MyCargo B744 at Maastricht on Nov 11th 2017, runway excursion on takeoff

Last Update: March 16, 2020 / 15:42:24 GMT/Zulu time

Bookmark this article

Incident Facts

Date of incident
Nov 11, 2017

Classification
Incident

Cause
Runway excursion

Airline
MyCargo Airlines

Departure
Maastricht, Netherlands

Destination
Jeddah, Saudi Arabia

Aircraft Registration
TC-ACR

Aircraft Type
Boeing 747-400

ICAO Type Designator
B744

Airport
Maastricht Aachen Airport, Maastricht

Airport ICAO Code
EHBK

A MyCargo Airlines Boeing 747-400 freighter on behalf of Saudia Cargo Airlines, registration TC-ACR performing freight flight SV-916 from Maastricht (Netherlands) to Jeddah (Saudi Arabia) with 4 crew, was accelerating for takeoff from Maastricht's runway 21 at 23:37L (22:37Z) when the crew rejected takeoff at low speed and deployed the thrust reversers. The aircraft veered right off the runway and came to a stop with all gear on soft ground to the right of the runway about 180 meters past the displaced runway threshold and about 360 meters after the begin of takeoff run (about abeam the first touch down zone markers runway 21). There were no injuries.

Saudia Cargo's parent Company Saudi Arabian Airlines reported the aircraft suffered an engine problem causing the runway excursion while attempting to takeoff. There were no injuries and no damage.

The Dutch DSB reported they opened an investigation into the runway excursion of a cargo aircraft at Maastricht Airport.

On Nov 16th 2017 the DSB reported that the aircraft of ACT Airlines was accelerating for takeoff when the aircraft suddenly pulled to the right, the crew could not correct, the aircraft exited the runway and came to a stop in the grass. The 4 crew remained uninjured, the aircraft sustained damage. The cargo and fuel were unloaded, then the aircraft was recovered and moved back onto paved surface. The investigation is ongoing with the assistance from Turkish, American and British accident investigation authorities.

The aircraft is still on the ground in Maastricht as of Feb 2nd 2018, a tent has been built around the aircraft to conduct the needed aircraft repairs, which suggests that other than initial information of no damage suggested, the aircraft sustained substantial damage (editorial note: which is why we changed the occurrence rating to Accident on Feb 2nd 2018).

ACT Airlines were renamed to MyCargo Airlines in 2011.

On Mar 16th 2020 the DSB released their final report concluding the probable cause of the accident was:

The runway excursion was caused by the pilot’s inability to maintain directional control under the conditions of prolonged asymmetric thrust that resulted from the loss of thrust on engine #4 at low speed. The loss of engine thrust was caused by a compressor stall.

Contributing factors The thrust levers were not retarded immediately after the loss of thrust. Such delays and not following trained procedures have been associated with the phenomenon known as the ‘startle effect’.

During training courses in flight simulators, the lessons learned from unexpected situations, such as engine failures, are quite limited as the crews know what to expect.

The DSB summarized the sequence of events:

After the aircraft was loaded, the captain, who was pilot flying, taxied to Runway 21 and initiated a rolling take-off. He advanced the thrust levers and, when the engines had stabilised, he pushed the TO/GA switches, causing the engines to accelerate to the selected take-off thrust. The aircraft had accelerated to approximately 30 knots when the outboard engine on the right side (engine #4) suddenly lost power. The aircraft veered to the right due to the resultant asymmetric thrust. The thrust levers were not immediately retarded to idle, so this asymmetric thrust continued. Attempts to steer the aircraft back to the centreline by means of nose wheel steering and differential braking, were unsuccessful. The aircraft could not be controlled. It veered off the runway and continued on into the soft ground on the right-hand side of the runway. The resistance of the soft ground and the eventual retardation of the thrust levers caused the aircraft to come to a standstill. None of the crew was injured. The aircraft sustained substantial damage.

The captain (63, ATPL, 18,550 hours total, 3,030 hours on type) was pilot flying, the first officer (42, ATPL, 4,462 hours total, 2,042 hours on type) was pilot monitoring.

The DSB analysed:

The crew received take-off clearance while approaching Runway 21. This enabled them to commence a rolling take-off, maintaining a ground speed of approximately 5 kts during the right turn to line up on the runway. At 22.36:34 hours, the captain advanced the thrust levers to approximately 70%. As thrust increased to the selected take-off power, a minor difference developed between engine #4 and the remaining engines. As a result, some pedal steering inputs were required to keep the aircraft aligned with the runway centreline. At this time, engine #4 (and, to a lesser extent, engine #3) was producing noticeably more thrust than the other engines. The steering inputs were to the right, as the difference in thrust was causing the aircraft to veer to the left.

When the engines had stabilised at 70% N1 the captain pressed the TO/GA switches, thus engaging the autothrottle. The captain’s call-outs and actions, and those of the first officer, were all in accordance with the SOPs. However, the first officer’s saying ‘Set thrust’, followed by the captain’s ‘Take off’ are not mentioned in SOPs or checklists. No explanation could be given for these calls. The whole sequence of the crew’s actions was in keeping with a routine flight, involving nothing out of the ordinary.

During the start of the take-off roll, the tasks were divided up as prescribed in the manuals; the captain was looking outside and keeping the aircraft in the middle of the runway, while the first officer monitored the instruments. Four seconds after the autothrottle was engaged, at a ground speed of 30 kts, a loud bang was heard and the N1 of engine #4 dropped to around 20%. In his interview, the captain stated that he did not hear the bang because he was the only one who was wearing an active noise cancelling headset over both ears and because the noise originated from the right side of the aircraft. The first officer had only one of his ears covered by his headset, while neither the technician nor the load master were wearing headsets. The fact that only these three individuals heard the bang confirms the assumption that the captain did not hear the noise because it was muted by his headset.

Unable to hear the loud bang, the captain could not understand why the aircraft was suddenly yawing to the right-hand side of the runway. This triggered the normal response of trying to counteract this yaw by steering to the left, by means of nose wheel rudder pedal steering. As the captain was still applying the right rudder pedal, it took some time before he could switch to the left rudder pedal once the right yawing occurred. The captain also stated that he used the nose wheel steering tiller to steer the aircraft to the left.

The DSB analysed the control inputs:

As he aligned the aircraft with the runway centreline, the captain increased thrust to perform a rolling take-off. Thrust is set using the right hand, so the captain’s right hand was holding the thrust levers. At that time, he was operating the tiller with his left hand.

This suggests that he was not holding the control column. This was confirmed by FDR data, which showed that zero force was being exerted on the captain’s control column at that point. The thrust difference between the engines during spin-up was such that the N1 of engine #4 was higher than the N1 of the other three engines. This resulted in a yaw to the left. As a result, the captain had to make corrections with the rudder pedals, steering to the right during the initial phase of the take-off.

The captain then felt the aircraft yaw suddenly to the right. Unaware of the cause, he tried to arrest this movement by using nose wheel steering inputs. As stated, he used both pedal and tiller nose wheel steering. However, tiller inputs override rudder pedal inputs.

When it became clear that nose wheel steering was not sufficient, he applied brake pressure. This occurred three seconds after the beginning of the event, as shown by an increase in the main gear brake torques. FDR data showed that, initially, both the righthand and left-hand main gear brakes were engaged. Shortly thereafter, braking of the right-hand gear decreased, which is indicative of differential braking. Although the aircraft had initially been heading a little to the left, it subsequently experienced an increasing deviation to the right. As take-off power was still engaged, the combined effect of tiller nose wheel steering and differential braking was insufficient to counteract the yaw to the right.

According to FDR data a control wheel input (steering to the left) was recorded shortly after the engine failure occurred. Thus, while the captain was holding the tiller with his left hand, he had most probably moved his right hand from the thrust levers to the control wheel. The captain was of the opinion that he had retarded the thrust levers almost immediately after noticing the deviation. However both FDR data and CVR data showed that the thrust levers were retarded to idle approximately eight seconds after the N1 drop of engine #4. This means that asymmetric thrust was acting on the aircraft for eight seconds.

The captain stated that he had tried to steer the aircraft to the left, using the tiller to keep the aircraft on the runway, while the first officer stated that he did not operate any controls. Thus, it was concluded that the captain’s left hand remained on the tiller until the aircraft had come to a full stop at its final resting place.

The DSB analysed the Rejected Takeoff Procedure:

During the take-off procedure, the first officer’s task is to monitor the engine instruments, call out any abnormal indications, and adjust take-off thrust as necessary until the aircraft reaches a speed of 80 kts. Initially, all engine instrument readings were normal. Thus his attention was focused on the air speed indicator, as he was required to call out the speed when the aircraft reached 80 kts. Immediately after the aircraft yawed to the right, the first evidence that an anomaly had occurred was the loud bang that he heard. No caution or warning was audible or visible on the EICAS, but the combination of the bang and the yaw indicated to him that the situation was very serious. That was probably why – almost immediately after these events – he called “off, abort, abort, abort”, in an effort to get the captain to abort the take-off. The interval between the bang and his abort call was a little more than one second. Partly because there was no caution or warning, he did not notice the drop in engine #4’s N1. As a consequence, he did not call out ‘engine failure’ as a reason for rejecting the take-off (the procedure specified in the manuals). Because the first officer did not announce the reason for the RTO, the captain did not know what had prompted the first officer’s abort call.

Despite the call to abort the take-off, the captain initially tried to keep the aircraft on the runway by a combination of steering and braking. The thrust levers were retarded to forward idle about eight seconds after the N1 of engine #4 dropped, and around six seconds after the first officer's call to abort. The selection of reverse thrust resulted in automatic disconnection of the autothrottle and extension of the speed brakes. At that point, the aircraft had already veered off the runway to the right and had rolled onto the grass. The thrust levers were not closed immediately, so they were still at the take-off power setting when the aircraft veered off the runway and rolled onto the soft ground.

By the time the thrust levers were moved to forward idle, the speed of the aircraft had already dropped almost to zero, due to the combined effects of the applied brakes and of the drag created by the soft ground.

Information obtained from the FDR, the CVR and the pilot interviews showed that the standard rejected take-off procedure had not been followed. Although the first officer was aware of the need to reject the take-off, the captain’s actions were consistent with an attempt to keep the aircraft on the runway by means of rudder pedal, braking, control column and tiller inputs. It can be concluded that, during the event, the captain had been using his left hand to operate the tiller and that he had moved his right hand from the throttles to the control wheel. This means that the captain was no longer holding the throttles, in contravention of the take-off procedure.

As the throttles were not retarded, a runway excursion became inevitable. This is because, if there is an outboard engine failure at low speed, while the remaining engines are at take-off thrust, it is impossible to keep the aircraft on the runway.

The DSB analysed they were able to rule out foreign object damage or weather conditions causing the engine compressor stall.

The DSB analysed that noise cancelling head sets as used by the captain cause concern because "the pilot may be unaware of environmental sounds and audible warning annunciations in the cockpit that do not come through the intercom system. Noise cancelling headsets are most effective over a narrow frequency range, but the specific frequencies may vary by make and model. Therefore, it is difficult to assess any effects the headsets may have on discerning environmental sounds ..."

The DSB analysed the "startle effect":

This investigation revealed that the flight crew's actions corresponded to some symptoms of the startle effect, as described above. In the interval until the aircraft had come to a complete stop, no mention was made of engine failure and the captain’s RTO actions were not monitored, nor was there any communication between the members of the flight crew. Such actions are consistent with the impact of a startle event. Someone asked ‘what happened?’ after the aircraft had stopped, indicating that their actions were prompted by surprise.

Metars:
EHBK 120025Z AUTO 23005KT 3100 BR FEW003 04/04 Q1010=
EHBK 112355Z AUTO 21005KT 1700 BR BKN002 OVC044 05/05 Q1010=
EHBK 112325Z AUTO 21005KT 2300 BR BKN032 OVC037 05/05 Q1010 REDZ=
EHBK 112255Z AUTO 22004KT 200V260 3600 -DZ FEW029 SCT033 OVC038 05/05 Q1010=
EHBK 112225Z AUTO 25003KT 210V300 8000 BKN040 OVC044 05/05 Q1011=
EHBK 112155Z AUTO 26004KT 230V310 7000 NSC 05/05 Q1010 TEMPO 4000 BR=
EHBK 112125Z AUTO 25005KT 220V300 8000 NSC 05/05 Q1010 TEMPO 4000 BR=
EHBK 112055Z AUTO 24004KT 210V280 3200 BR NSC 05/05 Q1010 TEMPO 2500=
EHBK 112025Z AUTO 23005KT 2500 BR FEW002 05/05 Q1010 TEMPO 1400 BKN003=
EHBK 111955Z AUTO 23005KT 1100 R21/1800D BR OVC000/// 06/05 Q1010 TEMPO 1800 BKN003=