the stick all the way back in a near-stall is what you're supposed to do in an Airbus. That tells the flight control system to do all it can to reduce the sink rate but to avoid a stall. Lacking engine power, the control system was able to provide a controlled crash. (If the engines had been providing power, the control system would have throttled them up.) Making the decision to ditch in the Hudson, rather than trying for, say, Teterboro and running out of altitude, was the important pilot action.

The real miracle of that episode was that there were enough boats nearby, including large ferries, to rescue everyone before the aircraft sank.

SO basically a quick-icing on one side/unevenly with the potential for pilot error is truly what caused it?

At the end he says...

"So… icing by itself likely cannot be the cause of the recent accident in Brazil."

Airline accidents almost always have multiple causes. Only when all the swiss cheese slices line up holes can the accident slip through.

Mentour Pilot four years ago on superstall (deep stall): https://www.youtube.com/watch?v=Kg6jV5mkf-8&t=0s

T-tail aircraft (like this one!) are at risk. ATR-72 specifically mentioned.

Thats why I specifically said with potential pilot error (I was going to say mechanical failure, but that is highly less likely given the age of the aircraft.)

>"Right, it was that bolded "400'" climb that clearly is what hes pointing at - as the ice caused the drop, then the pilot sharply climbed, then ice-stall, causes plane to yaw pilot corrects the wrong direction causing spin - all in the course of what seems like ~30 seconds?"

Is what my thoughts are.

> "The plane had been cruising at 17,000 feet until 1:21 p.m., when it dropped approximately 250 feet in 10 seconds. It then climbed approximately 400 feet in about eight seconds."

> The last sentence suggests that they were actively trying to get back to their assigned altitude of 17,000′.

I believe the author is implying that if the pilots attempted to maintain speed rather than altitude, the aerodynamic stall wouldn't have occurred - if they descended below 12000', the icing would be reversible.

Right, it was that bolded "400'" climb that clearly is what hes pointing at - as the ice caused the drop, then the pilot sharply climbed, then ice-stall, causes plane to yaw pilot corrects the wrong direction causing spin - all in the course of what seems like ~30 seconds?

It seems that the pilot error here was to fly into severe icing, or to staying in icing conditions beyond what the aircraft was capable of handling.

Or to not recover properly from icing.

Its early to make assumptions, but the weather report confirms the theory, and at the plane previous flight, from the inverse route, a passenger filmed ice at the wings.

Reply on news.ycombinator.com