Wx Watch: Icing Insights

Every pilot knows—or certainly ought to know—that airframe icing represents one of flying's greatest hazards. Textbooks and the more thorough accident reports address these subjects frequently, but a seasonal reminder is always in order. Besides, researchers and accident reconstructions are always coming up with new insights into the icing problem. Their findings deserve to be widely disseminated, but for one reason or another many pilots aren't exposed to late-breaking information and safety advice. Let's review some icing basics, then look at some icing issues that have been in the news but somehow haven't become—but ought to be—knowledge resident in all pilots.

First, the basics. Airframe icing comes in six main types—clear, rime, mixed, large-droplet, freezing rain, and frost. Here are the characteristics and typical environments identified with each type:

• Clear. As the name implies, clear ice is transparent. It's most often found in cumulus clouds, in clouds associated with marine or lacustrine (lake) air masses, and in the zero-to-minus 10-degree Celsius temperature range. Runback is the biggest danger here. By runback, we mean that the ice flows aft of the leading edges of the wing and other lifting surfaces. This lets clear ice flash-freeze on the areas of the wing most important to safe flight—those portions of the wing that ordinarily experience laminar flow and create the most lift. Clear ice is a large-droplet phenomenon. It's these large droplets that are the root cause of runback; they hit the leading edges and have enough momentum to keep moving aft on the wing chord before finally freezing solid.
• Rime. Rime ice is a whitish, opaque type of formation that doesn't run back from the wing leading edges. Instead, it builds with time and juts forward of the wings' impingement points—the thin, spanwise zone on a leading edge that first receives the oncoming icing droplets. Rime ice typically forms in the minus 10-to-minus 20-degree Celsius temperature range, and is associated with smaller water droplets of the kind found in stratus clouds. Statistically speaking, it's the most commonly encountered sort of icing condition.
• Mixed. This is a mixture of both clear and rime ice that can happen in temperatures just below freezing. Formations appear pebbly and can morph into horrendous-looking double-horn accretions.
• Large-droplet. This is the type of icing that apparently felled an ATR–72 commuter over Roselawn, Indiana, in October 1994. Another accident involving a turboprop twin commuter airliner, on approach in Detroit, has also implicated large-droplet icing. The Roselawn accident accelerated research into what has come to be called supercooled drizzle drops, or super-large-droplet (SLD) icing conditions. This represents the worst kind of clear icing in that runback can extend as far aft as 30 percent of the wing chord—well aft of any leading-edge areas protected by deice boots. This form of icing is associated with temperatures just a few degrees below the freezing mark, and the kind of sopping-wet air masses found around the Great Lakes and Northern Europe. Before Roselawn, no one knew about SLD icing. Now it's been "discovered."
• Freezing rain. This is really a subcategory of the clear-icing phenomenon. It, too, lives right around the freezing mark, and there's plenty of runback. It can accrete with shocking speed, and is most often found in warm frontal conditions where rain aloft falls through the frontal boundary and chills in the subfreezing air of the retreating cold air mass ahead of it.
• Frost. This forms on cold clear nights when the temperature and dew point are within five or fewer degrees of each other and the dew point is below freezing. Even the lightest coating of frost can radically change airfoil behavior, so it's imperative that all frost be removed before takeoff.

Micron-speak

How big are the water droplets that cause icing? This question is answered in terms of diameters measured in microns. A micron is one-thousandth of a millimeter. If you twirled a pencil point on a piece of paper, the resulting dot would have a diameter of about 500 microns. So when we talk about icing droplets, we're talking small.

Rime icing happens with cloud droplets in the 15- to 40-micron droplet size. Clear icing droplets are slightly larger—in the 15- to 50-micron range. Those are the numbers used in icing certification tests, and which define the icing environment in the eyes of the FAA. But the certification envelopes were developed back in the late 1940s. Now, it seems, nature doesn't always agree with the government. Studies have shown that most documented icing encounters—80-plus percent, whether clear or rime—involved droplets in the 10- to 15-micron range.

As for large-droplet icing, droplet sizes are in the area of 1,000 microns or more in diameter—as big as small raindrops. This helps further explain the runback problem.

Tailplane ice

The horizontal stabilizer is an airfoil, too. It just provides negative lift—lift that balances out the lift produced by the wings and gives the airplane pitch stability. In the past few years, there has been a good deal of study on tailplane icing. What experts have found is that the small radius of a tailplane's leading edge—like any small-radius object on an airplane, such as a temperature probe—is a very efficient ice collector.

When a tailplane ices up and stalls, negative lift is eliminated and the airplane pitches nose-down. This is now suspected as the cause of several previously unexplained airline accidents.

Adding to the tailplane icing problem is the fact that it's most likely to occur on final approach, when the airplane is being slowed and flaps are being extended. When flaps are extended there is a significant increase in the horizontal stabilizer/tailplane angle of attack, and wing downwash can further increase the tail's angle of attack, causing it to stall. Ice accretions on the tail leading edge cause this type of premature stall because they've reduced the tail's lifting force.

Pilots can tell if a tail stall is impending. Here are the warning signs, which may be masked if an autopilot is flying:

• Elevator/stabilizer control pulsing, oscillation, or vibration
• Abnormal nose-down pitch trim changes
• Unusual or abnormal pitch anomalies, which could result in pilot-induced oscillations
• Reduction or loss of elevator effectiveness
• Sudden elevator force change
• Uncommanded nose-down pitch.

To recover from a tail stall, test pilots recommend doing the following:

• Immediately retract flaps to the previous setting
• Apply nose-up elevator/stabilizer force
• Increase airspeed to an amount appropriate for the reduced flap setting
• Apply sufficient power for configuration and conditions
• Make nose-down pitch changes slowly—even in gusty conditions
• If deice boots are available, inflate them several times in an attempt to clear the tailplane of ice.

The tough part about the recovery procedures is that they're the diametric opposite of what our training recommends for a conventional stall—that is, pulling on the control column instead of pushing.

We'll talk in more depth about strategies for dealing with in-flight icing next month, in a second installment of this story.

Links to additional information on icing can be found on AOPA Online ( www.aopa.org/pilot/links/links0010.shtml). E-mail the author at [email protected].