Vous êtes sur la page 1sur 10

AIRCRAFT ICING

Sker de Salis, Roger Dunn and Ole Henriksen present the theory and reality of aircraft icing.

First off, an article based on Sker de Salis' lecture given at a PPL/IR Europe meeting Basics of cloud formation and associated icing types
The production of a cloud requires Saturation (humidity at or near 100%; temperature and dew point being equal) and the presence of Aerosols, e.g. solid particles of natural or artificial origin (pollens, spores, salt particles from evaporating sea spray, combustion processes in industry, traffic, etc.). Aerosols act as Condensation Nuclei when the air is saturated with humidity, thus producing Water Droplets which make the cloud. Depending on other factors, there are different types of clouds. As far as the basics of icing are concerned, two types of cloud are of interest: Convective clouds (cumulus, etc.) Layered clouds (stratus, etc.)

Observation has shown that water droplets exist in clouds at temperatures as far below freezing as - 40 C. Fig. 1 shows a typical distribution of Liquid Water Content (LWC) in a convective cloud against height and temperature.

Fig. 1

The average size of the water droplets depends on the type of cloud: larger in convective clouds, smaller in layered clouds. When hitting an airframe, the larger droplets of the convective cloud produce Clear Ice. Clear ice is strong and adheres well to the airframe. Under certain circumstances, it can "run back" along the wings, changing its profile.

The smaller droplets found in layered clouds produce Rime Ice. Rime ice is white and brittle; it does not adhere to the airframe as strongly as clear ice. I have never seen rime ice "running back". A mixture of clear and rime ice is also possible: this is Mixed Ice. Depending on water content, air temperature and temperature of the airframe, the icing will be light, moderate or severe. The definitions from the American Aeronautical Information Manual (AIM) are: Light: The rate of accumulation may create a problem if flight is prolonged in this environment (over 1 hour). Occasional use of deicing/anti-icing equipment removes/prevents accumulation. It does not present a problem if the deicing/anti-icing equipment is used. Moderate: The rate of accumulation is such that even short encounters become potentially hazardous and use of deicing/anti-icing equipment is necessary. Severe: The rate of accumulation is such that deicing/anti-icing equipment fails to reduce or control the hazard. Immediate flight diversion is necessary.

The debilitating effect of mixed ice on an unprotected leading edge

As the aircraft climbs, its airframe will gradually cool down as the air becomes colder. But this cooling will usually be slower than the air's temperature, because the heat that is stored in the aircraft's structure must be dissipated: the onset of icing will be delayed for some time. Of course, if the aircraft is already at sub-zero temperature on the ground, this delaying effect will not happen. In descent, if the aircraft did fly long enough in sub-zero temperatures, the reverse will apply: the icing might continue for some time, even though the air's temperature is positive. The pilot must bear that in mind especially when preparing to land: icing residues can increase the stall speed or cause control problems. tailwing icing is particularly critical: a tailwing stall due to ice might cause a pitch-down moment - not a very pleasant experience just before landing! When the aircraft ices up, it slows down: to maintain altitude, a greater angle of attack is needed. This in turn increases the surface that is exposed to icing. The effect of leading edge icing increases substantially with the angle of attack. See Fig. 2.

Interaction between airframe and atmosphere

Fig. 2

In flight, the pilot has only a very limited view of where the ice accumulates on the airframe. The best clue to the severity of the phenomenon is the Airspeed indicator. If you know your usual values, you'll immediately find out how much speed you are losing. The rate of speed decay is important: if it is high, then you are in trouble!

Special problems
Icing on the ground
This happens when the aircraft stayed overnight in sub-zero temperature, and there is freezing fog, or snow, or freezing rain. It is then imperative that the airframe be thoroughly deiced very shortly before take-off. There is a well publicized airliner accident where there was a long delay between deicing and take-off: the B-737 was unable to gain height and crashed shortly after take-off (Washington, 13.1.1982). Another problem is slush on the runway at temperatures near zero: during the take-off roll, slush may be projected in the wheel wells, flap hinges, etc. It can then freeze in flight, and cause various malfunctions of essential components.

Freezing rain/drizzle
This is a very dangerous situation that, luckily, does not happen frequently. It is caused by a warm air mass that comes over a cold air mass and precipitates. The rain falling in the cold air mass ices as clear ice. The cold air mass can very well be cloud free, so this type of icing is also of concern to VFR flights. When this kind of icing occurs, it is almost always of the Severe type. Freezing rain is really a No-go item.

Wing upper side icing at positive air temperatures


This happened to me once. Two factors must be considered:

1) True Air Temperature versus Indicated Air Temperature Due to air friction, Indicated Air Temperature is higher than True Air Temperature. This effect becomes noticeable at true air speeds above 140 knots - about 2 C for a true air speed of 160 knots. So when you have an OAT of + 3 C, you are already almost at the freezing level. 2) The cooling effect of under pressure on the wing's upper side As you know, lift is generated mainly by an under pressure on the wing's upper side. See Fig. 3. This under pressure also causes a slight cooling of the air. This is a consequence of the relationship between Pressure, Volume and Temperature for air (or other gaseous materials): P x V = T x Constant We can assume air to be incompressible at the speeds considered, so the Volume remains constant, and a decrease in Pressure causes a decrease in Temperature, which I have computed: the Temperature drop is about 2 to 3 C for a typical wing loading of 17 lbs/sqft.

Fig. 3

The combined effect of the two factors means that wing upper side icing is possible when your OAT indicator shows a positive temperature of up to about +4 C. In the icing encounter I experienced, I was able to see water steadily condensing on the leading edge of the wing, not icing there. It then ran back on the upperside where it started to freeze. This was in a stratus cloud, at Fl 110, with an Indicated Air Temperature of about +4 C. There was no turbulence. The effect on performance was quite dramatic. As soon as I noticed what was going on, I activated the TKS deicing system and requested a higher level, where I got out of clouds. The TKS system then needed several minutes to clear the ice completely.

Other things ice can do to your aircraft


Things one does not think of in the first place! Antennas: ice accumulating on antennas may get them to vibrate and break: loss of communication and/or navigation information.

Pitot tube: if the heater has not been turned on, or is defective, the airspeed indicator will work like an altimeter: as the aircraft climbs, so does the indicated airspeed! Windshield: if it is iced up, you won't see the runway! Fuel vents: ice may block them, causing fuel starvation and loss of engine power. Control surfaces: ice may bind or jam them, you'll lose control authority.

And, finally, a question:


What is the rate of weight increase when ice accumulates on the wings of a light aircraft? Assuming: An average wing thickness of 8 inches A wingspan of 33 feet A True Air Speed of 150 Knots And a LWC of 0.5 g/m3

Then, the wings will collect ice at a rate of about one pound per minute.

Why does the tail get more ice than the wing?
Experienced pilots know that smaller objects - temperature probes, pitot tubes, antennas, etc. - get more ice than larger ones. Ole Henriksen and I have had a longer correspondence about whether the tail would get more ice than the wing. Ole was very certain that the tail would get more ice than the wing, whilst I had trouble understanding why - assuming tail and wing would have exactly the same profile. So I did some research and a bit of thinking, and I believe I have found out why Ole is right. I would like to share this with members. Looking into Professor Ludwig Prandtl's book "Strmungslehre", first published in 1931, I found this interesting Wind Tunnel photograph (Figure 4). As can be seen, flow deflection starts ahead of the wing's leading edge. This is because, at subsonic speeds, the pressure wave generated by the movement of the wing through the air travels faster than the wing itself. I then asked myself: what happens to a supercooled water droplet when it arrives at the point where flow deflection begins? Obviously, it is going to be deflected vertically but, since it has its own mass, it won't be able to follow exactly the path of the flow lines. In Figure 5, I sketched this for a wing, and for a tail with exactly the same profile as the wing: the tail is scaled down from the wing by a factor 0.7.

Fig. 4

Fig. 4 Fig. 5

Whilst total similarity would require the Reynolds numbers to be the same in both cases, which they are not, I believe that the flow geometry at least will be similar for both cases, because the Reynolds numbers are high (that is well above the value for which a laminar flow can begin to turn turbulent), and not very different in both cases. As can be seen, the onset of flow deflection for the wing is at a distance D1 from the leading edge, and at a distance D2 for the tail: whilst the proportions are maintained, the absolute distance is shorter for the tail than it is for the wing. As wing and tail travel in the same air and at the same speed, the supercooled water droplet which is shown by a black dot on both sketches at the onset of flow deflection has less time to be deflected in the case of the tail than in the case of the wing. Of course, the size and mass of the water droplet must be assumed to be the same for wing and tail; it is further assumed that the tail's airflow is not disturbed by the wing's airflow, which is true for most airplanes in level flight. I have shown the path of the water droplet as a dotted line which just avoids the wing. The same path collides with the leading edge of the tail, because the onset of flow deflection happens later and nearer to the leading edge in this case. The consequence is that more water droplets are likely to hit the tail than the wing, so the tail will get more ice than the wing.
References: - AOPA: Aircraft Icing, March 1998 - http://www.aopa.org/asf/publications/sa11.html - Barthazy, Eszter, lab. for Atmospheric Physics, Swiss Federal Institute of Technology, Zrich: correspondence, Dec. 97 / Jan. 98 - Eichenberger, Dr. Willy: Flugwetterkunde, 1977 - Kohli, Jean-Pierre: Connaissance des avions, 1975 - Schmidt, Prof. Dr. Ernst: Thermodynamik, 1960 - Trunov, Dr. O.K., State Research Institute for Civil Aviation, Russia: Aircraft Icing must never be underestimated, June 1985

Following up Sker's article, our Deputy Chairman Roger Dunn, who has a degree in aeronautics, comments as follows:
Thank you for the interesting dissertation from Sker. I agree completely with his conclusions and broadly with his analysis which leads him to them. However there is one missing link which I believe could provide the final proof. Sker shows the droplets following the streamlines of the airflow. If the droplet was a molecule of gas this would be a reasonable assumption due to the effect of the neighbouring molecules. However in this case we are looking at a droplet of supercooled liquid which has a significant mass and is therefore subject to the basic laws of mechanics. For the droplet to move clear of the approaching airfoil, whether wing or tail plane, it is necessary for the droplet to acquire a vertical velocity. A vertical velocity will require a vertical acceleration which in turn requires a vertical force to be applied to it. The only source of such a force in the vertical plane is drag from the vertical movement of the droplet relative to the airflow. What this means is that the path of the droplet will be somewhere between the path of the streamline and the horizontal. The magnitude of the vertical velocity at any point ahead of the airfoil will depend on the horizontal velocity and the angle the streamline makes to the horizontal. Based on Sker's aerodynamic assumptions the vertical velocities will be identical for the large and small airfoils. However a droplet initially slightly displaced vertically from the centreline will be subject to the vertical velocities for a proportionately longer period

when approaching a large airfoil than a small one. One of the basic equations of motion is: S =1/2*a*t^2 Where S = distance, a = acceleration, t = elapsed time. It will be seen from this equation that if time is doubled and the acceleration remains constant the vertical distance travelled is quadrupled. If the assumption of constant acceleration was valid, a small airfoil would pick up the same quantity of ice as a large airfoil. Since the airfoil is smaller the thickness of ice would be greater. The assumption of constant acceleration is not quite accurate for the following reasons 1. The streamlines follow curved paths ahead of the airfoil and not straight lines. 2. As the droplet accelerates vertically the difference in vertical speed between the streamlines and the droplet will fall and the acceleration will reduce. The second effect will be much more pronounced for small droplets whose drag to mass ratio will be much higher than that of large droplets. The above would suggest that due to the droplet effect a heavy build up on the tail plane relative to the main plane is more likely to occur in cumulus cloud with large droplets than in stratus cloud where the droplets are smaller. Do we have a volunteer to test this theory?

Finally, Ole Henriksen joins the icing discussion with some personal observations
Apart from basic training, most of my flying has been in deiced aircraft, so my perspective on icing is a little different from the majority of PPLs. My overall experience from more than 3,000 hours of flying without actively avoiding forecast icing (except if severe) is that there is generally a lot less ice out there than we are led to believe - but there is no way to forecast exactly where it occurs, or how severe it is. I operated a Partenavia for a couple of years and had several icing encounters. My experience from those years of flying below 10,000 feet, and therefore spending a lot of time in potential icing is that some degree of icing is likely to occur whenever you fly in visible moisture in freezing temperatures, or even just above freezing; but particularly in the range 0 to -10C. I chose a Twin Commander turboprop, so I now fly faster and usually cruise above the weather. Both are important Clean boots but more than an inch of mixed ice built up factors in avoiding ice, and icing is therefore now mainly a on unprotected surfaces during just 10 minutes in moderate icing on a night approach in January factor during climb and descent. But I still meet it, and when I do, it is not always forecast, and it is very often more than I would be comfortable with on an aircraft not equipped and approved for flight into known icing.

Planning for ice


My personal recommendation is never to fly in potential icing conditions in any aircraft that is not approved for it, and to use approved de-icing and anti-icing equipment only as an escape route, i.e. avoid cruising or holding for any extended period in actual icing conditions because no ice system can cope with severe icing, and there is no reliable way to forecast where or when that might occur. However, since this would prevent many IFR flights in much of today's GA fleet for much of the year, a lot of pilots apply certain tactics to minimise the icing risks. But this requires very careful planning, because such tactics can expose you to other risks. For example, having sufficient clear, warm air below as an escape route sounds as safe as you can get. The caveat is making absolutely certain that you do have clear, warm air below for the entire route - this is not easy to ensure in the types of weather which are most likely to produce ice. Furthermore, descending may compromise safety in other ways; for example when flying over water in a single. The crucial things to remember about icing is that it: Increases drag Increases weight Decreases lift Decreases thrust

It all adds up to a very serious loss of performance. As performance is eroded, speed is often reduced to maintain altitude, and this exacerbates the ice buildup which is stronger the slower you fly and the higher the angle of attack. As a result, performance can quickly reduce to very little, or disappear altogether; particularly on lower powered aircraft. When it does, there is no way but down. To add to your troubles, ice will alter the shape of your wing, and the only thing we know for certain is that this causes your stall speed to increase. By how much is anybody's guess, but you definitely want to keep your speed well above normal stall, and not rely on the stall warner which will be misleading even if it is not iced over yet. Legalities aside, I feel all this makes "planned escapes" from icing in non-deiced aircraft a highly risky business. Not quite Russian roulette - odds are certainly better than 1 in 6, but they're still too short for my taste. However, I am well aware that many disagree and apply "ice tactics" successfully. AOPA (US) have even taken the bold step to publish such techniques, probably on the principle that since pilots do it anyway, at least they will be safer if they are well informed.

De-icing and anti-icing equipment


As for the use of de-icing and anti-icing equipment, it is important that they are used correctly. There have been accidents due to incorrect use of otherwise serviceable ice equipment. Since formal training in icing for most PPLs is sadly limited to the "check ice no ice" charade in IR training, our best source of information is our aircraft manual. Study it carefully, understand it and then follow it.

Pneumatic de-ice boots before and during inflation

Some may be aware of a recent FAA ruling on the use of pneumatic de-ice boots following a couple of high-profile commuter accidents in the U.S. The received wisdom (generally supported by flight manuals) is that you should allow 1/4"-1/2" of ice to build before activating the boots, because early activation could push a thin sheath of ice forward without breaking it off. More ice would build on top of that, and subsequent boot activation would be ineffective as the boots would just move within the envelope they created when the ice was first pushed forward. This so-called ice bridging is now being questioned, in fact rubbished, by the FAA who claim that it has no scientific base, but is based on folklore and Gann's novels, in particular Fate is the Hunter. I am taking the brave position to strongly disagree with the FAA, for I have seen ice bridging on several occasions. I recently accumulated a thin layer of mixed ice in climb (1/16" to 1/8" / 1.5 to 3.0 mm), and once I was on top in clear air, I decided to conduct a little experiment. I cycled the boots, and the result was exactly as described above: the ice was pushed forward, and subsequent boot cycles did nothing further. Some ice subsequently broke away in the airflow, but most of it took some time disappearing through sublimation. The boots were completely useless as long at this ice sheath was present, and had I been in icing conditions, I would have lost my deicing ability. This is probably what the FAA derides as "folklore", but I will nevertheless continue to follow my flight manual (and my own experience) and activate the boots when around 1/4" (6mm) ice has accumulated. It then breaks off completely in a few cycles and leaves a clean leading edge - but it is important to keep the boots in good condition. I apply Icex regularly to keep the surface smooth; if the boots are allowed to go matt, the ice will stick and become more difficult to break off. Anti-icing equipment also has to be used with care. Electric systems often have OAT limitations, and some must not be used on the ground because they would overheat without an airflow to cool them. Remember that electric prop boots are an anti-ice device and should be switched as soon as you enter potential icing. If you leave activation until you have ice buildup, the cold-soaked propeller will delay the effect of the heating elements, and you might get "behind the icing curve", effectively unable to shed ice as it accumulates. The TKS anti-ice system has its own strengths and weaknesses, but I have never used it and will leave its discussion to others who have.

Pitot tube troubles

If the pitot heat is not applied, or it fails or proves to be of inadequate wattage as has happened to me, the pitot tube may easily ice over and block. The dynamic pressure at the time of blockage gets trapped, but the the static pressure continues to change as you climb or descend. The result will be a rising airspeed indication in the climb and a falling airspeed indication in the Effects of faulty or forgotten pitot heat descent. The first is especially dangerous. It has caused, and continues to cause, accidents, even among airliners. The gut reaction is to pull back on the controls to check the rising airspeed. This causes an even greater rate of climb, and an even greater rise in indicated speed. The real speed, of course, decays rapidly with the steep climb, and if not checked, this condition

rapidly leads to a stall. And don't forget: you're already partially disorientated and in IMC... the outcome is sadly predictable. The correct technique faced with a runaway airspeed indication is to fly power and attitude. As long as you have set normal climb power and attitude, there is no way your trusty Cessna will do 200 knots - unless your AI has failed, and the nose is actually pointed down (or your Cessna happens to be a Citation). This is a good time to do some crosschecks (which, of course, are all part of your instrument scan anyway): if the VSI and altimeter show a climb, then you can trust your AI, and you are climbing. The only reason for a rising airspeed, therefore, is a faulty ASI. It doesn't matter so much if you can work out on the spot exactly what the trouble is. The important part is to realize that your AI is telling the truth, and then fly normal power and attitude - that is bound to give you something very close to normal airspeed. If the pitot blockage happens in the cruise, the indicated airspeed drops when you start your descent. This situation may lead the pilot to push the nose down and eventually overstress the airframe through overspeeding in what may well be turbulent conditions. While not developing as rapidly as the stall, this is still a serious situation which should be avoided at all cost, and if you fly normal power and attitude, chances are that your airspeed will not get out of hand.

Summary

My icing experience is that: some degree of icing is likely to occur whenever flying in visible moisture at or below freezing and this leads me to the following two icing policies: never to fly in potential icing conditions without approved icing equipment and: use even approved icing equipment only as an escape route.

Other Reference Reading

Aircraft Icing: A Pilot's Guide by Terry T. Lankford: From two-seat single-engine aircraft to jumbo 777s, icing is a silent creeper and one of the industry's hottest concerns. Covers all aspects of aircraft icing-ground, airframe, and induction. Provides solid foundation of icing theory and physics without superfluous meteorological concepts. Teaches pilots how to interpret icing forecasts and reports and integrate specific aircraft and engine considerations. Discusses practical procedures to deal with icing conditions throughout flight, from taxi to touchdown. Publisher: McGraw-Hill ISBN: 0071341390

In-Flight Icing by Porter Perkins and William Rieke: This book is the result of over 50 years of combined icing research experience by authors Porter J. Perkins and William Rieke at the NASA Glenn Research Center. Covers meteorological and operational factors including the probability of icing encounters, ice shapes and types, tailplane stalls, recovery techniques and limitations of iceprotection systems. Illustrated charts and black and white photos.

10

Vous aimerez peut-être aussi