Peter Garrison – Sometimes it’s not the speed that breaks aeroplanes, it’s the slowing down.
IN AN ARTICLE ON THE 2012 crash of a Pilatus PC-12, I faulted the National Transportation Safety Board for mixing up indicated and true airspeeds. Actually, it was I who misread the report. I am indebted to reader Timothy Burtch, an accident investigator with the NTSB, for pointing out that the maximum speed of 338 kt that the aeroplane reached in a spiral dive before it broke apart was, in fact, an indicated airspeed, not a true one, and that the aeroplane did, therefore, exceed its manoeuvring speed by 175 kt as the report stated.
The Pilatus, with a family of six aboard, was climbing through FL250 in IMC. It was at 109 KIAS, in a 25-degree right bank, deviating to avoid an area of rain, when the autopilot disengaged for unknown reasons. Presumably a chime sounded and a warning light illuminated, but the pilot seemingly did nothing to take over control of the aeroplane. A baffling aspect of the accident was the pilot’s apparent failure to act even when the aeroplane was vertically banked and plunging downward at an horrific rate.
Within 10 seconds of the autopilot disconnect, the angle of bank had increased to 50 degrees and the plane had begun to descend. After 30 seconds, the bank angle was 100 degrees and the plane had lost 2,600 feet. In the next 13 seconds, it lost another 5,900 feet while the positive load factor increased to 4.6 G. At 36 seconds the indicated airspeed, having peaked at 338 kt – 430 true — dropped to zero, suggesting that the breakup – the aeroplane lost its horizontal stabiliser and portions of both wings – had taken place somewhere around 15,000 feet.
Almost any aeroplane will enter a spiral dive, sooner or later, if no attempt is made to control it. Some may fly hands-off for many minutes in smooth air, but if they are disturbed by a gust, or if they are banked in the first place, they will inevitably bank more and more steeply, the nose will drop, they will pick up speed and the turn will continue to tighten, and the rate of descent to increase, without limit.
‘reluctant to bank and fatiguing to fly’
Builders of free-flight models may object that their planes can remain right side up indefinitely; but that is because models have much more dihedral effect than piloted planes do. If your plane had the lateral stability of a free-flight
model, it would be reluctant to bank and fatiguing to fly. For the sake of manoeuvrability, therefore, we accept the necessary evil called “spiral divergence.”
In VFR conditions hardly anyone gets into a developed spiral dive unless he literally falls asleep at the wheel. It is the pilot lacking instrument flying skills who stands the greatest chance of experiencing a spiral dive if he loses the visual horizon. In this case, however, what made the sequence of events doubly puzzling was that the pilot was both instrument qualified and current.
The procedure for recovery from a spiral dive is simple: power off, level the wings, and slow down. But there are potential pitfalls even in this simple recipe, and the Pilatus may have encountered one of them.
The aeroplane was trimmed for 109 KIAS. The dynamic pressure of air at that speed is 39 pounds per square foot (190 kg/m2). The wing loading of the aeroplane, which was probably near its gross weight, was about 36 psf (176 kg/ m2), and so the lift coefficient, which is the ratio between the lift force and the dynamic pressure, was about 0.9.
The dynamic pressure at 338 KIAS is 337 psf (1,645 kg/ m2) and the lift coefficient in level flight, if level flight were permissible or even possible at that speed, would be about 0.1.
Normally we don’t talk much about lift coefficients in the context of piloting, as opposed to designing, an aeroplane, but they are of interest in this case because G loads can be thought of as the ratio of two lift coefficients. If, for example, the Pilatus is trimmed for 109 KIAS and a lift coefficient of 0.9, it would experience an acceleration of over 9 Gs if its indicated airspeed were, by magic, instantly increased to 338 KIAS. Another way to
put it is that the acceleration is the square of the ratio of the high speed to the trimmed speed. Thus, if you were trimmed for 100 and you were suddenly indicating 300, the speed ratio would be 3 and the resulting G load, as the aeroplane seeks to return to its trimmed speed, would be 9 G.
‘limp home with wrinkled wing skins’
That is not physically realistic, however, because in the real world speed cannot instantly change from 100 to 300. But you don’t need 9 Gs to break an aeroplane, either. Six Gs will do it, 5 in some cases, and that would have occurred at 270 KIAS or less. In other words, if you are trimmed for 109 KIAS when some sort of upset occurs, you pick up a lot of speed in a spiral dive, and when you level the wings the airspeed indicator is showing over 270 kt, you are entering the neighbourhood of a normal-category aeroplane’s ultimate load factor.
The ultimate load factor is not the one where you limp home with wrinkled wing skins. That’s called the limit load factor. The design limit load factor for most general aviation aeroplanes is 3.8 G. The ultimate load factor, which by a handy convention is deemed to be 50% higher than the limit load factor, is the one where big pieces break off. Pilots will sometimes speak admiringly of an aeroplane that pulled far more G than it was supposed to and still held together. From the point of view of the occupants, that is clearly a good thing, but from an engineering standpoint it isn’t; it means that the aeroplane is more heavily built than it should be.
So a caution has to be appended to the recovery procedure for spiral dives. After you roll the wings level, the aeroplane is going to try to return to its trimmed speed. Normally, this will not be a problem. If you are cruising at 150 KIAS and you look up from a chart or your iPad to find yourself in a 60-degree bank and the airspeed indicator winding through 220, you are not going to take the wings off the aeroplane by just levelling out. You’ll pull a couple of Gs, perhaps, but nothing will break.
‘lower the gear – never mind the doors’
It’s when you’re flying at low speed, for example climbing, as the Pilatus was, and then get into a dive at very high speed, as it did, that the recovery may involve dangerously high load factors, especially if you try to help it along with a pitch-up command. What is actually called for is nose-down trim and a forward push on the yoke. Otherwise the aeroplane may overstress itself even with no help from you at all. Fortunately, when a pilot feels G building up a forward push is the instinctive reaction.
In-flight breakups are rare, but they occur regularly enough to be worth talking about.
Sometimes an aeroplane emerges from the bottom of the overcast already in pieces, but I suspect that some breakups occur only after the aeroplane has emerged into the clear. Then the pilot sees the ground in some highly unfamiliar position and instinctively overcontrols while trying to put it back where it belongs. This is a danger even when the overcast is quite high and there is ample room underneath for a more gradual recovery.
‘the aircraft was surprisingly intact’
A common piece of advice for VFR pilots who have blundered into IMC is to lower the landing gear. As a young pilot I used to wonder what in the world the landing gear could do to keep the aeroplane upright, but I was barking up the wrong tree. The authors dispensing the advice evidently considered it so obvious that they did not bother to explain that the purpose was to increase drag and thus to limit the speed gain in an inevitable spiral dive.
In the extreme situation in which the Pilatus found itself, lowering the gear could be part of the spiral dive recovery: power off, level the wings while trimming nose-down, lower the gear – never mind the doors – and don’t let the nose rise too rapidly. In a reciprocating-engine aeroplane, drag can be further increased by pushing the prop to high rpm.
The essential point, however, is to limit G while reducing speed. If the plane were going to come apart because of sheer speed, it would already have done so. It’s the G forces that pilots apply in their urgency to stop pointing downward that end up breaking aeroplanes.