In case you missed it, a plane carrying 62 people crashed in Brazil yesterday:
There’s not much official speculation about the cause yet. Video footage from the ground (you can find lots of clips on YouTube) shows the plane coming pretty much straight down in something close to a flat spin, and it sounds like it may have been descending like this from near cruising altitude until it pancaked on the ground. Crash scene photos show the fuselage pancaked from top to bottom, consistent with having very little horizontal speed and a lot of vertical speed, and striking the ground with the fuselage in an approximately level orientation.
According to Wikipedia, there was an advisory for severe icing that included the plane’s cruising altitude (17K feet) and flight path. Until now, I wasn’t aware that icing could be an issue in anything other than sloppy winter weather. Even so, the ATR-72 (the aircraft type involved here) reportedly includes anti-icing systems, so wouldn’t the pilots be expected to turn it on? Even if they hadn’t, wouldn’t ice have melted off once they descended to lower/warmer altitudes?
Assuming ice is what knocked the plane out of a trouble-free cruise, my own speculation is that this is another instance of a cockpit crew panicking, pulling back on the yoke until the plane stalled, and then holding it there until they arrived at the crash scene, similar to what happened to AF447. It’s weird that there’s no reports of any communication between the pilots and ATC.
Input from other speculators - especially professional pilots - is most welcome.
I have no idea. It sure looked like a flat spin and there are several ways I could imagine it happening, but they are all pretty unusual. These days, planes do not generally fall out of the sky from their cruising altitude.
I’ve never flown an ATR but I did used to fly Saab 340s, another turboprop that’s vaguely similar. In the simulator we did stalls but you don’t really practice spin entries or recoveries in that sort of plane. The idea is to never allow something like that to even begin happening, which starts with never allowing the plane to stall. Once a big plane like that enters a flat spin you’re beyond test pilot territory and into the oh s**t flight regime.
I’ll be interested to know what happened, if it is discoverable. Terrible way to die - I don’t want to imagine their last moments.
I am not a pilot and this is purely speculation, but the mention of icing conditions immediately made me think of American Eagle 4184, an ATR-72 that crashed in Indiana due to icing back in 1994. That said, I am pretty sure ATR improved their anti-ice system after that crash, so I could be completely off base.
Not a pilot, but I’m wondering if the crew simply got into a situation where no amount of applying rudder in a certain direction or the other usual counter-spin measures would suffice to stop the spin. Perhaps also simply just having such low airspeed that any airflow that went over the control surfaces wouldn’t exert enough effect.
Well, yes. But the idea is to not allow a stall to develop which can lead to a spin - a stall is required for that to happen. Most airliners these days have several layers of stall protection including stick shakers (warning) and stick pushers (which give a shove forward if you ignore the shaker).
For reference, in my entire career I’ve never had a stick shaker activate. I’ve done a bazillion stalls in prop trainers and also spins when I was flying aerobatics. But it takes some real mishandling to trigger a stick shaker / pusher in a transport category aircraft. Now I’m not at all saying this crew was incompetent because we don’t know what happened. Just reiterating that this kind of plane entering a stall / spin from cruise altitude is very, very unusual.
One difference would also be that AF447 stalled in pitch-black night darkness, so the crew might have been clueless about what the orientation of the plane was (although I wonder if the artificial horizon would have been the telltale sign.) But this Brazilian plane stalled in plain daylight.
That’s why instruments like the artificial horizon exist: so you can understand the position and location of your aircraft without having to look out the window.
On AF447, Bonin responded to a minor icing of the pitot tube (and loss of airspeed measurement) by inducing a stall so deep that even the plane didn’t understand what was happening: the angle of attack became so large, and the airspeed so low, that the plane’s computer decided that this couldn’t possibly be a real stall, and so it stopped sounding the stall alarm. At some point when he stopped pulling back on the stick and the other pilot pushed forward on his, the angle of attack decreased and airspeed increased, and the plane’s computer decided that it was a real stall afterall, and sounded the stall alarm. This confused the hell out of the pilots, right up until the pancaked into the ocean.
Any chance yesterday’s ATR-72 had something similar going on, i.e. the stall condition was so severe that the plane’s instruments and warning systems might have been giving the cockpit crew conflicting/confusing information?
Suppose the crew had been able to stop the spin and push the nose down to build up airspeed and resume controlled flight. How much altitude would they have needed in order to successfully recover and pull out of the dive before crashing?
The American Eagle ATR accident was interesting in that what stalled due to icing wasn’t the wing, but rather the tailplane. In ice-free operations airplanes are designed so that can’t happen; the angles of incidence on wings and tails are set up such that the wing will always stall first.
In a tailplane stall, the wing is still flying. With zero warning from cockpit alerts, and little to no aerodynamic buffet in the airframe or control, the airplane abruptly noses over, generating a lot of negative G which is real hard on the people and disorienting on the pilots. Pulling back to recover from the dive is generally the right move given those symptoms. Except if the tailplane is stalled, in which case pulling back makes it worse and there’s nothing you can do to resolve your plight.
Meanwhile your AOA (calibrated for your wing, and an ice-free one to boot) and airspeed may well be (and probably is) fully in the slow-but-unstalled regime. So your instruments suggest a normal recovery from the unexpected dive ought to work. And that’s exactly how we train for unexpected dives due to [whatever] random cause: roll to wings level, then apply aft yoke commensurate with available airspeed and AOA to return to level flight. With any significant nose-down angle, airspeed will be building rapidly and if you do have a low speed situation right now, in just a few seconds you may be approaching the upper airspeed limits instead.
That was a mightily confusing and hopeless situation the AE crew found themselves in. As with many things in aviation, the only way to win that particular game is not to play.
No real answers in this article (gift link), but a fair amount of uncertainty from experts since (as also described above) it’s not the easiest scenario to have happen.
A flat spin is just that situation. Video makes it clear that airflow near the tail would have been tolerable close to vertical, meaning the elevator would be almost completely ineffective (horizontal stab fully stalled) and the rudder about the same.
I’ve seen speculation that the right amount of differential engine power could have broken the spin, but I think that’s far outside anything the pilots would have trained for.
T-tail aircraft, like the ATR-72 in this crash, are said to be particularly vulnerable to getting stuck in a deep stall because the horizontal stabilizer and elevator end up being in the aerodynamic shadow of the wing:
The exerpt above is from the beginning of the article, but the link directs you to a particular point further down the page where there’s a lengthy discussion of deep stall.
I’ve read that some copilots choose to never get promoted to captain because they think the first-officer life offers the ideal balance (enough pay to be happy, while not having the pressure or burden of being the commanding captain.)
Or, it could be that Silva had spent most of his career at another airline, which would mean him having to start from the bottom of the rung all over again when he joined Voepass, while Romano had spent his whole career with Voepass.
Yeah. One of many unofficial ideas discussed over beer that you hope never to have to try in desperation.
Which also raises the possibility that if one engine/prop went into flat pitch or reverse while in cruise due to mechanical malfunction, that could have driven them into that situation and also prevented them from recovering. Doubly so if they never correctly diagnosed why the airplane suddenly went (technical term) apeshit.
The various warning systems on engines and airframes are designed around the failures the designers anticipated. And collectively have a poor track record of producing confusing or misleading alarms when faced with an unanticipated failure. Thereby leading an unsuspecting (or especially unimaginative) crew down a wrong branch of failure analysis.
There’s no assurance something like this is what happened here. But my general prejudice is that when faced with something seemingly mysterious, assume the trigger was something thoroughly unanticipated.
Voepass Linhas Aéreas - Wikipedia is the sort of small 3rd rate operator that doesn’t represent a lifetime career for most of its employees. Such that age and seniority are very much not correlated.
I don’t mean by that “3rd rate” comment to cast any aspersions on the crewmembers’ ability as pilots, just on their luck/skill as employees. I too had a really shitty career by US big jet standards. Which said nothing whatsoever about my flying but everything about my skill at standing in the right / wrong hiring line at the wrong / right time. And being born in the wrong year.
Somebody mentioned on CNN that the lever to feather the props is right next to the lever to lower the flaps. Said it has caused an incident with this model aircraft at some time previously.
That sounds like a Bell 47 crash up here several years ago. The pilot’s wife (who may have also been a pilot – I don’t remember) wanted to pull on the carb heat, and pulled the mixture control (right next to it) instead to Idle Cutoff. They ended up in the drink off of Orcas Island (or one of the islands there). Fortunately, the helicopter was equipped with floats. Unfortunately, it turned over and was towed in upside-down. (IIRC, neither occupant was injured.)
If you’re handling the flaps or the prop levers in cruise you’re already having a bad day. Either one is going to cause you heartburn and a talk with the boss. At best.
OTOH, there was a passenger fatality event with a DC-10 back in the 1970s(?) where there was an inadvertent flap deployment in cruise and the airplane went (technical terms again) apeshit until the startled pilots recovered. But not before somebody fatally broke their neck bouncing off the ceiling. A bunch of other folks were injured, some pretty badly.
What happened was the FO was adjusting his seat upwards. The seats have several inches of range of motion and are electrically driven. He was switching from a seating position for eating to one for taking the controls. Anyhow, somehow his clipboard ful of paperwork, that was commonly left between the seat and the center pedestal became caught up in the seat and was raised along with the seat. The clipboard’s upper edge ended up directly under the flap handle and pushed that upwards in turn. Which handle came out of the “flaps up” detent enough to migrate unnoticed to trigger the deployment of the leading edge slats at cruise Mach, some M0.82 or so.
They are very fortunate they didn’t lose the whole airplane. If one more piece had broken off they probably would have.
Oops.
The punchline being that inadvertent or mistaken inputs to the wrong switch, knob, or lever certainly happen out there. Whether that’s bumping something completely by accident or doing the intended thing but not with the intended control, but rather the one next to it. Oops.