Narrative:

Climbing in vertical and horizontal navigation modes of autoplt (normal operation) on scheduled flight in MD88 from bos to atl. Upon leaving cloud deck, engine anti-ice was turned off at FL290. Additional thrust from this action (apparently) caused acceleration to mach .78, about .02 above normal climb mach. Aircraft climb rate increased to 5000-6000 FPM and VNAV capture appeared on 'flight mode annunciator panel,' indicating autoplt capture of FL350. This occurred at FL310. Mach was approximately .76 at this point (approximately 260 KIAS) and decreasing, however, autoplt did not pitch over to maintain climb speed -- it continued on a capture profile that was apparently based on the climb rate at capture initiation. The airspeed bled off with full climb power applied until the autoplt was disconnected at 214 KIAS above FL340. A descent of nearly 1000 ft was required to gain enough airspeed for the aircraft to accelerate to normal climb speed of approximately 260 KIAS at this altitude. Light prestall buffet was experienced. I have no doubt the aircraft would have stalled at nearly FL350 if the autoplt had not been disconnected. 'What really caused the problem, and what can be done to prevent a reoccurrence, or correct the situation?' this is the same problem that caused the A330 flight test accident at toulouse. Given different circumstances and distractions it could have caused serious injury or an accident. Pilots are terrible monitors of system that work well for 99.99% of the time, especially, system that reduce workload so that 2 pilots can do the job. If we can't trust autoplts to work correctly or tell us when they are malfunctioning, we cannot operate well in high workload sits. Certification standards should be changed to require safe operation in all possible sits and not rely on pilot intervention to insure safety. Sits beyond normal control of the autoplt or conflicts between different input devices (eg, pitch wheel and altitude set window) should cause error messages and inaction until the conflict is resolved, or disconnection of the autoplt and high level disconnect warnings. We must be able to trust automation!! Automation certification standards must be improved.

Google
 

Original NASA ASRS Text

Title: MD88 ACFT IN CLB INCREASED PITCH AFTER ENG ANTI-ICE WAS TURNED OFF DUE TO INCREASED PWR AND VNAV MODE. CLB RATE VERY HIGH AND ALT CAPTURE WAS ANNUNCIATED ON MODE PANEL AT 31000 FT FOR A 35000 FT LEVELOFF. AIRSPD BLED OFF, AUTOTHROTTLES DIDN'T CORRECT NOR DID PITCH ATTITUDE. FLC HAD TO DISCONNECT AUTO SYS AND PUSH OVER TO REGAIN AIRSPD.

Narrative: CLBING IN VERT AND HORIZ NAV MODES OF AUTOPLT (NORMAL OP) ON SCHEDULED FLT IN MD88 FROM BOS TO ATL. UPON LEAVING CLOUD DECK, ENG ANTI-ICE WAS TURNED OFF AT FL290. ADDITIONAL THRUST FROM THIS ACTION (APPARENTLY) CAUSED ACCELERATION TO MACH .78, ABOUT .02 ABOVE NORMAL CLB MACH. ACFT CLB RATE INCREASED TO 5000-6000 FPM AND VNAV CAPTURE APPEARED ON 'FLT MODE ANNUNCIATOR PANEL,' INDICATING AUTOPLT CAPTURE OF FL350. THIS OCCURRED AT FL310. MACH WAS APPROX .76 AT THIS POINT (APPROX 260 KIAS) AND DECREASING, HOWEVER, AUTOPLT DID NOT PITCH OVER TO MAINTAIN CLB SPD -- IT CONTINUED ON A CAPTURE PROFILE THAT WAS APPARENTLY BASED ON THE CLB RATE AT CAPTURE INITIATION. THE AIRSPD BLED OFF WITH FULL CLB PWR APPLIED UNTIL THE AUTOPLT WAS DISCONNECTED AT 214 KIAS ABOVE FL340. A DSCNT OF NEARLY 1000 FT WAS REQUIRED TO GAIN ENOUGH AIRSPD FOR THE ACFT TO ACCELERATE TO NORMAL CLB SPD OF APPROX 260 KIAS AT THIS ALT. LIGHT PRESTALL BUFFET WAS EXPERIENCED. I HAVE NO DOUBT THE ACFT WOULD HAVE STALLED AT NEARLY FL350 IF THE AUTOPLT HAD NOT BEEN DISCONNECTED. 'WHAT REALLY CAUSED THE PROB, AND WHAT CAN BE DONE TO PREVENT A REOCCURRENCE, OR CORRECT THE SIT?' THIS IS THE SAME PROB THAT CAUSED THE A330 FLT TEST ACCIDENT AT TOULOUSE. GIVEN DIFFERENT CIRCUMSTANCES AND DISTRACTIONS IT COULD HAVE CAUSED SERIOUS INJURY OR AN ACCIDENT. PLTS ARE TERRIBLE MONITORS OF SYS THAT WORK WELL FOR 99.99% OF THE TIME, ESPECIALLY, SYS THAT REDUCE WORKLOAD SO THAT 2 PLTS CAN DO THE JOB. IF WE CAN'T TRUST AUTOPLTS TO WORK CORRECTLY OR TELL US WHEN THEY ARE MALFUNCTIONING, WE CANNOT OPERATE WELL IN HIGH WORKLOAD SITS. CERTIFICATION STANDARDS SHOULD BE CHANGED TO REQUIRE SAFE OP IN ALL POSSIBLE SITS AND NOT RELY ON PLT INTERVENTION TO INSURE SAFETY. SITS BEYOND NORMAL CTL OF THE AUTOPLT OR CONFLICTS BTWN DIFFERENT INPUT DEVICES (EG, PITCH WHEEL AND ALT SET WINDOW) SHOULD CAUSE ERROR MESSAGES AND INACTION UNTIL THE CONFLICT IS RESOLVED, OR DISCONNECTION OF THE AUTOPLT AND HIGH LEVEL DISCONNECT WARNINGS. WE MUST BE ABLE TO TRUST AUTOMATION!! AUTOMATION CERTIFICATION STANDARDS MUST BE IMPROVED.

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.