Narrative:

During a training flight (set up as a mock 'evaluation'); the pilot in training failed to disengage the autoplt at the missed approach point and subsequently landed the aircraft without having disengaged the autoplt. He thought he had disengaged it at the decision altitude on the practice ILS. As part of the training scenario I; as the flight instructor; was pretending to be 'incapacitated' to evaluate the student's automation management and instrument approach to safely get the airplane back on the ground. Although I was still monitoring the approach and landing; I did not notice that the PF had failed to disengage the autoplt because he was hand flying on the landing. He was simply overpowering the autoplt manually. Immediately after landing; I noticed the command bars and 'ap on' annunciation and realized he hadn't disengaged it. Upon discussion with the pilot; he said he thought he had disengaged it; but in fact had not. Though this was my first flight with him; he mentioned previously having trouble with the disconnect switch on this type of aircraft (cirrus). The disconnect switch is of the 'hat type' which is used for pitch trim; roll trim; and ap disconnect by pushing straight down. I believe cirrus has rectified this poor design by adding ap disconnect switches on newer models.callback conversation with reporter revealed the following information: the reporter stated that the 'hat' type switch he referenced is a five position button switch on top of the control joy stick incorporating nose up/down pitch; aileron left/right; and a center downward push to disconnect the autopilot. The reporter's experience is currently teaching pilots who have 'lots' of flight time but very little experience in advanced technology aircraft such as the SR20. In many cases the past experience in less advanced aircraft brings with it negative training as his students attempt to apply old behaviors to new technology that in fact require different skills. He has seen students fail to disconnect the autopilot in cruise as well as on approach because the downward disconnect motion is very precise and no other joystick motion will cause a disconnect. Manually trimming the aircraft with this switch will not disconnect the autopilot and so can result in either additional or counter trim from what the autopilot requires. Additionally; if a pilot is hand flying the aircraft and feels it out of trim; using the joystick he/she can make control inputs causing the autopilot to further trim opposite the pilot's input. Subsequently; when the autopilot disconnect button is finally pushed; the aircraft can be far enough out of trim to cause a pitch up or down. This event could be an issue at a low altitude with the aircraft in a landing flare. Newer cirrus aircraft have a separate red button forward of the 'hat' switch used to positively disconnect the autopilot.

Google
 

Original NASA ASRS Text

Title: STUDENT LANDS ACFT WITH AUTOPLT ENGAGED. AN SR20 INSTRUCTOR COMMENTS THAT THE AUTOPILOT DISCONNECT FUNCTION OF THE MULTIFUNCTION JOYSTICK CTLR ON OLDER ACFT IS DIFFICULT TO ACTIVATE.

Narrative: DURING A TRAINING FLT (SET UP AS A MOCK 'EVALUATION'); THE PLT IN TRAINING FAILED TO DISENGAGE THE AUTOPLT AT THE MISSED APCH POINT AND SUBSEQUENTLY LANDED THE ACFT WITHOUT HAVING DISENGAGED THE AUTOPLT. HE THOUGHT HE HAD DISENGAGED IT AT THE DECISION ALTITUDE ON THE PRACTICE ILS. AS PART OF THE TRAINING SCENARIO I; AS THE FLT INSTRUCTOR; WAS PRETENDING TO BE 'INCAPACITATED' TO EVALUATE THE STUDENT'S AUTOMATION MANAGEMENT AND INSTRUMENT APCH TO SAFELY GET THE AIRPLANE BACK ON THE GND. ALTHOUGH I WAS STILL MONITORING THE APCH AND LNDG; I DID NOT NOTICE THAT THE PF HAD FAILED TO DISENGAGE THE AUTOPLT BECAUSE HE WAS HAND FLYING ON THE LNDG. HE WAS SIMPLY OVERPOWERING THE AUTOPLT MANUALLY. IMMEDIATELY AFTER LNDG; I NOTICED THE COMMAND BARS AND 'AP ON' ANNUNCIATION AND REALIZED HE HADN'T DISENGAGED IT. UPON DISCUSSION WITH THE PILOT; HE SAID HE THOUGHT HE HAD DISENGAGED IT; BUT IN FACT HAD NOT. THOUGH THIS WAS MY FIRST FLT WITH HIM; HE MENTIONED PREVIOUSLY HAVING TROUBLE WITH THE DISCONNECT SWITCH ON THIS TYPE OF ACFT (CIRRUS). THE DISCONNECT SWITCH IS OF THE 'HAT TYPE' WHICH IS USED FOR PITCH TRIM; ROLL TRIM; AND AP DISCONNECT BY PUSHING STRAIGHT DOWN. I BELIEVE CIRRUS HAS RECTIFIED THIS POOR DESIGN BY ADDING AP DISCONNECT SWITCHES ON NEWER MODELS.CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING INFO: THE RPTR STATED THAT THE 'HAT' TYPE SWITCH HE REFERENCED IS A FIVE POSITION BUTTON SWITCH ON TOP OF THE CONTROL JOY STICK INCORPORATING NOSE UP/DOWN PITCH; AILERON LEFT/RIGHT; AND A CENTER DOWNWARD PUSH TO DISCONNECT THE AUTOPILOT. THE RPTR'S EXPERIENCE IS CURRENTLY TEACHING PILOTS WHO HAVE 'LOTS' OF FLT TIME BUT VERY LITTLE EXPERIENCE IN ADVANCED TECHNOLOGY ACFT SUCH AS THE SR20. IN MANY CASES THE PAST EXPERIENCE IN LESS ADVANCED ACFT BRINGS WITH IT NEGATIVE TRAINING AS HIS STUDENTS ATTEMPT TO APPLY OLD BEHAVIORS TO NEW TECHNOLOGY THAT IN FACT REQUIRE DIFFERENT SKILLS. HE HAS SEEN STUDENTS FAIL TO DISCONNECT THE AUTOPILOT IN CRUISE AS WELL AS ON APCH BECAUSE THE DOWNWARD DISCONNECT MOTION IS VERY PRECISE AND NO OTHER JOYSTICK MOTION WILL CAUSE A DISCONNECT. MANUALLY TRIMMING THE ACFT WITH THIS SWITCH WILL NOT DISCONNECT THE AUTOPILOT AND SO CAN RESULT IN EITHER ADDITIONAL OR COUNTER TRIM FROM WHAT THE AUTOPILOT REQUIRES. ADDITIONALLY; IF A PILOT IS HAND FLYING THE ACFT AND FEELS IT OUT OF TRIM; USING THE JOYSTICK HE/SHE CAN MAKE CONTROL INPUTS CAUSING THE AUTOPILOT TO FURTHER TRIM OPPOSITE THE PILOT'S INPUT. SUBSEQUENTLY; WHEN THE AUTOPILOT DISCONNECT BUTTON IS FINALLY PUSHED; THE ACFT CAN BE FAR ENOUGH OUT OF TRIM TO CAUSE A PITCH UP OR DOWN. THIS EVENT COULD BE AN ISSUE AT A LOW ALTITUDE WITH THE ACFT IN A LNDG FLARE. NEWER CIRRUS ACFT HAVE A SEPARATE RED BUTTON FORWARD OF THE 'HAT' SWITCH USED TO POSITIVELY DISCONNECT THE AUTOPILOT.

Data retrieved from NASA's ASRS site as of January 2009 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.