The circling approach should be the easiest maneuver in an instrument-rated pilots repertoire of skills.
Its really just an IFR descent to get beneath the clouds and then a visual approach to the landing runway – in essence a VFR pattern, but at roughly half the usual pattern altitude. Yet NTSB data shows that circling approaches account for about 30 accidents per year, most of them fatal. Since 1983 there have been 577 accidents due to circling instrument approaches. A review of the record shows that three major causes stand out: circling below minimums, steep turns to stay within the runway environment, and beginning the final descent before acquiring the proper glide path.
Recall that the FARs require the aircraft to be continuously in a position from which a descent to a landing on the intended runway can be made at a normal rate of descent using normal maneuvers. However, in many of the mishaps the pilots lost sight of the runway and racked the airplane around trying to regain a landing position. All too often, the pilot actually hits the ground with a wing while turning base to final, representing a fundamental misunderstanding of the difference between a normal traffic pattern and a circling approach.
During a dark night arrival to Sidney Municipal Airport in Sidney, N.Y., the pilot of a Piper Comanche flew a VOR approach and was circling to land on runway 7. The adjacent terrain is mountainous and wooded, with both factors providing ideal conditions for spatial disorientation, or black-hole effect.
The weather was relatively good, with scattered clouds at 1,500 feet, and a 5,500 foot overcast. Visibility was 7 miles, with winds from the southeast at 5 knots. The published minimum descent altitude was 2,280 feet – 1,253 feet agl, which should have put the airplane below even the scattered clouds. The aircraft crashed and burned about a mile north of the airport while on the downwind leg. The impact with trees occurred at 1,550 feet.
One of the prime rules in a circling approach is not to descend below the published minimum altitude until intercepting the VASI glideslope on final approach. Lacking a VASI, hold the minimum altitude until reaching a point where a normal glide path can be followed to the runway.
Noteworthy too is that there is no need to descend to the published minimums if the weather allows a higher circling pattern altitude. In this case the ceiling was a 5,500-foot overcast, yet the aircraft crashed 730 feet below the circling minimums.
In another example a Beech Baron 58P was cleared for the ILS/DME Runway 2 approach, circle to land runway 20 at Durango, Colo. The circle was forced by a gusty 12-knot southwest wind. The pilot told ATC he had the runway in sight and was observed on a left downwind for runway 20.
As the airplane passed abeam the runway 20 threshold, it disappeared into a small bank of clouds. Approximately five seconds later it broke out of the scud and witnesses saw a large green fireball. The circling MDA was 7,100 feet with an airport elevation of 6,685 feet, yet the airplane had struck power lines about 35 feet agl while on downwind.
One of the most highly publicized circling approach mishaps occurred last March when a Gulfstream III on a charter flight crashed at Aspen, Colo. This airport is one of 11 designated airports that require special alternate airport requirements because they have MDAs of more than 2,000 feet agl or landing visibility minimums of more than three miles. Because it is the mandate of professional pilots to please their passengers, this captains dilemma is, in the cold light of Monday morning quarterbacks, quite understandable.
Several factors were involved, all of which related to pilot/crew decision-making.
For starters, the pilot was racing the clock to arrive before a mandatory nighttime curfew. The airport directory states that any operations 30 minutes after sunset have special aircraft and pilot requirements. This requirement had been deemed necessary because of past accidents stemming from the very high mountainous terrain surrounding the airport. They neared the airport in a virtual dead heat with the clock, and the accident occurred about two minutes after the airport restriction became effective.
The VOR/DME or GPS approach requires a circle to land procedure for airplanes that fail to pick up the runway environment before reaching the missed approach point. The runway is equipped with PAPI, which provides a stabilized, three-degree glide path to the touchdown zone markers.
The approach minimums are unusually high – almost 2,400 feet agl – and the missed approach point is 1.4 miles before the runway threshold.
The approach also is not available to Category D aircraft because of the tight geography and the increased turning radius involved in Category D. The G III is a Category D aircraft until its gross weight is less than 46,000 pounds, however when flying into Aspen this landing weight would not leave adequate fuel reserves for flying to an alternate airport. In addition, the SimuFlite-recommended procedure for circling in the G III calls for using the greater of 160 knots or Vref plus 20 knots. In this case, both the fuel requirements and the circling speed would have placed the aircraft into Category D, making the approach illegal. There was also a Notam, which was provided to the captain before the flight began, that stated circling was not authorized at night.
Even if the pilots had somehow concluded the airplane fell into Category C at the time of the approach, the weather was still conspiring against them. The approach procedure calls for three miles visibility for a Category C airplane, and in fact when the airplane was 20 minutes out the visibility was reported as 10 miles.
However, at the time of the accident a snowstorm had reduced visibility to two miles. Just prior to the jets arrival, three aircraft had attempted instrument approaches into the airport and only one made it, on its second attempt at that. Nevertheless, the crew reported the runway in sight. As they turned from base to final they struck terrain. All 18 aboard were killed.
Trying to salvage the approach with an aggressive base-to-final turn is another common trap. There is seldom a valid reason to exceed 30 degrees of bank while maneuvering on instruments or in the traffic pattern. Steep banks result in a sharp increase in stall speed unless you also set up a rather prodigious rate of descent – which runs counter to the reduced need to descend while circling after an instrument approach.
Yet the accident record shows it happens all too frequently. Probably every pilot has at one time or another rolled the airplane into a sharp, steep turn in an effort to salvage poor runway positioning. Most get away with it a time or two. The problem usually begins when you set up your downwind too close to the runway or strong winds make you overshoot the final approach track.
It costs money to go around and repeat the procedure. And there are passengers to consider. They dont want an extended flight. To miss the approach due to weather or an airport restriction requires diverting to an alternate.
In the case of the G III at Aspen, diverting would have meant a trip to Colorado Springs or Grand Junction, then a very long trip in several rental cars with a late-night arrival. They might even get stuck in the alternate city overnight because of highway conditions. For the short weekend ski trip, a chartered jet to the destination was doubtless meant to avoid such delays.
Then too, there are human factors to consider when pondering why pilots dont abandon poor approaches rather than trying to salvage them. Foremost among them is the pilots ego. A missed approach, especially if made early in the approach, has somehow become an unstated failure. Most pilots go missed very reluctantly and often too late.
In theory, a skillful pilot should be able to execute every approach right the first time. Yet strong gusty winds, optical illusions, varying visibility at various points, momentary spatial disorientation or just a good ol oops can put the best pilot in a bind. If youre lucky, you bite the bullet and admit the temporary failure. Then, weather permitting, try it again.
Too often, a pilot tries to salvage a bad approach with steep turns. A tragic example involved a Piper Aerostar 602P. The weather was forecast to be 1,000 foot ceilings with reasonable visibility, but witnesses said when the flight arrived the ceiling was about 400 to 500 feet agl. The pilot completed a VOR approach to runway 22 at Georgetown, Del., then canceled his IFR flight plan. As he circled to land witnesses saw him make a steep left banking turn about 250 feet above the ground.
Upon roll-out the airplanes wings rocked back and forth then the nose and right wing dropped below the horizon. The airplane crashed, killing the pilot and both passengers.
In another case, the pilot of a newly purchased Cessna T337G was attempting a circling approach at Fullerton, Calif. Witnesses said the airplane entered an 80-degree bank angle in the base-to-final turn. The Skymasters POH stall speed chart shows that the aircraft as loaded would have a wings-level, full-flap stall speed of about 60 knots indicated. With a 60-degree bank the wings stall at 85 knots. With the estimated 80-degree bank angle the airplanes stall speed would be about 120 knots.
As the pilot attempted to recover and go around, he began retracting both landing gear and flaps. Then witnesses saw the airplane stall and crash. Retracting the flaps raised the stall speed again. Both the pilot and his son survived with serious injuries.
It happened also to a Learjet 35A at Aspen-Pitkin County Airport. It was at eight minutes before sunset and the flight had made a VOR/DME approach and was circling to land. A snow squall had passed over the field and still obscured the mountains to the east.
The terrain was snow-covered and featureless, possibly contributing to spatial disorientation. As the pilot circled to land, the aircraft was banked steeply to the right as it overshot the extended runway centerline. Witnesses reported a variety of indications of loss of control, including a flutter.
The last recorded transmission on the cockpit voice recorder was Oh no! youre … (unintelligible) stall. Impact was about one mile north of the runway 15 threshold. Both pilots were type-rated for the Learjet.
In yet another case, a Baron 58P on a VOR/DME-A circling approach to Brookhaven, Miss., crashed on the base-to-final turn. Human factors were involved in this one in that the pilots father and brother – both pilots themselves – were at the airport watching. This added an element of compulsion to the equation.
The weather was ideal for a circling approach, with a 500-600 foot ceiling and 3 miles visibility. The pilot communicated with his father on the advisory frequency when five miles south of the VOR. The father advised that winds were easterly at 7 knots. The pilots brother told investigators that when he first saw the airplane it was south of the airport paralleling runway 04/22. The landing gear was extended and flaps were in the approach setting. When the airplane was about a third of the way down the runway on downwind, the brother noted it was close-in; closer than normal.
When the Baron was approximately 20 degrees past the approach end of runway 22, the brother noted the aircraft began banking to the left on a base-to-final turn. The bank angle continued to increase until reaching 70-75 degrees. Suddenly the nose pitched down and the airplane crashed. Assuming the airplane weighed about 4,000 pounds and was carrying a typical approach speed of 90 to 100 knots in the turn, a 60 degree bank would have put the Barons stall speed at about 85 knots indicated, while a 75-degree bank would have put it past 120 knots.
As these accidents clearly illustrate, an airplanes bank angle has a direct and dramatic effect on stall speed. While the actual numbers vary with aircraft type and wing design, the stall speed of a typical airplane – if such a thing exists – increases a negligible 8 percent with 30 degrees of bank. With a 45-degree bank the stall speed increases 20 percent. Thus a 60-knot stall becomes 72 knots.
A 60-degree bank becomes serious business. Stall speed has increased 40 percent and that docile 60-knot stall comes at 84 knots, which is perhaps faster than many pilots would fly that portion of the approach.
In a 4,000-pound Baron, for example, the 63-knot flaps-down stall speed increases to 88 knots. Now the landing approach and stall speeds are about equal. Bank a bit more and the stall is imminent.
Instrument pilots away from the heat of battle will acknowledge that they should not descend below circling approach MDA until they have intercepted a normal glide path to the runway, generally as depicted by a VASI or PAPI. FAR 91.175(c)(3) specifically outlines what parts of the runway environment must be visible before you can start down.
Unfortunately, despite this guidance, the accident record is littered with pilots who pushed too far and got it wrong.
The problem arises because a normal visual pattern includes beginning the descent abeam the numbers on downwind or during the turn to base, depending on the airplane and the traffic pattern altitude. The descent continues on base and final approach until touchdown.
A circling approach is different. If weather has you at circling minimums, you are essentially flying the pattern at half the usual pattern altitude. Consider that many circling approaches have minimums of about 500 to 600 feet agl. Recall your standard VFR pattern and youll probably turn base to final at least 400 feet agl and possibly more, depending on the airplane and how wide the pattern is.
Thus on a circling approach you should not begin the descent on base leg until you acquire guidance from the VASI. In fact, if the pattern was wide you may have to make a level turn to final and still remain at circling minimums until intercepting a normal 3-degree glideslope. The following are some examples as to why.
On a night instrument approach into Marfa, Texas, the Queen Airs crew of two ATP-rated pilots was cleared for the VOR Runway 30 approach. Because of easterly winds they were circling to land on runway 12. For some reason the captain elected to make a right-hand pattern, which put the runway on the co-pilots side.
As the airplane was abeam the landing threshold, the co-pilot lost sight of the runway. The captain leaned over and looked out the co-pilots window then pointed it out to him.
In a descending turn on base leg, the captain asked if the co-pilot still had the runway in sight, but the co-pilot failed to respond. The captain again looked out the co-pilots side but, as he told investigators, all he saw was blackness. Still he continued the descent.
While completing the base to final turn, the right main landing gear and right wing-tip hit the ground. The rest of the landing gear sheared and the airplane slid to a stop in the rough desert terrain. Fortunately none of the 10 people aboard the airplane was injured.
In another example, the pilot of a Piper Lance had gone missed on his first approach and was attempting a second NDB circling approach. While turning base to final, the airplane dragged a wing tip and crashed, even though the landing runway was equipped with PAPI. The pilot and his passenger were killed.
The airport manager had advised the Lance pilot over the radio of a 700- to 800-foot ragged ceiling shortly before the Lance began its approach. The manager, who had just landed, also reported one mile visibility in fog. The Lance pilot began his approach almost immediately, but by the time the airplane had reached circling minimums the fog had intensified and visibility had diminished to one-quarter mile. Why the airport manager failed to relay this information is not covered by the report, especially after the manager witnessed the Lances first missed approach in below-minimums visibility.
The circling approach can be a valuable way to get into an airport that might otherwise be inaccessible, but inherent in the procedure are traps that can catch the unwary. Resisting the temptation to drop below minimums is common among all instrument approaches, but is more critical on circling approaches because of the maneuvering that follows.
Holding off on descending until ready to ride down the glideslope and watching the bank angle on the base-to-final turn are common practices during visual traffic patterns, but all too often they get lost in the shuffle in the aftermath of a low-weather circling instrument approach.
-by John Lowery
John Lowery is a former Air Force pilot, accident investigator and corporate pilot.