Narrative:

We were at FL280. ATC had given us 'pilot's discretion to FL240.' it was the first officer's leg to fly. ATC directed us to maintain 300 KIAS. At that moment we were cruising at 280 KIAS. I visually confirmed that the first officer had dialed 300 KIAS into the MCP airspeed window. Approximately 1 min later, ATC said to us, 'call sign, say your altitude.' since we had been given a frequency change between the time we were given pilot's discretion to FL240 and now, I thought maybe there had been a failure to communication between the last sector controller and this one. I glanced at the FMC to confirm we were not yet at our top of descent point and replied, 'call sign was given pilot's discretion to FL240.' then ATC replied, 'call sign, say your altitude.' well, I was sure we were still at FL280 (because that had been our cruising altitude, and we had not yet started our descent), but a quick glance at my instruments told me I was dead wrong, and we had a problem. With no warning whatsoever, and for no apparent reason, the aircraft had begun a very slow climb, and was now nearing FL290. At the same instant I saw our altitude deviation, I also saw that the pitch mode of the autoplt had reverted from 'altitude hold' to 'control wheel steering pitch.' for those not completely familiar with the B737-500 autoplt, control wheel steering pitch is an autoplt mode designed to maintain a constant pitch attitude. The aircraft was at the moment of reversion to control wheel steering pitch. Well, that part of the system worked as advertised. But remember, we had accelerated from 280 KIAS to 300 KIAS. And (lesson from basic airmanship 101) a constant pitch at an increasing airspeed results in -- that's right -- a climb. I immediately took control of the aircraft and began a descent, not just back down to FL280, but continuing down to FL240. Simultaneously I keyed the microphone and told center, 'call sign is at FL290, descending now to FL240. Our autoplt appears to have malfunctioned.' center replied, 'roger, copy you descending to FL240.' we had no TCASII alert of any kind, and when I asked the controller if I needed to call him when we got on the ground, he replied 'no, don't worry about it.' in my 12 1/2 yrs and 11000 hours on B737-300/500 aircraft, I have never seen this happen. I have seen the autoplt on the B737-300/500 revert to control wheel steering pitch for only 3 reasons: 1) manually deselecting all other pitch modes. (This did not happen.) 2) turbulence so bad the autoplt cannot maintain altitude. (We were in perfectly smooth ride conditions.) 3) a force being applied to the flight controls significant enough to cause the selected autoplt pitch mode to disengage. This is the only possibility I can think of. What may have happened is that the nose-up pitch forces resulting from the power advancing to accelerate from 280 KIAS to 300 KIAS were significant enough to cause the autoplt to revert from 'altitude hold' to 'control wheel steering pitch.' normally we would get a warning horn if the aircraft deviated from the selected cruise altitude by more than 150 ft. But in this case, because we had dialed FL240 into the altitude window on the MCP to comply with our clearance, when the aircraft departed FL280, even though it was climbing instead of descending, we got no warning horn and no altitude alert light. Also, because the autoplt had not completely disengaged, we got no 'autoplt disengaged' warning tone or light. What apparently happened was the control wheel steering pitch appeared on the EFIS display. I had my head down at the moment this occurred and was not looking at the instrument panel. The first officer, with no reason to expect any anomaly, did not notice the control wheel steering pitch display appear. And the climb was so slow that neither the first officer nor I had any 'seat-of-the-pants' indication we had departed altitude. Recommendations: human factors: 1) increased vigilance. 2) get the word out how this can happen. Mechanical: conduct inspections to determine if the 'pendulum effect' resulting from large power changes in B737's is significant enough to cause autoplt reversion to control wheel steering pitch in a statistically significant number of aircraft.

Related reports:

Google
 

Original NASA ASRS Text

Title: B737 CREW HAD FLT GUIDANCE CHANGE FROM ALT HOLD TO CTL WHEEL STEERING PITCH. THIS RESULTED IN A 1000 FT ALT EXCURSION.

Narrative: WE WERE AT FL280. ATC HAD GIVEN US 'PLT'S DISCRETION TO FL240.' IT WAS THE FO'S LEG TO FLY. ATC DIRECTED US TO MAINTAIN 300 KIAS. AT THAT MOMENT WE WERE CRUISING AT 280 KIAS. I VISUALLY CONFIRMED THAT THE FO HAD DIALED 300 KIAS INTO THE MCP AIRSPD WINDOW. APPROX 1 MIN LATER, ATC SAID TO US, 'CALL SIGN, SAY YOUR ALT.' SINCE WE HAD BEEN GIVEN A FREQ CHANGE BTWN THE TIME WE WERE GIVEN PLT'S DISCRETION TO FL240 AND NOW, I THOUGHT MAYBE THERE HAD BEEN A FAILURE TO COM BTWN THE LAST SECTOR CTLR AND THIS ONE. I GLANCED AT THE FMC TO CONFIRM WE WERE NOT YET AT OUR TOP OF DSCNT POINT AND REPLIED, 'CALL SIGN WAS GIVEN PLT'S DISCRETION TO FL240.' THEN ATC REPLIED, 'CALL SIGN, SAY YOUR ALT.' WELL, I WAS SURE WE WERE STILL AT FL280 (BECAUSE THAT HAD BEEN OUR CRUISING ALT, AND WE HAD NOT YET STARTED OUR DSCNT), BUT A QUICK GLANCE AT MY INSTS TOLD ME I WAS DEAD WRONG, AND WE HAD A PROB. WITH NO WARNING WHATSOEVER, AND FOR NO APPARENT REASON, THE ACFT HAD BEGUN A VERY SLOW CLB, AND WAS NOW NEARING FL290. AT THE SAME INSTANT I SAW OUR ALTDEV, I ALSO SAW THAT THE PITCH MODE OF THE AUTOPLT HAD REVERTED FROM 'ALT HOLD' TO 'CTL WHEEL STEERING PITCH.' FOR THOSE NOT COMPLETELY FAMILIAR WITH THE B737-500 AUTOPLT, CTL WHEEL STEERING PITCH IS AN AUTOPLT MODE DESIGNED TO MAINTAIN A CONSTANT PITCH ATTITUDE. THE ACFT WAS AT THE MOMENT OF REVERSION TO CTL WHEEL STEERING PITCH. WELL, THAT PART OF THE SYS WORKED AS ADVERTISED. BUT REMEMBER, WE HAD ACCELERATED FROM 280 KIAS TO 300 KIAS. AND (LESSON FROM BASIC AIRMANSHIP 101) A CONSTANT PITCH AT AN INCREASING AIRSPD RESULTS IN -- THAT'S RIGHT -- A CLB. I IMMEDIATELY TOOK CTL OF THE ACFT AND BEGAN A DSCNT, NOT JUST BACK DOWN TO FL280, BUT CONTINUING DOWN TO FL240. SIMULTANEOUSLY I KEYED THE MIKE AND TOLD CTR, 'CALL SIGN IS AT FL290, DSNDING NOW TO FL240. OUR AUTOPLT APPEARS TO HAVE MALFUNCTIONED.' CTR REPLIED, 'ROGER, COPY YOU DSNDING TO FL240.' WE HAD NO TCASII ALERT OF ANY KIND, AND WHEN I ASKED THE CTLR IF I NEEDED TO CALL HIM WHEN WE GOT ON THE GND, HE REPLIED 'NO, DON'T WORRY ABOUT IT.' IN MY 12 1/2 YRS AND 11000 HRS ON B737-300/500 ACFT, I HAVE NEVER SEEN THIS HAPPEN. I HAVE SEEN THE AUTOPLT ON THE B737-300/500 REVERT TO CTL WHEEL STEERING PITCH FOR ONLY 3 REASONS: 1) MANUALLY DESELECTING ALL OTHER PITCH MODES. (THIS DID NOT HAPPEN.) 2) TURB SO BAD THE AUTOPLT CANNOT MAINTAIN ALT. (WE WERE IN PERFECTLY SMOOTH RIDE CONDITIONS.) 3) A FORCE BEING APPLIED TO THE FLT CTLS SIGNIFICANT ENOUGH TO CAUSE THE SELECTED AUTOPLT PITCH MODE TO DISENGAGE. THIS IS THE ONLY POSSIBILITY I CAN THINK OF. WHAT MAY HAVE HAPPENED IS THAT THE NOSE-UP PITCH FORCES RESULTING FROM THE PWR ADVANCING TO ACCELERATE FROM 280 KIAS TO 300 KIAS WERE SIGNIFICANT ENOUGH TO CAUSE THE AUTOPLT TO REVERT FROM 'ALT HOLD' TO 'CTL WHEEL STEERING PITCH.' NORMALLY WE WOULD GET A WARNING HORN IF THE ACFT DEVIATED FROM THE SELECTED CRUISE ALT BY MORE THAN 150 FT. BUT IN THIS CASE, BECAUSE WE HAD DIALED FL240 INTO THE ALT WINDOW ON THE MCP TO COMPLY WITH OUR CLRNC, WHEN THE ACFT DEPARTED FL280, EVEN THOUGH IT WAS CLBING INSTEAD OF DSNDING, WE GOT NO WARNING HORN AND NO ALT ALERT LIGHT. ALSO, BECAUSE THE AUTOPLT HAD NOT COMPLETELY DISENGAGED, WE GOT NO 'AUTOPLT DISENGAGED' WARNING TONE OR LIGHT. WHAT APPARENTLY HAPPENED WAS THE CTL WHEEL STEERING PITCH APPEARED ON THE EFIS DISPLAY. I HAD MY HEAD DOWN AT THE MOMENT THIS OCCURRED AND WAS NOT LOOKING AT THE INST PANEL. THE FO, WITH NO REASON TO EXPECT ANY ANOMALY, DID NOT NOTICE THE CTL WHEEL STEERING PITCH DISPLAY APPEAR. AND THE CLB WAS SO SLOW THAT NEITHER THE FO NOR I HAD ANY 'SEAT-OF-THE-PANTS' INDICATION WE HAD DEPARTED ALT. RECOMMENDATIONS: HUMAN FACTORS: 1) INCREASED VIGILANCE. 2) GET THE WORD OUT HOW THIS CAN HAPPEN. MECHANICAL: CONDUCT INSPECTIONS TO DETERMINE IF THE 'PENDULUM EFFECT' RESULTING FROM LARGE PWR CHANGES IN B737'S IS SIGNIFICANT ENOUGH TO CAUSE AUTOPLT REVERSION TO CTL WHEEL STEERING PITCH IN A STATISTICALLY SIGNIFICANT NUMBER OF ACFT.

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.