Some thoughts about the ATR-72 VOEPASS Flight 2283 accident

Recently, a Voepass ATR 72, Flight 2283 from Cascavel to Viacopos in Brazil, went out of control and crashed 7 miles off the airport. The aircraft was flying at an altitude of 17,000 feet (5,200 m) when it went out of control and began a rapid descent at approximately 1:22 p.m. local time. All 62 occupants on board died. It was the first fatal accident in Brazilian commercial aviation since the 2011 crash of Noar Linhas Aéreas Flight 4896 and the first fatal air accident involving Voepass Linhas Aéreas since its founding in 1995. The accident is also the worst in Brazil since TAM Flight 3054 in July 2007.

Every time there is an accident on an ATR, the debate about the safety of this aircraft is reopened, especially in icing conditions. The videos show the aircraft descending into a flat spin, very unusual for this category of airplanes. Any aircraft, before going into a spin, must obviously stall and when the stall occurs, there must be a yawing force (such as the application of the rudder pedal) and this will cause one half of the wing (the internal one) to stall first as the speed will be further reduced, while the tip of the external wing will tend to have greater speed due to the yawing itself (making it still maintain a slight lift unlike the internal half of the wing). The consequence is the famous wing drop. The angle of attack will increase more and more (in profiles like that of the ATR after about 12° of AOA the wing can be considered stalled) which will lead the aircraft to a rotation induced precisely by this wing drop. And so the plane goes into a spin.

In light aircraft, the spin recovery procedure is often summarised with the acronym PARE:

• P - Power to idle,
• A - Ailerons to neutral,
• R - Rudder opposite of the spin,
• E - Elevators forward, down.

Some aircraft will almost automatically exit a spin by idling and releasing the controls, while others are a bit more difficult and require special procedures. Multi-engine aircraft with the engines installed on the wings are more difficult to get out of a spin. One reason is that the engines are further away from the aircraft's center of gravity. Engines are also heavy and also create significant gyroscopic forces. This could lead to the development of a so-called "flat" spin (a classic spin tends to bring the nose down anyway, which does not happen in a flat spin). A flat spin could also be generated when the engines are not idled or the ailerons are applied in the opposite direction to the rotation (even if it is instinctive, in reality you are only stalling the inner wing even more by lowering the aileron, paradoxically it is more convenient to do the opposite), or the center of gravity is shifted too far towards the tail, etc.

There are many conditions that can lead to a flat spin. When we are facing a flat spin, the normal recovery technique does not work. This is because the tail is practically stalled too (both horizontal and vertical surfaces). Which means it  has no longerany effect. Transport aircraft such as ATR (but also Boeing and Airbus) are not certified and tested for the spin, but they have a whole series of systems, equipment, mechanisms and sensors that prevent it, first of all by avoiding the stall which is the primary condition that could lead to a spin.

This image also shows the yawing moments due to the rotation of the propellers, highlighting the left engine as the critical engine. Another reason why engines should be set to idle when you want to avoid yawing moments.

The ATR is equipped with a Stall Warning system that informs the crew before the stall, first through the stick shaker system when the speed drops below a certain value and then through the stick pusher when the AOA (Angle of Attack) sensor reaches a value that is too high. These systems also have audible warnings, disengage the autopilot, etc. (see this video for more details). Among other things, the Stick Pusher is tested on the first flight of each day and when it starts working it is so strong that it even overcomes the strength of the pilot. This prevents the aircraft from stalling.

In the case of the ATR the Apha Probe activates the stick push (after the Stick Shaker has activated) unless we are at 500 ft AGL.

According to the SIGMET of the Sao Paulo area, Severe Icing conditions were expected between FL120 and FL210. The ATR was at FL170 (I remember that the maximum altitude of the ATR is FL250, this means that it cannot normally fly above the usual icing conditions). What is meant by Severe Icing? It is that condition in which the rate of ice accretion on the surfaces of the aircraft is such that both the Anti-ICE and the De-ICE systems may not be able to reduce this hazard. Unfortunately the ATR does not have a good reputation in icing conditions, especially since the accident of American Eagle 4184 in 1994. This was followed by other accidents for Severe Icing, the last two were due to the delay of the ATC clearance to descend and during the wait they accumulated ice, losing speed and, consequently, stalling, losing control. ATR has demonstrated that descending to 3000 ft solves the problem in 90% of cases.

After the American Eagle accident, ATR implemented some changes and improved the anti-icing systems, demonstrating that the aircraft can fly even in severe icing conditions (see this video of the tests carried out at Edwards after the accident). The worst danger that can happen is that the stall warning system does not activate before the wing is contaminated with ice to the point of stalling first. For this reason, an additional system called APM (Aircraft Performance Monitoring System) has been installed on ATRs, which warns the crew of an abnormal resistance that reduces the normal performance of the aircraft (due to the ice accretion). The procedure to apply on the ATR in the event of severe icing is to maintain a speed of at least 30 kts above the minimum speed in icing conditions with flaps 0 (on the speed tape it is indicated with an amber bug) and to achieve this you almost certainly have to descend (in fact the aircraft does not have enough excess power to be able to do this at altitude in level flight with an accumulation of ice on the surfaces). For those who want to have an idea on the logic of the ATR speed bugs, you can click on this link. However, in this case we are faced with a case that must be recognized as an emergency, and therefore you must not wait for ATC authorization to descend (making an exaggerated example, it is like waiting for ATC authorization if you lose an engine). The most important thing in this situation is to maintain the right speed, it is the priority task (“speed is life”). Reconstructing the data from the ADS-B, the flight profile is truly anomalous, the plane loses and regains altitude several times with sudden changes in speed and unusual rates.

Obviously we have no further data so far and it is not right to speculate on what is not yet known, there will be an investigation and a Flight Safety report to do that for us.

If you enjoyed this post you can follow me on this BLOG or just subscribe to my Youtube Channel.

You can also support me on Instagram