Accident DC-8 in Anchorage (1970)

The Douglas DC-8-63CF ^{[* 1]} freight-passenger with the registration number N4909C (factory - 46060, serial - 472) that was involved in the accident was released around May 1969 , since on July 2 it received a flight certificate. The board N4909C was sold to the corporation CIT which in turn leased it to the American airline Capitol International Airways . The last periodic inspection of the aircraft took place in Wilmington ( Delaware ) on November 26, 1970, the day before the crash. During this test, six wheels of the main chassis ( No. 1 , 2, 4, 6, 7 and 8) were replaced, and the No. 3 wheel brake was replaced. After that, the Douglas made four flights, including landing at Anchorage before the fatal takeoff, with no comments on it. In the past, comments on the board N4909C were only in relation to its steering wheel. So, from September to November 26, 1970, there were eight requests that in the process of taxiing the plane sometimes leads to the left, while it is difficult to turn right. To eliminate these requests, the left wheel was replaced on the front landing gear, the steering mechanism was missed, the right and left steering cylinders were replaced, and the steering trimmer was adjusted. November 24 was the application "It is difficult to turn right," then replaced the left-hand steering cylinder. There were no comments on the other systems ^{[1] [2]} .

The cabin of the aircraft could be converted for the carriage of goods and passengers, and in the latter case it was possible to install up to 250 seats for passengers with a 31 inch (790 mm) pitch. In this flight, in accordance with the requirement of the charterer, the passenger capacity of the cabin was 219 seats with a spacing of 38 inches (970 mm) ^[3] . The total operating time of the N4909C board at the time of the disaster was 4944 hours 49 minutes, including 11 hours and 11 minutes from the last periodic check (in Wilmington) ^[1] .

The flight crew (in the cockpit) consisted of two pilots, a flight engineer and a navigator:

• The aircraft commander is 48-year-old William G. Reid ( born William G. Reid ), who worked at Capitol International Airways on January 1, 1955. He was qualified as the commander of the aircraft Lockheed Constellation , Curtiss-Wright C-46 Commando and Douglas DC-8 , as well as single-engine aircraft. Had a total plaque of 14,650 hours, including 5740 hours on the DC-8 type. 90 days before the incident, flew 257 hours, and 30 days - 87 hours. In his own words, over the past 60 days, he flew to Anchorage Airport 10 times, and all flights were carried out on DC-8-63 aircraft ^[4] .
• The co - pilot is 55-year-old James A. Downes ( born James A. Downs ), who worked at Capitol International Airways since May 28, 1962. He had the qualifications of the commander of the aircraft Lockheed Constellation , Douglas DC-3 and DC-4 , as well as single-engine aircraft. Wore contact lenses. In May 1970, he began training for the commander of the DC-8, but after six training flights the classes were discontinued, as he did not learn the material badly, which required additional flights on the DC-8, while these planes were badly needed on passenger routes. Therefore, on June 9 Downes returned to the position of co-pilot DC-8. Had a total raid of 13,500 hours, including 2,057 hours on the DC-8 type. 90 days before the incident, flew 227 hours, and 30 days - 83 hours ^[4] .
• Flight engineer - 41-year-old Edward W. Fink ( born Edward W. Fink ), who worked at Capitol International Airways since May 12, 1964. He had the qualifications of a flight engineer with piston and turbojet engines. His total flight time was about 10,000 hours, including 2,000 hours on the DC-8 type. 30 days before the incident, flew 69 hours ^[5] .
• The navigator is 53-year-old Robert D. Leonard ( born Robert D. Leonard ), who worked at Capitol International Airways since February 28, 1966. Wore contact lenses. Had a total plaque of about 14,000 hours, including 2,500 hours on the DC-8 type ^[5] .

All four at the time of the incident had been working for 7 hours and 20 minutes, including 3 hours and 45 minutes of flight time, and before that they had rested 24 hours ^[5] .

6 flight attendants worked in the cabin ^{[5] [6]} :

• Marlin Faystoer ( Eng. Marlene Faistauer , in the airline since June 11, 1968)
• Alexandra Plommer ( eng. Alexandra Plommer , in the airline since June 11, 1968)
• Barbara M. Ogden ( eng. Barbara M. Ogden , in the airline since June 9, 1969)
• Alice B. Mendez ( Eng. Alice B. Mendez , in the airline since June 9, 1969)
• Britta E. Thonsen ( eng. Britta E. Thonsen , in the airline since May 23, 1970)
• Birgitta I. Ekeland ( Eng. Birgitta I. Ekelund , in the airline since May 23, 1970)

Arriving in Anchorage

The board N4909C was chartered by the Military Transport Command for a charter flight C2C3 / 26 from Tacoma ( WA ) to Cam Ranh ( South Vietnam ) for the transfer of a group of 213 military and 6 more members of their families. Also on the route were scheduled stops for refueling in Anchorage ( Alaska ) and Tokyo ( Japan ). At 12:04 ^{[* 2]} with 219 passengers and 10 crew members aboard the Douglas flew out of McCord airbase (near Tacoma), and after three and a half hours at 15:32 in the evening twilight (sunset was 14:59 ^[7] ) landed safely at Anchorage Airport . This flight took place in the usual monotonous mode, although during landing it was necessary to use reverse and maximum braking in order to stop on the ice-covered strip. The braking effect was weak, so when taxiing to the parking lot, only a slight slowdown was used. Next, to stop on the platform, parking brakes were applied, which were then released after the engines stopped ^{[8] [3] [9]} .

The aircraft was inspected by a technician, who also checked the condition of the tires, due to the fact that maximum braking was applied during the landing. But the surface of tire shells was normal, and no abnormal temperatures in the area of ​​the wheels were noticed ^[8] .

The crew on the airliner had a remark about the inflated readings of the air pressure ratio (EPR) in engine No. 1 (far left), and also that the readings themselves were unreliable. Then the aircraft for the period of refueling opened the casing of this engine, but found no deviations in this engine. After this inspection, the engine cover was closed again. During the test, it was found that the EPR sensor of engine No. 1 failed, but since the sensors on the other three engines worked normally and did not notice any deviations in the operation of those three engines, they decided to continue the flight. Refueling went without comment, and only 117,227 pounds (53,173 kg) of A-type jet fuel was poured into the tanks. The total estimated weight of flight 3/26 was 349,012 pounds (158,309 kg) with a maximum allowable 350,000 pounds (158,757 kg) ^[8] .

Going to the lane

At that time overnight Anchorage was already standing, the sky covered with overcast with a lower boundary of 1,700 feet (520 m ) and some clouds with a lower boundary of 400 feet (120 m ), visibility on the strip of 5 miles (8.0 km ), wind eastern 8 knots, air temperature 23 ° F (−5 ° C) , dew point 22 ° F (−6 ° C) , light overcooled rain. Subsequently, it was found that due to precipitation, the airfield was covered with a thin layer (1-2 mm) of ice ^[7] .

Before towing and taxiing, both half-wings of the aircraft, horizontal stabilizer and control surfaces were treated with heated ethylene glycol solution to clear them of ice. At about 4:54 pm, the N4909C side departed from the ramp, after which the crew received permission to follow the runway "06 right" (06R, length 10,900 feet (3,300 m ), currently 07R). During the flight to the strip, the flight crew performed a reading of the checklist before taking off, with the exception of the points on switching on the respondents and checking the ignition of the engines. At 17:00:25 the dispatcher gave permission to occupy the 06R lane, after which the Douglas drove to the executive start, being at the same time slightly to the right of the center line. According to the aircraft commander, the parking brakes from the moment of departure from the apron, including waiting at the executive start, were not used ^[10] .

The crew made a decision among themselves that the takeoff would be carried out under the control of the co-pilot, and the commander would monitor the brakes. At 5:02:40 PM, the dispatcher gave permission to take off, and at about 5:03 pm the crew finished reading the remaining items on the checklist. For current conditions, the decision making speed (V ₁ ) was defined as 138 knots , the lifting speed of the nose strut (V _R ) was 153 knots, and the safe take-off speed (V ₂ ) was 163 knots. The commander transferred the engine control levers to 80% of the nominal, while the brakes were still pressed and the plane remained in place. Then the brakes were released, after which the commander told the co-pilot “Let's go” ( eng. Lets go ) and also increased the engine mode to 1.87 EPR (the mode of the first engine due to the inoperative sensor was set according to the position of its lever flush with the others) ^{[10 ]} .

Take off

After 1 minute 34 seconds after the executive start was taken, the plane began to run along the runway, while the co-pilot immediately leveled it along the center line, and the commander periodically reported speed values, which initially grew normally. But at a speed of 130-135 knots, the acceleration suddenly decreased, that is, the speed increase began to occur more slowly, because of which the speed V ₁ was reached at the 59th second, later than usual, but since the engine parameters were normal, the decision on no takeoff was taken. Then, at a speed of 145 knots, the acceleration returned to normal, so the speed began to grow at a normal pace. At 72 seconds of takeoff and 1500–1800 feet (450–480 m) to the end of the runway, the commander, in his words, reported reaching the speed V _R , after which the co-pilot raised his nose at a pitch angle of 9 °. Next, the liner was supposed to take off, but instead continued to rush along the strip. When the commander realized that they were now rolling out of the strip to the ground, then acting at the level of self-preservation instinct , at a speed of 152 knots he immediately removed the engine mode in order to stay on the ground and “save his plane, passengers and himself”. After a moment, they rolled out of the band ^{[10] [11] [12]} .

All the interviewed passengers described that the Douglas ran 2000-3000 feet (600-900 m), and rather slowly, after which a series of three pops rang out like tires exploded. Then the plane lifted its nose, as all the engines stopped. With a crash, a heavy car rolled out from the runway to the ground, after which, at a distance of 675 feet (206 m ) from the end of the runway, a wooden fence was demolished, and at a distance of 1002 feet (305 m ) from the runway, the left wing crashed into the course beacon. On the left side of the fuselage, the passengers at that time saw fire. At this moment, one of the passengers on the move opened the emergency exit door on the port side, through which the fire penetrated the cabin. At 5:05 pm, at a distance of 3,400 feet (1,000 m ) from the Douglas strip, it crashed into a deep drainage ditch and stopped, with the fuselage ripped in half, and the right wing plane separated ^{[11] [12] [13]} .

Rescue work

Tons of aviation kerosene leaked from damaged fuel tanks and formed a huge puddle around 6-8 inches (15-20 cm) around the aircraft, then they ignited, causing an extensive fire that was especially strong in the tail section. Within three minutes, the first firefighters arrived and began to extinguish the fire with dry reagents. After 5 minutes from the moment of the incident, all the airport's rescue services were already working at the site, as well as the city calculations. Then on the left side there were two powerful explosions, as a result of which many people trying to escape, were thrown back into the salon engulfed in flames. Among them was one stewardess, who at the same time lost consciousness and the passengers had to carry her in their arms. Then another series of smaller explosions occurred, which also significantly hampered rescue operations ^{[7] [13]} .

The flight crew (4 people) managed to evacuate through the windows of the cabin and then began to help evacuate people from the right side, as the door from the cabin to the cabin was jammed. Most of the passengers escaped through the main exits, which were opened 3 of 4 (one of the exits in the tail was jammed with torn seats), emergency exit over the right wing, as well as a rift in the fuselage. As a result of the fire, the cabin of the aircraft burned out completely, while 47 people were killed: 46 passenger soldiers and one flight attendant, Birgitta Ekelund . It was established that they were mainly in the tail section, or at the fault line of the fuselage and died from exposure to fire and smoke. The lost stewardess sat in the 33rd row at the left emergency exit. 49 people (43 passengers and 6 crew members) were non-fatally injured, and the remaining 133 (including 3 crew members) were not injured relatively ^{[7] [14] [9]} .

The voice recorder was badly damaged in the fire and its data could not be recovered, so I had to rely on the testimony of the pilots ^[15] . The parametric recorder survived and the investigators were able to decipher the data of the last few minutes ^[12] .

The study of the band 06R, from which the fatal takeoff was carried out, was started only on November 29, that is, two days later, when part of the prints were destroyed. However, then managed to find traces of the wheels of the left main chassis. When rubbing against the tire strip, the ice was broken and a brake mark was left on the strip, which stretched from the start of the run and even along the yellow lines of the strip. After 560 feet (170 m ) from the start of the left chassis run, partially melted and partially torn pieces of rubber from the wheels were found. Also at a distance of 5,000 to 6,000 feet from the start of the takeoff run (1,500–1,800 m), pieces of rubber were found, which mostly lay on the right side of the runway. Further along the strip, the investigators continued to find other pieces of cord from the wheels of the main chassis. Thus, the claps that the passengers heard were really caused by the explosions of the wheels of the chassis ^[16] . In the snow, on which the liner rolled out from the strip, only continuous tracks from the main landing gear were found, and tracks from the left landing gear at this moment were to the right of the continuation of the runway center line. Also at a distance of 545 feet (166 m ) from the strip between the tracks of the wheels, an imprint of 71 feet (22 m ) appeared — at this moment the Douglas raised its nose so much that it struck the ground with its tail ^[17] .

A check of the parking brake indicated that it was off, and no indication of a fault could be detected. The position of the usual brakes at the time of take-off could not be determined, because their cables were destroyed. The flaps were released at 23 ° at the moment of take-off, and the stabilizer was set at an angle of 4.2 ° for tailing. The chassis lever was fixed in the lower (chassis release) position ^[18] .

Specialists from NASA were involved in the investigation, with the participation of whom the following picture of events was established. After obtaining permission to occupy the lane, the Douglas left for the executive start, taking a position two miles from the front, then stopped. Waiting for permission to take off was one and a half minutes, during which the airliner stood with braked wheels. При этом лёд под левыми шасси оплавился, так как они нагрелись примерно до 60 °F (16 °C) , когда во время поворота экипаж использовал раздельные левые тормоза, а также из-за повышенной нагрузки на лёд вследствие высокого взлётного веса. Затем когда разрешение на взлёт было получено, экипаж отпустил тормоза и увеличил тягу двигателей, после чего самолёт начал выполнять разбег. Однако за всеми четырьмя колёсами левого шасси стал наблюдаться тормозной след, то есть колёса не вращались. Была версия, что раз коэффициент трения резины по льду очень мал, то колёса потому и не стали вращаться. Но проведя дополнительные тесты, в НАСА определили, что тормоза всё-таки не отпустили до конца, а зацепления колёс по льду не хватило, чтобы провернуть их и тем самым отделить тормоза. Почему колёса правого шасси не оставили тормозного следа, следователи точно определить не смогли, но есть вероятность, что лёд на самой полосе оказался неравномерным, то есть на правой стороне его было больше, да и шины на правом шасси не были так нагреты, как на левом. Коэффициент трения корда по ледяной полосе в момент начала движения был оценён как 0,025, но после того, как шины нагрелись, упал до 0,019, что в принципе соответствует нормальному трению качения, а потому самолёт и начал словно нормально разгоняться, из-за чего взлёт не стали прерывать ^{[19] [20] [21]} .

Разгон с заклиненными колёсами приводит к тому, что от трения по асфальту шины начинают всё сильнее тереться и перегреваться, после чего колёса стали взрываться. Уже спустя 560 футов (170 м ) от начала разбега происходит первый взрыв колеса, после чего последовали остальные. Спустя 4300 футов (1300 м ) от начала разбега все колёса на левом шасси оказываются разрушены, то есть теперь по асфальту стали скользить их металлические диски. На удалении 4300 футов (1300 м ) от начала разбега лопается последнее целое колесо на правой стойке. Экспертиза всех восьми колёс основного шасси подтвердила, что ни одно из них не вращалось в момент разгона по полосе ^{[21] [22]} .

Компания McDonnell Douglas провела тесты по взлёту с заторможенными колёсами на покрытой льдом полосе, по результатам которых полученные данные примерно совпали с фактическими. При этом был отмечено, что разница с набором скорости при нормальном взлёте и взлёте с заторможенными колёсами относительно небольшая, а потому экипаж не смог определить заклинивание колёс. Таким образом было точно установлено, что борт N4909C не смог взлететь из-за того, что его колёса не вращались. Причины этого могли быть следующие ^{[23] [24]} :

• Возник отказ в тормозах или их гидросистеме, из-за чего сохранялось тормозное нажатие, достаточное для заклинивания колёс.
• Колёса не могли вращаться из-за высоких сил трения в подшипниковых узлах, вследствие перекоса, вызванного неверной установкой колёс.
• Экипаж мог забыть отпустить тормоза перед взлётом.

Следователи разослали запросы 12 американским операторам DC-8 с просьбой сообщить, были ли у них случаи, чтобы на этих самолётах отказывала или работала с нарушениями тормозная система. В ответ сообщили, что отказов тормозов ни разу не наблюдалось, хотя в чуть ли не единичных случая бывали задержки с отпуском, либо медленная реакция тормозов на команды. Возможной причиной таких нарушений в работе подозревали падение давления в тормозной системе из-за заедания распределительного клапана, либо попадание воздуха в тормозную жидкость. Данные нарушения могли наблюдаться как на тормозах всех шасси разом, так и только на тормозах основных шасси ^[20] .

Проверка элементов тормозной системы разбившегося самолёта не выявила на них признаков неисправностей или отказов. Стояночный тормоз удалось восстановить и определить, что в момент взлёта он был отпущен. Зазоры между тормозными дисками были в пределах нормы, а на самих дисках признаков перегрева, заклинивания, навара или какой-нибудь другой неисправности найти не удалось. Рычаг аэродинамического тормоза находился в убранном положении, но цилиндр пневмопривода был разрушен, а потому определить, действительно ли аэродинамический тормоз не был выпущен, установить не удалось. Большинство компонентов гидропривода тормозов оказалось уничтожено в пожаре, а потому следователи не смогли изучить эту часть тормозной системы. Из истории самолёта было обращено внимание на заявку, что при рулении по аэродрому иногда уводит влево, в чём подозревали нарушения в приводе носового шасси. Уже во время расследования была высказана версия, что на самом деле авиалайнер уводило в сторону из-за заедания тормозов на левой стойке основного шасси. Однако этому противоречило то, что такую серьёзную неисправность, как отказ тормозов только на одной стойке достаточно быстро должны были обнаружить и устранить ^{[24] [25]} .

The version about wheel jamming on the axles due to mounting violations was based on the fact that in Wilmington , Delaware , before leaving for Takoy to perform the fateful flight, 6 out of 8 wheels of the main chassis were replaced on board the N4909C. All wheel bearings were found to be in working condition and none of them had color or surface state changes that were supposed to indicate abnormal performance. Wheel disks also had no signs of bias and distortions ^[25] .

When studying the version that the brakes were forgotten to be released, or inadvertently pressed the brake pedals at the time of acceleration, the investigators interviewed both pilots. The commander said that during the waiting period, he held the brake pedals with his toes, and after he increased the engines' mode before acceleration, he released the brakes, shifting his legs to the steering pedals of the front desk. The co-pilot said that his feet were only on the taxi pedals, heels resting on the floor. None of the pilots felt that the brake pedal was pressed during takeoff ^[25] .

For the current conditions (taking off from the icy lane), when the coefficient of friction of the wheels on the lining surface was very small, it was enough to have a slight brake pressure so that the wheels did not rotate during movement. And then the investigators paid attention to the parking brake, which could create this pressure. The crew claimed that he did not use the parking brake after driving away from the apron, including waiting on the runway. But after all, they used it earlier, and then they could simply forget it, since their attention was also taken away by the inactive sensor of the leftmost engine. The lack of data from the voice recorder did not allow us to find out how the checklists were actually carried out before taxiing and before taking off. Before taxiing, the pilots could think more about leveling the operation mode of engine No. 1 relative to number 2 , after which they could start taxiing with wheels that were actually slowed down, which allowed ice at the aerodrome. The parking on the lane was 1 minute 30 seconds and slowly, with the implementation of prejudicial procedures, the crew could again miss the point on the brakes, and then begin to take off with the parking brake applied. But after interviewing the crew members, the investigators were forced to reject this version. The fact is that the warning lights of yellow color, which are bright enough not to notice them, especially at night, warn about the activation of the parking brake, and besides they are visible not only to both pilots, but also to the flight engineer. No one in the cockpit saw that this indicator was on ^{[26] [27]} .

Thus, the investigators could not confirm any of the versions about the reasons why the wheels did not rotate during takeoff ^[28] .

On March 29, 1972, the National Transportation Safety Board (NTSB) issued a final report, AAR-72-12, in which it concluded that the Anchorage crash was due to the fact that the plane during take-off could not pick up the speed necessary for takeoff . The reason for this was the reduced acceleration caused by blocking all the wheels of the main chassis, which because of this did not rotate, which remained unnoticed by the crew until reaching the decision-making speed. It was not possible to establish the reason why the wheels of the main chassis were braked. The probable reason for this may be the occurrence of pressure in the hydraulic actuator of the brakes due to the failure or malfunction of the braking system, or the unintended activation of the parking brake ^[30] .

According to the results of the investigation of the crash of the N4909C crash, the following was recommended ^{[30] [31]} :

1. Add an item to the take-off procedures so that the crew checks the acceleration relative to normal in time or distance.
2. Revise the design of fuel systems and improve their protection against fire.
3. Revise the layout of the cabin of transport aircraft, including the location of flight attendants.

It is worth noting that the first proposal was sent to the US Federal Aviation Administration on March 10, 1963, but in both cases it was rejected as unacceptable. The second and third sentences were introduced in 1978-1983 ^[32] .

• Yak-42 crash near Yaroslavl - taking off with braked wheels

Comments

Sources

• Capitol International Airways, DC-8-63F, N4909C, Anchorage, Alaska, November 27, 1970. (English) . National Transportation Safety Council (29 March 1972). The appeal date is April 28, 2015.

• Capitol N4909C air crash (English) . General aviation portal. The appeal date is September 28, 2015.