Confusion?
By Richard L. Collins
April 2007
 I usually stay away from preliminary NTSB reports but every now and then one offers up so much food for thought that it is worthy of discussion.
It was dark out and the weather at Charlotte, North Carolina, was advertised as 10 miles visibility with an overcast at 800 feet. For an IFR flight, that hardly qualifies as “bad” weather, but all the maneuvering for the approach would be in the clouds.
The Cirrus SR22 was vectored for a downwind leg for the ILS approach to Runway 18R. The pilot was then given a heading of 090 and was cleared for the approach. The pilot passed through the final approach course for 18R and continued toward the final approach for 18L. The controller alerted the pilot and issued a heading of 230 to rejoin the final to 18R.
From the report: “As the airplane maneuvered, it descended through an altitude of 2,200 feet, the controller issued a ‘low altitude warning’ and instructed the pilot to ‘climb and maintain 2,300 feet.’ During the following 20 seconds, the airplane climbed from 1,800 feet to 3,800 feet and maneuvered from a heading of 180 degrees to an approximate heading of 330 degrees. During this time, the airspeed decreased from 183 to 90 knots, until the final radar return was observed at 1944, approximately a quarter of a mile from the accident site.
Witnesses reported hearing engine sounds that seemed to “phase in and out.” A witness reported that the engine “wound up at a very high pitch” and it sounded as if the airplane “banked sharply, and began to nose dive.” The witness heard the impact about four seconds later.
The airplane crashed into trees about 10 miles from the approach end of 18R. The wreckage path was 80 feet long and the airplane was well shattered. The airframe parachute remained packed and there was no indication that the pilot had tried to deploy the chute.
The airplane was virtually new, having been certified only about six weeks before the accident. The pilot, who was alone in the airplane, was appropriately rated and reported 1,600 hours of flight time on a medical form a little more than a year before the accident.
So, what should have been a routine arrival turned into a disaster. The NTSB will be a while issuing a probable cause, but enough similar accidents happen for us to try to learn something from what has been reported about this particular accident.
The airplane had an Avidyne Entegra flat glass primary flight display (PFD) system that got navigational information from Garmin 430s. The pilot was likely trained on the use of this but a simple mistake could be the first domino in a chain of confusion.
The Garmin has a procedure button that, when pressed, first lets you select the airport and then the specific approach to that airport. I don’t know whether or not this pilot had experience at airports with parallel runways, but you could sure screw up your situational awareness by selecting the incorrect runway. That map that is so good at helping situational awareness would, with the wrong runway selected, be misleading.
When I used to fly to Atlanta Hartsfield a lot, on an IFR day the controllers would make double, triple sure that any arriving general aviation pilot knew which of the parallel runways he would be using and if you strayed even a dot off toward the other runways they would yell at you. In other words, they didn’t trust us. And I am sure that enough mistakes had been made over the years to teach them to not trust us.
If you select the wrong approach on a GNS 430 you have to go back through the process to select the proper approach. This takes some concentration, a lot of button pushing, and a firm knowledge of how to make the change. I have watched other pilots falter while doing this and had problems with it myself when first learning the Garmin.
One other item: The report says the pilot was assigned a heading of 90 degrees and cleared for the approach to 18R. No word on a heading change for an intercept, though it seems likely to me that the controller would have followed procedure and given an intercept heading about 30 degrees off the final approach course. If that did not happen it could have been a further source of confusion.
It seems likely that for this pilot to so thoroughly lose control, as evidenced by the wild gyrations, the autopilot had to have either been turned off or become inoperative.
It is not a good idea to turn the autopilot off when confusion is setting in. In fact, I have always wished that pilots would embrace a “time out” procedure to battle confusion. The best one is with the autopilot on and set to fly the airplane on a safe heading and altitude. Then the pilot can back off, let the autopilot fly, settle down, and work on reorientation. Certainly if a pilot told a controller he was confused and needed to regroup, the controller would rather deal with that than with an airplane careening around the sky.
We also need a “time out” procedure if the autopilot has a problem. Mac, Dick Karl and I have all had autopilot roll servo failures in the past few months, so things like this do happen.
If an autopilot has a problem, hopefully the real pilot has been monitoring, notices that the autopilot is no longer doing its thing, and disconnects it and takes over manual control of the airplane. In this case, again the first thing to do would be to fly a safe altitude and heading and not do much of anything but fly until the ducks are in a neat row. Tell the controller, and even ask for a no-gyro approach (even though the instruments are fine), where the controller assumes the navigational thinking process and tells the pilot to turn or go straight and what altitudes are proper for the location of the airplane.
There are many ways the controller can help when a pilot has a problem. He can’t fly, but he can navigate and provide information. I remember a vacuum failure at FL 210 with clouds all the way down to 2,000 feet. I didn’t have any standby instrumentation in my P210 then. It was not a time to act like the ace of the base, either. I told the controller what had happened and that he was now the navigator and procurer of weather information. All I wanted to do was hunker down and keep the airplane under control using partial panel as I descended.
Discuss this article in our forums
| 
Copyright @ 2009 Bonnier Corp. All Rights Reserved. Reproduction in whole or in part without permission is prohibited.
