A wing makes lift by keeping the air attached to it, bending the flow smoothly over a shape built for that purpose. Ice changes the shape. It does not have to be thick, it does not have to be visible from the cockpit, and it does not take much of it. That is why ice on the wing is its own entry in the science of injury: it is a way an aircraft can be lost in a few minutes at an altitude where the ground is a long way down.
What the ice is. The FAA’s pilot guide defines clear ice as “A glossy, clear, or translucent ice formed by the relatively slow freezing of supercooled water drops,” and says what makes it: “large amounts of liquid water, high aircraft velocities, and large drops are conducive to the formation of clear ice.” Supercooled large drops, which the guide abbreviates SLD, are “Water drops with a diameter greater than 50 micrometers (0.05 mm)” “that exist in a liquid form at air temperatures below 0 °C”; they are liquid until they hit something, and the wing is the something. Rime, the other kind, “is generally rougher in appearance than clear ice.” The guide’s word for the ice that freezes behind the heated or booted surfaces, from water “running back to unprotected surfaces,” is runback (FAA, AC 91-74B).
What it does to the wing. The numbers are plain. “A reduction of CLmax by 30 percent is not unusual,” CLmax being the most lift a wing can make, and “a large-horn ice accretion can result in reductions of 40 percent to 50 percent.” Drag goes the other way: “An airfoil drag increase of 100 percent is not unusual,” “for large-horn ice accretions, the increase can be 200 percent or even higher,” and, of the drag, “Note that the effect is significant even at very small AOAs.” It does not take a large accretion. “A significant reduction in CLmax and a reduction in the AOA where stall occurs” “can result from a relatively small ice accretion.” Rough is worse than big: “Wind tunnel testing indicates that if such accretions are particularly rough” “they can have more adverse effects than larger accretions that are relatively smooth.” The result for the pilot is a wing that quits earlier than the one he trained on. “Even a thin layer of ice at the leading edge of a wing, especially if it is rough,” “can have a significant effect in increasing stall speed,” because “the Angle of Attack (AOA) at which a stall occurs (the stall angle) is much lower with ice than without ice.” A stall is the wing letting go of the air; with ice, it lets go at a smaller angle and a faster speed than the airplane’s warnings may be set for (FAA, AC 91-74B).
The tail. The part of the airplane that collects ice fastest is the part no one can see. “An aircraft’s tailplane will be a better collector than its wings” “because the tailplane presents a thinner surface to the airstream,” and “On most aircraft, the tailplane is not visible to the pilot,” who “cannot observe how well it has been cleared of ice by any deicing system.” The tail is what holds the nose up, and the guide warns that “With ice on the tailplane, it may stall after deployment of flaps,” so pilots are told to “be alert to the possibility of tailplane stall, particularly after full flap deflection” (FAA, AC 91-74B).
What certification promises. Less than people assume. “An aircraft that is certificated for instrument flight rules (IFR)” “is not necessarily certificated for flight in icing conditions.” An airplane approved for known icing is built to “safely operate throughout those icing conditions encompassed by the icing envelopes specified by the FAA,” and then the guide adds the clause that matters: “It is always possible to encounter an unusual condition for which the aircraft has not been certificated,” its example being liquid water content outside the envelopes, which may show as a very rapid rate of accumulation. Severe icing is defined by failure, not by thickness, though the guide gives a reference rate of more than 3 inches an hour: “The rate of ice accumulation is such that ice protection systems fail to remove the accumulation of ice,” and “Even airplanes approved for flight into known icing conditions should not fly into severe icing.” “Airplane certification for flight into known icing conditions does not include freezing drizzle and freezing rain.” When severe icing is met, “By regulation, immediate exit is required” (FAA, AC 91-74B).
The other edge. High up, a jet flies inside a narrow band. On the slow side is the stall, which ice moves faster. On the fast side is a limit of its own. Every jet has a maximum operating speed, written as VMO, “maximum operating speed expressed in terms of knots,” and MMO, “maximum operating speed expressed as a Mach number,” and those two “always define the maximum operating speed of the airplane.” “At high/cold altitudes, the speed of sound is lower so the aircraft is limited by Mach,” and “The handling qualities of a jet may change significantly at speeds higher than the maximum allowed.” A shock wave forms on the wing well below the speed of sound, and far enough past the limit “the separation of air behind the shock wave may result in severe buffeting and possible loss of control,” and the wing’s center of pressure moves aft, “causing the start of a nosedown tendency or” tuck. “Mach tuck develops gradually, and the condition should not be allowed to progress,” “to prevent entry into a steep, sometimes unrecoverable, dive.” The handbook has a name for the altitude where the slow-side buffet and the fast-side buffet meet: coffin corner (FAA, Airplane Flying Handbook, Chapter 16). An overspeed is not the opposite of a stall. It is the other way out of the same narrow room.
The upset. This is the general mechanism the handbook teaches, not an account of any flight. “Spatial disorientation has been a significant factor in many airplane upset accidents,” and “Disoriented pilots may not always be aware of their orientation error.” In cloud, “A pilot distracted by other sensations can easily enter a slightly nose-low, wing-low, descending turn,” which is how a spiral dive begins, and the recovery has an order the handbook writes out: “Reduce power (throttle) to idle,” unload before rolling, “Roll to wings level using coordinated aileron and rudder inputs,” then “Gently raise the nose to level flight.” The errors it lists are “Attempting to pull out of dive without rolling wings level” and “Failure to reduce power first,” and a dive like that “might be at or even beyond VNE (never exceed speed) speed,” where a startled pull on the yoke creates loads “capable of causing airframe structural damage and/or failure” (FAA, Airplane Flying Handbook, Chapter 5).
The arithmetic. Descent rates in aviation are given in feet a minute, and the numbers hide how fast that is. A descent of 10,000 feet a minute is about 167 feet a second, roughly 114 miles an hour straight down; 20,000 feet a minute is about 227 miles an hour; 30,000 is about 341. Those are vertical speeds alone, before the airplane’s forward speed is added, and a stop from any of them against the ground is the shortest stop there is, which The Fixed Object and Seven Times the Energy explain for the ground vehicle: the energy rises with the square of the speed and the distance to spend it is inches. The figures are arithmetic, not a measurement of any flight.
What the record holds. A crash in cloud leaves a longer paper trail than a crash on a road. The air traffic control recording, with its timestamps, is public within hours. The radar and tracking data put altitude, speed and descent rate on the same clock. The weather is on file: the forecasts, the pilot reports from other aircraft, the radar returns. The airplane’s own records say what it was certified for, what its flight manual allowed in icing, how its anti-ice systems were maintained and what the crew was trained to do. If it carried a cockpit voice recorder and a flight data recorder, those say the rest, when they survive. And the question the law asks of all of it is not only why, which is the investigators’ question, but what the people aboard knew and for how long, which is the question Conscious and Mayday take up. None of this page is about any one flight. What ice does to a wing is the same on every wing; what happened on a particular one is for the record of that flight to say.
FAA, Advisory Circular 91-74B, Pilot Guide: Flight in Icing Conditions (Oct. 8, 2015) (paragraph 1-3g, page 2, and 3-1a(1), page 15, clear ice; 1-3ll, page 5, supercooled large drops; 1-3ee, page 5, rime; 1-3ff, page 5, and 2-1d, page 9, runback; 3-2b(1), page 18, the lift and drag figures and the small accretion; 1-4d, page 7, roughness; 3-2a, page 17, the thin layer and the stall angle; 3-1a, page 15, and 3-7a and 3-7b, pages 22 to 23, the tailplane; 1-4a, page 6, 3-12a, pages 26 to 27, 1-3gg, page 5, 5-2a, page 39, and 4-2d, page 36, certification, severe icing and exit; 1-1, page 1, the guide is not regulatory). FAA, Airplane Flying Handbook, FAA-H-8083-3C, Chapter 16, Transition to Jet-Powered Airplanes (page 16-4, VMO, MMO and handling qualities; page 16-6, Mach buffet, Mach tuck and coffin corner). FAA, Airplane Flying Handbook, FAA-H-8083-3C, Chapter 5, Maintaining Aircraft Control: Upset Prevention and Recovery Training (page 5-3, spatial disorientation; page 5-26, the spiral dive, its recovery and VNE; page 5-27, the common errors). The feet and miles are arithmetic: feet a minute divided by 60 gives feet a second; feet a second times 3,600 divided by 5,280 gives miles an hour. This is general information, not legal advice. Figures and text as published at publication.
