Public Service

From the Ray Prouty Archives: Icing and Deicing

By Ray Prouty | December 21, 2017

In many technical fields, the helicopter lags the airplane by 20 to 30 years. This isn’t for want of trying but because, both in configuration and self-made environment, rotary-wing aircraft are much more complex than their fixed-wing cousins.

Aircraft deicing is certainly one of these lagging fields. Whereas many airplanes can effectively detect and remove ice from critical components, very few helicopters (outside of the Soviet Union) are so equipped. The reason is because icing conditions are encountered so rarely in normal helicopter operation that it is hard to justify the weight and expense of systems that on any one helicopter may never be needed. (Note: In Europe, icing conditions appear to be more prevalent than in the U.S.)

On the other hand, during those times when icing is a factor, the presence of ice protection may save the aircraft and all aboard — or at least permit a flight to be competed that would otherwise have to be aborted. The philosophy here, of course, is similar to that which dictates carrying fire-extinguishing systems around.


Flight limitations

A review of helicopter flight manuals gives a pretty consistent picture: “Flight during icing conditions is prohibited” (Hughes 500D); “Flight into known icing conditions is prohibited” (Boeing Vertol CH-46E); “Continuous flight in light icing not recommended” (Bell Helicopter AH-1G); “The helicopter is restricted from flying in moderate or heavy icing conditions. Flight in light icing conditions is limited to 30 minutes duration, due to the probability of damage from shedding ice” (Sikorsky CH-53E).

The severity of icing depends both on temperature and the “liquid water content” of the air — that is, on how many grams of water there are in each cubic meter. In practice, the airplane people define severity by the distance traveled while picking up half an inch of ice on a small probe. If 40 miles are required, the icing is “light”. If only 10 miles are needed, then it is “heavy icing”.

RWP Icing Deicing 1Ice formation

Even at below-freezing temperatures, it is possible for water droplets in the form of fog, drizzle or ram to remain in the liquid state — as long as there is nothing for the first ice crystal to form on, such as a speck of dust. When an aircraft flies by, it provides the required foreign body and the “super-cooled” droplets will freeze to any surface on which they impinge.

Naturally, it is the forward-facing surfaces that receive the brunt of the impingement, especially on the very leading edge of a component at the “stagnation point,” where one streamline of air is brought to a complete stop. Further back, the airflow is parallel to the surface and so diverts most of the droplets. Small objects are more efficient in picking up ice than larger ones.

Figure 74-1 shows the results of exposing three cylinders to the same icing conditions. The reason for the significant difference is that the larger the cylinder, the more warning it projects ahead to distort the approach flow field and to deflect the oncoming droplets to either side. Since blade leading edges are similar to small cylinders, they are efficient catchers and the smaller a blade — stabilizer leading edge, or inlet lips — the more sensitive it will be to ice.

Engine protection

One of the most serious potential problems concerns the engine installation. A turbine engine may be prone to compressor stall if large distortions of the air coming into it produce regions of very high local angles of attack on some of the blades. If the separated-flow region becomes too large, the airflow through the engine decreases to the point where the fuel-air ratio in the combustion chamber is too rich to burn and the engine flames out.

Designers work hard to shape the engine inlets to obtain the most uniform flow in all flight conditions but the accretion of a blab of ice on an inlet lip or on a screen may completely undo their efforts. In some installations, the inlet duct has bends or a plenum chamber where wet snow can be trapped.

And, as if these possibilities weren’t bad enough, if that chunk of ice or wad of frozen snow should come loose before it has become large enough to cause a flameout, it may still cause physical damage to the compressor blades as the engine tries to gulp it down.

Part of this latter problem has to do with the nature of ice, which can be either brittle or tough, depending on the temperature. For example, ice cubes right out of the freezer are brittle, but they toughen up as their temperature rises to the melting point. This phenomenon once caused some red faces at a major turbine-engine manufacturer. A press tour of the plant was planned and the test-cell people had successfully practiced for their part by throwing handfuls of fresh ice cubes into the intake of their new engine. As usually happens, the tour group was delayed and the bowl of ice cubes sat at room temperature for some time. They say the result was an unexpected exploded view of the engine.

To avoid such problems, one current technology for engine-inlet protection involves burying electric heating blankets in the skin of the inlet leading edges and other critical locations to prevent ice from forming. Alternate methods include ducting hot compressor bleed air between double skins to keep the surfaces warm. These systems are modeled on those used in airplanes and, as such, are not so much of a challenge to the designer as that required to protect the rotor blades.

Carburetors of reciprocating engines can ice up even in clear air if the humidity is high enough. The pressure drop as the air enters the intake may lower the temperature enough to first cause the moisture to condense and then freeze. Under conditions of very high humidity, carburetor icing can occur at outside temperatures as high as 30 C (100 F). To prevent this, the pilot is given a carburetor heat control that ducts warm air from around the exhaust manifold to the inlet. Turbine inlets can experience similar clear-air icing but because the pressure drop is less, the critical temperature is below about 8 C (46 F).

The problem with rotor ice

Just as in the engine inlet, ice on the main and tail rotor blades can have bad effects while building up and possibly catastrophic effects when it breaks loose. Since the blades are going faster than any other part of the helicopter, they encounter more droplets per second and, because of their collection efficiency, accrete ice at a faster rate. On the other hand, compression of the air at the stagnation point on the leading edge of the blade raises its temperature — as much as 200 C (40 F) at the very tip. For this reason, tips often remain free of ice while it builds up on inboard sections.

RWP Icing Deicing 2Ice degrades the aerodynamic characteristics of the blade airfoil by forming irregular shapes that increase drag and decrease maximum-lift capabilities. Figure 74-2 shows some of the odd lumps and bumps that ice has been observed to construct on airfoils during tests in icing and wind tunnels.

The power increase due to the higher drag may give an alert pilot the first indication of ice. The degraded aerodynamic characteristics will also jeopardize the ability to autorotate in case of an engine flameout. Autorotation with ice on the blades puts the helicopter into a multiple bind:

Bind 1: Because the power required is higher, the autorotative rate of descent must be higher to produce that power from the loss of potential energy. This gives the pilot less time to get the engine restart if that is in the cards.

Bind 2: Because the rate of descent is higher, the collective pitch required to maintain normal rotor speed is lower. If the designers did not foresee this possibility, the collective downstop may be too high to get the rotor speed up to a desired level.

Bind 3: Because the maximum-lift capability is lower, the safe minimum rotor speed will be higher. With lots of ice, the minimum safe rpm may actually be higher than the rpm on the downstop.

Bind 4: Because the drag is higher, the ability to build up excess rotor speed in a cyclic flare will be lower.

Bind 5: Because the maximum lift is lower, the load factor that can be generated in the final collective pull will be lower.

Blade ice will eventually be shed because of centrifugal forces, airloads, blade flexing or fling into warmer air. When it does, it often goes from one blade at a time, or “asymmetrically” so that the rotor goes out of balance and out of track. This can produce severe vibration in the cockpit and damaging cyclic loads in the main and/or tail rotor support structures. At the same time, damage to the tail rotor may result from ice thrown off the main rotor or vice versa.

Blade protection

I was once told that the early Russian Mil helicopters had blades constructed from metal spars and wooden ribs like the early Sikorskys — but with sealskin coverings instead of doped aircraft fabric. The sealskin was kept pliable and ice-free by periodically swabbing with walrus oil.

Perhaps because of the current shortage of walrus oil, almost all rotor deicing systems — including those on recent Russian helicopters — now use electrical heating blankets. These are built into the leading edges with heat supplied through slip rings at the rotor hub from a generator mounted on the transmission. The blankets are divided into zones and each is periodically heated. This melts the bond between the skin and the ice that has accumulated since the last heating cycle, allowing centrifugal forces to fling the ice off. Corresponding zones of every blade are heated at the same time to minimize possible unbalance. The interval between heatings of any one blade segment is determined by the icing severity measured with an ice detector, sometimes mounted in an engine inlet to ensure a flow over it even in hover. It is generally accepted that ice can be allowed to build up to about a quarter of an inch halfway out on today’s blades before it becomes dangerous. By popping the ice off quickly from the leading edge, there will be little water to run back and freeze on the trailing edge. This deicing system takes much less power than would an anti-icing system that would try to continuously keep the whole blade above freezing. Of course, on some experimental pressure-jet helicopters, the effects on an anti-icing system come free, as surface temperature of the blades can run at several hundred degrees.

Proof of the pudding

A big problem in the development of any ice-protection system is proving that it works after it is installed. Sending the test aircraft out to look for natural icing conditions of various kinds and severities could take years. For this reason, two artificial icing flight-test facilities are no in operation.

One is operated each winter by the Canadians in Ottawa and costs of a tower that sprays a cloud of water from calibrated nozzles to give the desired droplet size and cloud density. A helicopter hovering downwind can stay in the cloud until it demonstrates that its system is working or until a dangerous amount of ice builds up. The procedure is relatively safe since a landing can be made quickly. It also gives the engineers a chance to observe the ice formations lose up before they melt. The limitation of the tower is that it does not simulate forward flight.

That limitation is not a factor with the other facility. A U.S. Army Chinook called the “Helicopter Icing Spray System” (HISS) emits a cloud from a 60-foot-wide spray bar.

Looking for Mr. Goodsystem

No one likes the cost and complexity of an electrical deicing system, nor the weight of its electrical generator. Therefore high motivation persists to find other methods of ice protection and a large number have been tried in the past. One is to try breaking the ice up with blade bending induced by sudden cyclic, collective or rotor-speed changes. This seemed to work in hover on the hingeless Lockheed Cheyenne attack helicopter at the Ottawa spray rig but has been less than successful on other rotor designs. In forward flight, this method has often led to severe asymmetrical shedding and resultant high vibration.

Another approach that was once used on propellers is to allow antifreeze fluid to run out the blades through holes in the surface from a distribution system at the hub. Engineers have tried it but were soured on the idea for two reasons: There is difficulty in getting uniform distribution all the way out, and all those holes affect blade fatigue life.

“If we could only coat the blade with something that ice wouldn’t stick to, we would be in like Flynn.” That has been the motivation behind many experiments. Except perhaps walrus oil, no one has yet to come up with that magic paint, paste or tape that would be “ice phobic” yet could still stand up to the erosive effects of dust, sand and rain. There is a fortune waiting for the person who does.

Receive the latest rotorcraft news right to your inbox