Landing: The Time of Maximum Hazard

The final descent and landing is, statistically, the most dangerous segment of an airline flight. In recent years, the number of such accidents is high, accounting for hundreds of passenger and aircrew deaths. Bad weather is often a contributing factor, as is poor pilot training.

The latest crash during the landing phase involves a FlyDubai B737-800 with 55 passengers and seven crew aboard on 19 March 2016 at the city of Rostov-on-Don in Russia. Everyone died as flight FZ981 dove right onto the runway, smashing the airplane and its occupants into bits. It was the pilots’ second landing attempt, aborted at the last second but too late to avoid a fiery crash during the heavy winds that prevailed. Severe turbulence and wind sheer may well have violently tossed about the airplane in the last moments of flight.

All that's left of FlyDubai flight FZ981

All that’s left of FlyDubai flight FZ981

From what is known about the airline, founded in 2008, poor pilot decision-making was involved, although details remain to be fleshed out in coming weeks and months by Russia’s Interstate Aviation Committee. However, enough details have surfaced to allow some informed commentary.

Task fixation. When aircrews are fatigued, there is a human tendency to focus on the task at hand, repeatedly, even if these actions are revealed by the investigation to be questionable. A great contributor to task fixation is fatigue; an operator who is significantly rest-deprived will tend to focus on one task, to the exclusion of other actions.

News reports assert that both pilots may have been experiencing fatigue late in the day of the crash. The pilot-in-command of the flight, 38-year old Captain Aristos Socratous of Cyprus, had filed paperwork with the carrier to leave because of the unbearable schedule. A former colleague and captain with the airline said, “The reason that the captain was resigning is because of the schedules. He just couldn’t do it anymore. He was too tired, going to work fatigued, and that is why he had resigned.”

The co-pilot, 37-year old First Officer Alejandro Álava, a Spanish national, had worked 11 days with only one day off before the fatal flight. According to his flight log, he had been transferred from daytime to nighttime flights without being given adequate transition time to adjust his sleep pattern.

“There is no doubt has was fatigued and exhausted for this flight,” the former FlyDubai captain asserted. “That definitely was a contributing factor, no matter how [FlyDubai] may try to deny it.”

In cases of chronic fatigue, pilot behavior becomes more erratic. Thinking is foggy, instruments are not monitored, and errors go unmissed or not acted upon. Task fixation is one of the many insidious effects of fatigue.

FlyDubai is the low-cost sister of Emirates Airline. Its lower cost may be the result of packing more seats aboard and flying a demanding schedule with fewer aircrews. There are only a few ways a low-cost operation can trim expenses. Flight FZ981 departed from Dubai International Airport for the nighttime flight to Rostov-on-Don Airport, one of the airline’s 11 destinations in Russia. Upon nearing the airport, air traffic control warned the crew about “severe turbulence and moderate wind shear.” Wind speed was approximately 30 mph gusting to 40 mph. Wind shear involved …..

Broken cumulonimbus clouds scudded by at 3,600 feet.

Two airliners successfully landed immediately before flight FZ981 made its descent to land. The crew aborted the landing at 1,700 ft above and 4 miles from the runway because of the windy conditions. They put their B737-800 into a holding pattern.

Soon after, an Aeroflot airliner made three failed attempts to land and diverted to nearby Krasnodar Airport.

FluDubai flight FZ981 orbited for approximately two hours, the pilots hoping the weather would improve. It is not publicly known what the pilots discussed during their time in the holding pattern, or what communications, if any, were held with the dispatcher at FlyDubai’s operations center. Surely, the possibility of diverting to Krasnodar, or elsewhere, was discussed. While the B737-800 was consuming fuel, after the four hour flight from Dubai it still had substantial fuel to divert — the airplane took off with eight hours fuel pumped aboard.

The aircrew decided to make a second landing attempt at Rostov-on-Don. Here was task fixation with a vengeance. The airplane got within 3.4 miles of the runway threshold, but the crew decided to abort the landing for the second time because of the gusty wind conditions.

Vasily Golubev, the governor of the Rostov region, and the home of most of the 55 passengers, said, “In all likelihood, the cause of the air crash was heavy winds approaching hurricane strength.”

It is probable, but not known with certainty, that Captain Socratous was the handling pilot; in which case First Officer Álava would have assumed the duties of pilot monitoring. According to one report, either Socratous or Álava turned off the autopilot and activated the take off/go around (TOGA) button. (See https://www.rt.com/news/

337400-pilots-conflict-boeing-rostov/) In any event, the airplane was put into a steep climb with speed declining.

“Wait! Where are you flying? Stop! Stop!” a voice on the cockpit voice recorder apparently shrieked.

By the time crew actions were coordinated, it was too late. The airplane was plummeting downward at nearly 200 mph at an angle of approximately 45 degrees, then 60 degrees Immediately off the approach end of  Runway 22 a mighty explosion marked the impact point.

Would a well rested aircrew not functioning in the window of circadian low (when the body’s demand for sleep is most pronounced) even attempted landing under these conditions? Probably not. The fatal wages of task fixation.

Both pilots had received their training on Western-built aircraft.

Reacting wrongly to instruments. There are other cases where the pilots have momentarily misread their instruments — especially if they received training and gained experience on Russian-designed aircraft.

Case in point: the fatal crash in 2008 of Aeroflot-Nord flight SU821 while descending to land at Perm Airport, Russia, killing all 88 of the occupants. The domestic flight originated at Moscow. The Russian pilots were flying a Boeing B737-500. In the late afternoon, the weather at Perm was cloudy.

The air traffic controller radioed the pilots that the aircraft was too far to the right of the glide slope, and shortly thereafter the aircraft was viewed on the ground radar as climbing.

“Roger, we are descending,” the aircrew radioed. However, the pilots failed to comply and the air traffic controller ordered a go-around.

The report by Russia’s Air Accident Investigation Commission of the Interstate Aviation Committee noted, “After the base turn, approaching the landing course at 600 m with both autopilot and autothrottle disengaged, the aircraft started climbing up to 1,300 meters, rolled 360° over the left wing and collided with the ground.”

The final investigation report cited the following reasons for the crash: the loss of spatial orientation by the crew (Captain Rodion Medvedev was the pilot flying; First Officer Rustam Allaberdin was the pilot monitoring) and chiefly by the captain who was the handling pilot during the landing phase. The plane banked left, rotated onto its back and went into a rapid descent from an altitude of 600 meters. The loss of spatial orientation in the night, while flying in the clouds, with autopilot and autothrottle switched off led to the crash. Poor crew resource management and insufficient training for using the Western type of attitude indicators contributed to the accident, the report said.

Both pilots had previously flown Tupolov Tu-134 and Antonov An-2 aircraft, which featured Russian designed instruments. In the Russian design of the primary flight display (PFD) the airplane’s bank angle is shown by a moving aircraft symbol, while the horizon remains fixed.

In the Western design of the PFD, the airplane symbol is fixed and the artificial horizon moves. A pilot trained and accustomed to the Russian-designed PFD must be constantly conscious of the difference; a lapse while looking at the Western PFD and momentarily assuming the Russian convention can lead to loss of control of the aircraft. At an altitude of just 600 meters — 1,968 feet — and in an extreme attitude, there was insufficient height to regain control.

Captain Medvedev had only 452 hours in the B737; First Officer Allabertin had just 219 hours in the B737. In the waning hours of the day, in clouds, and a momentary confusion over what the PFD was displaying could easily lead to incorrect control inputs and an unrecoverable attitude. Add in the “inadequate practices” by Aeroflot-Nord regarding the operation of the B737 and the “mild intoxication” of Captain Medvedev, according to the investigation, and the set-up for disaster is complete.

It was not the first time, and probably not the last, where the difference between the Russian-designed PFD and the Western-engineered counterpart played a role in an accident.

 

The Western-designed primary flight display for aircraft attitude has the airplane fixed, with the horizon moving as the airplane maneuvers (on left). The Russian design (center) has a fixed horizon with the airplane symbol moving. A new design (at right) features both moving horizon and airplane symbol

The Western-designed primary flight display for aircraft attitude has the airplane fixed, with the horizon moving as the airplane maneuvers (on left). The Russian design (center) has a fixed horizon with the airplane symbol moving. A new design (at right) features both moving horizon and airplane symbol

In other cases, the wholly Western design and pilot training in its use may lead to fatal consequences. The case of the fatal August 2000 crash of a Gulf Air A320 while attempting a nighttime landing at Bahrain comes to mind. On the first approach the airplane was descending much too fast. The captain declared a go-around, but he did not perform a standard go-around. Instead, he circled at 500 ft. altitude, with flaps not adjusted and a positive rate of climb not established.

Coming out of the orbit, the jet was not properly configured for a second approach. The captain applied TOGA thrust but only 5° nose-up pitch, instead of the requisite15° nose-up pitch. As a result of the 5° nose-up pitch, the airplane at TOGA thrust accelerated rapidly.

This rapid acceleration can lead to a phenomenon known as false climb or “pitch up” illusion. The pilot’s brain misinterprets the stimuli, perceiving incorrectly that the airplane is climbing. Consciously or subconsciously, the pilot experiencing this illusion might apply forward side stick to correct his perceived nose-high attitude. This causes the aircraft to accelerate even more, causing a worsening of the illusion. The pilot suffering false climb illusion can rapidly “correct” his way right into the ground or, in this case, the inky dark waters below.

Then the master warning sounded, indicating a flap overspeed condition. The so-called “barber pole” would appear in the speed portion of the PFD. At the time the overspeed warning occurred, the airspeed was 191 knots. The maximum speed for the “flaps three” configuration was 185 knots.

Indicated airspeed is shown on the left with red "barber pole" descending to show an overspeed condition

Indicated airspeed is shown on the left with red “barber pole” descending
to show an overspeed condition

Two seconds after the warning, the captain pushed forward on the side stick controller, thus pushing the nose down and increasing airspeed even further. He held the side stick forward for the next 11 seconds, pushing the nose down -15°. At low altitude, it was positively the worst response. The airplane struck the Persian Gulf waters at a point three miles northeast of Bahrain International Airport. All 143 aboard the A320 were killed.

Why did the captain push the side stick forward to drop the nose? With an overspeed — especially at low altitude — the correct response would have been to pull back on the side stick to increase pitch, hence angle of attack, and thereby decrease the airplane’s speed. At fixed TOGA thrust, speed is controlled by pitch.

But to increase pitch would have meant moving into the descending “barber pole” on the speed tape. By dropping the nose, in a direction away from the barber pole, the pilot actually increased airspeed during the overspeed condition. The captain pushed when he should have pulled back on the side stick.

As a general precept, people are accustomed to “move away from the red.” For example, the correct response when the needle on an automobile tachometer goes into the red is to back off on the accelerator. However, this situation required the pilot to fly into the red in order to bleed off excess airspeed.

One surmises that the barber pole is moving in the wrong direction, inviting the wrong response from the pilot who may have lost situational awareness (the infamous “pitch up” illusion).

To be sure, it is difficult to believe that this seasoned captain would pay less attention to the primary attitude display while attempting, incorrectly, to deal with the slow moving “barber pole” warning on the adjacent speed tape display and its insistent “ding, ding, ding” aural warning.

This writer suspects that the barber pole, instead of descending, should have been rising. The pilot then would have been induced to increase pitch, not decrease it.

From these examples of landing catastrophes, two succinct observations come to mind:

  1. Fatigue and its attendant symptom of “task fixation” can rob an aircrew of all available options. They’ll keep trying to land when prudence dictates a diversion.
  2. Different instrument designs can, in the moment of confusion, lead pilots into irrevocable loss of control mistakes.

Both problems emanate from complacency. The FlyDubai pilots were operating under a perfectly legal schedule, albeit one at the limits of permissibility. The Aeroflot-Nord and the Gulf Air pilots were set up to fail by non-standard and pernicious instrument designs.

Jetliners frequently fly above foul weather. But to safely get back on the ground, everything should be done to ensure the crew is well-rested for critical decision making, and the instruments on which they rely should be standardized and intuitively logical.

Neither is the case.