Wx Watch: Tailplane Icing

The airplane was on a night visual approach in freezing rain. It turned base, then final. After rolling out, witnesses saw the airplane pitch up slightly, then nose 90 degrees down and crash short of the runway.

Another time, another airplane, also on approach in icing conditions. It pitched over immediately after glideslope intercept, then recovered at 100 feet agl. It then flew along at that altitude for 2 miles. There were some airspeed changes, apparently due to a flap configuration change, and the pilot initiated a missed approach while passing over the runway. Then the airplane accelerated, and the pilot began to retract the flaps. As the airplane climbed out of ground effect, it suddenly pitched over and crashed.

The causes of both these fatal crashes were still unknown when another airplane, also flying an approach in icing conditions, experienced what its crew described as a series of two or three pitch oscillations above and below the glideslope. The crew said they never felt like they had total pitch control; nevertheless, the airplane landed without incident. Post-flight investigation showed that ice on the horizontal stabilizer's leading edge — together with specific airspeeds and flap settings — caused the tail to stall. The riddle of the first two accidents was solved.

Those three cases took place in April 1958, January 1963, and February 1963, respectively, and all involved the Vickers Viscount, a popular four-engine, 40-seat Transport-category turboprop of the day.

Though the dangers of ice-induced tail stalls have been known for years, a more recent spate of 18 fatal accidents involving several different types of newer turboprops — principally the Saab 340, ATR 42, Embraer EMB-110 Bandeirante, Jetstream 31, and Mitsubishi YS-11 — created the impetus for the first International Tailplane Icing Workshop in 1991, a joint FAA/NASA effort. A second workshop was held last year, and from these two meetings have come several revelations and recommendations of major importance to pilots, airframe manufacturers, and various regulatory bodies. In short, the workshops are wake-up calls to the entire aviation community.

By the time the second workshop was convened, several incidents involving general aviation aircraft had come to light. Tail icing was strongly suspected in the crash of a Cessna 210 in Stockholm, Sweden. The pilot of a Piper Cherokee experienced a stabilator stall after he extended the flaps. The stall was so violent, the airplane pulled negative Gs, and when the startled pilot released the flap handle, the airplane recovered. Cessna 414s are alleged to have pitch control anomalies when the tail is iced up. A Piper Malibu pilot experienced an uncommanded pitchover with flap extension. When he retracted the flaps, he regained pitch control. It's interesting to note that at the time of this incident, the pilot reported very little ice on the wings but a half-inch, double-horn ice formation on the horizontal stabilizer's leading edge.

When we think of airframe icing, we tend to focus our concerns on ice adhering to the wings, propeller, pitot tube, and windshield. Perhaps it's because icing of these components is emphasized so heavily in training curricula, or maybe it's just that these items are in our immediate field of view. The truth, of course, is that every surface exposed to the relative wind can ice up. This includes icing of the horizontal stabilizer or stabilator, depending on the design of the airplane. (For simplicity's sake, we'll simply refer to these structures as tailplanes from now on.)

Tailplane icing is every bit as critical as wing icing. If ice forms on a tailplane's leading edge, the tail can stall — just as surely as ice accretions on a wing can cause that airfoil to lose lift. Because the tailplane generates negative lift — producing a downward force — a tail stall can cause an abrupt rising of the tail, a sudden pitchover, and loss of control. If this happens at low altitude, the consequences are usually fatal to all aboard.

Under what conditions are ice-contaminated tailplane stalls most likely to occur? From a meteorological point of view, large supercooled water droplets are most conducive. Porter Perkins, a well-known figure in the icing research community and contributor to the FAA/NASA workshops, identified droplets ranging from 40 to 5,000 microns in diameter as the worst culprits. The 40-micron droplets are most often found in cumulus clouds and create clear ice formations. The 5,000-micron (about 5 millimeters) droplet is characteristic of freezing rain conditions.

Large droplets are bad because they adhere so quickly and easily run back from the initial point of impact. This is especially true of small-radius objects such as antennas, probes, and, for that matter, tailplane leading edges. In one incident, after a tailplane stall occurred and the airplane landed, there was 1 inch of ice on the wing, 2 inches on the windshield wipers, and 3 inches on the tailplane leading edge. There have been many other cases — such as the Malibu incident mentioned earlier — that further illustrate the tailplane's great ice-collecting efficiency.

Headwind gusts are also bad news, as are downdrafts. These phenomena can momentarily increase the tailplane's angle of attack and bring it closer to the stall.

Most tailplane icing incidents occur in the approach phase of flight, when work load is highest and flaps are deployed. The more effective the flaps, and the more they're extended, the more likely a tailplane stall. Higher than normal airspeeds, a forward center of gravity, and sideslips add even more danger. All of these factors increase the tailplane's angle of attack, beginning with the nose-down pitching moment that comes with initial flap deployment on a low-wing airplane. (High-wing airplanes tend to pitch up with flap extension.) Airplanes with trimmable stabilizers, it should be noted, have less of a problem because the entire tailplane can be trimmed to lower angles of attack for any given airspeed.

Perhaps most problematic is the effect of higher than normal approach speeds. The greater the airspeed, the greater the need for a compensating nose-down elevator force. This has the effect of reducing the angle of attack on the wings and increasing the angle of attack on the tailplane. So the traditional remedy of adding airspeed to protect against a stall may hold true for the wing, but not for an ice-contaminated tail.

Extra airspeed also helps build more ice. The faster you fly, the more directly a water droplet will splatter against a leading edge and run farther back on an airfoil — away from any areas protected by deice boots or thermal ice protection systems.

Workshop participants listed several warning signs of a tailplane stall. These include:

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

*Bear in mind that the pilot may not even notice these warnings if using an autopilot.

Though recovery procedures are seldom published, the last workshop published a recommended, generic sequence of actions. At the first sign of a tailplane stall, the participants said to:

1. Immediately retract the flaps to the previous setting.
2. Apply appropriate nose-up elevator/stabilator force.
3. Increase airspeed appropriate for the reduced flap setting.
4. Apply sufficient power for configuration and conditions. High engine power settings may be adverse to tailplane stall conditions on some designs at high airspeed. Observe the manufacturer's recommendations for power.
5. Airspeed may be increased to eliminate wing buffet (airplane vibration sometimes accompanied by lateral instability), but observe the manufacturer's recommendations for airspeed at the flap configuration.
6. Make nose-down pitch changes slowly — even in gusting conditions, if circumstances allow.
7. If a pneumatic deicing system is used, operate the system several times in an attempt to clear the tailplane of ice.

John Dow, project officer of the FAA's icing certification steering group, foresees a host of necessary changes as a result of the workshops' disclosures. For example, certification flight testing of tailplane stalls — both in and out of icing conditions — ought to be better researched.

"I'm not sure we totally understand what's going on back there," Dow said, referring to tailplane aerodynamics. "This tailplane icing study is a new thing — that's why it's taken so long to get this project going.

"We're all very familiar with the way wings behave and have developed reliable certification criteria to identify and test their aerodynamics," said Dow. "For example, we have to determine 130 percent of the stall speed of a wing in order to meet the certification rules and publish a safe approach speed. But that's for the wing's stall speed. There's no language requiring us to find out the tailplane's stall speed, or how the tailplane flies at various margins above the stall — with or without ice on it."

Flight in icing conditions is bad enough without pondering the upside-down world of tailplane stalls — a world where you pull on the yoke to recover, and more power and airspeed can work against you. But if the researchers can convince the powers that be, we can all expect to see and learn more about tailplane icing and tailplane stalls — including new technologies that will allow us to detect tailplane icing before it becomes dangerous (see "Tailplane Icing Workshop Recommendations").

For those of us who elect to fly on instruments in winter weather, the more information, the better. Those of us who remain VFR are certain to have even greater reason to take comfort in that decision.

Tailplane Icing Workshop Recommendations

A synopsis of concluding remarks

Information

1. Continue to distribute information to aviation industry publications like Flying and AOPA Pilot, ensuring that the industry understands that although the turboprop may have initiated work in ice-contaminated tailplane stalls (ICTS), small general aviation and corporate airplanes are also susceptible.
2. Propose that the airplane manufacturers distribute information about ICTS to their owner networks.
3. Attempt to use national weather channel TV media for dissemination of ICTS information because of its large pilot viewer population, accessibility, and the ability to schedule the broadcast to coincide with periods of forecast or reported icing conditions.
4. Because of the possible severe icing caused by large drops in freezing rain, freezing drizzle, and ice fog, flight crews should be made aware of the seriousness of this hazard and, where possible, be advised of the possibility or existence of these conditions.

Regulations and Guidance Material

5. Develop and distribute advisory material on stability, control, and performance testing similar to guidance from Transport Canada and the Joint Aviation Administration.
6. Harmonize the definitions and information on ice that appear in the Airman's Information Manual and 14 CFR Parts 91, 121, and 135.
7. Include in the airplane flight-test guides the use of the pushover maneuver [ a negative-G flight-test procedure used to measure tail stalls — Ed.] to evaluate handling qualities with ice-contaminated empennage.

Technology — Research and Development

8. Speed up the implementation of ice detector technology as a practical solution to ICTS.
9. Help the manufacturer by better defining critical ice accretion so that the manufacturer is not unnecessarily burdened in the certification process.
10. Improve computer codes for ice accretion and computational fluid dynamics solutions of the aerodynamic effects of ice accretion so that they will be available to the small-airplane manufacturer.
11. Evaluate possible remedial modifications to existing and new designs that would lessen the adverse effect of ice accretion on the tailplane stall angle.

Training

12. Distribute ICTS information to the FAA's accident prevention specialists for training.
13. Consider use of the AOPA/BFGoodrich/FAA video for pilot training. — TAH