NTSB Factual Report

HISTORY OF THE EVENT

On December 26, 2008, at 0859 pacific standard time, an American Airlines McDonnell Douglas MD-80, (N9617R) overran and impacted a tug that was positioning the airplane during pushback at Los Angeles International Airport. No injuries were reported to the 124 passengers, 3 flight attendants, 2 pilots, or the tug operator; however, the airplane sustained substantial damage to the forward fuselage.

The airplane had been cleared by ground control to push back onto taxiway C from its position at Gate 48B. The airplane was connected to the tug and was under power and control of the tug during the push back. Soon after the initial pushback, the tug operator cleared the flight to start its engines, and the captain initiated the start sequence for the left engine. The engine start was normal.

As the pushback continued and the airplane was turned onto taxiway C, ground control instructed the flight to pull forward slightly because of concerns that it would obstruct an adjacent taxiway. The pilots reported that the tug operator transitioned to forward motion just as the airplane began stopping from the pushback. The airplane at this point was offset slightly to the left of the taxiway centerline.

The pilots reported that as the tug began pulling the airplane forward and to the right, the tug began lurching right and then left with greater amplitudes. This continued for a couple of cycles for about 3-4 seconds until the tug jackknifed and became stuck under the right side of the nose of the airplane. Two flight attendants were thrown to the floor during the airplane's movement. The airplane briefly came to a stop but then continued to drift slightly to the left. The captain indicated that he didn't think they were attached to the tug anymore and he applied the brakes and brought the airplane to a stop.

DAMAGE TO AIRPLANE AND TUG

The airplane sustained impact damage to the radome and the forward fuselage. The fuselage damage consisted of gouges and buckling of the skin and internal structure.

The tug sustained minor damage during the impact sequence, although the tug's tow bar was substantially bent.

PERSONNEL INFORMATION

American Airlines (AA) reported that the captain had a total of 14,877 hours flight experience, with 7,299 hours in the DC-9/MD-80. He was hired at American Airlines in 1987 after retiring from the U.S. Navy. AA reported that the first officer had a total of 12,000 hours, with 2000 in the DC-9/MD-80. He had been with AA since the transition from Reno Air in 1999.

The two pilots had previously flown together on several occasions, including previous flights earlier in December 2008. The accident flight was the first flight on the current three-day trip.

CREW INTERVIEWS

The two pilots were interviewed independently by the NTSB.

First Officer:

The first officer stated that he went to bed about 2230 the night before the flight and slept well. He awoke a little after 0600 and arrived at the airport about 0705, which was his scheduled show time for the flight.

He stated that their pushback was a bit abrupt and fast. After they were cleared to start the engines, the captain said that he was starting the left engine and he proceeded to select the engine start valve. The first officer stated that he moved the fuel lever to "on" once the engine's N2 reached 20% (i.e., 20% of the engine core's maximum speed) and continued to monitor the engine readings. He made a callout of 40% N2 as that threshold was reached, and the captain released the start switch. The first officer said that the engine start was normal except that he recalled that the N2 indication "stayed low…a little over 40%" after starter release.

He indicated that a "stable" range for N2 would be 50-52%. He indicated that a low N2 was generally due to a "stuck" 13th-stage bleed valve and that he had had to increase engine power on previous occasions to clear the problem. In these situations, he stated that he usually pushed the throttle forward about an inch and that he would pull the throttle back when N2 had reached 57%-62%. He stated that pushing the throttle forward was something he "did on his own" to achieve normal parameters after a start. He stated that information about clearing a problem with the 13th-stage bleed valve was located in the operations manual. He said there were no callouts associated with the procedure.

On the accident flight, he stated that he was anticipating the end of the pushback and that they would be setting the parking brakes, and so he pushed the left engine throttle forward to help with any problem with the 13th stage bleed valve and bring the engine into normal parameters. He didn't recall how far he moved the throttle or what N2 value was reached during the pushback. He said that his movement of the throttle occurred around the same time that they were requested by ground control to have the tug pull the airplane forward to "Charlie 9". He stated that the volume of the radio call was loud and that he reached with his right hand to adjust the volume of the radio. He said he wasn't sure if his left hand was on the throttle when the radio call occurred.

He said that after they started moving forward, the tug's movement became erratic and it looked like it was "overcorrecting". He said that the tug eventually "jackknifed" and impacted under his seat on the right side of the airplane. He stated that after the airplane stopped, the captain checked to make sure the tug operator and everyone in the airplane was ok.

He stated that after the tug started moving forward, he put his hand on the throttles to ensure they were retarded. He stated that he observed the left throttle forward "about an inch" and that he pulled it back to idle. He said he didn't know how long the throttle was forward, but said he guessed it was for a few seconds.

Captain:

The captain stated that he went to bed about 2100 on the night before the flight and fell asleep soon afterward. He awoke about 0430 and arrived at the airport about 0645.

He said it was a normal operation before departure. After they had been cleared by ground control for the pushback, he told the tug driver "it's your aircraft." He stated that the pushback was quick but smooth. Once the pushback was underway, they were cleared to start the engines. The captain stated that he then pushed the start valve for the left engine. He released the start valve after the first officer announced that the engine's N2 reached 40 percent. He confirmed the N2 reading and did a quick scan of the engine instruments. After confirming that everything was within limits, he said that he was "outside the airplane" and focusing on the tug's movement of the airplane.

He said that ground control called them around this time as the airplane was still in pushback and requested that they pull forward slightly in order to maintain clearance from taxiway C10. The captain informed the tug operator and said that the operator began moving forward without coming to a complete stop. The captain stated that there was a "little jolt" as they changed direction. He said that the airplane had not backed fully onto the centerline for taxiway C, such that the tug and the airplane were slightly offset to the left of taxiway C as the tug began pulling the airplane forward.

The captain stated that after the tug began to pull forward, it began to lurch right and left with increasing intensity. He said that the tug seemed "out of control" and that after a couple of oscillations to the right and to the left, the tug became impinged under the right side of the airplane's nose.

The captain stated during the interview that following the tug impact, the first officer said, "I think I moved the throttle." He said that he asked the first officer why he moved the throttle, but he didn't recall the first officer's response. The captain said that he wasn't aware of any first officer having moved a throttle after engine start.

He was not aware that the first officer had moved the left throttle forward. He said that he didn't hear the engine spooling up and that it is difficult anyway to hear the engines on an MD-80, especially with other ambient noise from the tug and other taxiing aircraft. He stated that he could only envision moving the throttles if there were a low idle after engine start, but that this wouldn't be done until after the brakes had been set. He stated that he would expect a first officer to announce low idle but that he (the captain) would generally be the one to advance the throttle if it were necessary.

Tug Operator:

The tug operator stated that as he began pulling the airplane forward, the airplane began to accelerate forward. He reacted by downshifting and applying the brakes, but the airplane continued to push the tug and he was unable to stop. He stated that he ducked down as the airplane impacted the tug.

The AA DC-9 Operating Manual specified that the captain should check that N2 stabilizes at 50-61% after N2 has achieved 40% and he has released the engine start switch. The procedure included the following note regarding "low idle RPM":

Lower than normal idle RPM (less than 50% N2) accompanied by a

higher than normal idle EGT [exhaust gas temperature] and possible generator cycling may be an indication that the 13th stage start bleed valve has failed to close. When conditions permit, momentarily advance the Throttle to approximately 65% N2 and then retard to idle. This action should close the bleed valve resulting in increased idle RPM and decreased EGT. If the original low idle or generator cycling persists after the throttle is retarded to idle, Maintenance action is required.

• Do not advance throttles until parking brakes are parked.

• At high altitude airports throttle may have to be advanced beyond

65% N2.

The airplane's cockpit voice recorder and flight data recorder were removed at the request of the NTSB and transported to the NTSB laboratory for readout. No CVR data from the event remained because the CVR had continued to run and overwrite the accident sequence.

FDR data revealed that the throttle for the left engine was moved to a high thrust position near the throttle's forward limit about 15 seconds after the engine start was initiated. At the time of throttle lever movement, the left engine's N2 was increasing through approximately 44% and was accompanied by corresponding increases in other readings such as the engine pressure ratio and the exhaust gas temperature. Seven seconds after throttle movement, N2 was increasing through approximately 53%. The throttle remained at the high thrust setting for approximately 28 seconds before being reduced, and N2 reached as high as 91% during the engine spool-up.

As part of the investigation, engine runs were subsequently arranged in order to correlate throttle positions and engine readings. Several engine runs were conducted, and the data correlating throttle positions and engine parameters showed consistency to the throttle and engine data from the accident FDR.

NTSB Probable Cause

The first officer's unannounced advancement of the left engine throttle, which led to a sustained application of high thrust that caused the airplane to impact the tug.