Narrative:

We were climbing on the departure approaching level off at FL360. The climb in VNAV (autopilot 2) was normal. However; at FL355; at .74 mach; an under speed symbol (a) appeared on the MCP; and the aircraft pitched over and started a slight descent. My first officer (pilot flying) attempted to re-establish climb with V/south mode (V/south wouldn't work in under speed mode). The aircraft accelerated in the descent approaching max mach with the under speed symbol still in view on the MCP. At some point; my first officer manually overrode the autopilot with control wheel steering inputs; and started a manual climb. With a manual climb going; we re-attempted to engage VNAV; and the aircraft again went into a descent with the under speed symbol still present. We repeated the manual override and then attempted to use autopilot 1 with the same results. I instructed my first officer to disengage the autopilot and hand fly the climb to level off; which we did. At level off; we re-engaged the autopilot; with no further abnormalities the rest of the flight. Essentially; we had climbed; leveled off; and descended between about FL356 and FL354 at least three times in short order. Complicating this issue was the weather (we were IMC) with multiple thunderstorms around us with concurrent turbulence. Further complicating this issue was head on traffic at FL350 approaching us. After the first or second cycle; center asked us when we would be leveling off. I told him we were having an autopilot malfunction and if we could get it under control; about 30 seconds. However; it took longer while we were still wrestling with the issue. ATC notified the approaching traffic that he didn't know what we were doing; climbing and descending below FL360. The traffic was six miles ahead when we finally got to FL360. Even further complicating the issue was the irrationality of what was happening. We should never have gotten the under speed symbol in the conditions we had. We should have subsequently gotten the over speed symbol (8) when the aircraft accelerated in the descents; but the under speed symbol remained. We were wondering if we had some kind of pitot heat issue; but could not confirm it; and we had no irregularities with the main and standby airspeed and altitude instruments. Because the aircraft accelerated in the descents; my first officer pulled the power back to avoid over speeding; which further delayed the level off. One last complication was that somewhere in this incident; the LNAV disengaged; which we re-engaged immediately after level off when we noticed it was disengaged. We were only slightly off course by this time. After landing; we called dispatch and maintenance and discussed the issue with them and the oncoming pilots. I don't think there is anything we could have done to prevent the initial erroneous under speed indication from happening. However; my first officer and I discussed at length what we could have done better in response to it. He attempted to use the MCP modes to correct the initial issue; which was not possible; and that greatly delayed correcting our flight path to the proper position. I frequently hand fly the aircraft for some time after takeoff. Had I been flying the aircraft; my first response would have been to disengage the autopilot and hand fly the aircraft to level off; instead of trying to use any MCP modes to fix the issue. That way; our level off would have occurred expeditiously and normally; instead of taking so long to accomplish.

Google
 

Original NASA ASRS Text

Title: A B737-300 climbing to FL360 developed an un-speed condition with A symbol in the SPD window and pitched over causing an airspeed increase to near MMO. Pitch control with the autopilot failed so it was disconnected to finish the climb after three osculations. An airspeed issue as suspected by the crew.

Narrative: We were climbing on the departure approaching level off at FL360. The climb in VNAV (autopilot 2) was normal. However; at FL355; at .74 Mach; an under speed symbol (A) appeared on the MCP; and the aircraft pitched over and started a slight descent. My First Officer (pilot flying) attempted to re-establish climb with V/S mode (V/S wouldn't work in under speed mode). The aircraft accelerated in the descent approaching max Mach with the under speed symbol still in view on the MCP. At some point; my First Officer manually overrode the autopilot with control wheel steering inputs; and started a manual climb. With a manual climb going; we re-attempted to engage VNAV; and the aircraft again went into a descent with the under speed symbol still present. We repeated the manual override and then attempted to use autopilot 1 with the same results. I instructed my First Officer to disengage the autopilot and hand fly the climb to level off; which we did. At level off; we re-engaged the autopilot; with no further abnormalities the rest of the flight. Essentially; we had climbed; leveled off; and descended between about FL356 and FL354 at least three times in short order. Complicating this issue was the weather (we were IMC) with multiple thunderstorms around us with concurrent turbulence. Further complicating this issue was head on traffic at FL350 approaching us. After the first or second cycle; Center asked us when we would be leveling off. I told him we were having an autopilot malfunction and if we could get it under control; about 30 seconds. However; it took longer while we were still wrestling with the issue. ATC notified the approaching traffic that he didn't know what we were doing; climbing and descending below FL360. The traffic was six miles ahead when we finally got to FL360. Even further complicating the issue was the irrationality of what was happening. We should never have gotten the under speed symbol in the conditions we had. We should have subsequently gotten the over speed symbol (8) when the aircraft accelerated in the descents; but the under speed symbol remained. We were wondering if we had some kind of pitot heat issue; but could not confirm it; and we had no irregularities with the main and standby airspeed and altitude instruments. Because the aircraft accelerated in the descents; my First Officer pulled the power back to avoid over speeding; which further delayed the level off. One last complication was that somewhere in this incident; the LNAV disengaged; which we re-engaged immediately after level off when we noticed it was disengaged. We were only slightly off course by this time. After landing; we called Dispatch and Maintenance and discussed the issue with them and the oncoming pilots. I don't think there is anything we could have done to PREVENT the initial erroneous under speed indication from happening. However; my First Officer and I discussed at length what we could have done better in response to it. He attempted to use the MCP modes to correct the initial issue; which was not possible; and that greatly delayed correcting our flight path to the proper position. I frequently hand fly the aircraft for some time after takeoff. Had I been flying the aircraft; my first response would have been to disengage the autopilot and hand fly the aircraft to level off; instead of trying to use any MCP modes to fix the issue. That way; our level off would have occurred expeditiously and normally; instead of taking so long to accomplish.

Data retrieved from NASA's ASRS site 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.