APPENDIX R. ICE SHAPES (CONTINUED) from calculated ice shapes will show more adverse effects on aircraft handling qualities and performance (section R.4 of this appendix). You should show that ice shapes selected from i cing codes and analytical methods are conservative. You should also evaluate the effects on aircraft handling, performance, structural vibration, ice shedding, and airflow disturbance affecting downstream components and systems. Such information may be difficult to produce withou t providing a suitable ice shape for reference. This may result in the need to verify the calculate d ice shape with an empirically determined ice shape (preferably from natural icing flight tests).
R.5.5 Prediction of Landing Gear and Strut Ice Shapes.
Applicants may conservatively estimate or use artificial (tanker) icing tests to determine ice shapes for the landing gear.
R.5.6 Prediction of Protected Surfaces Ice Shapes.
Applicants may use full-scale natural icing flight test or simulated icing tests (tankers and icing wind tunnels) to determine ice buildups on protected surfaces, such as intercycle ice for deicers (sections R.5.1, R.5.2, and R.5.3 of this appendix). You may test full-scale protected surfaces in icing wind tunnels using the hybrid-model method if the IPS conforms with the aircraft’s type design. Compare time histories of pressure or temperature increases during activation to the aircraft. With pneumatic deicing boots, this may require adding air volume in the system test set-up. The test should include longer times in Continuous Maximum conditions to evaluate th e stability and cyclic nature of intercycle and residual ice. Certain unique design features (for example stall strips mounted on deicing boots) may not readily shed ice.
Confidence has not been shown for the use of scaled models with scaled IPSs to predict intercycle ice, runback ice, and to evaluate other ice protection design considerations.
R.6 CORRELATION OF PREDICTED (SIMULATED) AND NATURAL ICE SHAPES.
To verify safe flight in icing conditions, compare the flying qualities of the aircraft with simulated critical ice shapes to those of the aircraft with natural ice shapes. Comparison of the simulated ice testing with that in natural icing may be difficult and require engineering judgmen t.
There may be a limited amount of quantitative flight characteristic data available for comparing the natural and simulated ice shape test results. Natural icing conditions are variable, especially during lengthy exposures or during a series of brief exposures required to build up the desired ice thickness. Close examination of the resulting natural ice shapes on the flight test aircraft is often impractical. The icing conditions used to determine the simulated ice shapes may not match the natural icing conditions encountered. The simulated ice shapes also may reflect a composite of critical ice shape considerations (such as horn length, location, angle, and mass). Natural icing tests, however, are required to provide overall checks of the aircraft’s safe flight in icing conditions, IPS analyses, simulated ice shape analyses, and unforeseen icing anomalies. Aircraft performance and handling qualities shown during natural icing flight tests should be equivalent to or less affected by ice accretion than those shown with the simulated ice shapes.
R-38
APPENDIX R. ICE SHAPES (CONTINUED) Applicants may use the lists of ice shape and water catch evaluation parameters in tables R-1 and R-2, ranked against their adverse airplane effects, to compare simulated and natural ice shapes.
These lists are from SAE ARP5903 (Reference R20).
Table R-1. Ranking of Ice Shape Evaluation Parameters Rank Parameter Units Conservatism Criteria 1 Upper (suction surface) horn height - Equal or greater horn peak thickness (height) 2 Upper Horn Angle Degrees Criticality of location (at upper peak thickness) 3 Lower (pressure surface) height - Equal or greater horn peak thickness (height) 4 Lower Horn Angle Degrees Criticality of location (at lower max. thickness) 5 Total ice cross-sectional area - Equal or greater area 6 Leading edge minimum thickness - Equal or smaller thickness 7 Upper accretion limit % x/c* Equal or greater x/c 8 Lower accretion limit % x/c* Equal or greater x/c NOTE: The first four parameters assume that icing horns exist. This is not always the case.
* Percent of local component chord Table R-2. Ranking of Water Catch Evaluation Parameters Rank Parameter Units Conservatism Criteria.
1 Upper impingement limit (suction surface) % x/c* Equal or greater x/c 2 Lower impingement limit (pressure surface) % x/c* Equal or greater x/c 3 Total water catch efficiency, E - Equal or greater magnitude 4 Maximum local water catch magnitude - Equal or greater efficiency, ȕmax 5 Water catch efficiency (Beta) curve - Equal or more adverse ȕ ҏdistribution * Percent of local component chord Reasons why ice shapes produced by various icing wind tunnels, computer codes, and other analytical methods may vary include:
x Calibration of the icing wind tunnel.
x The uniformity and qualities of the tunnel’s flow and icing cloud.
x Other tunnel testing considerations.
Reasons why ice shapes produced by various ice accretion computer codes, and other analytical methods, vary include:
x Differing algorithms and assumptions used in the computer codes and analytical methods.
x The use of various computer code versions and inputs.
x The use of empirical ice shapes from different sources to “tune” computer codes and other analytical methods (to account for unknown icing physics and other effects).
R-39
APPENDIX R. ICE SHAPES (CONTINUED) Once you have satisfactory correlations between icing wind tunnel, computer codes, and analytical methods with the verification database, use version control of these simulation ice shape methods to keep confidence in the correlation. Understanding the limits of the correlation requires engineering judgment. For example, a correlation may continue for changes in a singleelement airfoil’s thickness, thickness distribution, and camber. But, you should question applying a correlation for a single element airfoil to a multielement airfoil. Also, a correlation for 2D airfoils may not apply for 3D surfaces.
R.7 REFERENCES.
R1. “A Study of Inter-cycle, Residual, and Preactivation Ice,” Riley, J. T., et al., AIAA-2001-0089, 39th AIAA Aerospace Sciences Meeting and Exhibit, January 8-11, 2001, Reno, Nevada.
R2. “Effect of Intercycle Ice Accretions on Airfoil Performance,” Broeren, A.Pl, Addy, H.E. Jr., and Bragg, M.B., AIAA-2002-0240.
R3. “NPA 25F-219; Flight Characteristics in Icing Conditions,” Joint Aviation Authorities Letter JAA/SEC/3-9-2, dated 23 April 1993.
R4. “Empennage Deicing System Deletion,” Malone, Frank, Boeing Airliner, May-June 1962.
R5. “Effects of Large-Droplet Ice Accretion on Airfoil and Wing Aerodynamics and Control,” Bragg, Michael B. and Loth, Eric, Aeronautical and Astronautical Engineering, University of Illinois at Urbana-Champaign, August 16, 1999.
R6. “Report of the 12A Working Group on Determination of Critical Ice Shapes for the Certification of Aircraft,” FAA Report Number DOT/FAA/AR-00/37, Office of Aviation Research, Washington, D.C. 20591.
R7. “Effects of Ice Accretions on Aircraft Aerodynamics,” Lynch, F.T. and Khodasdoust, A., Progress In Aerospace Sciences, January 2002.
R8. “DES For Airfoils With Upper-Surface Ice Shapes,” Pan, J. and Loth, E., Department of Aeronautical and Astronautical Engineering, University of Illinois at Urbana-Champaign, September 2001.
R9. “Detached Eddy Simulations of An Iced-Airfoil,” S. Kumar and E. Loth, AIAA Paper AIAA 2001-0678.
R10. “Airborne Icing Tankers,” SAE ARP5904.
R11. von Glahn, U.H ., “Use of Truncated Flapped Airfoils for Impingement and Icing Tests of Full-Scale Leading-Edge Sections,” NACA RM E56E11, July 1956.
R-40
APPENDIX R. ICE SHAPES (CONTINUED) R12. Saeed, F., Se lig, M.S., and Bragg, M.B., “Design of Subsonic Airfoils with Full-Scale Leading Edges for Ice Accretion Testing,” AIAA-96-0635, January 1996 and J. Aircraft, vol. 34, no. 1, January-February 1997, pp. 94-100.
R13. Saeed, F., Se lig, M.S., and Bragg, M.B., “Experimental Validation of the Hybrid Airfoil Design Procedure for Full-Scale Ice Accretion Simulation,” AIAA-98-0199, January 1998.
R14. Ruff, G.A., “Analysis and Verification of the Icing Scaling Equation,” AECD-TR-85-30, Vol 1 (Rev), March 1986.
R15. Anderson, D.N., “Manual of Scaling Methods,” Ohio Aerospace Institute (to be published).
R16. Anderson, D.N., “Effect of Velocity in Icing Scaling Tests,” AIAA-2000-0236, January 2000.
R17. Kind, R.J., “Assessment of Importance of Water-Film Parameters for Scaling of Glaze Icing,” AIAA-2001-0835, January 2001.
R18. “Calibration and Acceptance of Icing Wind Tunnels,” SAE ARP5905.
R19. “Ice Accretion Formations on a NACA 0012 Swept Wing Tip in Natural Icing Conditions,” Vargas, M., Giriunas, J.A., and Ratvasky, T.P., AIAA-2002-0244.
R20. “Ice Accretion and Droplet Impingement Codes,” SAE ARP5903.
R21. “ Ice Accretion Simulation Evaluation Test,” NATO RTO Technical Report TR-038, November 2001.
R-41 (and R-42)
APPENIDX S. ICE-CONTAMINATED HORIZONTAL STABILIZER (TAILPLANE) STALL S.1 ICE-CONTAMINATED HORIZONTAL STABILIZER (TAILPLANE) STALL.
As discussed in section 6.1, of this AC, safe flight of airplanes in icing conditions requires compliance with 14 CFR §§ 23.143 and 25.143. Many accidents have resulted from stalling of an ice-contaminated horizontal tail. This phenomenon is commonly called ice-contaminated tailplane stall (ICTS). The accidents typically occur during the approach phase of flight with the wing flaps extended.
The horizontal tailplane produces down-lift to balance the pitching moment created between the wing lift and the center of gravity, as shown in figure S-1. The down-lift required by the horizontal tail to balance the pitching moments varies with the airplane flap configuration and flight condition (thrust and airspeed (angle-of-attack (AOA))). When extending the flaps, the wing center of lift moves further aft of the center of gravity. This results in increased nose-down pitching moment (figure S-1). Also, the extended wing flap increases the wing camber, resulting in an increased wing lift for a given AOA and an increased circulation of chordwise airflow around the wing. The flight condition also has an effect on the pitching moment. Decreasing airspeed (increasing aircraft AOA) requires an increase in tail down-lift. Increasing the thrust may also affect the pitching moment. This depends on the position of the thrust line with the center of gravity. All of these effects require an increase in the tail down-lift to balance the airplane in the pitch axis.
Trailing Edge Flaps ExtendedCruise Configuration Figure S-1. Airplane Forces and Pitching Moments.
An increase in the AOA at the tailplane or deflecting the elevator position to modulate the tail down-lift balances the moments in the pitch axis. The wing trailing edge flap position strongly affects the AOA at the tailplane. Extending the trailing-edge flaps significantly increases the downwash at the tailplane. This increased downwash increases the tailplane’s AOA. The increase in the AOA at the tailplane effectively increases the tail down-lift. A trimmable horizontal stabilizer can compensate for the change in the tailplane’s AOA. Otherwise, deflecting the elevator trailing-edge up can provide the down-lift required for balance.
The tail down-lift requirements remain the same when ice builds up on the horizontal tail.
However, the aerodynamic effects of the ice buildup degrade the ability of the horizontal tail to produce the down-lift. Figure S-2 shows that light roughness, such as that resulting from thin ice accretion on the leading edge of a horizontal stabilizer, may cause the horizontal tail to stall at a lower AOA (when compared with the uncontaminated tail). Flight-testing with ‘sandpaper’ ice S-1
APPENDIX S. ICE-CONTAMINATED HORIZONTAL STABILIZER (TAILPLANE) STALL (CONTINUED) shows that the light roughness can be more severe than the larger critical ice shapes on some aircraft. Applicants should evaluate the acceptability of using the sandpaper ice during the certification program. Wind tunnel testing of a 36-inch NACA 23012 airfoil at a Reynolds number of 7.5x106, as shown in figure R-8, points out that intercycle ice surface roughness may be more adverse than sandpaper ice.
Figure S-2. Aerodynamic Effects of Standard Roughness (0.011-inch Carborundum Grains) for a NACA 652-215 Airfoil. (The Carborundum grains were uniformly distributed at a density of 5 to 10 percent of the area that extended from the airfoil’s leading edge to 0.08c of the 24-inch model.) (Reference S1).
In ICTS, the flow separation of interest typically starts at the leading edge and progresses aft over the lower surface of the tailplane to some line where the flow reattaches. As the AOA increases, the line-of-reattachment moves further aft. If the flow separation moves aft to the elevator control surface, a significant hinge moment in the trailing edge down direction occurs because of redistributed pressures on the elevator. If this occurs with an unboosted elevator control system, the increased hinge moment may snatch the control column forward from the pilot and require significant pilot strength to oppose. The elevator will automatically deflect trailing edge down until the hinge moment reaches zero (as the air loading moments balance on elevator’s upper and lower surfaces). This reduces the pitch controllability and stability.
ICTS occurs when the iced stall AOA is exceeded. The tail stall AOA may vary with different ice shapes. The primary driver for increasing the AOA at the tail is the wing trailing-edge flap as discussed above. However, other reasons that can lead to exceeding the iced tail stall angle are nose-down pitch motion (to capture a glide slope), reduced airplane AOA to increase airspeed, wind gusts, and changes in engine power. The local flow interactions emanating from the S-2
APPENDIX S. ICE-CONTAMINATED HORIZONTAL STABILIZER (TAILPLANE) STALL (CONTINUED) juncture of the horizontal and vertical stabilizers for cruciform and T-tail designs may adversely affect the airflow around the tailplane during high sideslip maneuvers.
Accidents caused by ICTS have occurred when the airplanes were on approach and at low altitude. Flight near the ground does not allow much time for the crew to distinguish between a wing stall and tail stall. Subsequently, the crew will not be able to perform the proper piloting technique to recover control of the airplane. If ICTS occurs, the pilot technique to recover control is to reduce the AOA at the tailplane by raising the flaps and pulling back on the control column to raise the nose.
You should consider the aerodynamic effects of ice-contaminated horizontal stabilizers for all airplanes. This includes those with tab driven controls and those with powered controls. You should perform an evaluation to determine if this unsafe flight condition is likely to occur.
Airplanes susceptible to this phenomenon are those that have little or no margins to stall with uncontaminated horizontal stabilizers. ACs 23.143-1 and 25-7A provide acceptable flight test procedures for determining the susceptibility of an airplane to ice contaminated tail stall.
AC 23.1419-2C and AC 25-1419-1 provide more guidance on this subject.
S.2 REFERENCES.
S1. “Theory of Wing Sections,” Abbot, Ira H. and Von Doenhoff, Albert E., Dover Publications, Inc., 180 Varick Street, New York, N.Y., 10014, 1959, p. 626-627.
S-3 (and S-4)
APPENIDX T. SIMULATED ICING FLIGHT TESTS: ICING TANKERS AND
SPRAY RIG TESTS T.1 ICING TANKER TESTS.
Simulated icing conditions developed by icing tankers have been used successfully to check analyses required for showing regulatory compliance. These simulated icing methods have provided drop impingement information, checks on and information to develop simulated ice shapes, heat transfer measurements, and ice-shedding information. Applicants may use icing tanker test information to check small-scale tunnel test results. You may also use this information to extend natural icing tests results to icing not found during natural icing flight tests. You must show that the results are both accurate and conservative. Icing tankers, especially those that have hydraulically aspirated nozzles, may not reproduce all the icing conditions in 14 CFR parts 25, Appendix C and 29, Appendix C (Reference T1).
T.2 SPRAY RIG TESTS.
Applicants may install spray rigs on their aircraft to evaluate the icing characteristics of some surfaces. Some of the limits that apply to tanker tests also apply to spray rig tests. Test rigs of this nature are expensive to develop and install on test aircraft. An advantage of this testing is the ability to control the distance between the spray rig and the test surface. You should ensure that the drop transit time is long enough for the drops to become supercooled.
The major disadvantage of this method is that the aerodynamic disturbance caused by the spray rig may disrupt the flow field around the test surface. This feature may produce unrealistic impingement characteristics that are difficult to evaluate. You may need to perform studies, such as wind tunnel tests, to ensure that the spray rig does not aerodynamically affect the surface ice buildup. The size and weight limits of the spray rig also limit the area that can be sprayed.
Because of these limits, small spray rigs are typically not used for certification. This does not eliminate the use of this test method as a development tool for evaluating the icing of small surfaces.
T.3 REFERENCES.
T1. “Airborne Icing Tankers,” SAE ARP5904.
T-1 (and T-2)
APPENIDX U. ICING WIND TUNNEL TESTS U.1 ICING WIND TUNNEL TESTS.
Several types of wind tunnel designs (including closed and open loop, refrigerated and atmospherically cooled) have the ability to simulate icing conditions. These facilities have the ability to produce controlled, steady state plumes of supercooled water drops for a range of temperatures, LWC, drop size, and airspeeds to simulate various icing conditions. They allow access to examine ice accretions and to measure IPS performance accurately. Instrumentation for measuring the icing cloud is more extensive and accurate than flight test instrumentation.
Also, the icing wind tunnel provides direct control of the icing conditions. The icing tunnel should be in calibration, following the practices described in SAE ARP5905 (Reference U1).
Consider SAE ARP5905 when judging the acceptability of information obtained in icing wind tunnels. The disadvantages of ice tunnel tests are their limited size, their inability to simulate altitude effects (except the CIRA IWT), and their inability to simulate the variability of natural icing conditions.
Most tunnels are small when compared with the full-scale test articles. For models of large articles having tunnel blockage of about 10 percent or more, getting accurate aerodynamic and thermodynamic similarity may be difficult. Applicants should also address wind tunnel flow characteristics, icing plume uniformity, wall effects, model effects, model support, and model scaling issues. Perform dimensional analyses of the aerodynamic and thermodynamic parameters that characterize the full-size article before conducting a small-scale model test. Use this information to ensure similarities between the full-scale article and small-scale model, including considerations of Reynolds and Weber numbers. Also, you should perform error analyses to ensure that test results are useful after considering expected experimental errors.
Appendix I and appendix R, section R.5.3 of this AC contain more information on using scaled models for drop impingement and ice shape investigations. You can determine full-scale data from natural icing flight tests, dry air flight tests, spray rig tests, tanker tests, or a combination of these tests.
Design test conditions and models to ensure that Reynolds number and other scaling parameters are close to full-scale values. You should mount your test models to simulate the flight attitude associated with the most critical condition. If you use flaps or other devices to produce the proper flow field conditions, provide instrumentation to show that test and full-scale design parameters agree. Liquid IPSs tested in an icing wind tunnel should prevent ice formation on the protected surfaces at the design fluid flow-rate.
U.2 REFERENCES.
U1. “Calibration and Acceptance of Icing Wind Tunnels,” SAE ARP5905.
U-1 (and U-2)