Autunno 1999

Siamo alle soglie dell’inverno ed ancora una volta riteniamo utile riproporre la problematica della formazione di ghiaccio sulle superfici e sui motori degli aeromobili.

Abbiamo utilizzato l’indagine sul disastro dell’Antonov AN-24 accaduto a Verona il 13 dicembre 1985 raccontata da uno dei Membri della Commissione di Civilavia istituita per l'investigazione, il Comandante Arturo Radini.

Dalla dinamica dell'incidente si possono trarre molte indicazioni per rafforzare un atteggiamento difensivo nei confronti di un "nemico" subdolo e spesso sottovalutato che si manifesta in maniera diversa ed imprevista.

I risultati dell’indagine dell'NTSB relativa all’incidente del ATR-72 (31 Ottobre 94) della Simmons Airline (American Eagle) in fase di attesa per l'avvicinamento a Chicago O'Hare evidenziarono la presenza di uno specifico tipo di formazione di ghiaccio. Si trattò di SLD che sta per Super cooled Large Droplets.

L'SLD è un tipo di ghiaccio che necessita di almeno 3 condizioni meteorologiche particolari:

1. Freezing rain
2. Freezing drizzle
3. SCDD una forma di freezing drizzle descritta appunto come (Super Cooled Drizzle Droplets).

Nel caso dell’incidente ATR-72 ci si trovava in presenza di grande quantità di acqua nell’atmosfera, di temperatura prossima al freezing point e di "goccioline" SCDD che misuravano 400 micron di diametro, ben 10 volte più grandi di quelle previste in progetto dalle norme FAA.

In effetti gli impianti anti-ice degli aerei sono disegnati per coprire le parte di superfici sulle quali si può formare ghiaccio purché le famigerate goccioline che compongono il ghiaccio siano di misura inferiore ai 40 micron.

Quando ci troviamo in presenza di questo tipo di ghiaccio l’ala si ricopre di uno strato che oltre ad estendersi lungo la corda, quindi al di là della zona protetta, presenta caratteristiche di ruvidezza e estrusione (creste, protuberanze (fingers), ecc.) che aumentano significativamente la resistenza aerodinamica con ovvie conseguenze.

Il documento lo proponiamo in lingua inglese.

E con l'occasione auguro a tutti i professionals una "buona ri-lettura" anche delle sezioni Special Operations nel Technical Manual e nell'Operations Manual.

Silvano Silenzi

 

ROLLING UPSETS IN SEVERE ICING

Now you have an entirely new reason to avoid icing.

By DAN MANNINGHAM.

Three years ago, the aviation community was surprised by research that explained the existence of serious threats to pitch control caused by icing on some airplanes with flaps fully extended. Now, additional research has revealed that some airplanes are susceptible to uncommanded and uncontrolled roll excursions in severe icing. The causes and effects of airborne icing still are not fully understood, but keep this one fact in mind: this season you have at least one new and serious reason to avoid icing. A newly understood effect of inflight icing emerged from the investigation of a serious commuter accident.

On october 31, 1994, an ATR-72 suffered a roll upset during descent after holding in severe icing conditions. The crew was unable to recover from this roll upset, and the airplane crashed into a field in Roselawn, Indiana, killing all 68 persons on board. Prior to the upset, the airplane was holding in an icing environment that was almost more intense than the crew realized. Further, while holding with the flaps extended, this severe icing conditions caused ice to form aft of the deicing boots. Thus, this critical ice could not be shed even though the ice-protection system was funcyioning normally. In addition the crew elected to use the autopilot during this holding period, and that decision likely obscured the development of control problems due to severe icing. Then, when the flaps were retracted, the airplane suffered a roll upset from which it never recovered.

During the accident investigation, the NTSB discovered other accidents and incidents that occurred in conditions of "super cooled large droplets"(SLD). SLD is an icing category that includes, at least, three conditions:

• freezing rain, freezing drizzle and a little known form of freezing drizzle described as "super-cooled drizzle drops" (SCDD).

In the case of ATR-72, the prevailing conditions likely included a high degree of liquid water in the air, a temperature near freezing and droplets size up to 400 microns or 10 times larger than those assumed for normal certification requirements.

FAA icing certification assumes a maximum droplet diameter of 40 microns to determine the aft limit of ice system protection coverage (one micron is one millionth of a meter or approximatly 0.00004 inches). At or below 40 microns, droplet inertia and droplet drag normally limit ice accumulation to the area covered by the area covered by the ice protection system. Stated another way, ice protection systems are designed to cover that portion of the airfoil on which ice would be expected to form if the droplets were no larger than 40 microns.

When they occur, drizzle drops may be 10 times that 40 microns diameter assumed by the aircraft certification policy of the FAA. Freezing raindrops may be as large as 4000 micron -100 times the assumed size. At a size of 400 microns, drizzle drops would have 1000 times the inertia to carry them further aft on the chord line, but only 100 times the aerodinamic force (drag on the droplet) to inhibit the movement.

The combined effect is that drizzle drops would tend to strike the airfoil and adhere farther aft along the chord. Droplets larger than drizzle drops have exponentially greater inertia than drag, compounding the effect even further.

Thus, SCDD impinge aft of the area covered by the ice protection system and accumulate as ice that the operating system cannot shed. Ironically, the extreme droplet sizes associated with freezing rain may, in some cases, actually be less hazardous, because they tend to form an extended layer. That layer, sometimes, evenly covers the whole aircraft with a thin, uniform coating of ice. SCDD, like freezing drizzle, are particularly hazardous because they tend to form ridges, feathers and fingers that disturb the airflow, causing greatly increased airplane drag and higher stall speeds. In a different scenario caused by slightly different circumstances, these drizzle drops can disturb the flow over control surfaces, with less obvious increases in drag.

How roll control is affected.

Two scenarios are possible for ice accretion leading to roll control degradation or loss in the presence of SCDD or freezing drizzle. At moderate angles of attack associated with the slower speeds used in holding or loitering without flaps, ice from drizzle drops can form sharp fingers or feathers perpendicular to the surface of the airfoil. These protruding ice formation can be up to 10 mm.(0.39 inch) high over a large chord-wise expanse of the wing’s lower surface, and the fuselage. The ice increases drag, requiring an increase in angle of attack to maintain altitude. If not corrected, such adjustements in angle of attack lead to a conventional stall, although possibly at indicated airspeeds somewhat above those normally associated with the prevailing configuration. This is due to the airfoil changes created by the ice accumulation. The stall, of course, is likely followed by a roll-off perceived as a roll upset. The second scenario is very dissimilar to the first. If the angle of attack at which the ice forms is somewhat less than in the scenario above (as it would be in higher speed flight or at some slower speeds with flight extended), the cosequences can be enterely different. In the case of lower angles of attack with some forms of SLD present, the inertia of the droplets will carry them beyond the ice protection area, allowing ice to build on the suction surface of the airfoil forward of the ailerons. There it can block airflow over those critical surfaces, and two very dissimilar consequences can result. First, ice forward of the ailerons can unbalance the sum of the aerodynamic forces that tend to keep the ailerons at a neutral position. In extreme cases, those forces can become sufficiently unbalanced to "snatch" unpowered ailerons into a fully deflected position, imparting a significant rolling motion to the airplane. In most cases of the ailerons snatch, control can be regained by normal control-wheel motion, althoug more force will be needed to maintain control. When aileron snatch, the aerodynamic forces across them will have changed.

So, instead of requiring force to deflect the ailerons, force is required to return them to neutral. Depending on the geometry and design of the ailerons, the force required to recover and control the airplane can range from minor to very large.

You may observe some signs of this impending aileron instability in the form of vibration or buffeting in the control wheel in the roll axis. However, those indications may be masked if the autopilot is in use. And, if the ailerons do snatch, you likely be able to regain control by muscling the control wheel back where it belongs.

The second possible result of ice buildup is perhaps more hazardous. Ice can form ahead of the ailerons (whether powered or unpowered) and/or in front of the roll control spoilers. This ice can form in the way that blocks the airflow so that those controls cannot produce the same rolling moment as an uncontaminated airplane. Then, if the airplane is displaced in a roll for any reason, the control and stability characteristics are degraded, and the airplane will not respond normally. In extreme cases, this ice ahead of the ailerons/roll spoilers could prevent recovery.

Ice Accretion Patterns

Ice accumulates at different rates for a number of reasons: airspeed, temperature, droplet size, moisture density, etc. And ice generally accumulated at a faster rate on smaller-radius surfaces than on larger-radius surfaces. All other things being equal, ice will tend to accumulate on the wing tips at a greater rate than it does on the wing roots. In addition, the ice accumulation is likely to extend along a greater percentage of the chord line. Further, even the same thickness of ice will often have a more adverse effect on airfoils of a smoller chord. These effects are compounded in multiengine, propeller-driven airplanes, since airflow in the prop wash tends to maintain a more constant angle of attack. Taken together, these two effects may produce a stall that begins at the wing tips, a reversal of normal design criteria. When it does happen, a tip stall would produce at least two effects:

1 - The wing would generally continue to "fly", althoug roll control and stability would be partially or completely lost.

2 - The tip stall probably would not produce the sharp "G-breaking" sensations of a traditional stall, thus creating an insidious situation in which the pilot is left whithout critical cues of airplane flying conditions. And, just when the pilot needs to immediately reduce the angle of attack with an aggressive pitch change in order to break the stall, he or she would be preoccupied with correcting the roll oscillations in an airplane with significantly different "feel" and degraded roll authority and stability.

Visual Cues of SLD

Forecasting techniques cannot provide complete accuracy for avoiding encounters with SLD icing conditions. You will have to rely on avoidance, or exit strategies, of any atmospheric conditions that produce icing conditions greater than those for which your airplane is certificated. It is difficult to determine the droplets size in a cloud until it is penetrated, of course. If, however, you do inadvertently encounter a situation which you believe may envolve exposure to SLD, take any possible action to depart the area immediately. You should declare an emergency because, in fact, it is one. At all times, when operating in areas of possible icing conditions, watch for the following signs of abnormal icing conditions:

• Any ice accumulation - especially ridges of ice - on the upper or the lower wing surfaces aft of the area protected by the anti-ice/de-ice system. (At night this will require appropriate wing-illumination lighting. Lower-surface wing ice often creates high airplane drag and a concomitant increase in engine-power requirements for any particular regime of flight)

• Any ice accumulation on the propeller spinner that develops noticeably aft of normal ice-accumulation patterns.

• Unusually extensive ice coverage of parts of the airframe not normally affected by icing, or visible ice "fingers" or "feathers".

When they are associated with temperatures near freezing, the following atmospheric sometimes can herald SLD icing:

1. Visible rain is a clear indication that dropled size is well above the 40 microns certification limit. Visible rain in freezing conditions is a seriious threat. Turn an exterior light on periodically to look for rain, and listen for every sound of rain impact on the airframe.
2. Droplets that splash or splatter on impact with the windshield are well above 40 microns. Droplets of the size covered by icing certification limits are too small to be seen, let alone splash.
3. Droplets or rivulets that stream across the airframe or airfoils probably suggest a high liquid water content that can produce rapid ice growth.
4. Any weather radar returns of precipitation suggest droplets sizes well above those associated with icing certification and, thus, likely to produce ice outside of the protected airfoil areas.

What to do

Icing prevention begins at the flight-planning table. Know the forecast situation and avoid those areas and/or altitudes that are predicted to produce icing conditions.

Monitor outside temperature. You are unlikely to encounter SLD icing unless the OAT is within the range of approximatly 0°C down to –18°C. Remember, a great percentage of icing-related accidents and incidents occur at temperatures close to freezing. Hand fly the airplane in any known icing conditions because the autopilot can mask those important indications of severe ice accumulation. If you do encounter a roll control anomally:

• Use whatever human force is necessary to overpower aileron forces in order to regain and maintain level flight.

• Reduce the angle of attack by lowering the nose and increasing power. Recovery by a stall occasioned by heavy airfoil icing may require a significant reduction in angle of attack in order to allow the airflow to reattach. When they are confronted by a stall from an ice-contaminated wings, higher-performance airplane pilots who have been taught to recover their airplane from a stall with minimal pitch change and maximum power may need to revert, if altitude allows, to basic light airplane stall-recovery techniques.

• Consider extending the flaps to their first increment if below flaps-extended speed. But do not retract the flaps if they are extended, since that would increase the angle of attack for any given airspeed.

• Use basic airmanship to maintain airspeed and altitude, in that order, assuming that you have enough altitude. Due to the ice accumulation, you may need to maintain an indicated airspeed well above that associated with a normal stall. Also, you may have to accept a controlled descent, but that is better than a stall or loss of control.

• Be sure to advise ATC if you are unable to maintain altitude, and do not hesitate to declare an emergency, if appropriate.

The Aviation Community is just beginning to understand the causes and effects of roll upsets in severe inflight icing. By its very nature, severe SLD icing can occur even with all ice-protection systems operating normally. Still, if you avoid icing conditions in general and diligently watch for any signs of SLD icing, an episode is unlikely to occur. If you should encounter any rolling anomaly, fly the airplane using whatever forces are required, use whatever combination of power and pitch is necessary to avoid a stall, and leave the area/altitude immediately.

In some cases a mere few degrees of temperature change may be all that is necessary to melt the offending ice.

Cloud drop sizes

When clouds first form, the cloud drop sizes are typically from 10 to 20 microns in diameter. Twenty–five microns is one thousandth of an inch, which is also about the diameter of a human hair. Meteoreologists have defined drizzle drops to be from 50 to 500 microns in diameter. Five hundred microns is 0.5 millimeters or 0.020 inch, and it is the diameter of those new fine-line mechanical pencils with the lead that always breaks. To get some idea of the size of 0.020 inch, ask your mechanic friend to show you a feeler gauge of 0.020 inches. Rain drops are 500 to 5000 microns, or 0.5 to 5 millimetres in diameter. An Eversharp pencil has lead that is one millimeter in diameter. When raindrops get to be larger than five millimeters, they break up from aerodynamic forces. Thus, cloud drop sizes vary over three orders of magnitude (from 5 to 5000 microns).

The mass of 5000 micron drop exceeds the mass of a five-micron drop by nine orders of magnitude, or one billion times. The concentration of natural cloud drops also varies by many orders of magnitude. For example, the concentration of cloud droplets varies from 100 to 1000 per cubic centimeter, while the concentration of rain drops is about one per liter.