When the automation heads out to lunch

The following article was written by an Air Facts reader. To submit your story, click here. –Ed.

Over the past 37 years of flying GA aircraft, I’ve become a strong proponent of totally understanding and using the available automation in the cockpit. I use the autopilot in our Aerostar 601P/700 a lot and make sure that I understand how the A/P or other automation works in every airplane I fly. I just don’t like surprises. But once in awhile, surprises still happen.

Piper Aerostar
Piper Aerostar 700

On a recent weekend, we were headed back to Chicago for a conference and another opportunity to visit our son. Since there were four of us on board, we decided to fly into Midway (MDW) instead of Chicago Executive (PWK) and just catch a cab into downtown. It saves nearly 30 minutes over PWK and our son’s car only holds four. And there was only about a $.20 per gallon difference in the fuel cost between MDW and PWK. Midway is a GA friendly airport with three FBOs on the field and several runways that are GA oriented.

It was an OK day in Chicago, hazy with about 4 miles visibility at MDW. They were using the ILS to 4R for both airliners and GA aircraft. There was quite a bit of traffic heading into MDW. The Aerostar can hold 170-185 knots indicated in level flight down low, so we mix in pretty well with jet arrivals. The challenge is to get slowed down to 156 knots to lower the gear sometime before the marker and to do it without shock cooling the cylinders.

When mixing with higher speed traffic, it’s very common that we’re asked what kind of speed can the Aerostar hold to the marker. In decent weather, my response is: “170knots till about 3 outside of the marker slowing to 150 knots at the marker.” That gives me a bit over a minute to get slowed down and gives the controller a good idea of what I’m planning to do.

To make all this work, I’ll start pulling the power back about 6-7 miles from the marker. Typically, I’ll be in a descending turn onto the final approach course and altitude, 2500’ MSL in the case of MDW. I’ll then drop 10-20 degrees of flaps as the speed goes below 175, all happening at around 3-4 miles from the marker, pulling more power back so that as I hit the marker at 155. Once the gear is down, the plane is right on the localizer and glideslope, and I let the airplane slow to around 130 knots till landing assured. We practice all this quite often, to make sure it all works well when we head to bigger airports.

As you can tell, a lot happens pretty quickly to pull this off smoothly, so I use the autopilot to help me out. Normally, the A/P captures the localizer and then also captures the glideslope which is supposed to happen right at the marker. If all is going well, we’re just at gear speed (156knots or less) at the marker. My job is primarily to monitor what’s going on and manage the power, which in turn manages the energy.

So, all was going just as advertised. We were about three miles from the marker on the 4 Right localizer to MDW. I had just pulled back the power to get slowed down to drop the gear at the marker. Suddenly the airplane starts climbing from 2500 ft. That really caught me by surprise. I hit the autopilot disconnect while glancing at the A/P display which showed that the glideslope had “captured.” I quickly leveled the airplane and kept slowing down. Of course, with the climb of a couple of hundred feet, we flew right through the glideslope. I got the speed back to 155 knots, dropped the gear and added more flaps, and raced down the localizer at 145 knots (max speed for full flaps) all the while pulling back the power. To say the least, it wasn’t quite a “stabilized approach” but with fairly decent visibility and a view of the runway, we made it work. On landing, everyone commented on the smooth touchdown, but once I shut down, I got thinking as to what happened. (After every flight I insist on some “quiet time” to reflect on the flight and develop an “after action” evaluation!)

Over the years, I had heard about “ghost” or “phantom” glideslopes. They usually occur many miles from the airport and typically at much higher altitudes. Occasionally, you’ll see a NOTAM about them. None of that happened in this flight. The only thing I could think of was that the glideslope “captured” for a moment while on the way to the marker. Once captured and with the GS needle going back to normal, we were “below” the glidepath, so the A/P commanded a “climb” to capture. Annoying to say the least. Disconnecting the A/P and hand-flying the approach was the only reasonable alternative at the time.

I talked the whole scenario over with David St. George, Executive Director and Chief Pilot at East Hill Flying Club in Ithaca, NY. David is a Master Instructor as well. We were both a bit baffled at exactly what happened, but sometimes strange stuff can occur at “big city” airports. (BTW, I flew an A/P coupled ILS into Ithaca (ITH) on the way back and everything worked perfectly.)

Midway 4R
Final approach to runway 4R at MDW

David had an interesting suggestion. At MDW as at many other airports including ITH, there are now LPV (localizer performance with vertical guidance) approaches to the same runways as the ILS approaches. He suggested flying the coupled LPV approach using the A/P and to “monitor” the ILS approach on the #2 radio. Since LPV is totally GPS based, there should be no “ghosts” or “phantoms.” And since you’re “monitoring” the ILS on the #2 radio, you don’t have to ask ATC for the LPV approach.

In any weather other than right to ILS minimums where most ILSs have lower minimums than LPVs, this combination should work fine. The result will be a smoother glideslope and one that is more tightly coupled than any ILS. We’d also have two totally independent systems that should agree. Redundancy like this in aviation is a source of comfort and safety. If for some reason we don’t have agreement, a missed approach might be in order to sort things out before trying again.

So, the next time, we go flying for practice, we’ll be trying out LPV approaches with the #2 tuned into the ILS. As pilots we must always be ready for the unexpected!

Tags from the story
, ,

2 Comments

  • Partly because of the existence of some ILS approaches booby-trapped with altitude restrictions that put you ABOVE the glideslope outside the marker, and just build-in a final equipment cross-check, I personally prefer (since I’m ‘just’ an FO, so I can never say I always just DO anything…) not to even couple an autopilot to a glideslope until I reach the marker at or near the correct altitude for it (as published on the profile view) with a valid (common-sense test – PASS) indication on the glideslope needle.

    Outside the marker, I’m only comfortable coupling to the localizer or other lateral modes (minding tighter localizer-based tolerances), and using vertical speed or other vertical modes that correspond to, but don’t just blindly follow, the glideslope.

    It’s smoother, it’s safer, and precludes “surprises” such as the one you experienced. It only requires awareness, which, sadly, is something some pilots seek to replace, not enhance, with automation.

Leave a Reply

Your email address will not be published. Required fields are marked *