Narrative:

Aircraft had to return to gate during taxi because some cargo had to be removed form aft compartment (to keep center of gravity within limits). Flight had already been delayed to move all passenger to first class because of extreme aft center of gravity. Even after doing these items, aircraft was apparently just inside the center of gravity allowable envelope, except that stabilizer trim setting was much farther forward than I had ever seen previously. Engaged autoplt after leveling for cruise at 35000 ft. (Performance management computer indicated optimum and maximum altitude was FL360). Gradually, the aircraft altitude began to oscillate 25 ft to 100 ft, and the autoplt was moving the stabilizer trim to compensate more and more. As I had not been trained or seen this situation before in the medium large transport, I thought there was some sort of jet stream or mountain wave activity, and asked the captain to notify ATC we were having trouble maintaining FL350 and request a lower altitude. Before he could make the request, the aircraft pitched over. I disengaged the autoplt, added back pressure, and stopped the descent at 34500, then gradually climbed back to FL350. I did this gently, because, by this time, I realized we had a stability problem and didn't want to stall the airplane. The captain confirmed that it was our extreme aft center of gravity that was the problem, and from his experience, the medium large transport is very unstable by design. He requested FL310, which we were cleared with no further problems. I would suggest that simulator training programs, airline or otherwise, include some of these 'off nominal' sits, so that pilots new to the aircraft are not caught 'off guard.' callback conversation with reporter revealed the following information: reporter assured analyst that no takeoff warning horn sounded during takeoff. The center of gravity was within limits. He said the center of gravity moved forward (he thought) during flight. Descending from FL350 to FL310 made the aircraft more stable. Passenger had been moved forward before takeoff. The first class cabin was full. When the rate of descent increased, it was a smooth pitch-over. He thought the passenger never noticed. He thinks this was one time that the human recovered better than the machine.

Google
 

Original NASA ASRS Text

Title: LOSS OF ACFT CTL IN ACR MLG CREATES AN ALTDEV ALT EXCURSION. CTR OF GRAVITY TOO FAR AFT.

Narrative: ACFT HAD TO RETURN TO GATE DURING TAXI BECAUSE SOME CARGO HAD TO BE REMOVED FORM AFT COMPARTMENT (TO KEEP CTR OF GRAVITY WITHIN LIMITS). FLT HAD ALREADY BEEN DELAYED TO MOVE ALL PAX TO FIRST CLASS BECAUSE OF EXTREME AFT CTR OF GRAVITY. EVEN AFTER DOING THESE ITEMS, ACFT WAS APPARENTLY JUST INSIDE THE CTR OF GRAVITY ALLOWABLE ENVELOPE, EXCEPT THAT STABILIZER TRIM SETTING WAS MUCH FARTHER FORWARD THAN I HAD EVER SEEN PREVIOUSLY. ENGAGED AUTOPLT AFTER LEVELING FOR CRUISE AT 35000 FT. (PERFORMANCE MGMNT COMPUTER INDICATED OPTIMUM AND MAX ALT WAS FL360). GRADUALLY, THE ACFT ALT BEGAN TO OSCILLATE 25 FT TO 100 FT, AND THE AUTOPLT WAS MOVING THE STABILIZER TRIM TO COMPENSATE MORE AND MORE. AS I HAD NOT BEEN TRAINED OR SEEN THIS SIT BEFORE IN THE MLG, I THOUGHT THERE WAS SOME SORT OF JET STREAM OR MOUNTAIN WAVE ACTIVITY, AND ASKED THE CAPT TO NOTIFY ATC WE WERE HAVING TROUBLE MAINTAINING FL350 AND REQUEST A LOWER ALT. BEFORE HE COULD MAKE THE REQUEST, THE ACFT PITCHED OVER. I DISENGAGED THE AUTOPLT, ADDED BACK PRESSURE, AND STOPPED THE DSCNT AT 34500, THEN GRADUALLY CLBED BACK TO FL350. I DID THIS GENTLY, BECAUSE, BY THIS TIME, I REALIZED WE HAD A STABILITY PROB AND DIDN'T WANT TO STALL THE AIRPLANE. THE CAPT CONFIRMED THAT IT WAS OUR EXTREME AFT CTR OF GRAVITY THAT WAS THE PROB, AND FROM HIS EXPERIENCE, THE MLG IS VERY UNSTABLE BY DESIGN. HE REQUESTED FL310, WHICH WE WERE CLRED WITH NO FURTHER PROBS. I WOULD SUGGEST THAT SIMULATOR TRAINING PROGRAMS, AIRLINE OR OTHERWISE, INCLUDE SOME OF THESE 'OFF NOMINAL' SITS, SO THAT PLTS NEW TO THE ACFT ARE NOT CAUGHT 'OFF GUARD.' CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING INFO: RPTR ASSURED ANALYST THAT NO TKOF WARNING HORN SOUNDED DURING TKOF. THE CTR OF GRAVITY WAS WITHIN LIMITS. HE SAID THE CTR OF GRAVITY MOVED FORWARD (HE THOUGHT) DURING FLT. DSNDING FROM FL350 TO FL310 MADE THE ACFT MORE STABLE. PAX HAD BEEN MOVED FORWARD BEFORE TKOF. THE FIRST CLASS CABIN WAS FULL. WHEN THE RATE OF DSCNT INCREASED, IT WAS A SMOOTH PITCH-OVER. HE THOUGHT THE PAX NEVER NOTICED. HE THINKS THIS WAS ONE TIME THAT THE HUMAN RECOVERED BETTER THAN THE MACHINE.

Data retrieved from NASA's ASRS site as of July 2007 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.