Flight Director Systems — Page 352, Lesson 410

Let's start with the core idea: gain scheduling. This is the varying of the gain of the pitch and roll demands of the FDC — that's the Flight Director Computer — in relation to the task. So the computer doesn't always command the same strength of correction; it adjusts how forceful its commands are depending on what the aircraft is doing. This has many parallels with autopilot gain scheduling, or gain adaption. The FD mode of operation that uses gain scheduling is the FDA — the Flight Director Approach mode. Now, why do we need this? Think about how the ILS works. The ILS is a beam set at a certain angle, diverging from the runway — normally about 3 degrees. The ILS equipment displays the received signal as an error in degrees from the ideal. So the raw glide slope indication shows you, in degrees, how far off the ideal path you are. Here's the key problem. At 6 NM, one degree of error equates to about 608 ft of vertical distance. But at ½ NM, that same one degree of error equates to only about 54 ft of vertical distance. So the raw indication on the glide slope could look identical — one degree of error — but the actual physical distance off the path is vastly different. At 6 NM you're 608 ft off; at half a mile you're only 54 ft off. So although the indications on the raw ILS glide slope will be the same, a less forceful correction is required as the aircraft nears the ILS transmitter. If the Flight Director commanded the same aggressive pitch change at half a mile as it did at 6 NM, you'd be overcorrecting wildly. So the FD computed pitch information must be modified as the approach progresses, to reduce the commanded corrections. As the aircraft approaches touch-down, the magnitude of the pitch changes required to follow the glide slope reduces. Gain scheduling reduces the magnitude of the commands as the aircraft proceeds on the FDA. So initially, the FDS — the Flight Director System — can demand manoeuvres almost to the full authority of the system. But as the threshold approaches, the gain is reduced to perhaps a half or a third of the original value. Now, how does the system know when to reduce the gain? The initiation of this scheduling can come in many forms, depending on the age of the system. Early systems simply used time — for example, 45 seconds after glide slope capture, the gain is reduced. That's a simple timer-based approach. The next systems used the marker beacons, to try to actually match the scheduling required to the approach being flown. The idea was that as you passed each marker beacon, the system would know your position along the approach and adjust the gain accordingly. But this system has fallen into disuse, because of the loss of the marker beacon systems at many airfields. So that method is largely gone now. Let me just make sure the picture is clear. The whole point is that the Flight Director's job is to tell you what pitch and roll to fly. But a one-degree error near the runway is a much smaller physical deviation than a one-degree error far out. So the FD must scale its commands down as you get closer — that scaling is gain scheduling, and it's what makes the FDA mode usable all the way down to the threshold.