Icing: Accretion, Detection, Protection

LS- )

4-I

ICING: ACCRETION, DETECTION, PROTECTION

John J. Relnmann* National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 1. INTRODUCTION freezing precipitation covers all the upper surfaces of the Icing conditions present an adverse environment to aircraft. aircraft. Ice on grounded aircraft can take several forms: In flight, the most common icing hazards are clouds contain- frost, wet snow that freezes, freezing rain that turns into clear ing supercooled water droplets, clouds with mixtures of ice, slush that freezes, or moisture from humid air that supercooled droplets and ice crystals, and freezing rain. On condenses on cold-soaked wings and freezes to clear ice.

the ground, icing hazards include freezing rain, freezing drizzle, freezing fog, falling or blowing snow, frost, slush, and Five methods are used to protect against ice humid air.

1. Keep water wet—apply heat continuously The adverse effects of icing on aircraft operations are described below. Ice contamination on wings and tails 2. Evaporate water—apply more heat continuously reduces maximum lift coefficient and stall angle of attack, and increases stall speed and drag. During takeoff, ice on wings 3. Melt ice—apply heat intermittently has caused wing stall and serious stability and control problems with nearly every kind and size of aircraft, resulting 4. Mechanically remove ice—crack, debond, and expel the in pitch up, rolloff, and crash. During approach or landing, ice with pneumatic deicers or impulse deicers the combination of extended wing flaps and ice on the horizontal tail has caused tailplane stall, resulting in 5. Chemically prevent ice or melt ice—apply freezing point uncommanded pitchovers with some aircraft. Ice contamina- depressant fluids tion on propulsion system components—such as air intakes, engine nacelles, inlet ducts, propellers, fan blades, spinners, In that ice protection systems have been used successfully inlet guidevanes, and helicopter rotorbiades—reduces since the 1940's, it would seem reasonable to expect that all propulsive efficiency and adds to aircraft drag. For smaller the icing technology problems have been solved by now.

fixed-wing aircraft, the combination of reduced lift, increased However, new problems continually arise, because the use of drag, and reduced propulsion efficiency can result in the loss new technologies in modem aircraft have a ripple effect on ice of the capability for level flight, and the aircraft will execute protection. Also intense global economic competition forces an uncommanded descent. For helicopters, increased rotor airplane manufacturers to optimize overall airplane perfor- drag caused by ice can result in required torque exceeding mance while minimizing airplane capital costs, operating available engine torque, and the helicopter will execute an costs, and maintenance costs, and ice protection must be uncommanded descent. Iced rotors can also cause retreating included in this optimization process.

blade stall, resulting in an uncontrolled rolloff. Furthermore, iced rotors will cause more rapid descents during autorotation. The global aircraft industry and its regulatory agencies are Pieces of ice shed from wings, propellers, rotors, or engine currently involved in three major icing efforts: ground icing; nacelles can cause structural damage to the airframe or advanced technologies for in-flight icing; and tailplane icing.

engine, or cause engine flameout. Ice on aircraft instrumenta- These three major icing topics correspondingly support the tion can give wrong airspeed indications or wrong engine three major segments of any aircraft flight profile: takeoff; pressure ratios that lead to improper engine power settings. cruise and hold; and approach and land. In this lecture, we Ice accretion on antennas, struts, wheels, and external stores will address these three topics in the same sequence as they adds weight and drag to the aircraft. Ice can cause destructive appear in flight, starting with ground deicing, followed by vibration of parts such as antennas and wing struts. advanced technologies for in-flight ice protection, and ending with tailplane icing.

Anytime an aircraft flies through visible moisture at outside air temperatures below about 5 °C, there's a good chance that 2. GROUND OPERATIONS AND HAZARDS IN ice will form on the aircraft components. In-flight icing CONDITIONS CONDUCIVE TO ICING conditions normally occur from ground level up to 22,000 ft, Aircraft ice contamination caused by freezing precipitation but pilots have reported icing at altitudes as high as 40,000 to during ground operations poses a potential hazard for takeoff 50,000 ft. On aircraft in flight, ice forms on the leading edges and subsequent flight. Airplane manufacturers do not design of wings, tails, engine nacelles, spinners, etc. At temperatures their airplanes to take off with ice on critical surfaces, nor near freezing (0°C), in-flight ice is clear (glaze) and does the FAA/JAR certify them to take off under such horn-shaped, while at colder temperatures, it is white and conditions. Airplane manufacturers fully support the "Clean opaque (rime) and spear-shaped. On grounded aircraft, Aircraft Concept" and warn that it is imperative not to attempt takeoff unless the pilot is certain that all critical surfaces of the aircraft are clear of ice. The Federal Aviation Regulations *Distinguished Research Associate.

Presented at an .4GARD Lecture Series an 'Flight in an Adverse Environment', November 1994.

4-2 a pilot Figure 4 shows that even a small nondimensional roughness (FAR) Sections 121.629, 91.209, and 135.227 prohibit height of 5x10 4 , which is comparable to about a 0.2 mm from taking off with ice contamination. Section 121.629 roughness height on a small jet transport wing, can reduce Operations in Icing Conditions states: maximum lift of a hard wing by 35 percent and angle of attack for maximum lift (stall angle) by about 60 . Percent increase in No person may take off an aircraft when frost, snow, or ice is stall speed, the operationally more significant parameter, can adhering to the wings, control surfaces, or propellers of the be estimated as about half the percent loss in maximum lift aircraft.

coefficient. So for the present example, a 35 percent loss in maximum lift translates to an 18 percent increase in stall Deicing and anti-icing fluids are available to protect aircraft speed. For a normal takeoff run with this amount of contami- from ground icing. Deicing fluids remove ice from aircraft nation on the wings, V2 (takeoff safety speed) would be less but do not prevent refreezing. Anti-icing fluids prevent than stall speed, which means that the wings could not precipitation from freezing on the aircraft for a limited period generate enough lift to take off.

of time (holdover time). The FAA requires operators to develop and use an FAA-approved aircraft ground deicing Several times thus far, we have mentioned that ice contamina- program and specifies checks and inspections to ensure a tion on the lifting surfaces (i.e., wings and tails) reduces clean aircraft at takeoff.

maximum lift and stall AOA and increases stall speed and drag. Typically, a modern transport is required for certifica- 2.1 Effects of Roughness on Wing Aerodynamics tion to have about a 13 percent stall speed margin at takeoff, On takeoff, the predominant effect of ice contamination is on which is to say that its normal safe takeoff speed, V 21 is the lifting characteristics of the wing. Figure 1 shows 13 percent higher than its I g stall speed for the clean wing conventional plots of lift coefficient versus angle of attack takeoff configuration. Landing speeds are typically about (AOA) for a clean wing and a contaminated wing. The plots 23 percent higher than stall speed for the clean wing landing show that contamination reduces both the maximum lift configuration.

coefficient and the angle of attack for maximum lift (stall angle).

Although ice contamination increases form drag it does not (from Ref. 1, 2, and 3, respectively) appreciably affect drag on large transports. However, if the Figures 2, 3, and 4 angle of attack is high enough to stall the wing, the wing form provide a comprehensive and valuable collection of wind drag becomes appreciable and may double the aircraft drag.

tunnel and flight test data that quantifies the percent loss in On smaller aircraft, ice on exposed landing gears and wing maximum lift as a function of nondimensional roughness struts could contribute appreciably to airplane drag.

height, k/c. Percent loss in maximum lift is defined as: 2.2 Effects of Ground Dc/anti-icing fluids on Wing lOOx(maximum lift clean—maximum lift contaminated) Aerodynamics maximum lift clean When ground deicing or anti-icing fluids are present on the wing during the takeoff run, the fluid surface becomes unstable and develops a waviness which is, in effect, a form of These last three figures present somber evidence that roughness that contaminates the wing. So, even though these contamination causes a significant loss in aerodynamic fluids can protect against the large aeroperforniance losses performance for both slatted and unslatted wings. Unslatted caused by ice roughness, the fluids themselves could wings refer to wings without leading edge devices extended.

potentially cause performance penalties during takeoff. In the Slatted wings refer to wings with extended/deflected leading 1980's and early 1990's, the Boeing Airplanes Company edge devices such as a slat. (In this paper, we will denote demonstrated both in wind tunnel tests and in flight tests that unslatted wings as "hard" wings.) Brumby's correlation for these fluids do cause measurable losses in maximum lift the entire upper surface covered with roughness (Fig. 2) coefficients (Ref. 4). These results were confirmed by the von brackets the data from all three figures and can be considered Karman Fluid Dynamics Institute in Brussels, Belgium, under an upper limit on percent loss in maximum lift for hard wings a grant from the Association of European Airlines (AEA) and tails. On the last three figures, the data from Boeing, (Ref. 5).

Fokker, and McDonnell-Douglas for slatted wings at higher k/c values show much lower losses in maximum lift coeffi- Remarkably, these studies demonstrated that the loss in cient than does Brumby's correlation.

maximum lift coefficient correlated with the boundary layer displacement thickness measured at the trailing edge of the Figure 5 (Ref. 3) shows that loss in angle of attack to stall wing's fixed element. Professor Mario Carbonaro from the varies nearly linearly with nondimensional roughness height, von Kaman Institute used this correlation to develop a cost k/c. Figures 2 to 5 together contain enough information to get effective Aerodynamic Acceptance Test for the qualification a representative estimate of the effects of roughness on wing of Type I and 1)'pe II fluids. The test measures the growth in or tail aerodynamics. They should prove useful to those boundary layer displacement thickness at the trailing edge of a concerned about icing problems during both takeoff and flat plate covered with the fluid and located on the test section landing.

floor of a specially designed wind tunnel. A correlation exists that defines an acceptable upper limit on displacement The data presented in Fig. 2 to 5 include data taken at both thickness over a range of temperatures. Fluids that exceed the subscale and flight Reynolds numbers. Reference 3 presents upper limit fail the acceptance test and are rejected. (Keep in data that demonstrates that testing must be done at chord mind that the acceptable upper limit on displacement Reynolds numbers of 5x106 or higher to achieve percent loss thickness is directly correlated with the acceptable upper limit of maximum lift results representative of full-scale flight.

on loss in maximum lift.)

4-3 Fokker Aircraft engineers acquired data from wind tunnel Only two wind tunnel facilities are currently approved to tests of wing sections and airplane half models with roughness conduct this test: one is located in Europe at the von Kamm Institute and the other is in North America at the University of distributed uniformly over the entire wing upper surface and also from flight tests with simulated rime ice and sandpaper Quebec at Chicoutimi. References 4 and 5 discuss the Boeing and von Kaman Institute aerodynamic studies on ground roughness on the leading edge of the wing. They made use of this data in an engineering flight simulator of the Fokker 100 deicing fluids. And Ref. 6 describes the aerodynamic (Ref. 2). Figure 6 shows the lift versus AOA curves for the acceptance test and its rationale, which was approved by the Aerospace Industries of America (AlA) and the Association clean and contaminated wing, along with stick shaker AOA and roll control boundaries. When the clean aircraft is rotated Europeenne des Constructeurs de Materiel Aerospatial 3° per second the peak AOA was approximately 10.5 0. The (AECMA).

clean aircraft would still have about a 2.5 0 margin before stick The acceptable upper limit on loss in maximum lift derives shaker activation and a 5.5° margin to stall AOA. The aircraft with contaminated wings will stall at about 9°, or about 1.5° from the criterion for stall speed margin at takeoff safety speed, V2 The airplane manufacturers and the regulatory below the aircraft's target angle of attack.

agencies recognized that the fluid imposed a transitory loss in maximum lift, since the fluid completely flows off the wing Van Hengst (Ref. 2) pointed out that the clean airplane gives a shortly after rotation (except for a very thin residual film that slow progression of wing flow separation that starts inboard remains for much longer). Although authorities require a and moves toward the wing tips as AOA is increased, thereby 13 percent stall speed margin for clean wings, they accepted a ensuring exceptionally good roll control throughout a stall test maneuver. The manner in which a contaminated wing will ten percent stall speed margin for fluid-covered wings because the fluid effect is transitory. To meet the minimum 10 percent stall is unpredictable and, therefore, extremely dangerous because the inherent good stalling characteristics of the clean stall speed margin, the percent loss in maximum lift coeffi- cient must not exceed 5.24, which establishes the acceptance wing are lost. Unequally distributed contamination over both wings will most likely further aggravate the situation, causing test criteria. We shall work through the arithmetic to show an asymmetric stall accompanied by violent roll. In addition, that a 5.24 percent loss in maximum lift coefficient results in a significant increase in drag develops during rotation as the 10 percent stall speed margin. Using the rule of half, a wing goes into stall.

5.24 percent loss in maximum lift coefficient causes a 2.62 increase in stall speed. Therefore, we have the identity As Ref. I and 2 noted, stall of a contaminated wing is usually fluid = 1.0262 * V accompanied by either a pitch up or a pitch-down tendency of ig stall clean' Ig stall the aircraft, both of which tendencies will likely lead to The required safe takeoff speed, V 21 is over-rotation of the aircraft. A pitch-up tendency directly leads to over-rotation, driving the wing deeper into the region V 2 = 1.13 * v of stall where airframe buffet occurs. A pitch-down tendency Ig stall clean' is noticed by the pilot after rotation when the aircraft fails to which means that the airplane takes off with a 13 percent stall gain sufficient climb rate and customary height. The pilot's speed margin when the wing is clean and dry. normal response is to increase the elevator input, which action will over-rotate the aircraft and again lead to airframe buffet.

Assuming no adjustment is made to V 2 for the presence of The moment of airframe buffet is the pilot's first indication fluid on the wing (i.e., V 2 remains constant), use the first equation to substitute V is fluid'1 .0262 for in that something is wrong. Fokker studies showed that the pilot stall V1g stall clean would not notice the reduced acceleration caused by the the second and obtain contamination drag or the accompanying slight increase in takeoff run, and therefore, the pilot would not be alerted that V2 = 1.13 * ( Vig stall fluid/1.0262) something was wrong with the aircraft.

or With clean wings, aircraft drag is low enough to ensure climb V2 ll0*V capability at the required climb angle at V 2 (takeoff safety 2 Ig stall fluid speed) with one engine inoperative. However, with contami- Thus the fluid-contaminated airplane takes off with a nated wings, the stalled wing may double aircraft drag, and transitory 10 percent stall speed margin. Nearly all aircraft in even with all engines operative at take-off thrust, climb the jet transport category were found to be able to accept this capability may be lost.

transitory loss in stall speed margin without any takeoff adjustments, but there were a few aircraft for which adjust- As mentioned above, the Fokker 100 wing is designed for ments had to be made, such as by offloading passengers or flow separation to first occur inboard and then, as angle of cargo when fluids were used. attack increases, progress towards the wing tip. For the clean wing, inboard wing flow separation occurs at 16° AOA when The Type II anti-icing fluids were designed for use on jet maximum lift is reached, and flow separation does not affect transports that have rotation speeds of about 110 kt, and they roll control until an AOA of 19° is reached (Fig. 6). For the are not recommended for aircraft with rotation speeds below contaminated wing, the slow progression of flow separation kt. This is explained further in section 2.4.2. towards the wing tip is lost, and uncontrollable roll may develop at an AOA as low as 10°, just I ° beyond the AOA for 2.3 Effects of Wing Contamination on Takeoff maximum lift of the contaminated wing.

Characteristics References 1 and 2 give good descriptions of the typical In the Fokker 100 engineering simulator studies for a effects of contamination on airplane takeoff characteristics. symmetrically distributed roughness of a thickness that caused 4-4 wing stall at 90, an altered takeoff technique was found that States, has a low viscosity over the range of expected achieved a successful simulator takeoff if the peak angle of operating temperatures. Diethylene, triethylene, and propy- attack were 8.5 0. It was also found, however, that this was lene glycol-based fluids are used in Europe and are becoming achieved at the expense of significantly increased runway more common in North America. Compared with distance. monoethylene glycol, when these latter fluids are used undiluted, their viscosity is higher and increases faster with Van Hengst concluded from these simulator studies that: decreasing temperatures. If applied undiluted to the wing at "With the lack of any means of relating the amount of the colder temperatures, their viscosities are high enough to contamination in ground icing conditions to its effect on the cause unacceptably high aerodynamic penalties at takeoff.

aerodynamics of the aircraft, this flight simulation study Diluted with water, these latter fluids have acceptable aero shows that NO TAKE-OFF SHOULD BE ATTEMPTED penalties; therefore, they should always be used in the diluted UNLESS IT IS FIRST ASCERTAINED THAT ALL CRITI- formulation for deicing aircraft.

CAL SURFACES OF THE AIRCRAFT ARE FREE OF ICE, SNOW OR FROST DEPOSITS." 2.4.2 SAE/ISOIAEA Type II Fluids Type II fluids have markedly improved anti-king capabilities 2.4 Characteristics of Ground Deicing and Anti-icing compared with Type I fluids. These fluids contain at least Fluids 50 percent glycol in their neat form. They exhibit We shall begin this section with definitions of the terms non-Newtonian behavior, which means that their viscosity strongly depends on shear as well as on temperature. Their "aircraft deicing" and "aircraft anti-icing". "Aircraft deicing" is a the procedure that removes frost, ice, snow, or slush from viscosity decreases strongly with increasing shear stress. This non-Newtonian behavior is achieved by adding thickeners the aircraft in order to provide clean surfaces. Deicing composed of long polymer chains. When the airplane is involves spraying the surfaces with hot water or hot water/ glycol mixtures. stationary and wind speeds are low, the Type II fluid film on the wing is gel-like and therefore thicker than 'I)'pe I films.

"Aircraft anti-icing" is the procedure that pmtects clean Its greater thickness allows it to absorb more freezing surfaces of the aircraft against the formation of frost, ice, and precipitation before ice crystals begin adhering to the wing.

accumulations of snow or slush for a limited period of time During the takeoff run and climbout (at rotation speeds over (holdover time). Anti-icing involves spraying the clean 85 kt), air flowing over the wings shears the fluid and reduces its viscosity to near that of a Type I fluid, and it readily flows surfaces with thickened glycol-based fluids that can protect against freezing precipitation for a limited time (holdover off the wing.

time). The international aviation community has accepted the Type II fluids are sensitive to storage tank materials and AEA's coding of these deicing and anti-icing fluids as 'I)'pe I and Type Il, respectively. Type I and1'pe II fluids must meet handling equipment. Therefore, special tank materials are rigorous physical, chemical, and aerodynamic acceptance used to prevent fluid vapors from corroding the tanks; and specifications before they can be qualified for use in aircraft pumps, nozzles, and piping are designed to avoid degrading operations. Although the AEA, SAE, and ISO have cooper- the fluid (i.e., the polymer chains must not be broken up by ated to develop a comprehensive set of specifications and the shearing action of pumping and pressure drop) before it qualification tests for these fluids, some minor differences settles on the aircraft surfaces.

exists between the three organizations' specifications.

Therefore the fluid is prefixed by the letters AEA, SAE, or Type II fluids were developed for typical commercial ISO, as for example, SAE '13'pe I or ISO Type II, etc. transports that have rotation speeds of about 110 knots.

During the takeoff run, high viscosity fluids, such as Type II Unfortunately, some confusion could develop over fluid fluids, develop a wavy surface that in effect is a form of terminology because there are also Mil Spec Type I and surface roughness that degrades aerodynamic performance.

The high takeoff run speeds of the large transports help shear 'I)/pe II fluids, which are similar, but not identical, to the SAE/ ISO/AEA 'I)/pe I fluids. These Mil Spec fluids will not be the fluid and reduce its viscosity, and the long takeoff runs (about 25 sec) provide time for most of the fluid to flow off discussed herein. (For further clarification, see the discussion by M.S. Jarrell starting on p 243 of Ref. 7.) the wings before rotation. These fluids are not intended for commuters and general aviation aircraft whose rotation speeds are usually less than 85 knots and whose takeoff run times are 2.4.1 SAE/ISO/AEA Type I Fluids about 15 sec. One commuter manufacturer found that its Type I fluids, in their undiluted (neat) formulation, usually aircraft would have to be held on the ground for about 30 sec contain a minimum of 80 percent glycol. These fluids are during the takeoff run to ensure adequate fluid runoff from the easily stored and handled. They are eutectic with the minimum freezing point occurring approximately at a mixture wings and tails; otherwise, the aircraft would not rotate of 60 percent glycol and 40 percent water by volume. Their because the residual fluid caused excessive lift loss on the tail viscosity is a function of temperature but not of fluid shear, (Ref. 8). THE OPERATOR OF AN AIRPLANE SHOULD and therefore they are said to exhibit Newtonian behavior. CONSULT THE AIRPLANE MANUFACTURER FOR Type I fluid viscosity is relatively low except at very cold RECOMMENDATIONS REGARDING THE USE OF TYPE I AND TYPE II FLUIDS.

temperatures, where the viscosity depends significantly on the type of glycol used. This low viscosity allows Type I fluids to 2.4.3 Holdover Time readily flow off aircraft surfaces, leaving only a thin layer of "Holdover time" is the estimated time the anti-icing fluid will protection against freezing precipitation. They have limited prevent frost, ice, snow or other forms of freezing precipita- effectiveness when used for anti-icing purposes. Mono- tion from forming or accumulating on the protected surfaces ethylene glycol, which has been widely used in the United 4-5 airline, and air traffic control officials to jointly develop of an aircraft. Holdover time is estimated to be the time deicing plans tailored to their specific airport.

interval between when the fluid was applied and when ice crystals became visible in the fluid, for a given intensity and On November 1, 1992, revised EAR 121.629 became type of freezing precipitation and outside air temperature or effective. It requires the operator to develop an wing surface temperature.

FAA-approved aircraft ground deicing program and imple- ment it when weather conditions are conducive to ground Tables of holdover times were first developed by the ABA on icing. These plans are highly individualized to the particular the basis of tests in the laboratory and in real winter condi- operator at the given airport and must be approved by the tions, and from years of experience of several European FAA's Principal Operations Inspector or Principal Mainte- airlines. These holdover time tables have served as a nance Inspector for the airport.

guideline to pilots in Europe, where there have been no takeoff accidents attributable to ground icing for over The FAA-approved aircraft ground deicing program must 20 years. The original ABA tables have been modified by the include (Ref. 2, ID: international SAE G-12 Committee on Aircraft Ground Deicing Fluids, which had access to additional test results in I. Procedures to determine the existence of conditions real winter conditions in the United States and Canada. These conducive to icing of aircraft on the ground.

adjusted tables are presented in Figs. 7 and 8 for Type I and 'IS'pe II fluids, respectively (Ref. 9). While the AEA tables 2. Sound management, training of flight and ground crews, give only one protection time, the SAE tables give two qualification of all affected personnel, and assignment of protection times: a lower time and an upper time. This range specific responsibilities.

serves to remind the user that protection times depend on many factors.

3. Specific checks and inspections during the deicing process.

The SAE tables should not be separated from the procedures within 5 min of takeoff.

document (SAE ARP 4737), since the holdover times depend 4. Apre-takeoff check or inspection Using supportable holdover time tables, the operator must upon following the proper procedures, cautions, and caveats establish a holdover time for the applied deicing or anti-icing given in that document. The two cautionary notes given in the fluid under the existing precipitation conditions and outside tables are worth repeating here: air temperature.

CAUTION: The times of protection represented in this table FAR 121.629 allows operators to: are for general information purposes only and should be used only in conjunction with a pre-takeoff inspection.

I. Develop and use FAA-approved alternative procedures such as ice detectors.

CAUTION: The time of protection will be shortened in heavy weather conditions, high wind velocity and jet blast may 2. Elect to not operate in ground icing conditions if so stated cause a degradation of the protective film. If these conditions in its Operations Specifications.

occur, the time of protection may be shortened considerably.

This is also the case when the fuel temperature is significantly 3. Dispatch and take off with slight amounts of frost (up to lower than OAT.

3mm) on underwing surfaces in the vicinity of cold-soaked These two cautions reveal the complexity and challenge the fuel cells if approved by the FAA Aircraft Certification Office.

pilot in command faces in making a final determination as to Operational procedures acceptable to the FAA as set out in whether it is safe to take off.

FAR 121.629 are summarized below: 2.5 FAA Rulemaking on Ground Icing I. A pre-flight external aircraft icing check must be per- On July 21, 1992, the FAA announced that it would issue a formed by qualified ground personnel immediately following Notice of Proposed Rulemaking that would require each applications of aircraft de-icing and anti-icing fluids. This airline to have an FAA-approved ground deicing program in check determines whether the critical surfaces are free of place by the next winter (Ref. 10). The proposed rule would frost, ice or snow before push-back or taxi, and the results of require airlines to provide training for pilots and other the checks are communicated to the pilot in command. The personnel on the detection of wing ice and provide for aircraft should be released for take-off as soon as possible.

establishment of limits on how long an airplane can be exposed to snow or freezing rain before it had to be inspected 2. A pre-takeoff check is required within 5 min of takeoff or deiced again. The FAA would also change operational anytime conditions conducive to ground icing exist and/or procedures for controlling the flow of aircraft on the ground to anytime the aircraft has been deiced or anti-iced and a reduce the time aircraft have to wait in line for takeoff after holdover time established.

being deiced. The FAA would also encourage the use of the longer-lasting AEA Type Il anti-icing fluid, which is thicker within the holdover time, and stays effective longer than Type I. The FAA would also 3. if the pre-takeoff check occurs the pilot or designated crew member (co-pilot or flight help finance the construction of deicing pads on taxiways to engineer) normally checks from inside the cockpit or cabin, further reduce the time between deicing and takeoff. For whichever provides the best vantage point. The pilot in airports that historically had experienced takeoff delays due to command may require the assistance of qualified ground heavy winter operations, the FAA would encourage airport.

personnel to assist in the pre-takeoff check.

4-6 4. If the pre-takeoff check occurs after the holdover time is 2. Initial and recurrent training and testing for pilots regard- ing procedures and ground operations in icing conditions.

exceeded, the pilot in command must make a pre-takeoff contamination inspection. Depending upon the agreement 3. A pre-takeoff contamination check within 5 min of takeoff.

between the FAA and the operator (which would take into account the type of aircraft and other factors) this inspection FAR 125 and 135 allow: may range from observing the wings from some vantage point inside the aircraft to an external inspection by a licensed 1. Voluntary application of FAR 121 rules, summarized inspector. The FAA prefers external inspections, which might include observation from a high vantage point using binocu- above, to FAR 125 and 135 operations.

lars, or actual touching of the aircraft surfaces.

2. Use of supportable holdover time tables with anti-icing An alternative action that could be taken if holdover time is fluids to assist in departure planning.

exceeded is to re-deice/anti-ice the wings, control surfaces and other critical surfaces and establish a new holdover time. 3. Takeoff with slight underwing frost formations if FAA approved.

5. An Airworthiness Directive (AD) on pre-takeoff ground icing inspections has been published for each of three specific As a result of these revised FAR's, FAA personnel, airline operators, traffic controllers, and airport authorities have been aircraft types. Actions to be taken for these aircraft if aggressively developing and implementing procedures to holdover time is exceeded are as follows: minimize aircraft takeoff hazards in icing conditions. The SAE G-12 Committee, airplane manufacturers, and fluid a. On the F28 and DC-9-10 (hardwing aircraft) and on the MD 80/88 (aircraft with cold-soaked wings), conduct manufacturers have also supported the activity. These actions the check from outside the aircraft in accordance with an have resulted in significant improvements in the ground deicing/anti-icing technologies as summarized below FAA-approved method as set forth in the AD. The operator must include a tactile check of selected portions (Ref. II): of the wing leading edge and upper wing surface.

1. New and improved AEA/SAE/ISO Type If anti-icing fluids and procedures are now in prevalent use in North America and b. On those aircraft for which an AD exists for pre-takeoff in Europe. If used properly, these fluids give longer protection.

ground icing inspections, the manufacturer may offer alternative methods to establish that the critical surfaces 2. Holdover time tables now exist that can be used in concert are not contaminated.

with other methods of assuring the clean aircraft concept.

Although not yet fully validated, these tables, with proper For the hardwing aircraft, such methods might include: on-ground operation of the wing thermal anti-ice system; training and guidance, can reduce flight crew confusion and an abrasion strip which is rough when no ice is present workload.

and smooth when covered with ice, such that when an inspection rod is ran over the strip, vibrations are felt 3. New aerodynamic test data and experimental qualification when the strip is clean, but no vibrations are felt when the methods ensure that deicing/anti-icing fluids do not them- selves impose unacceptable aerodynamic penalties during strip is covered with ice; a paint stripe or special reflective takeoff.

surface for a background that clearly shows up ice when it is present; improved lighting; surface ice detectors; or 4. Many operators and airports are now using either perma- surface boundary layer flow sensors.

nent or mobile deicing and anti-icing facilities located very near the departure end of runways. This method offers For cold soaked wings, existing methods include: wing enormous benefits, including last minute assurances of a clean tufts placed near the fuel cells such that when the tufts are probed with a long stick, the tufts will move when aircraft at takeoff, minimized operations time and fuel ice is absent, but will be frozen in place when ice is consumption, and avoidance of aircraft having to return to a present; or surface ice detectors placed near the fuel maintenance or service area for re-deicing/anti-icing if ice cells. formations were detected during pre-takeoff inspections. At many airports, where it is not yet feasible to locate spray If the FAA approves an alternative method, they will facilities at the departure end of runways, other alternatives issue a replacement AD, which will define the allowable exist.

alternative inspection methods. A replacement AD has been issued for the P28 that allows the external tactile There is still room for improvement of the SAE holdover time tables. Better scientific methods are needed to obtain and inspection to be replaced by the use of black paint analyze holdover time data, and to quantify the weather stripes on the wing at selected locations in conjunction conditions. Ideally, the pilot needs a way to quantify the with an external visual inspection.

weather conditions and to put this quantified information into FAR's 125.221, 125.287, 135.227, 135.345, and 135.351 were a computer program that will output a more accurate estimate revised and became effective December 1, 1993. They of holdover time.

require: Instruments are needed that will help the pilot determine if ice has formed on the the wings at the time of the pre-takeoff 1. The operator to develop and use FAA-approved, airplane check. The instruments should be able to survey the entire type specific procedures for performing required pre-takeoff upper wing surface; ice detectors that sample ice at discrete contamination checks or an approved alternate procedure for points on the wing are probably not sufficient.

assuring the clean aircraft concept.

bleed air to anti-ice engine nacelles and critical sections along And finally, operators should be encouraged to locate deicing the wing span. Because today's high by-pass-ratio engines facilities near the departure end of runways so that the aircraft can deliver only a limited amount of compressor bleed air, can be deiced, and without further delay, start the takeoff run.

aircraft manufacturers anti-ice as little as 40 percent of the Fortunately, there is a trend toward airports opting for end of wing span, and allow the other 60 percent to accrete ice runway deicing when it is feasible.

during an icing encounter. Reference 12 gives an illustration that shows the percentage of wing span that is anti-iced on 3. ADVANCED TECHNOLOGY FOR IN-FLIGHT ICE each of Boeing's aircraft.

PROTECTION Even though aircraft ice protection technology matured in the Jet transport manufacturers use wind tunnels to test aircraft 1940's, icing technology problems still continue to arise.

models with simulated ice shapes attached to the leading edge Most of these problems have arisen either because modem of the wings and tails. From these tests they learn how the ice aircraft have incorporated new technologies that have a ripple affects handling characteristics and stability and control, and effect on the ice protection systems or because global determine which parts of the wings and tails can be left economic competition has intensified the need to further unprotected. To verify their wind tunnel results, they apply optimize aircraft performance and to minimize development, simulated ice shapes to a real aircraft and flight test it in clear capital, operating, and maintenance costs. And the aircraft ice air.

protection system enters into these optimization strategies. In this section we briefly discuss in-flight icing and the advanced Some airplane manufacturers have found from their airframe icing technologies being globally pursued by researchers, integration studies that fuel burn during cruise can be reduced manufacturers, and regulatory agencies. Among these by eliminating the ice protection on the empennage (thus advanced technologies are advanced ice protection concepts avoiding heavy and complex ducting that carries bleed air and advanced computer codes.

from the engines to the tail) while making the tail sections larger to tolerate the expected ice. Other manufacturers have Figure 9 shows the components of an aircraft that require ice found that the best way to reduce fuel bum is to electrother- protection.

mally deice the tail, which allows them to reduce tail size, and in turn, reduce weight and drag penalties.

3.1 Protection Against In-Flight Ice Before discussing approaches to ice protection, we need to Anti-icing systems Business jets usually employ the same approach to ice define "anti-icing" and "deicing" systems.

protection as the larger jet transports, but turboprop and prevent ice from forming either by using an evaporative general aviation aircraft must employ a significantly different system that applies enough heat to evaporate all the surface approach. Their power margins are so small that only their water deposited by cloud droplets, or by using a running wet propellers and engine intake lips are electrothermally anti-iced system that applies just enough heat to prevent the water from and the remaining critical components are deiced either with freezing. With the running wet system, the water would run back in the form of rivulets that would cover the entire upper expandable pneumatic boots or with electrothermal deicers.

surface of the wing. To prevent the rivulets from freezing, the Pneumatic boots are attractive because they require very little entire upper surface would have to be heated; but this power, are lightweight, and are reasonably priced. One arrangement would require far more energy than an evapora- drawback usually cited for pneumatic boots is that for tive system and would greatly complicate the design of the effective ice removal, they must not be activated until about wing. Therefore, running wet systems are usually reserved for one quarter to one half inch of ice accretes on them. This use on engine inlets with short duct runs. On some aircraft, procedure prevents "ice bridging," which sometimes occurs anti-icing is accomplished with freezing point depressant when boots are expanded with smaller thicknesses of ice.

fluids (usually mixtures of glycol and water) which are oozed Several inflations may be required to remove the bridged ice, out through a porous panel on the leading edge of the wing or during which time ice continues to accumulate on the cap and other component.

further degrades aerodynamic performance. Airplanes that are systems allow ice to build to some prescribed certified for flight into icing with pneumatic boots must be Deicing designed to tolerate the additional one-quarter to one-half inch thickness, and then the system is actuated to remove the ice.

This is normally a repetitive or cyclic process of ice growth of ice.

and ice removal. Thus, those wings and tails protected with Advanced Impulse Deicers deicing systems must be designed to tolerate the aerodynamic 3.2 The last decade has seen the development of alternatives to penalties imposed by the expected maximum thickness of ice the conventional electrothermal and pneumatic boot deicers.

that would accrete before actuation.

These are the electromagnetically and pneumatically actuated mechanical impulse deicer systems: EIDI (electromagnetic The ideal protection against icing would be to anti-ice all impulse deicing); BESS (electro-expulsive separation system); components that collect ice. The simplest way to do this EDI (electromagnetic deicing strip); and FlIP (pneumatic would be to heat the surface and evaporate all the water.

impulse ice protection) (Ref. 13). These systems produce a Unfortunately, this approach is not practical because no rapid impulse that cracks, debonds, and dynamically expels aircraft can economically provide the required thermal energy the ice. Unlike the slowly expanding conventional pneumatic from the available on-board heat sources, which are hot boots, which rely on aerodynamic forces to remove ice, the compressor bleed air, engine waste heat, and electricity.

impulse systems accelerate the iced surface up to 1000 g's, and inertially eject the ice as the surface snaps back. The A more realistic approach is to protect only critical compo- inertial ejection process can remove ice layers as thin as nents, and design the airplane to tolerate some ice on the other 0.75 mm (Ref. 14). With these thinner ice layers, the components. Today's modem jet transports utilize compressor 4-8 aerodynamic penalties for ice contamination are correspond- discussion and clarification by all the major aircraft and ingly reduced, and in addition, the ejected ice particles are engine manufacturers. And perhaps there is need for flight very small. The small particles make the impulse deicers tests of a modern jet transport with simulated roughness attractive for application to engine inlets, where ingested applied on its wings and tails to correlate two-dimensional particles must not damage the engine components. Their wing section results with full-scale three-dimensional flight power requirements are quite low—about equal to the power test results.

consumed by the aircraft's landing lights or about one percent of an electrothermal anti-icing system or ten percent of an 3.3 Physical Characteristics of Ice Accretion electrothermal deicing system. Although somewhat heavier As mentioned earlier, the shape of ice accreted on the than conventional pneumatic deicers, their weights still appear unprotected portion a wing or other component depends on competitive.

the atmospheric environment (outside air temperature, liquid water content, and droplet sizes), the flight conditions The P11?, invented and manufactured by BFGoodrich Deicing (airspeed and AOA), and the component geometry (size, Systems, Inc., is the only impulse system which is being cross-section, and sweep). Rime ice and glaze ice were applied commercially today—on only one aircraft, the Grob mentioned in the Introduction as two extremes of icing GF-200. The electromagnetic impulse systems have proved shapes, but actually there is a continuum of icing shapes that effective in removing ice, but their relative complexity and range from rime at the coldest temperatures to glaze at the uncertain life expectancy have apparently discouraged any warmest. In time ice formation, the droplets freeze upon airframer from using them thus far, but manufacturers can impact and trap air in between the frozen droplets, causing the produce them for the aviation market right now. ice to appear white and opaque. In glaze ice formation, the droplets impact the surface and form a water film that partly Lynch, et. al., have expressed concern about the aerodynamic freezes at the droplet impact site and partly runs back along penalties that would be imposed on jet transports by the use of the chord to freeze farther aft. Glaze ice is clear like the advanced impulse deicers, which can limit ice thickness to refrigerator ice. Between the extremes of rime and glaze, the only 0.75 mm (Rd 3). In their experimental studies on the ice shape gradually changes from a pointy shape to a single- effects of roughness on airfoil aero performance, they found or double-horned shape, and these in-between shapes are that".., reductions in maximum lift capability on configura- referred to as mixtures of rime and glaze, or as mixed icing.

tions without leading-edge devices extended are very large, In the mixed regime, the ice formed near the stagnation region even for extremely small leading-edge ice (roughness) is usually clear glaze while the ice farther aft, where heat buildups. For example, roughness heights of around 0.005 in. transfer is higher, has an opaque rime appearance. Figure 10 [0.127 mm] would result in reductions in the maximum-lift (Ref. 15) illustrates how total temperature affects the ice cross capability of 20 percent at the critical spanwise stations on the section when all other atmospheric and flight conditions are wing or tail of a representative 200-seat transport. Obviously, held constant.

the concern is even greater for smaller aircraft. Increasing the leading-edge roughness size to at least 0.03 in. [0.762 mm], Liquid water content, droplet size, and air speed also affect perhaps the minimum ice buildup that can be reliably the ice shape, and increasing any or all of the above will eliminated by a deicing system, would result in losses in increase the amount of water deposited on the surface and maximum lift capability of up to 40 percent for the 200-seat move the ice shape towards the glaze end of the spectrum.

aircraft. If the wing or tail surface areas for a particular Increasing airspeed presents two opposing influences on ice configuration are sized by maximum lift capability, then shape: the increased convection heat and mass transfer corresponding increases in surface areas would be required, encourages freezing while the increased kinetic heating with all the attendant performance penalties (drag, weight, discourages freezing. But ultimately, a speed or Mach etc.)." number will be reached, beyond which kinetic heating will dominate and ice will not form. This explains why the wings For slatted airfoils, Lynch, et. al., say: "Lower percentage of fighter aircraft are not ice protected.

losses in maximum-lift capability due to leading-edge ice buildups are experienced if the ice buildup occurs on an Component size and shape also affect the ice shape. Relative extended/deflected leading-edge device such as a slat. The to their size, smaller components accrete more ice than do maximum penalty for the 0.03 in. [0.762 mm] ice buildup on a larger components. This is illustrated in Fig. 11. This size slat would be about 10 percent at typical landing flap settings. dependence has great significance to aircraft ice protection However, the penalty could well be near 20 percent on the system design requirements. For example, the wings of a wing for lower takeoff flap settings, or for a tail (without C-5A aircraft are so large that only a small strip of ice would deflected flaps). Again, these penalties would all be increased accrete on them, and as a result, the C-5A's wings do not for smaller aircraft." require ice protection.

It appears that because of the engine manufacturers' continu- Size dependence is very important to smaller aircraft flying in ing quest to improve engine performance, their next genera- icing conditions. It is frequently the case that even though the tion of high by-pass-ratio turbofan jet engines will provide wing appears to be free of ice, the tailplane has collected little or no excess bleed air for thermal anti-icing. Therefore, enough ice to adversely affect its aerodynamics, particularly those involved in ice protection technology have been forced during approach and landing. To deal with this problem, to consider efficient deicing systems as a possible alternative pilots often look for ice accretion on the smallest object they can see, for example, the windshield wiper blade. If the to the conventional thermal anti-icing systems. The above conclusions, by a major aircraft manufacturer, will likely windshield wiper is picking up ice, pilots know they are in create controversy and confusion about the future markets for icing conditions and should either turn on the ice protection impulse deicers—:heflrst new ice protection concept to be systems or get out of the icing clouds.

demonstrated in about 40 years. This issue needs further 4-9 Then the flowfield is recomputed for the new ice covered Another parameter affecting ice shape is sweep on wings or airfoil, the droplet trajectories are recomputed, and a new any other component. Sweep causes spanwise flow of air layer of ice is computed for a time step. This process is along the leading edge, which in turn causes ice to form repeated until all the of time steps add up to the total exposure scallops or lobster tails. These ice shapes have a spanwise time.

periodicity that has not been satisfactorily explained or predicted by analysis (Ref. 16). Although they look gro- LEWICE is most accurate in predicting the colder ice shapes, tesque, their effect on aerodynamic performance may be no and less accurate for the warmer ice shapes. At the colder worse than that of the glaze horns on unswept wings. In fact temperatures, the droplets freeze upon impact and the it has been suggested that the scallops may act as turbulence accuracy of the ice shape prediction is determined primarily generators and help keep the flow attached.

by the accuracy of the droplet trajectory prediction. Fortu- nately, droplet trajectory codes have good accuracy. At the 3.4 Advanced Computer Codes warmer temperatures, the water only partly freezes at the As in every other aircraft technology area, computer codes are droplet impact site, and the unfrozen water runs aft and heavily used in aircraft icing to support design, development, eventually freezes. At these warmer temperatures, the shape and certification. Although the final proof in the icing of the ice is controlled by heat transfer to the surrounding air.

certification process will always be through flight testing in The lower prediction accuracy for the warmer temperatures is natural icing conditions, manufacturers hope that computer attributed primarily to the lack of good heat and mass transfer code calculations can replace some of the flight testing. Icing prediction models for ice-roughened surfaces. Another source flight testing is regarded by the entire aircraft industry as of inaccuracy is the limitation of the physical model of the risky, costly, lengthy, and resource intensive.

icing process.

Today, computer codes are used extensively throughout the As just noted, central to the heat balance on the water at the aircraft industry to design ice protection systems or to predict warmer temperatures is the prediction of convective heat and ice accretion shapes on unprotected surfaces. For example, mass transfer from the water surface to the air flowing over about 100 organizations in the United States are using the surface. It is helpful to keep in mind that the air flowing NASA's LEWICE ice accretion code. Codes developed by over the airfoil is the sink for heat and mass. Thus, when ONERA in France and the DRA in Great Britain give water freezes on the surface, its heat of fusion is transferred to comparable results and are heavily used throughout Europe.

the surrounding air by convection heat transfer and by Other organizations also have ice prediction codes in various evaporative cooling through convective mass transfer.

stages of development. Most of these codes are considered research codes, which means that while they are not fully The surface of ice is covered with roughness. This roughness validated, they are the best codes available and are very useful does not affect the heat and mass transfer in the laminar to those who have experience with them and understand their boundary layer, but it does affect the transition location and limitations. But, they are still being improved and validated the heat and mass transfer in the turbulent boundary layer.

through the development of advanced numerical methods, Thrbulent heat transfer correlations and analytical models through the development of advanced physical models exist for standard sandgrain surface roughness that is about obtained from fundamental physics experiments, and through 10 percent or less of the boundary layer displacement comparisons with new data from basic experiments and thickness. Unfortunately, ice roughness is usually thicker operational experience.

than the boundary layer displacement thickness, and there- fore, there are no validated correlations or analytical models The codes most often employed in aircraft icing include for heat transfer over ice-roughened surfaces. Lacking 1) flow codes, 2) droplet trajectory codes, 3) ice accretion anything better, the ice accretion prediction codes use the prediction codes, 4) electrothermal deicer design/analysis sandgrain roughness models and ignore the fact that they were codes, 5) anti-icing ice protection system design codes, validated only for roughness heights much less than the 6) iced airfoil aeroanalysis codes, and 7) helicopter rotor and displacement thickness.

propeller performance-in-icing codes.

3.6 Basic Studies in Support of Computer Codes Accretion Predictions 3.5 Ice In spite of their limitations, the computer codes are being used Since this lecture is time-limited, we are unable to discuss the successfully by those who have had experience with them and details of the computer codes. The interested reader should know their limitations. Yet, it is obvious that there exists a consult Ref. 17, which give more details and more references.

need for further basic studies to improve the numerical In this section we will briefly discuss the LEWICE ice techniques and the physical models of heat and mass transfer accretion prediction code. Figure 12 shows the modeling and of the ice accretion growth process. Basic numerical approach in LEWICE. The code consists of three main studies to improve the numerical stability of the LEWICE ice elements: 1) flowfield prediction; 2) water droplet trajectory accretion code are being conducted at NASA Lewis. For prediction; and 3) ice accretion prediction. The flowfield LEWICE, it has been found that predicted warm ice shapes code normally used is a potential flow panel code, although are sensitive to the number of time steps that make up the LEWICE can accommodate Navier-Stokes or Euler or total exposure time. For the current version of LEWICE, compressible potential flow solvers. The droplet trajectory about five time steps is optimal. More time steps can lead to code uses results from the flowfield code to calculate the flux instabilities in the ice shape, such that the results do not of water impinging on the leading edge region of the airfoil.

converge to a single ice shape as would be expected with finer The ice accretion code solves a thermodynamic energy and finer time increments. Bidwell (Ref. 18) has succeeded balance and a mass balance on surface control volumes that in writing a numerical algorithm that eliminates the shape coincide with the panel elements on the airfoil as shown on instability, and his results converge to a single ice shape as Fig. 13. LEWICE cycles through the three elements to time step increment is reduced.

calculate an ice shape for a given time increment or time step.

4-Ta NASA Lewis is conducting basic experimental research to aircraft accounted for an additional twenty accidents and better understand how ice shapes develop under various icing incidents in which ICTS was considered a likely factor.

conditions and how ice roughness develops and affects 4.1 Uncovering ICTS Problems laminar-to-turbulent transition and heat and mass transfer.

By early 1991, the European Joint Airworthiness Authorities One objective is to quantify ice roughness and correlate it (JAA) had developed and published (Ref. 21) a required flight with the cloud and flight conditions. Other objectives are to test maneuver that they believe identifies aircraft with Id'S conduct experiments that will aid in developing heat and mass problems. In this maneuver, the airplane (with specified tail transfer correlations and analytical models for flow over iced ice contamination) is pitched over at a prescribed pitch rate to surfaces and mass transfer correlations and analytical models a load factor of zero "g". If the airplane remains in control, for flow over smooth surfaces.

and the stick force characteristics are within defined limits, then the airplane stability and control characteristics with ice 3.7 Experimental Icing Simulation Good experimental icing simulation facilities such as icing on the tailplane are judged acceptable by the JAA. The FAA has adopted the JAA's zero "g" maneuver in principle and has wind tunnels and in-flight spray tankers are essential because they are more productive, more economical, and far safer than required it be performed as part of the icing certification process on a selective basis to date.

flight testing in natural icing conditions. Since smaller aircraft and helicopters have limited range, they must wait for the Some airplane manufacturers' test pilots believe the pushover icing weather to come to their test sites. This waiting has to zero "g" maneuver is too dangerous and not a maneuver extended icing flight trials over several winters, and has that pilots would intentionally do in normal operations. But driven up the cost for icing certification. Longer-range the airworthiness authorities have been successful in getting aircraft fly long distances to where ice is forecasted, but the compliance from the manufacturers and believe that it is cost per flight hour of these aircraft is very high, and the currently the best maneuver to uncover ICTS problems.

manufacturer works hard to control these costs.

Generally, the manufacturer will use a cautious approach by starting with, say, a pushover to one-half "g". And in some Figure 14 shows the closed-loop circuit of the NASA Lewis cases, the one-half "g" maneuver has uncovered the ICTS Icing Research Tunnel (IRT). Two unique components of the problem.

IRT are its heat exchanger that refrigerates the air and its water spray system that uses air-blast nozzles to produce Although acknowledging that the zero "g" maneuver poses supercooled clouds. The IRT can produce the desired test high risk for test pilots, aircraft operators point out that even conditions any time of the year regardless of the weather in normal operations, if the pilot's approach was high or fast, outside. Because of its uniqueness and versatility, the IRT is he might push over to get to the glide slope, or if the approach one of NASA's most heavily utilized wind tunnels, logging was slow, he might push over to pick up speed. In either case, about 1000 hr of test time annually.

that would be pushing the airplane towards zero "g". The operators would like these problems to be found by the highly 4. ICE CONTAMINATED TAILPLANE STALL (ICTS) skilled test pilots and fixed by the manufacturer rather than be We ordinarily associate the hazards of in-flight icing with wing ice, knowing from our earlier discussions that 2mm encountered unexpectedly by the journeyman pilot. But everyone agrees that it would be beneficial to better under- thick ice roughness can increase stall speed by 18 percent and stand what happens during the pushover maneuver and to use reduce stall angle by 60 . But we are not as aware that even that knowledge to develop lower risk methods to identify thinner roughness on the tailplane leading edge can cause ICTS problems.

potentially catastrophic tailplane stall during approach or landing, when the wing flaps are extended. Ice contaminated 4.2 Screening of Turboprop Airplanes for ICTS tailplane stalls during approach or landing have caused some Two recent international workshops sponsored by the FAA airplanes to go into a steep dive. And if this happened at low have alerted the aviation community to the seriousness of altitude, the chances of recovering from the dive were slim to ICTS. Numerous articles about ICTS have been published none.

subsequently in magazines read by pilots. Both the FAA and the manufacturers have responded with a new and concerted Figure 15 shows the catastrophic flight path and aircraft effort to I) identify the causes of tailplane stall, 2) prevent it attitudes of the Vickers Viscount that crashed at Stockholm in by design, 3) discover and fix it before the airplane is January 1977 (Ref. 19). "The broken curve" according to certificated for icing, and 4) educate pilots on how to avoid it Dr. Martin Ingleman-Sundberg, "shows the flight path that in flight operations, might have been attained, in spite of the stall, if the pilots had managed to keep the yoke back."

An important recommendation that came from the first workshop was that the FAA should screen all turboprops used According to Mr. John Dow (Ref. 20), from the FAA's Small in Part 121 or 135 operations for susceptibility to ICTS. The Airplanes Directorate, "Sixteen known or suspected ICTS FAA responded by contracting with Mr. Pete Hellsten, a accidents occurred worldwide to turboprop-powered transport consultant in aircraft design, to develop an analytical method and commuter category airplanes, resulting in 139 fatal and apply it to the thirty-one turboprop airplanes that the FAA injuries." The FAA has issued eight Airworthiness Directives identified in the Part 125 or 135 categories. In the study, the (ADs) against five airplane types in commercial service in FAA and Mr. Hellsten analyzed cruise, approach, and landing response to ICTS related accidents and incidents. Although configurations at speeds from stall to VFE; they assumed a turboprop aircraft had the highest number of ICTS accidents forward center-of-gravity and made calculations for both one and incidents, the problem is not limited to a specific size or "g" and zero "g" load factors; and they also analyzed each of configuration of airplane. Indeed, a worldwide survey the above combination of conditions for both clean tails and revealed that piston, jet, and non-U.S. T'pe Certificated 4-Il tails contaminated with standard roughness. In discussing stall margins as (CLmax - CL) and (CL - CLmin), respec- Helisten's work (Ref. 22) herein, we present results only for tively. These margins are given on Fig. 18 for the six points tails contaminated with standard roughness. on Fig. 16. (Figure 18 also shows stall margins for the zero "g" analysis, which will be discussed later). As expected, the 4.3 Wing and nil Aero Characteristics for Approach and tail stall margin is least at point 5 but does not seem exces- Landing sively small. It would be difficult to draw any conclusions For the purpose of explaining his analytical approach, from the results shown in Fig. 18 about the susceptibility of Mr. Hellsten used the average geometry of the 31 turboprop this generic airplane to ICTS. These stall margins, however, configurations to develop a generic or "paper" airplane. are for the trimmed airplane and do not account for transients Although not an actual airplane, the generic airplane was a such as the nose down pitch rates or downgusts or control good representation of the study. We will use the data from inputs that might temporarily increase tail AOA and push the Hellsten's analysis of this generic airplane to illustrate the lift margin to zero or negative.

changes in wing and tailplane operating characteristics during approach and landing (this viewpoint was first adopted by Just as the JAA had found that the zero "g" flight test maneuver was the best discriminator of ICTS, Hellsten also Dr. Ingleman-Sundburg in Ref. 19).

found that his calculated response of an aircraft to the zero maneuver correlated best with the airplane's actual ICTS Figure 16 shows plots of wing lift coefficient versus wing angle of attack for the cruise, approach, and landing configu- history. Hellsten's study of the thirty-one turboprops showed rations. Superimposed on these plot are the operating points that when his analysis predicted that an airplane had a for cruise, approach, and landing. Starting with a cruise speed negative orjust slightly positive tail stall margin for the zero "g" maneuver, that airplane was likely to be susceptible to of 200 kt (point 1), the pilot decreases speed to 126 kt (point 2) while increasing the wing AOA from 2.5 0 to 9.5°. ICTS. During the pitchover maneuver, the wing loading is Next, the pilot deploys half flaps and lowers wing AOA to 7° zero (zero lift) and the large nose-down pitching moment (point 3) and then further decreases speed to 114 kt while coming from the flaps is balanced by a large download or increasing wing AOA to 9.5 0 (point 4). Finally, the pilot negative lift on the tail. The predicted tailplane configura- extends full flaps and lowers wing AOA to 4.5 0 (point 5). tions, lift coefficients, and stall margins during the zero-"g" Eventually the pilot reduces speed and increases wing AOA maneuver are shown in the bottom of Fig. 18 for three while slowing to touch down. airspeeds: Vsiau, 1.3*VstaH , and VFE. The predicted stall margins for the zero "g" maneuver range from a low of 0.04 to a high of 0.07 at V. The tailplane was contami- Several aircraft responses accompany the deployment of at Vstau nated with standard roughness. These slightly positive stall flaps. First, when the flaps are initially deployed, say at margins put the generic airplane in the susceptible range for point 2, the wing AOA has not yet changed, and the airplane is lifted because it is temporarily operating on the half-flap lift 'cr5.

curve at an AOA of 9.5°. This effect is termed "ballooning".

The results of Hellsten's screening analysis are summarized in Second, the deployed flaps move the center of pressure farther aft on the wing, causing a nose-down pitching Fig. 19, which is a histogram of the calculated tailplane stall margins for each of the 31 airplanes during the zero "g" moment. Third, the extended flaps increase the wing downwash angle, which in turn increases the AOA on the tail maneuver for standard roughness contamination on the tail and produces a greater downward force on the tail. The pilot (Ref. 23). Hellsten's analysis predicted that eighteen of the compensates for these effects by moving the yoke to trim out 31 turboprop airplanes had stall margins that were either the airplane. During the trim adjustment, the nose pitches negative or just slightly positive. Of these 18 airplanes 13 down to a lower wing AOA and the tail pitches up to a higher have known histories of ICTS. The remaining 13 airplanes tail AOA. During the upward motion of the tail there is an outside the susceptible range (meaning they have substantial positive stall margins) have no histories of ICTS. These increased downward relative velocity on the tail which further findings gave the FAA confidence that they were heading in increases the tail AOA. The critical moment occurs when deploying full flaps because this results in the largest trim the right direction with this screening process. The FAA is currently working with the manufacturers of these 18 adjustment to get to the smallest wing AOA (point 5) and, conversely, to the largest tail AOA. This is the critical point potentially susceptible airplanes to verify the methodology where tail stall margin is least and tail stall might occur. developed by Hellsten.

The above discussion also illustrates a dilemma that the pilot The question naturally arises as to whether certain design can get into: If the pilot suspects that wing ice contamination features distinguished airplanes that are not susceptible to ICTS. Hellsten found that those airplanes having no history has increased the wing stall speed and he increases speed to compensate, the wing AOA will decrease, but the tail AOA of ICTS problems either had properly trimmed movable horizontal stabilizers or had tails with inverted camber. But, will increase, The increased tail AOA might possibly reduce Dow cautioned that not all aircraft with movable tailplanes or tail stall margin to the point where a sudden downburst or cambered tails were free of ICTS problems. Hellsten also nose-down pitch could stall the tail. IT IS IMPORTANT observed that the more effective wing flap systems (i.e., more THAT THE PILOT KNOW AND FOLLOW THE MANUFACTURER'S RECOMMENDATION IN THIS Fowler motion) cause more nose-down pitching moment, thus SITUATION AND FLY IT BY THE BOOK. requiring more negative lift on the tail to trim out the flaps and thereby driving the tail further toward stall, Although the By interpolating the data of Hellsten, we were able to deduce use of movable stabilizers or cambered tails appear to be steps the wing and tail operating characteristics for the six points in the right direction, Hellsten cautioned that there is no shown in Fig. 16. The wing and tail configurations for these simple answer to what works or doesn't work. The designer must go through the details of the design.

points are given in Fig. 17. Hellsten defines the wing and tail M .

When you observe ice on the wing, assume that there is even 44 Stick Forces Caused by Stall of Fixed-Incidence more ice on the tail and that it will have a more profound Thilpianes effect.

In a number of turboprop ICTS incidents or accidents, it was found that after the pilot extended full flaps, the stick lurched forward with such force that the pilot, or both the pilot and Use pneumatic boots and other deicing and anti-icing co-pilot, had to use all their strength to pull it back (on large components strictly according to manufacturers recommen- turboprops, as much as 400 lb were required). This occurred dations. Hangar tales and rumors never provide better operating procedures than those who make and test the on airplanes with fixed incidence tailplanes that had aerody- namically balanced elevators without power boosting. When equipment. If you have a question, talk to the manufacturen directly.

the tailplane stalled, the separated flowfield redistributed the pressure over the elevator and caused an enormous downward In icing conditions, make the landing approach with some- hinge moment on it (Ref. 19). The problem is not a complete thing less than full flaps. Halfflaps or less are about right.

loss of elevator authority when the tailplane stalls, but rather Ask your manufacturer Check the applicable ADs [Airwor- the problem is the inability to detect the stall (stick lightening thiness Directives]. [Have afinn hold on the stick, and if or vibration or tail buffeting) soon enough and to muscle the your airplane has a fixed incidence tailplane with aerody- stick back so that the elevator is moved from the down to the natnically balanced elevators and no power boosting, up position. Even though the tail is stalled, it apparently can anticipate the possibility that it could lurch forward with great develop sufficient downward lifting force to prevent a dive.

force.]

4.5 Operating in Known or Suspected Icing Conditions If you are in known or suspected icing conditions, you must In icing conditions be circumspect about adding speed for the final approach to compensate for ice on the wings. Every be keenly aware of the deleterious effects of ice contamina- tion on both wing and tailplane aeroperformance and the knot of speed added to prevent wing stall is a knot closer to tail stall.. Fly the approach by the numbers if your airplane is resulting reduction in airplane handling qualities and stability and control. Keep in mind that because the tailplane is one of those at greatest risk [Adding speed lowers wing ADA and raises tail AOA, thus reducing tail stall margin to the smaller than the wing, ice can accumulate on the tail before point where a sudden downburst or nose-down pitch could you can actually see it on the wings or elsewhere; and relative stall the tail.]

to its size, the tailplane ice coverage will be thicker and extend farther aft than the ice on the wing (see Fig. to). The bottom line is that the tail will accumulate more ice and be If you do encounter pitch problems on final approach in icing conditions, muscle the elevator to the position you want, and less tolerant of it than the wing. Also, on airplanes equipped it will provide adequate control to avoid a pitchover The with pneumatic boots for ice protection, remember that the problem is not elevator authority but hinge moment and, smaller leading edge radius of the tail renders the tail boot less effective than the wing boot in removing ice. Thus the therefore, control forces that you can overcome with muscle.

tail boot may have to be exercised more often than the wing Be alert and wary during flap changes. Make final flap boots; BUT HERE YOU SHOULD FOLLOW THE AIR- selection at least 1,000 above ground level so that any PLANE AND PNEUMATIC BOOT MANUFACTURER'S ft uncommanded pitchover will occur with enough altitude to RECOMMENDATIONS.

recover Several recommendations came from the two tailplane icing If you experience an uncommanded pitchover during or conferences and from subsequent articles published in pilot magazines. Some recommendations involved operational shortly after flap selection, immediately return the flaps to the previous setting.

strategies that should help prevent an ice contaminated tailplane stall or help recover from one. There is an important If the aircraft is high and fast on final approach, go around caveat to any recommendation or guideline published here or and try again. Several uncominanded pitchovers have been elsewhere: READ YOUR AIRPLANE FLIGHT MANUAL reported by pilots who attempted to slow rapidly with mar AND FOLLOW THE AIRPLANE MANUFACTURER'S RECOMMENDATIONS; THEY OVERRIDE ANYTHING flaps.

THAT IS PRINTED HERE.

Tailplane icing is a real and serious threat to all airplanes, but especially to mid-sized, propeller-driven airplanes.

The following partial, but representative, list of the guidelines Conversely, tailplane icing need not threaten your flight for avoiding or recovering from ICTS was published recently safety if you: (1) are aware of the potential, (2) limit final by Manningham (Ref. 24): After checking with the manufacturer and the FM, consider flap settings in icing conditions, and (3) maintain vigilance during the final approach.

the following guidelines to avoid tailplane icing and its worst consequences: It takes just the right combination of a number of factors that can momentarily increase the tail's AOA and trigger a tail Know the level of icing for which your airplane is certified stall. These factors include tail ice accumulation (and, maybe, and never intentionally fly into icing conditions which exceed not very much of it), deployment of flaps, higher airspeeds, that level.

low airplane AOA, nose-down pitch, forward center of gravity, headwind gusts, sideslips, downdrafts, etc. But by Never fly in known icing conditions with any anti-icing or studying the above guidelines AND FOLLOWING THE deicing components inoperable.

4-13 AIRPLANE AND ICE PROTECTION MANUFACTURERS 10. Farrar, F.: FAA Announces Action to Prevent Ice on RECOMMENDATIONS, you should be well prepared to Aircraft. FAA News Release FAA 36-92, July 21, 1992.

cope with ICTS.

We will conclude this section on ICFS by repeating 11. Adams, RI.: Personal communication.

Manningham's succinct advice: 12. Cozby, D.E.: Ice Protection/Detection. Report of the FAA International Conference on Airplane Ground Deicing, The classic tail icing pitchover occurs on final approach as the flaps move to an increased setting. Appropriate pilot Reston, VA, U.S.A., May 28-29, 1992.

action is to use all necessary force to pull back on the yoke while returning the flaps to their previous setting. If you can 13. Bond, TH.; and Shin J.: Advanced Ice Protection Systems Test in the NASA Lewis Icing Research Tunnel.

remember the contents of this paragraph, you will have NASA TM- 103757, May 1991.

retained virtually all of the important information regarding tail icing.

14. Shin, J.; and Bond, T.H.: Surface Roughness Due to REFERENCES Residual Ice in the Use of Low Power Deicing Systems.

AIAA Paper 93-0031, Jan. 1993. (Also, NASA TM- I. Brumby, RE.: The Effect of Wing Ice Contamination on 105971, Jan. 1993.)

Essential Flight Characteristics. Paper No. 2 in AGARD- CP-496, Dec. 1991. 15. Olsen, WA.; Shaw, Ri.; and Newton, J.: Ice Shapes and the Resulting Drag Increase for a NACA 0012 Airfoil.

2. van Hengst, J.: Aircraft Dc-Icing and Future Technical NASA TM-83556, Jan. 1984.

Developments-An Aircraft Manufacturer's Point of View. Wingtips (Fokker Aircraft), No. 26, Sep. 1993. 16. Reehorst, AL.: Prediction of Ice Accretion on a Swept NACA 0012 Airfoil and Comparisons to Flight Test Results. AIAA Paper 92-0043, Jan. 1992. (Also, NASA 3. Lynch, F.T.; Valarezo, W.O.; and McGhee, R.J.: The TM-105368, Jan. 1992.)

Adverse Aerodynamic Impact of Very Small Leading-Edge Ice (Roughness) Buildups on Wings and Tails. Paper No. 12 in AGARD-CP-496, Dec. 1991. 17. Potapczuk, M.G.; and Reinmann, ii.: Icing Simulation: A survey of Computer Models and Experimental Facilities. Paper No. 5 in AGARD-CP-496, Dec. 1991.

4. Hill, E.G.; and Zierten, T.A.: Flight and Wind Tunnel (Also, NASATM-105192, Oct. 1991.)

Tests of the Aerodynamic Effects of Aircraft Ground Deicing/Anti-Icing Fluids. AIAA Paper 91-0762, 18. Bidwell, C.: Personal communication Jan. 1991.

19. lngelman-Sundberg, M.: Why Icing Causes Tailplane 5. Carbonaro, M.: Aerodynamic Effects of Dc/Anti-Icing Fluids, and Description of a Facility and Test Technique Stall. Air Line Pilot, Jan. 1992.

for their Assessment. Paper No. 18 in AGARD-CP-496, Dec. 1991. 20. Dow, J. Sr.: Presentation at Ice Contaminated Tailplane Stall Program, FAA-NASA Planning Meeting. NASA 6. Anon.: Aerodynamic Acceptance Test for Aircraft Ground Lewis Research Center, May 2-3, 1994.

Deicing/Anti-icing Fluids. Boeing Commercial Airplanes Document No. D6-55573, Mar. 1992. 21. Cattaneo, C.: Evolution Reglementaire en Matiere de Certification des Avions Civil en Conditions Givrantes.

Paper No. 4 in AGARD-CP-496, Dec. 1991.

7. Jarrell, M.S.: SAE Type I Aircraft Deicing/Anti-Icing Fluids. Report of the FAA International Conference on Airplane Ground Deicing, Reston, VA, U.S.A., May 28- 22. Hellsten, P.: Aerodynamic Analysis of Susceptibility to 29, 1992. Ice-Induced Tailplane Stall. Conference Proceedings of International Tailplane Icing Workshop II, San Hose, CA, U.S.A., Apr. 21-23, 1993.

8. Ellis, N.; Lim, E.; Teeling, P.; and Zhu,S.: Wind Tunnel Tests of Aerodynamic Effects of Type I & 11 Ground Del Anti-icing Fluids on Small Transport & General Aviation 23. Dow, J. Sr.: Actions in Progress. Conference Proceedings of International Tailplane Icing Workshop II, San Hose, Aircraft During Takeoff. AIAA Paper 91-0763, Jan.

CA, U.S.A., Apr. 21-23, 1993.

1991.

24. Manningham, D.: Tails of Woe. Business & Commercial 9. Anon.: Aircraft Deicing/Anti-Icing Methods with Fluids, for Large Transport Aircraft. Society of Automotive Aviation, Jan. 1993.

Engineers (SAE) Aerospace Recommended Practice Document No. SAE ARP4737. Oct. 1992.

4-14 C C,

I

Drag coefficient Angle of attack, dog Figure 1.—Effect of wing ice contamination on lift and drag coefficients (ref. 1). (a) Lift coefficient.

(b) Drag coefficient.

rK

C

a

-

Entire upper C surface roughened C, a 8ao Roughness at a leading edge + E Slats extended E I.

j20 Slats E retracted C Slats extended ft - A• 1.Solid symbols indicate distributed roughness 2.Open symbols indicate singular disturbance is 3.Flagged symbols indicate swept wing data 4.• Indicates unpublished data retracted - 0 .- -1 10-1 10-2 io- 10- 1-6 Nondimensional roughness height. k/c Figure 2.—Correlation of the effect of wing surface roughness on maximum lift coefficient (ref. 1).

4-IS F28 MoSI 643 o WR4VH (Flap sr; BLF37$4) c2.15(10Ml9 Fo*. ICC 112 MoSt Il-S O WEll (flp 20';eLF 5200) RN,. 5.3 x 10' ;M. la FddwIOO MOM I5-25

I

0WTh(FlapI0;Bt.F5200;UCd.n) AN0-3.lxlO';M..19 FdF29112Modd10-10 C WE (RIP ir SkI I0 (B) or no Sat (IG) 'c_5'(b0;Mt9 Ha*ki, SIddSy TdStd 38 (flit I) o Wn1VH (Flap 2i:Itll5Imis2V) RN 0 . 3.6 x 10;M . IS snl-200A0v. RIP r) oD157-O, Pap 2r I(.2s1d3) o B767-200. Flap 20 ) Fr - FIgm Is Sf3 Figure 3.—Effect of roughness on maximum lift coefficient (ref. 2).

4-16

• 3-D Tall at ONERA Fl, M 0

= 0.134 0.0*5 Chord 0.025 Chord

N = 6.2 x 10')

(R

£ Current Test

Range 5-18 x 10)

(RN Chord

Brumby. Entire

0.075 Chord • 1985 L.TPTTeSI 3-12x 106)

Range

(RN

Upper Surface

Roughened

aGE a

e

a

a

I £ EE .9 V E C $ 0 (a) Q-3 10-2 '0-a 10- Nondimensional roughness height, k/c M 0 = 0.20 =

9 x 106

a FL E 20

IVA

Slats :---- a"

"

-LTPT Test Results -

Roughness on Slat

'F F, EdK I

(b= 35°/15°)

G') 'V 'I 10-2 10-6 10- 10-3 Nondimensional roughness height, k/c Figure 4.—Effect of roughness on maximum lift (ref. 3). (a) Single element airfoil and tail. (b) Four element airfoil.

4-Il 3-D Tail at ONERA Fl, M 0 = 0.134 (RN= 6.2 x 106) 106) = 0.20 (RN Range 5-18 x 2-0 Single Element at LIPT, M 0 = 0.20 (RN Range 5. 16 x 10) MuRlelernent at LTPT, M 0 S 2-0 -8 0) a,

£

a

3-4

S

V

10-2 10- 10-3 ia-5 Nondimenslonal roughness height, lc Figure 5.—Effect of roughness on loss of angle-of-attack margin to stall (ref. 3).

4-IS FLAP 15° Out of ground stied engine out C a-, wing

I

Thrust reserve in 2nd segment climb Drag equal to thrust; / .1 MI enalnes TO-thrust r. thrust—e Drag hrust—ø- Angle of attack Figure 6.—Effect of wing contamination on aircraft lift and drag (ref. 2).

THE RESPONSIBILITY FOR THE APPLICATION OF THESE DATA REMAINS WITH THE USER AND SHOULD ONLY BE USED IN CONJUNCTION WITH THE SAE METHODS DOCUMENT (SEE CAUTIONS) VP of SAP. Tvne I Fluid Mixture Must he at least 10 C (18 °fl below OAT Approximate Holdover Times Under Various Weather Conditions (hours:mirnites) OAT Rain on Cold Freezing Snow Freezing Frost Soaked Wing Fog Rain 0:06-0:15 0:12-0:30 0:06-0:15 0:02-0:05 32 and above 0:18-0:45 0 and above 0:06-0:15 0:06-0:15 0:01-0:03 0:18-0:45 below 0 to -7 below 32 to 19 0:06-0:15 0:12-0:30 0:06-0:15 below -7 below 19 = Degrees Celsius = Degrees Fahrenheit OAT = Outside Air Temperature = Freezing Point FP CAUTION: THE TIMES OF PROTECTION REPRESENTED IN THIS TABLE ARE FOR GENERAL INFORMATION PURPOSES ONLY AND SHOULD BE USED ONLY IN CONJUNCTION WITH A PRE-TAKEOFF INSPECTION.

CAUTION: THE TIME OF PROTECTION WILL BE SHORTENED IN HEAVY WEATHER CONDITIONS. HIGH WIND VELOCITY AND JET BLAST MAY CAUSE A DEGRADATION OF THE PROTECTIVE FILM, IF THESE CONDITIONS OCCUR, THE TIME OF PROTECTION MAY BE SHORTENED CONSIDERABLY. THIS IS ALSO THE CASE WHEN THE FUEL TEMPERATURE IS SIGNIFICANTLY LOWER THAN OAT.

Figure 7.—Guideline for holdover times anticipated for SAE Type I fluid mixture as a function of weather conditions and OAT (ref. 9).

'0 I-i THE RESPONSIBILITY FOR THE APPLICATION OF THESE DATA REMAINS WITH THE USER AND SHOULD ONLY BE USED IN CONJUNCTION WITH SAE METHODS DOCUMENT. (SEE CAUTIONS) OAT SAE Type II fluid Approximate Holdover Times Under Various Weather Conditions (hours:minutes) Concentration Neat-Fluid/Water Freezing Rain Rain on Cold Frost Freezing Fog Snow (Vol %/Vol %) Soaked wing 1:15-3:00 0:25-1:00 0:08-0:20 0:24-1:00 0 and above 32 and above 100/0 12:00 0:50-2:00 0:20-0:45 0:04-0:10 0:18-0:45 75/25 6:00 50/50 4:00 0:35-1:30 0:15-0:30 0:02-0:05 012-0:30 below 0 to 0:35-1:30 0:20-0:45 0:08-0:20 –7 below 32 to 19 100/0 8:00 75125 5:00 0:25-1:00 0:15-0:30 0:04-0:10 50/50 3:00 0:20-0:45 0:05-0:15 0:01-0:03 below –7 to –14 below 19 to 7 100/0 8:00 0:35-1:30 0:20-0:45 75125 5:00 0:25-1:00 0:15-0:30 below –14 to –25 below 7 to –13 100/0 8:00 0:35-1:30 0:20-0:45 Of) below –25 below –13 100/0 Use of SAE Type H for anti-icing below –25 °C (-13 must maintain °C (13 °F) buffer, and the fluid shall conform to the lowest operational use temperaturelaetodynamic acceptance limitation (see part 6.3.1.12). Consider use of SAE Type I where SAE Type II fluid cannot be used.

ac = Celsius = Degrees Fahrenheit OAT = Outside Air Temperature VOL = Volume * = For maintenance purposes CAUTION: THE TIMES OF PROTECTION REPRESENTED IN THIS TABLE ARE FOR GENERAL INFORMATION PURPOSES ONLY AND SHOULD BE USED ONLY IN CONJUNCTION WITH A PRE-TAKEOFF INSPECTION.

CAUTION: THE TIME OF PROTECTION WILL BE SHORTENED IN HEAVY WEATHER CONDITIONS. HIGH WIND VELOCITY AND JET BLAST MAY CAUSE A DEGRADATION OF THE PROTECTIVE FILM. IF THESE CONDITIONS OCCUR THE TIME OF PROTECTION MAY BE SHORTENED CONSIDERABLY. THIS IS ALSO THE CASE WHEN THE FUEL TEMPERATURE IS SIGNIFICANTLY LOWER THAN OAT.

Figure 8.—Guideline for holdover times anticipated for SAE Type 11 fluid mixture as a function of weather conditions and OAT (ref. 9).

4-2!

Wing ge tC I17 leadingV ges

edges _e\t1// I

Propeller Antennas WiBal Rahor Essential Figure 9.—AIrcraft ice protection.

40; airspeed, 209 km/hr; LWC, 1.3 g/m 3; time, 8 mm NACA 0012 chord, 0.53 m; AOA,

4Z E — <C- — (Q — C^ C---

-1°C 0°C -20°C -18°C -15°C -12°C -8°C -5°C -2°C -26°C Total temperature Figure 10.—Effect of total temperature on ice shape development (ref. 15).

4-22 NACA 0012 airfoil; AOA, 4 0; airspeed, 252 knvbr. LWC, 0.35 gJm 3 ; MVD = 20 gm; time, 10 mm Figure 11.—Effect of airfoil size on ice coverage.

4-23 Flowfield prediction Waterdrop trajectory prediction Ice accretion prediction Aeroperformance: rough surface and flow separation Figure 12.—Ice accretion modeling approach.

Heat and mass Evaporation/ transfer to air sublimation Cloud water droplet flux, water njnback V0n,

y_ç9r

Unfrozen \ water njnback lee Panel element / \ 71!': "Wet" ice < surface Figure 13—Control volume approach for ice accretion model.

4-24 Corner Corner B - Turning vanes Flow tan 2100 ton cooler Varlchron drive control room Balance chamber - 420 mph test Icing - section 6x9tt sprays - - -

I [[] Control / Corner

I

/ A Corner L.

Model Shop D

/

access door–' Spray bar room / control room

I

- Airlock chamber Figure 14.—Schematic of the NASA Lewis Research Center Icing Research Tunnel (IRT).

E 400 V C 2 300 a, ±100 0 100 200 300 400 500 Horizontal distance, m Figure 1 5—Catastrophic flight path of Vickers Viscount accident in Stockholm, Sweden on January 1977 (ref. 19).

4-25 3.0 boy 2.0h 13Vstjun-105kt V tall i

jfry

Cruise - 200 kt -10 -5 0 5 10 15 20 Wing angle of attack Figure 16.-Angle-of-attack conditions for P. 1-fellsten's generic turboprop airplane (ref. 22).

Condition Raps Wing Wing Tail Elevator Tall, Airspeed.

(fig. 16) AOA, downwash, AOA, deflection, CL deg deg deg deg Trimmed aircraft I Clean 2.3 1.4 -1.5 2.3 0 200 2 Clean 9.5 3.8 3.1 -3.4 0.04 3 Half 7.0 3.8 -2.0 1.1 -0.07 126 Half 9.5 4.8 -0.2 -0.8 -0.05 114 5 Full 4.4 4.8 -4.9 3.1 -0.22 114 Full 6 18.1 9.6 4.2 -7.5 -0.075 80 Zero "g" pushover VFE Full 0-14.8 15.1 -0.29 145 1.3 V Full 0 -15.8 16.3 -0.29 105 Full 0 -17.3 18.3 -0.29 80 CL = 0.29 g' pushover V=l.3Vr Downwash - Figure 17.-Wing and tail configurations for the flight conditions shown on figure 16 (ref. 22).

4-26 -J

I

Airspeed, kt Touchdown -J C) C LM E to zero VstWI 1.3 V 11 V (:i) Airspeed, kt Figure 18.—Stall margins for the flight conditions shown on figure 16 and for the pushover to zero 'g' maneuver (ref. 22). (a) Wing.

(b) Tail.

4-27 m CQ• (0 w C S 0.

(0 0) di D 1 2 3 4 7 5 6 8 9

p

Number of aircraft Figure 19.—Predicted zero 'g' 'rough' tailplane stall margins for thirty-one Part 1211135 Turboprop Aircraft (ref. 23).