A flight instructor since 1956, Barry Schiff retired as a captain for TWA in 1998.
In a recent magazine article, I read the author's view that many, and perhaps most, pilots with the capability do not execute coupled ILS approaches because they do not trust their autopilots to keep them out of trouble when low and slow. That concurs with my experience. During years of administering instrument flight tests, flight reviews, and competency checks, I found that most pilots are more comfortable hand-flying an ILS approach.
Not completely trusting an autopilot when approaching minimums is actually a healthy attitude. But it is not a reason to avoid coupled approaches. Conversely, those who have too much faith in their autopilot's capability are more of a problem. They can become excessively complacent and allow an errant autopilot to place them in jeopardy.
Never will I forget my simulator introduction to coupled approaches and automatic landings in the Lockheed 1011, training that involved approaching and landing without ever seeing the ground until after touchdown. The approach is routine until 150 feet agl, which is when the Lockheed would roll out of a crab and enter a slip to compensate for a crosswind. It was weird to see the instruments reflecting this maneuver and feeling the side load. The throttles began to retard at 50 feet; watching pitch attitude increase during the flare is a mind-blowing experience.
Nor can I forget my instructor's advice. "Keep your hand loose on the control wheel, your left thumb ready to disconnect, and your right hand ready to apply full power in case of the need to miss. Constantly monitor the autopilot to ensure that it is behaving properly with respect to raw-data [ILS] indications. As long as it makes appropriate corrections and everything appears normal, just hang on and enjoy the ride."
The same advice applies to a coupled approach in a general aviation airplane. Constantly compare raw data with the actions of the autopilot, and as long as "Iron Mike" continues to make appropriate pitch-and-roll corrections, leave it alone. But never trust the autopilot blindly (no pun intended); remain spring-loaded throughout the approach to "punch out" and go around.
Many pilots notice less than optimum performance during a coupled approach because they might not have given their autopilot a fighting chance. For example, they may attempt to intercept at too large an angle. For optimum performance, do not intercept the localizer at more than 30 degrees. Avoid intercepting a glidepath from above. Some autopilots won't do this at all while others have difficulty.
Extend the landing gear (if applicable) at glideslope intercept. This helps to prevent an unwanted airspeed increase when the autopilot pitches down to track the glidepath.
Don't strangle the control wheel; autopilots have difficulty overcoming pilot-induced control forces. Hold the wheel lightly; apply barely enough touch to feel what the autopilot is doing. Constantly monitor raw data throughout the approach and confirm that the autopilot is doing what you would do as the result of any given movement of the ILS needles. Keep your left thumb or forefinger near the autopilot disconnect switch and be ready to use it when anything suspicious occurs.
Don't chase airspeed by overworking the throttle to make airspeed changes; move it slowly. It's better to accept small airspeed excursions than to make rapid and fluctuating power changes that can result in autopilot-induced pitch oscillations and a misbehaving pitch channel.
Pilots are susceptible to two common errors that often cause them to believe that their autopilots are leading them astray.
The first occurs when breaking out of a low overcast and noting that the nose of the aircraft is pointed toward one side of the runway without realizing that this is the result of the autopilot crabbing during the approach to compensate for a crosswind. They quickly disconnect and turn to line up with the runway. The crosswind begins to have its way, and the aircraft drifts immediately downwind of the centerline. This can be avoided by correlating the tower-reported wind with crab angle developed by the autopilot during the approach. During crosswind conditions, expect the aircraft to be pointing toward the upwind side of the runway after entering visual conditions.
The second error is the result of breaking into visual conditions and sensing that the aircraft is too high at a time when it probably is not. This dangerous illusion is called "ILS tuck" and there is a natural tendency for a pilot to dive for the runway after breaking out of a very low overcast.
When the runway comes into view, do not be in a hurry to disconnect. Continue the approach visually and with the autopilot engaged to confirm that the aircraft is tracking and descending properly. Don't remove any crab unless truly necessary or until ready to flare. Concentrate on not yielding to ILS tuck. Hold the autopilot-established sink rate unless it is genuinely necessary to steepen the descent, and then make only gentle pitch changes.
Just remember that an autopilot is much less susceptible to visual illusions than a human pilot.
Practicing coupled approaches in VFR conditions can go a long way toward developing confidence in an autopilot's ability to keep you out of harm's way. But trust it blindly? Not on your life.
Visit the author's Web site ( www.barryschiff.com).
Related Articles
