posted on Nov, 11 2018 @ 08:36 AM
On several flights prior to the Mishap Flight, the aircraft experienced a slight rpm drop on the number one engine (approximately 1% lower than the #4
engine). The aircraft was flown to Savannah for scheduled fuel cell maintenance, and was originally scheduled to return to Puerto Rico afterwards.
Upon landing in Savannah, the Mishap Crew wrote the aircraft up for #1 engine RPM dropping to 96% upon switching the synchrophaser master switch off.
They believed it returned to normal when switching to mechanical governing. After completing repairs a crew arrived to troubleshoot the engine, and
make any required repairs.
The maintenance crew arrived without an Accu-Tach, which is required to verify engine RPM issues, so they borrowed one from the local unit. It was a
different unit than the one they normally used, and they were unaware that it had an adapter to fit their aircraft, so they failed to use it. During
the first engine run, the crew noticed the same problem with the engine, but were unable to get the engine over 96%, regardless of how it was
governed. The first step was to verify the gauge was operating correctly. This is usually done by changing the gauge with one known to be working.
Instead of shutting down the engines, the crew changed the RPM gauge with the gauge from the #2 engine while the engines were running. The swapped
gauge showed 96% still.
After completing the first engine run, the crew made the decision to adjust the engine one valve housing. They made a 12 click adjustment to the
housing, which should have increased the RPMs to 99%, which was within range of the T.O. requirements. During the second engine run, the maintenance
crew said they observed 99%, but the DFDR showed 96.8% sustained, and fluctuations between 95-98%. The T.O. calls for 99.8-100.2% on the precision
tach, or 98-102% on the aircraft gauge. The CVR showed the crew calling for an additional four clicks, but then discussing that doing so would
require another engine run, so it was never performed.
During the takeoff roll, engine #1 fluctuated between 94-98% RPM, and never reached the required takeoff RPMs. The Flight Manual calls for the crew
to abort the takeoff if any of the engines malfunction before decision speed is reached. The aircraft was at 64 knots when the FE called "Power Set".
Eight seconds prior to rotation, #1 dropped to 65%, and 1971 inch pounds of torque (normal for takeoff is over 18,000). Engine thrust had dropped to
under 300 HP, down from over 4,000. Just before rotation, the pilot applied 25 degrees of right rudder to keep the aircraft on the centerline. The
aircraft almost left the runway right before rotation.
The crew failed to follow the Engine Shutdown procedure, or follow the Takeoff Continued After Engine Failure, or After Takeoff Checklists. The flaps
were never retracted, and were left at 50 degrees. They began a left turn into the dead engine, but failed to maintain anything close to required
airspeed for flight. Three engine climb speed was calculated at 158 KIAS, and it was estimated that a left bank would require 166 KIAS. The MA never
achieved higher than 131 KIAS. The aircraft went into a 19 degree left bank, and the MP input left rudder, contrary to what the Flight Manual
indicated, resulting in the aircraft going into a skid, and departing controlled flight. The aircraft reached 52.2 degrees nose low, where it
performed a left barrel roll, and impacted the ground in an upright position, with a northerly heading.