Voepass Flight 2283, a large passenger plane, crashes in Vinhedo, Brazil

I'm confused at to why Juan Browne stated that turboprops (generalization) are "stuck right smack dab in the middle of it" (icing conditions) at FL 170.
The ATR-72-210 has a ceiling of FL 250. Why not stay above it, same as the passenger jet he compares it to?
 
I'm confused at to why Juan Browne stated that turboprops (generalization) are "stuck right smack dab in the middle of it" (icing conditions) at FL 170.
The ATR-72-210 has a ceiling of FL 250. Why not stay above it, same as the passenger jet he compares it to?
I doubt he's flown one...I haven't either. His content is entertainment.
 
I'm confused at to why Juan Browne stated that turboprops (generalization) are "stuck right smack dab in the middle of it" (icing conditions) at FL 170.
The ATR-72-210 has a ceiling of FL 250. Why not stay above it, same as the passenger jet he compares it to?
Just because it’s certified to that altitude doesn’t mean it’s practical.
 
The ATR-72-210 has a ceiling of FL 250. Why not stay above it, same as the passenger jet he compares it to?
Look at the duration of most turboprop flights. Usually a couple hours, and this one was 1h50min. Took them 30 min to climb from a field elevation of 2500ft to 17k. Around 500fpm climb rate.
Even assuming the plane could maintain that climb rate to FL250, by the time it reached that altitude it would have to start its descent.
 
Does anyone know at what altitude the plane exited the clouds? It's important because it gives us a sense of how long they had to get out of the spin.

The video is painful to watch because the spin remains flat throughout, and no apparent effort is made to get out of it (stick in center, push forward, and opposite rudder).

I also fly sailplanes and this recovery sequence is second nature to me.

Are airline pilots not trained for this simple maneuver? The Air France accident over the ocean in Brazil also hit the water with a flat spin. In this case they didn't trust their airspeed. Maybe the pilots in this accident refused to believe an airspeed of 20 kts too. GPS speed would be a clue.

Condolences to the families.

AF447 never entered a spin. It was held in a deep stall. It was recoverable with proper control inputs.
 
What about using differential thrust?
I've been told -- and it makes sense in hindsight -- that adding differential thrust will actually flatten the spin further.

Don't spin twins 'yall -- bad idea. Just because there's a gut wrenching delay between an action and death doesn't mean that death isn't guaranteed once that action is taken. Stalling a twin uncoordinated and initating a spin is near certain death, whether that's GA or transport category.

A lot of A/P manuals require disconnect in icing conditions. There's a reason for it. It isn't fun to increase workload right when you're also starting to manage the deice equipment and monitor the airframe (BTDT several times), but it sure beats not knowing the airplane is misbehaving / handling is changing in a really bad way...
 
I'm confused at to why Juan Browne stated that turboprops (generalization) are "stuck right smack dab in the middle of it" (icing conditions) at FL 170.
The ATR-72-210 has a ceiling of FL 250. Why not stay above it, same as the passenger jet he compares it to?
I haven't flown the ATR, but I did spend a some years on the Dash 8-300/400. The 400 had lots of power, but the 300 at times struggled to get much higher than FL180 in icing. Some colleagues had to divert because they couldn't get high enough to cross the mountains. You see a lot of weather flying at those levels.
 
I'm confused at to why Juan Browne stated that turboprops (generalization) are "stuck right smack dab in the middle of it" (icing conditions) at FL 170.
The ATR-72-210 has a ceiling of FL 250. Why not stay above it, same as the passenger jet he compares it to?
FL250 isn't "above it" at all!
 
The Air France accident over the ocean in Brazil also hit the water with a flat spin. In this case they didn't trust their airspeed.
Air France was a straight ahead stall. They lost their airspeed indications due to a problem with the pitot tubes, the autopilot kicked off and the crew couldn’t figure it out while falling 37,000 feet. And no, airline pilots are not trained in spin recovery.
 
Nope. See #68.
The aircraft was put by the crew into a deep stall. It was held there by control inputs from the right hand stick. AOA varied from 30 to plus 40 degrees. Most jet aircraft on approach have a AOA of 5 to 8 degrees. When those inputs were momentarily relaxed the aircraft began to recover. The stall was at such a high AOA that the system discontinued the stall warning. When the AOA began to reduce the warning came back on triggering the right seat pilot to once again pull back on the stick. There were literally dozens of other issues going on contributing. The aircraft was never in a spin! To understand everything you need to read this book..

Wikipedia also has a reasonably accurate explanation but greatly simplified.
 
Last edited:
The aircraft was put by the crew into a deep stall. It was held there by control inputs from the right hand stick. AOA varied from 30 to plus 40 degrees. Most jet aircraft on approach have a AOA of 5 to 8 degrees. When those inputs were momentarily relaxed the aircraft began to recover. The stall was at such a high AOA that the system discontinued the stall warning. When the AOA began to reduce the warning came back on triggering the right seat pilot to once again pull back on the stick. There were literally dozens of other issues going on contributing. The aircraft was never in a spin! To understand everything you need to read this book..

Wikipedia also has a reasonably accurate explanation but greatly simplified.
Still nope. An aircraft in deep stall, is impossible or at least very difficult to get out of. And a deep stall isn't held with control inputs. Once you're in a deep stall, the controls are mostly useless.

Peter Garrison covered this. I trust him a LOT more than Wikipedia.

Velocity had an issue with deep stalls when they developed a new wing. One of their aircraft entered a deep stall and the pilot rode it down into the Atlantic. He had a chute but elected to stay with the aircraft because he wasn't sure he would be able to get clear. The airplane was towed to shore and eventually flew again. They created a mechanism that moved a large weight fore and aft so they could do testing.
 
Garrison says the aircraft was deeply stalled! He says it was not a deep stall. I have no idea what he means by that. It appears he believes it’s a deep stall if it can’t be recovered but deeply stalled if it can. As I mentioned AF447 was recoverable but the crew held it in a deep or deeply stalled condition. If you want to argue semantics have at it.
 
Air France was stalled by aircrew because it was in some undefined alternate law regime, that happened because of the pitot system input failure.

Aircrew is not taught how to handle this at all because it was considered too unlikely to happen.

Exacerbated by the fact there was no visual reference.

Deep stall, shallow stall, whatev… not Germaine to the conversation. A basic failure. ANY of us are prone to doing the same thing, again not relevant.

ALL troubleshooting needs to start with the very basics. In this case, power plus attitude equals performance. Period.

In the case of an airplane likely icing up while IMC… different deal. The aircraft has known problems dealing with icing. They were likely doomed WAAAAY before they ever knew there was a problem. By the time it manifested itself, they were passengers. Very sad. The fact it was a flat spin made for a HORRIBLE video… doesn’t matter, the fact it can’t deal with the ice it encountered is what the REAL issue is. They still would have likely hit the ground in an unflyable state, what ever that was gonna be.
 
Garrison says the aircraft was deeply stalled! He says it was not a deep stall. I have no idea what he means by that. It appears he believes it’s a deep stall if it can’t be recovered but deeply stalled if it can. As I mentioned AF447 was recoverable but the crew held it in a deep or deeply stalled condition. If you want to argue semantics have at it.
It's not a question of belief (or semantics).

The defining characteristic of a deep stall, then, apart from the absence of a yawing component, is that you cannot get out of it by normal use of the controls.


Put another way, if the stall can only be held by use of the controls, it's not a deep stall.
 
Air France was stalled by aircrew because it was in some undefined alternate law regime, that happened because of the pitot system input failure.

Aircrew is not taught how to handle this at all because it was considered too unlikely to happen.
I've never flown an Airbus but find it hard to believe that there wasn't an unreliable airspeed procedure, probably with memory items.

It would have them disconnect the autopilot and auto-throttles, and turn off the flight director cues, then set a predetermined pitch and thrust setting that will produce stable, but not necessarily level, flight. That gives them time to work the problem and determine what data is good and what is bad.

On my current airplane that's 80% N1 and 10° ANU with flaps extended and 75% and 4° with flaps up.

Like the two MAX crashes, the crew did not follow the established procedures which would have allowed continued controlled flight.
 
Northwest/Delta had multiple airspeed failures in high altitude icing events. One in particular was a carbon copy of AF447 with one difference. The crew flew the aircraft manually until they exited the icing area. They did not have any issues controlling the aircraft. The AF447 copilot lacked even the most basic instrument scan beyond the flight director. I saw this more and more in the last 10 years. Tell someone they are off and they reply, “I am on the flight director”. If they were scanning they would have noted the altitude deviation and understood why it was happening.
 
ATR-72's rarely cruise higher than FL200, so Juan Browne is correct that they tend to spend a lot more time in the icing zone.

Just because their published ceiling is FL250 doesn't mean they can practically cruise at that, especially one that was close to fully loaded like this.
 
I am not aware of any non centerline thrust multi engine civilian aircraft certified for spins. Zero. All pilot training and higher level avionics (stick shakers and stick pushers) are to avoid a spin in the first place. There’s just not enough rudder authority to overcome the moment arm of the heavy engines (and outboard wing and tip fuel tanks) even if you can get the nose pointed down.
That does not make multi engine planes unsafe. Thousands are flown every day without spinning.
 
Startle response. All bets are off if you get there…
 
ATR-72's rarely cruise higher than FL200, so Juan Browne is correct that they tend to spend a lot more time in the icing zone.

Just because their published ceiling is FL250 doesn't mean they can practically cruise at that, especially one that was close to fully loaded like this.
its been a long time since I flew the ATR, but the alpha will go to 250 with most loads. its just that on shorter routes it doesn't make much sense to go that high due to the time to climb. fuel burn is really about the same anywhere above 180 in it. the ATR handles ice ok, it just does not handle severe ice well. everybody that flys it knows what happened at roselawn and the lessons learned from it. you do not stay in ice with it, you exit asap. if you see ice on the rear side windows you exit now, your in territory you do not want to be in. the book says turn off the A/P and increase speeds by 10 kts. most use more than 10kts.
in the ATR when entering icing conditions you bug icing speeds, the boxes will tell if they crew did that. the CVR will tell the tail of what they saw and how they dealt with it. it could be a case of instantaniouly severe icing that the airframe could just not handle. if it did depart like roselawn did, with aileron reversal, trying to bring the wing up with aileron just made it worse.

the boxes will be an interesting data set.
 
its been a long time since I flew the ATR, but the alpha will go to 250 with most loads. its just that on shorter routes it doesn't make much sense to go that high due to the time to climb. fuel burn is really about the same anywhere above 180 in it. the ATR handles ice ok, it just does not handle severe ice well. everybody that flys it knows what happened at roselawn and the lessons learned from it. you do not stay in ice with it, you exit asap. if you see ice on the rear side windows you exit now, your in territory you do not want to be in. the book says turn off the A/P and increase speeds by 10 kts. most use more than 10kts.
in the ATR when entering icing conditions you bug icing speeds, the boxes will tell if they crew did that. the CVR will tell the tail of what they saw and how they dealt with it. it could be a case of instantaniouly severe icing that the airframe could just not handle. if it did depart like roselawn did, with aileron reversal, trying to bring the wing up with aileron just made it worse.

the boxes will be an interesting data set.
Thanks for the first hand insight!
 
This commentator discusses the ATR's problems with icing and stall/spin recovery. He also mentions that starting with the one in 1994, there have been four previous accidents in the type in which icing was involved.

 
Last edited:
ATR might want to stick those NASA spin test tail parachutes on it.
Should also figure out how to get a hot wing anti ice setup.
This commentator discusses the ATR's problems with icing and stall/spin recovery. He also mentions that starting with the one 1n 1994, there have been four previous accidents in the type in which icing was involved.

 
Last edited:
After a little time in a King Air 90 putting around at FL180-200…I really appreciated the Citation II at FL350+, but still those tall T storms can be 40-50k’ high. Had to go around the tall stuff but was in VMC and no ice. Jets zip up through the icing altitudes and don’t linger. You slug it out in the turboprops. Juan Browne is correct.
 
Interesting. They didn't seem to be too concerned about the icing, at least not enough to even tell ATC they needed to get out of it NOW.
 
I need everything dumbed down, can someone abbreviate this event to a useable summary (if enough information is currently available)?
Something like (I have no idea if this applies) "Failed to turn on anti-ice/de-ice in the presence of abundant warnings".
I realize not all things are so simple but pages of reporting and comments leave me wondering.
 
I need everything dumbed down, can someone abbreviate this event to a useable summary (if enough information is currently available)?
Something like (I have no idea if this applies) "Failed to turn on anti-ice/de-ice in the presence of abundant warnings".
I realize not all things are so simple but pages of reporting and comments leave me wondering.
It’s almost exactly as you said, but it’s a series of things.

The flight was limited to FL170 due to an inop pack. That put it close enough to the freezing level for long enough to encounter some significant icing. There was a de-I’ve failure warning, but they don’t say why.

Then there’s a myriad of ice alerts from the wing sensor (and it should have also been noticed on the visual probe). Several more warnings in the several minutes prior to the accident, mostly ignored in favor of communications and the approach briefing. They were held at altitude longer because of traffic crossing below them.

The APM system activated, giving three subsequent, increasing warning levels of loss of performance, all of which indicate icing. These warnings and severe icing are called out with procedures in the AFM, to include keeping speed up, activating anti and de-icing, and getting the hell out of it. Just after the first officer finally noticed the ice, ATC gave them a turn, but the airspeed was rapidly decaying and it stalled.

There’s also a simulation video floating around that shows that they put pro-stall/spin inputs on the controls that may have contributed to flattening the spin. Poor show, all around.
 
Had to go around the tall stuff but was in VMC and no ice. Jets zip up through the icing altitudes and don’t linger. You slug it out in the turboprops. Juan Browne is correct.
and yet we still get people on placing a premium on piston fEekY. As I've said, stupid party tricks abound.

I gotta say, hell of a prelim. Maybe the NTSB should go down to Brasilia and learn a thing or two from the so-called global south. Leave clownshoes, bring sunblock. :rofl:
 
ATR-72's rarely cruise higher than FL200, so Juan Browne is correct that they tend to spend a lot more time in the icing zone.

Just because their published ceiling is FL250 doesn't mean they can practically cruise at that, especially one that was close to fully loaded like this.
Truth. And this particular flight was held lower (FL170) than normal due to an inoperative pack. And they “had to” stay there because of crossing traffic.

If I had to come up with a situation to make a plane ice up and enter a flat spin, it would be exactly all the things that happened on this flight.
 
Back
Top