MIA97IA050

As the first officer began the takeoff roll the aircraft began to drift to the left. Attempts to correct the drift by application of full right nose wheel steering through the rudder pedals was unsuccessful. The captain took control of the aircraft and aborted the takeoff. As engine power was reduced the aircraft turned sharply to the right and went off the right side of the runway where it came to rest on soft terrain. The aircraft center of gravity was near the aft limit. After the incident the nose landing gear was found to be over serviced and extended 2.35 inches higher than normal. DFDR data showed that as engine power was added for takeoff the nose of the aircraft pitched up extending the nose gear further. This allowed the centering cam to restrict nose gear steering to +/- 4 degrees. As the captain aborted the takeoff the nose pitched down and normal nose steering returned at which time the aircraft responded to the full right rudder input. DFDR and CVR data also show that as the captain aborted the takeoff he did not fully close throttles 1 and 2 and the asymmetric thrust contributed to the uncontrolled right turn.

On December 30, 1996, about 2226 eastern standard time, a Douglas DC8-62, N804AX, registered to ABX Air, Inc., ran off the right side of the runway when directional control was lost during takeoff roll at Orlando International Airport, Orlando, Florida, while on a Title 14 CFR Part 121 scheduled domestic cargo flight. Visual meteorological conditions prevailed at the time and an instrument flight rules flight plan was filed. The aircraft received minor damage and the airline transport-rated captain, first officer, flight engineer, and two passengers were not injured. The flight was originating at the time of the incident.

The flightcrew stated the captain taxied the aircraft on to the runway and after aligning it on the runway center line, gave control to the first officer, who would be flying the leg. The first officer advanced the thrust levers slowly toward takeoff power. After stabilizing engine power at 80 percent, the first officer continued to move the thrust levers toward takeoff power. The aircraft began to drift to the left. The first officer applied full right rudder; however, the aircraft continued to drift left. He reported this to the captain who took control of the aircraft. When the captain could not regain directional control, he aborted the takeoff. He applied reverse engine thrust and brakes in an attempt to stop the aircraft. The aircraft turned to the right and departed the runway where it came to rest.

Examination of the departure end of runway 18 right at Orlando International Airport showed that about 400-500 feet from the threshold, two parallel sets of light-to-moderate dark tire tracks about 20 feet apart were found beginning in the vicinity of the runway centerline. The tracks continued down the runway to the left of and away from the centerline at an initial angle of 20-30 degrees, then gradually arced back to the right. As the tracks crossed the runway centerline, a wide, very dark track appeared about midway between them. This track consisted of what appeared to be tire rubber residue, and was on average about 2 feet wide. All tracks continued off the right side of the runway to the point where N804AX had left the runway and come to rest. The aircraft went off the runway about 2,250 feet from the threshold and stopped on a heading about 50 degrees to the right of runway heading, about 75 feet from the runway edge.

Examination of the aircraft at the point it came to rest was performed by an FAA inspector. The aircraft had no obvious damage or mechanical failure/malfunction. The main landing gear and the nose landing gear had dug into the ground after the aircraft departed the runway, and the resultant trenches extended from the runway edge up to the respective landing gear. All of the landing gear tires were partially buried in the ground and only the tops of the tires were visible. All of the tires appeared to be intact. Examination of the nose gear revealed that it was turned about 45 to 55 degrees to the left.

Examination of the aircraft after removal from the runway showed the left nose landing tire was worn and had no tread remaining. The right nose landing gear tire had tread remaining. The nose landing gear strut was extended about 8.35 inches. The nose gear steering operated normally. At the request of NTSB the nose landing strut was serviced by ABX Air, Inc. personnel, using aircraft manufacturer procedures. After servicing the nose strut was extended about 6 inches. The nose landing gear strut fluid was found to be serviced to the proper level and the ground to air sensor was found to be adjusted within normal parameters.

Examination of the main landing gear showed that each strut was extended about 6.25 inches. After servicing the main gear struts were again extended about 6.25 inches. Each wheel turned freely and there was no evidence of failure or malfunction of the wheel brakes and anti-skid system.

Weight and balance calculations performed by ABX Air, Inc. showed the aircraft weighed 215,012 pounds and the center of gravity was at 30.11 percent mean aerodynamic chord at the time of the incident. The zero fuel weight was 167,812 pounds and the zero fuel center of gravity was 26.99 percent mean aerodynamic chord. The maximum allowable takeoff weight for the aircraft is 335,000 and the center of gravity limits, which are calculated for the zero fuel weight, are 18.20 percent mean aerodynamic chord forward limit and 31.66 mean aerodynamic chord aft limit. After the incident the cargo was removed from the aircraft and weighed under the supervision of an FAA inspector. The weights were found to be consistent with those used for the weight and balance calculations. (See attached weight and balance information and FAA Inspector Statement)

Readout of the digital flight data recorder (DFDR) from N804AX was performed by the NTSB Vehicle Performance Laboratory. The DFDR data indicate that the aircraft began its rolling takeoff on runway 18. About 2 seconds after being lined up with the departure runway the aircraft began a gradual turn to the left. At the same time a disparity in the Engine Pressure Ratios (EPRs) for engines 1 and 4 developed with engine 1 lagging the others in spooling up and engine 4 preceding the others in coming up to power. The left turn and engine disparity continued for another 7-8 seconds, at which point the power was reduced on engines 3 and 4 and the aircraft began a right turn which continued over the next 12 seconds finally resulting in a heading of 240 degrees. Engine 1 and 2 power was reduced slightly after engine 3 and 4 power was reduced but remained at about 1.40 EPR. After reduction of power on engine 3 power was increased momentarily to about 1.65 EPR. During the takeoff roll the maximum displacement of the control column was about 5 degrees nose down. The maximum airspeed attained by the aircraft was 68 knots. (See Flight Data Recorder Specialist's Factual Report of Investigation)

Readout of the cockpit voice recorder (CVR) from N804AX was performed by the NTSB Engineering and Computer Services Division. The contents of the CVR were transcribed. The transcript starts at time 2203:19 and continues to 2228:08 when electrical power was removed from the CVR. (See Cockpit Voice Recorder Group Chairman Factual Report)

A sound spectrum study of the CVR recording was performed by the NTSB Office of Research and Engineering. The purpose of the study was to determine the engine power settings during the takeoff roll. For most of the initial takeoff roll four distinct engine traces could be identified. At the start of the takeoff the engine traces show a continuous and smooth acceleration from about 41 percent N1 to about 94 percent N1. It took about 12 seconds for the engines to reach 94 percent N1. The signatures remained at 94 percent N1 for about 5 seconds. At this point all of the engines started to decrease. Two engine signatures decreased to a low of 44 percent N1. One of these traces then increased to 98 percent N1 for about 2 seconds and then decreased to 45 percent N1. The remaining two engines decreased to 85 percent N1 and remained there for 15 seconds before being reduced to about 43 percent N1. (See Specialist's Sound Spectrum Study)

Information about this incident was analyzed by personnel of Boeing Commercial Airplane Group-Douglas Products Division. The difference in power being produced by the No. 4 engine to that being produced by the No. 1 engine was determined to be about 2,700 pounds of thrust. They stated an EPR split of about 0.2 between two engines is not considered to be particularly unusual and is normally easily controlled via nose wheel steering through the rudder pedal from either pilot position.

Boeing personnel stated the finding of the nose landing gear being service to an extension of 8.35 inches is higher than the usual extension of 2 to 6 inches. They also stated that the nose up movement of .5 degrees as recor