ATR-72 Crash in Brazil (2 Viewers)

[XBe02Drvr]

ATRs and icing are a bad combination. Hang on to your hats, folks, this tale's going to be a long one. The ATR was the first aircraft to be globally certificated under the Joint Arcraft Airworthiness Standards Agreement, whereby all ICAO nations agreed to accept the airworthiness certification of an aircraft by it's originating nation at face value. No second guessing by other nations, as had been the case previously. Investigating the 1994 AmEagle crash at Roselawn IN, the NTSB determined that the unrecoverable spin was the result of a stall of the horizontal stabilizer, due to "bleed back" of ice behind the deicing boots, forming a ridge of ice on the (negative) lifting surface that detached the airflow and killed the horizontal stab's downforce. This caused the elevators to slam against the "nose down" stops, pitching the plane into a dive. When the speed increased and the stabilizer AoA decreased enough to reattach the air flow, and then stall negative, the elevators slammed nose up and drove the plane into an accelerated stall. A high G accelerated stall on a long slender wing like the ATR's is going to be super sensitive to the slightest asymmetry and highly likely to spin. With its long fuel filled wings and stretched fuselage, the ATR72 has a lot of distributed mass with a lot of rotational inertia, and it's narrow elevators are never going to get enough purchase in the roiled air of the rotation to lower the AoA and break the stall.
So what caused the bleed back icing that caused this whole event? Would you believe, droplet size? A lot of research by NTSB and FAA established that supercooled water droplets in superstable subfreezing air over the great plains of North America attained sizes seldom seen in the choppier air over Europe. These superstable droplets would remain as a flowable liquid longer after striking a fast moving airfoil, thus bleeding back farther before freezing than their European counterparts. In the case of the ATR, this was behind the boots. THE ATR HAD UNDERSIZED DEICE BOOTS!
This was news the Europeans didn't want to hear, and they insisted Roselawn was pilot error. The FAA revoked the ATR's icing certification, and the Canadians followed suit. All North American ATRs had to shift to sunnier climes, and Billy Clinton closed US airspace to foreign operated ATRs. In the end the US forced the ATR people to design and produce upgrade kits, and where feasible, to install them free of charge. This was accomplished by closing US airspace to all foreign airlines whose countries did not mandate the ATR fix. That policy was quietly maintained until the end of the Clinton administration. I don't think it's been maintained since. We may have just seen an outcome of that this week.
My ex went through AmEagle ATR72 initial training three years after Roselawn, and everybody in her class had to experience that event in the simulator. A real eye opener.

 

[Engineman]

The ATR has known issues with icing and the operating manuals reflect this. Stay within the limits in the operating manuals and the dangers are minimised.

Head in the sand response for flawed aircraft certification standards. TP aircraft are generally vulnerable to icing and are less safe than modern western jet airliners, and yet they are interlaced within the same mass transport operating system. This is unsatisfactory. TP aircraft could be much safer but, they are built down to a cost and have weaker systems, such as the rubber de-icing boots. They can be built to better standards but, that would erode their operating margins. They exist because of a veneer of safety, underlain with simple cheap operating costs and vulnerability. Unsatisfactory.

Eng

 

[RichardSuhkoi]

An ATR-72 with 62 people on board crashed today. The aircraft was filmed in a complete spin with virtually no forward airspeed.


View: https://x.com/AviationSafety/status/1821963083314835596?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1821963083314835596%7Ctwgr%5E83c54e82a2a719103f56d60a9b59a4893239a793%7Ctwcon%5Es1_&ref_url=https%3A%2F%2Fasn.flightsafety.org%2Fwikibase%2F409335

Here is a link to the Airworthiness Directive requiring the monitoring system to alert the crew to icing. The idea being that if the autopilot is flying it may mask icing and finally switch off and hand the pilot an un flyable aircraft. https://ad.easa.europa.eu/blob/easa_ad_2009_0170.pdf/AD_2009-0170_2?trk=public_post_comment-text
Any aircraft operated outside of its capability will suffer. I haven't looked it up but from memory the ATR72 is certified for 30 minutes operation max in severe icing. So any "flight into known icing" presumes the crew will strive to escape icing conditions in less than 30 minutes.
I fully expect that ALL of the de-ice equipment and the performance monitoring system referenced in the AD are on the Minimum Equipment List (cannot fly if these are broken). For these reasons I doubt that this accident occurred with a crew flying a serviceable aircraft within certification / Pilot Operating Handbook limitations.
At this point I am favouring the speculation regarding the recent structural repairs or hijack, or, less exciting, simple deviation from MEL / POH.

 

[RichardSuhkoi]

Head in the sand response for flawed aircraft certification standards. TP aircraft are generally vulnerable to icing and are less safe than modern western jet airliners, and yet they are interlaced within the same mass transport operating system. This is unsatisfactory. TP aircraft could be much safer but, they are built down to a cost and have weaker systems, such as the rubber de-icing boots. They can be built to better standards but, that would erode their operating margins. They exist because of a veneer of safety, underlain with simple cheap operating costs and vulnerability. Unsatisfactory.

Eng

I am curious why you find pneumatic "boots" weaker than other systems?
I have heard arguments about engine bleed air anti ice vs. Electrical resistive anti ice where the system cost and weights were compared. I recall seeing the duct losses as being very severe (the hot bleed air cools off on the way to leading edge), and the performance loss to the engine taking away from available power as may be required to maintain altitude on a heavy aircraft gaining ice weight. Due to high bleed temperatures, the ducts nearest the engine can be forced into Inconel to tolerate. Earlier stage air may not have the temperature required after duct losses.
The 787 system is all electric except for the engine inlet, and seems to be an excellent system but was integrated into the aircraft's "low loss high voltage architecture ". From memory the 787 uses a strategy where zones are heated in sequence as opposed to everything on at the same time, so is partly a de-ice system. I have read of ultrasonic and piezo shock systems in development.
I personally experienced astoundingly severe icing in a Beech 1900. Three people on board and we porpoised with both engines at max power (not TO) periodic violent roll (aileron still had authority to recover). Able to hold altitude but only at max throttle. The props were throwing heavy ice at fuselage, with impacts moving forward as the mass increased. When we landed, thick slabs of hoar ice were sliding off the aircraft, some maybe 10 cm thick. But, there was only modest localized ice on wing and tail surfaces. We stood silently on the tarmac just watching the ice fall off. The copilot was pale from stress. Due to the crew reporting ATC had moved all traffic out of that area.

 

[cammerjeff]

Here is a link to the Airworthiness Directive requiring the monitoring system to alert the crew to icing. The idea being that if the autopilot is flying it may mask icing and finally switch off and hand the pilot an un flyable aircraft. https://ad.easa.europa.eu/blob/easa_ad_2009_0170.pdf/AD_2009-0170_2?trk=public_post_comment-text
Any aircraft operated outside of its capability will suffer. I haven't looked it up but from memory the ATR72 is certified for 30 minutes operation max in severe icing. So any "flight into known icing" presumes the crew will strive to escape icing conditions in less than 30 minutes.
I fully expect that ALL of the de-ice equipment and the performance monitoring system referenced in the AD are on the Minimum Equipment List (cannot fly if these are broken). For these reasons I doubt that this accident occurred with a crew flying a serviceable aircraft within certification / Pilot Operating Handbook limitations.
At this point I am favouring the speculation regarding the recent structural repairs or hijack, or, less exciting, simple deviation from MEL / POH.

I agree with most of what you posted, and I agree that it is most likely going to found that Icing of the elevators will be found to be the cause. I used to work for an airline (Simmonds) that originally bought the ATR-42 & 72 aircraft that was purchased by American Eagle, and the ATR-72 involved in the Indiana crash was an aircraft I worked on. I left before the buyout by American Eagle. But I followed the investigation closely. And spoke with a couple of Pilots that flew the types that I stayed in touch with. They both had spoken of sluggish and erratic vertical performance of the types in icing conditions. But both had fortunately recognized the signs of icing before it became critical.
Also just to clarify most items on an MEL list can be inoperative. With conditions such as verifying that backup systems operate normally, and depending on the system can be inoperative for 10 or 3 days in most instances. There also can be flight restrictions such as Altitude, weight, or operating into know icing condition. Sometimes the restriction can be that it can be operated to a maintenance base for repair.

The other instance of the acceptance of international airworthiness approval I became aware of was the crash of a Fisher Brothers CASA C-212 at Detroit Metro Airport in the late 1980's. The aircraft was certified without a stall warning system installed by either the Spanish or European authority. The FAA accepts this if the aircraft has noticeable buffeting before the onset of a stall. In the case of the CASA C-212 did not have the warning horn, a stick shaker, or a natural buffet. And the international regulators approved it in that condition.
That approval cost the lives of 8 or 9 people that morning in Detroit. I watched it happen from the cockpit of a B-727 that I was Taxing from the Hangar to the Gate. I was standing by for clearance to cross the active runway that the CASA was landing on. It turned onto its back and if it would have gone to its right instead of its left it probably would have hit the 727 I was Taxing!

It continues to astound me that these events continue to occur decades after the issue is identified. At least the CASA's were soon out of service due to age and economic reasons. At least as far as regular passenger service is concerned.

 

[DerAdlerIstGelandet]

Pneumatic boots are just fine. They are a bitch to remove and install though.