systems when demonstrating compliance to §§ 23.1093, 25.1093, 27.1093, and 29.1093, see paragraph 11.p.(3)(a) of this AC.
d. Application of Cloud Extent Factors for § 33.68 Compliance Tests. Cloud extent factors are depicted in appendices C and O of part 25, appendix C of part 29, and appendix D of part 33.
The horizontal cloud extent factor is a dimensionless number, which relates the length of a cloud to an average LWC across the cloud. It is used in the CPA to assess the probable icing content during various aircraft mission and performance analyses.
(1) Typically, the average LWC of a cloud is less for longer clouds. The horizontal cloud extent factor is not intended to be used to limit the severity of exposure to icing conditions, assuming the aircraft will be required to operate in that condition. For example, when in a holding pattern which may require repeated passes through a severe icing environment, or continuously remain in that severe environment, the horizontal cloud extent factor should not be applied.
(a) Engines and their inlet systems should demonstrate continuous operation in icing conditions without regard to time. Typically, an applicant assumes multiple clouds in their CPA, with a cloud extent factor equal to 1.0. This is because actual cloud extent factors are not a consideration for most engine operations, particularly in an aircraft hold pattern. This approach assures proper engine and induction system operation within the atmospheric conditions depicted in appendices C and O of part 25, appendix C of part 29, appendix D of part 33, and as experience indicates, in actual icing environments. The following are some additional factors to consider:
1. These cloud extent factors are also applicable to airframe flight profiles. Airplane applicants use the cloud extent factors depicted in the appendices during their evaluation of the straight-line flight portion of their compliance demonstration. However, engines and induction systems have historically not been limited to or evaluated against a specific aircraft flight profile when considering icing environments. Instead, they are evaluated for unlimited operation in icing.
2. While appendix C of parts 25 and 29 define supercooled clouds, appendix O of part 25 includes freezing precipitation (that is, rain and drizzle).
3. Appendix D of part 33 depicts ice crystals and mixed conditions (meaning, a mixture of supercooled water droplets and ice crystals). The cloud extent factor depicted in appendix D of part 33 relies primarily on limited data derived from the reference in paragraph
5.k. of this AC. The data covers several geographical regions including eastern Asia where ice crystal-related service events are prevalent.
9. Induction System Icing and Operation Test (§§ 33.68 and 33.89(b)).
a. Critical Point Analysis (CPA). A CPA is an analytical method utilizing engine test data to show an engine meets the certification requirements of part 33. This method derives critical test points from the collection and analysis of data on icing effects on engine performance.
Compliance with § 33.68 includes identifying, through analysis, the critical operating test points for icing within the declared operating envelope of the engine. A CPA should include consideration of the icing conditions depicted in appendix C of parts 25 and 29, and as we gain experience, in appendix D of part 33. A CPA should relate these appendices with the aircraft speed range and engine powers defined by the engine manufacturer. It should also include a prolonged flight operation in icing (for example, in-flight hold pattern), or repeat icing encounters. These combined elements within the CPA should identify the most critical operational icing conditions.
(1) The applicant (you) should ensure that analysis is supported by test data. Your CPA should also include environmental and engine operational effects on accumulation, accretion locations, and the most critical engine operating conditions for ice shed and ingestion. You may propose conditions outside the requirements of the appendices (for example, recommending conditions that are more severe based on actual service experience). You may also supplement a CPA with development test data (for example, wet and dry testing with thermocouple components).
(2) Your CPA should include ice accretion calculations that account for freezing fraction and aerodynamic effects of the ice as it moves into the air inlet. For example, water ingestion into the fan module and core inlets, water impingement rates for critical surfaces, forward aircraft airspeed effects, engine configuration effects such as inter-compressor bleed, and altitude effects (such as bypass ratio effects). The CPA should also include an energy balance of critical engine surfaces (for example, latent heat and heat of fusion effects, metal-to-ice heat transfer effects, and ice insulating effects).
(3) For anti-iced parts, the CPA should identify a critical test point determined from energy balance calculations of required heat loads, and encompassing the range of possible combinations of icing condition and engine power. In glaze ice conditions, assessing the effects of non-aerodynamic ice formations and their shedding is more complex. FAA Report number FAA-RD-77-78, titled “Engineering Summary of Powerplant Icing Technical Data”, provides additional guidance on performing a critical test point icing analysis.
b. Test Versus Analysis. The CPA is not meant to replace testing. Rather, it provides a means to predict critical test points. The CPA test points can replace the standard table points defined in Tables 1 and 2 of § 33.68, when applicants can show they are equivalent. Otherwise they supplement the standard certification test points. Today’s analysis tools are greatly improved over their predecessors; however, experience has shown the CPA is best used to predict the critical icing conditions for a given design, then use them in conjunction with the standard certification test points for certification compliance test purposes. Please note, even if the FAA concurs with a type certificate holder’s generic CPA method, this does not constitute our acceptance of the resulting critical icing test points for future certification projects. The content of an icing certification program for any given certification project will be evaluated on a case-by-case basis.
c. Test Facilities. Facility limitations or weather can delay § 33.68 compliance demonstrations. These delays may often be avoided by using national and international
icing test facilities. National and international icing test facilities for accomplishing engine icing compliance testing include the United States Air Force (USAF) McKinley Climatic Lab in Florida, the Arnold Engineering Development Center (AEDC) in Tennessee, and the Canada’s National Research Council in Ottawa, the Global Aerospace Center for Icing and Environment (GLACIER), and GE Aviation Engine Testing, Research and Development Centre (TRDC) in Winnipeg. In Europe, there is the Delegation Generale Pour L’Armement (DGA) Aero Engine test facility in France. Other test facilities are likely to become available in the future as test capabilities improve.
(1) When considering any icing facility, the capability to provide simulated cloud conditions that meet the requirements of § 33.68 should be thoroughly considered. Critical items that that you should also consider are:
(a) A description of the icing facility, including the spray system arrangement, spray nozzle(s), water flow system, airflow system, and test facility operation.
(b) Instrumentation should include facility operation monitoring, cloud simulation, and cloud property determination.
(c) Air/water flow operating map.
(d) Psychrometrics.
(e) Pre-test icing rig functional checks.
(f) Facility test procedures.
(g) Test data measurement accuracy and capability description.
d. Elements of CPA. Your CPA should address, at minimum, the following icing issues:
(1) Ice Shed Damage. Shed ice can cause engine damage if it impacts an engine surface with sufficient mass and velocity. The following types of damage are common, and you should include them in your CPA with an assessment of each:
(a) First Stage of Compressor (example, fan). Various rotating and non-rotating parts of the fan module, or first stage compressor section for non-fan engines, are susceptible to ice shed damage. For example, acoustic panels, fan rub strips, and fan blade tips are susceptible to ice shed from the inlet sensor(s), spinner, and the fan blade root.
(b) Since engine icing compliance testing is performed in test cells and not actual flight conditions, the effects of ice density, hardness, and adhesion strength as it sheds should be assessed to realistic flight conditions. For example, in realistic flight conditions the ice shed cycle for rotating surfaces such as fan blades, or first stage compressor blades for non-fan engines, is strongly influenced by the rotor speed and the adhesive strength of the ice to the surface. The adhesive strength of ice generally increases with decreasing surface temperature.
The ice thickness and the rotor speed at the time of the shed, defines the impact threat. Available data typically shows that shedding from unheated rotating and static parts can be quite variable and difficult to predict. This is due to uncertain factors such as adhesion strength properties, local thermodynamic properties, and ice structure (such as ice bridging). In some cases, applicants have utilized consistently demonstrated shedding trends in their CPA, as opposed to absolute values, since there is often uncertainty in precise shed predictions.
(c) When determining the critical conditions for fan module damage, the surface temperature, exposure time, and rotor speed are important considerations, as well as atmospheric icing conditions and scoop factor. In particular, extended operation in a holding condition power level during very cold and continuous maximum icing conditions will maximize the adhesion of ice on rotating first stage compressor or fan components. This can result in large ice accretions and resulting sheds which can damage the engine or cause power loss.
(2) Compressor Damage. When ice formations on static components shed, they often result in damage. This type of damage generally occurs on the first blade set in the high-pressure compressor (that is, intermediate pressure compressor for three spool engines, or first stage of compression for conventional turboshaft engines). Establishing the critical conditions for these glaze ice accretions requires careful consideration, since they occur at specific limited conditions of low freezing fractions, over a range of local mach numbers and air densities. The critical conditions may not occur during any of the power settings discussed in this AC (for example, flight-idle, 50% and 75% of maximum continuous or 100% maximum continuous);
consequently, you should evaluate the power setting at the critical condition. Finally, since icing is a common environmental condition, you should evaluate any engine compressor damage that results from ice testing against the possibility of multiple occurrences.
(3) Engine Operability and Compressor Rematch. Ice shed from upstream components may enter the core compressor. The presence of ice, or water from melted ice, in the gas path may cause engine component cycle changes. The engine should be capable of accelerating from minimum flight-idle to takeoff power, and should demonstrate takeoff power set procedures on the ground at any icing condition, without unacceptable power loss or power loss instabilities.
Ice sheds should not result in flameout, rollback, or surge. Any anomalous engine behavior should be raised to the FAA Engine or Aircraft Certification Office (ECO or ACO) for evaluation, and if found acceptable, it should be documented in the engine’s installation manual.
You should consider as part of the CPA, the engine accelerations and decelerations relative to operability challenges (for example, surge and stall). Assume that the minimum engine bleed schedule allowed for the condition being tested minimizes the operability margin. CPA testing should demonstrate those conditions where the minimum operability margin is expected.
(4) Core and Booster Ice Blockage. Ice accretion on internal engine vanes from glaze ice accretions may affect airflow capacity and rematch of the engine cycle. This should be considered in your CPA. For engine powers that can sustain flight, you should reconcile ice accretion through a demonstration of several ice build shed cycles. Show that there are no adverse operating effects from ice builds or sheds.
(5) Sensor Fouling. Ice accretion and blockage of control sensors can result in erroneous engine pressure and temperature measurements. Critical sensors include inlet total pressure and temperature probes, and inter-compressor temperature probes. A power loss or power loss instability could result if erroneous measurements are used by the engine control system’s control law to establish power or thrust ratings, or if used to schedule other systems required to operate the engine (for example, variable stator vanes). Critical sensors should be designed to operate when exposed to the conditions defined in 33.68 and the associated appendices, with minimal ice accretion and no erroneous measurements that would result in an unacceptable operating characteristic, such as a power loss. In accordance with 33.68 icing requirements, turboshaft engines need only be evaluated against the environmental icing conditions depicted in appendix C of part 29. The effects of installation on the local icing conditions at the probe should be accounted for in all types of applicable installations. Additionally, ice accretion on upstream sensors can shed and cause engine damage to downstream rotating hardware. You should evaluate engine inlet probes for supercooled droplet icing susceptibility, as well as ice crystal icing susceptibility.
e. Test Point(s) Selection. Test points selected for a supercooled droplet environment should address the applicable icing envelopes depicted in appendices C and O of part 25, and appendix C of part 29. Test points for an ice crystal icing environment should be representative of the meteorological conditions depicted in appendix D of part 33. Typically, the supercooled droplet test points include those defined in the regulation (either § 33.68 or § 25.1093) and any additional test points identified as part of your CPA. You should consider pertinent service experience as well as the anticipated use of the aircraft when selecting critical icing test points.
Do consider the following when constructing an icing test matrix:
(1) Section 33.68, Acceptable Means of Compliance. The engine should be capable of operating acceptably under the meteorological conditions depicted in appendices C and O of part 25, and appendix C of part 29, covering the engine operating envelope defined in § 33.68(b)(2)(ii)(B), and under the conditions of ground icing defined in § 33.68(d). Section
33.68(c) permits eliminating specific standard certification test points so long as you choose to run a similar CPA test condition that is more severe in terms of ice accretion mass at critical engine locations or to produce an equivalent level of safety to the standard test points.
(2) Rotorcraft Turboshaft Engines. Engines installed on rotorcraft present their own unique conditions. Holding phase test conditions are not applicable to turboshaft applications.
Therefore, turboshaft engine testing should include test conditions that address prolonged hover or continued operation in a relatively small area of operation. For example, prolonged operations in the local vicinity of the airport or possibly local operations around oil rigs at sea. This may require additional test, analysis, or both, to show installation configuration dilution benefits in prolonged operations, based on the aircraft induction system design.
f. Table 1 of § 33.68 Icing Conditions.
(1) The supercooled LWC shown in Table 1 of § 33.68 represents ambient icing conditions for an open inlet ground test facility or equivalent, within the inlet duct of a direct connect test
facility. An analysis of local enrichment for flight effects, as would normally be done for a CPA LWC, is not necessary for the table points.
(2) For Table 1 test conditions 1 and 2, applicants should run the engine under both rime and glaze icing conditions. You should test under both conditions for at least 10 minutes each at 100%, 75%, and 50% maximum continuous power, and for 10 minutes at a flight-idle. Follow each power setting with a snap acceleration to takeoff power. If ice is building at the end of 10 minutes at the three higher power settings, continue running the test point until the engine demonstrates stabilized building and shedding, or until the engine no longer operates satisfactorily.
(3) Table 1, conditions 1 and 2 are intended to partially represent a broad test matrix of environmental and engine operating conditions to be used when showing compliance with § 33.68. This test matrix includes power settings from idle to 100% maximum continuous. They are representative of exposure to conditions typical of both high altitude where rime ice formations occur, as well as conditions typical of low altitude where glaze ice formations often occur.
(a) Icing conditions are normally run for a minimum of 10 minutes for all engine powers that can sustain level flight, or longer if the natural engine ice shed cycle is not established. Special consideration and tests should be conducted to adequately demonstrate engine inlet screens and inlet air passages that might accumulate snow or ice due to restrictions or contours. At low power, such as idle descent power, the required time period is limited to 10 minutes. The rime icing condition should be run at an engine speed associated with top of descent operation for fixed wing aircraft, as depicted in appendix C of part 25. The glaze icing condition should be run at the minimum engine speed associated with lower altitudes at the end of the descent phase of flight for fixed wing applications. Applicants for rotorcraft applications should propose an appropriate flight profile to address this low power test period duration.
(4) Conditions 3 and 4 (holding phase). The engine should continue to meet part 33 requirements while demonstrating it can operate indefinitely in a flight hold pattern under icing conditions. Guidance on test durations and procedures to achieve this are outlined throughout this AC. The test program for turbofan and turboprop applications should include test point conditions (for example, icing conditions and power settings) to address the effects of prolonged exposure in icing conditions that are typical of in-flight holding patterns.
(a) Condition 4 in Table 1 represents a rime icing condition that is typically encountered on transport category airplanes. Condition 3 in Table 1 represents a mixed rime and glaze icing condition that was originally derived from the European Union’s JAR-E ice requirements of LWC, temperature, and droplet size. The engine and inlet should be capable of prolonged exposure to the conditions specified in Table 1. For fixed wing aircraft, a 45-minute
test exposure will typically demonstrate several ice shed cycles. You should follow each accretion point for condition 3 and 4 with a snap acceleration to takeoff power. To demonstrate unlimited icing operation, the engine should exhibit stabilized operation at the conclusion of the 45-minutes test condition, prior to the snap acceleration.
(5) For All Conditions. Engine operation in all icing conditions should be uneventful, uninterrupted, and without any significant adverse effects. It should also include the ability to continue in operation and accelerate and decelerate normally without adverse operational effects.
Some power reduction is acceptable at idle power settings in icing conditions, due to the cycle effects of pumping ice and water; however, all other operation should be unaffected.
(a) You should determine what parameters need to be monitored to determine the stabilized operation of the engine during the icing test. Engine operation is considered stabilized either when measured engine parameters are not changing, or when a regular, repeatable shed cycle has been demonstrated through the recording of measured engine parameters. Variations in measured parameters are acceptable during the performance of the ice test, as long as the longterm trend (typically the duration of several shed cycles) is stable and not trending upwards or downwards.
(b) Engines with manually activated icing protection systems (including systems for probes), should be tested. You may test these systems by stabilizing the engine for at least 2 minutes in the icing atmosphere (1-minute for turbo shaft engines) with the protection systems off (that is, prior to turning the icing protection system on). Conduct this test at flight-idle and above. Doing so simulates the delay expected for the pilot to recognize the icing condition.
(c) Fully automatic systems may use an appropriate delay. Systems that are automatic and controlled by the full authority digital electronic engine control (FADEC) do not require the 2-minute delay in the ice protection system demonstration. Where the engine’s anti-icing system relies on an ice detector to indicate the presence of icing conditions, a delay in the anti-ice selection is likely. Therefore, delayed selection testing should still be demonstrated.
(d) Stable engine operation should occur under the tested icing conditions. A stable engine operation is intended to address both stabilized ice accretion and shed cycles during steady-state engine operation. Ice accretions are considered stabilized when either ice is not forming on any engine part or the accreting ice has demonstrated a regular shed cycle when viewed by a video camera, or as indicated by instrumentation. Engine operation is considered stabilized when the measured engine parameters are not changing, or a regular and repeatable shed cycle has been demonstrated through the recording of measured engine parameters. You may provide a justification to the FAA for a performance change while operating a steady-state test point. Momentary performance changes such as a thermodynamic engine response to shed ice ingestion may be acceptable. These momentary performance changes are due to an ice shed which can momentarily affect the thermodynamic cycle.
(6) Turboprop and Turboshaft Engines Equipped with an Inlet Screen. You may perform engine icing tests utilizing the aircraft inlet system or alternately with a bellmouth. Some airframe installations utilize an inlet screen, which in some cases is included as part of the testing
for the part 33 compliance demonstration. The inlet screen compliance demonstration for icing may be addressed separately from the § 33.68 engine ice compliance testing. If addressed separately, then you should note the requirement for an engine inlet screen in the engine installation manual.
(a) If the inlet screen is not addressed under § 33.68, then compliance could be demonstrated under the installation requirements of §§ 23.1093, 25.1093, for turboprop, and §§ 27.1093, or 29.1093, for turboshaft installations. Icing testing that utilizes inlet screens should demonstrate acceptable operation, including multiple shed cycles representative of longterm or unlimited operation in icing conditions. Protruding or unusual inlet features may require specialized tests or analysis to assure they don’t adversely affect operations in icing conditions.
(b) At the conclusion of each steady-state icing test point, the engine should be accelerated to takeoff power (throttle or power lever movement of one second or less) or shedding should be induced by other methods, depending on the critical shed methodology. The throttle or power lever motion should be the most critical when considering the ice shed effects on engine operability. In some cases, a quick deceleration before accelerating to takeoff power may be more critical to the ice shed effects on engine operation. Assess this effect and account for your assessment in your test proposal.
g. Section 33.68(d) – Ground Operation. Section 33.68(d) provides two icing test points that represent freezing fog conditions; a falling or blowing snow icing encounter during ground operation; and a supercooled large droplet (SLD) test point that represents freezing rain or drizzle. The SLD condition is similar to freezing fog in terms of ambient temperature range and the LWC levels. However, the larger droplets can penetrate farther back on the surface of the engine spinner with the potential for ice shedding into the fan blades.
(1) The ground-fog icing, SLD, and falling or blowing snow demonstrations of § 33.68(d) should continue for at least 30 minutes, or until acceptable operation is demonstrated. If you cannot achieve acceptable operation then you should demonstrate the periodic engine speed runups. These run-ups will become mandatory in the engine operating manual and aircraft flight manual for operating in these icing conditions, since they were required to comply with the icing testing requirements. Turboshaft engines do not need to address SLD conditions.
(2) Falling and Blowing Snow. Service experience has shown that compressor damage occurs as a result of exposure to prolonged periods of falling snow ingestion during ground operation. Based on our review of service events we have found that airports have continued to operate with falling snow concentrations that result in 0.25 mile or less visibility (about 0.9 gm/m3of snow). While visibility in snow can be a poor indicator of precipitation rates (see reference in paragraph 5.o. of this AC), the following calculation is based on equivalent rainfall rate and gives a similar result as the 0.25 mile visibility criteria:
(a) Visibility and Snow Fall Rates. The maximum precipitation rate for moderate snow is equivalent to 2.5 millimeters per hour (mm/hr) of rainfall over approximately 30 minutes.
Using a typical fall speed for snow of 0.8 meters per second (m/s), this translates into a snow/ice concentration of 0.9 gm/m3. From a Transport Canada data set of 338,000 minutes of snowfall data, the 95% and 99% values were 2 and 4 mm/hr respectively, showing that a 2.5 mm/hr threshold provides an extreme value of snowfall rates. Holdover time tables for de-icing fluids are only good for the extreme values of moderate snow, defined as 2.5 mm/hr (or a visibility of ¼ of a statute mile). This is equivalent to the engine certification levels described here.
(b) Engine core icing service events in snow conditions confirm the critical snow accretion temperature range is 250 to 320 Fahrenheit (-40 to 00 Celsius). The engine service events have demonstrated that a snow environment is conducive to ice accretion behind the fan in front compressor stages of the engine, at low engine power. The engine test for snow conditions was developed to represent the engine conditions where snow can form accretions aft of the fan on the core inlet and first stages. You should choose a temperature within the range provided above to achieve icing behind the fan.
h. Ground Idle Demonstrations. Section 33.68 requires operation at ground idle setting for at least 30-minutes under the icing conditions defined in § 33.68(d), followed by acceleration to takeoff power or thrust. Since a broad temperature range is provided within the regulation, you should identify the most critical temperature, as determined by the CPA, and target that range.
The CPA analysis can be used to demonstrate that colder conditions are less critical to the engine’s operation than the CPA condition that is demonstrated during ground icing tests.
Turboprop and turboshaft engine inlet screen icing compliance demonstrations may be addressed separately from the § 33.68 engine ice compliance testing. If so, you should note the requirement for an engine inlet screen in the engine installation manual. We recommend you give special attention to these inlet features:
(1) Unless data is provided indicating otherwise, the minimum ground idle engine speed will be considered the most critical engine operating condition. Unless a more critical ground operating speed is identified, you should show that the engine continues to perform acceptably at its minimum ground idle speed attainable within all icing conditions. This test demonstration generally establishes the maximum allowable ground icing operation time for the engine. In addition, this test usually establishes the maximum allowable time in the engine operation instructions between engine run-ups to shed ice. This time may be prescribed as a single duration or, if supported by test conduct and results, the maximum time interval allowed between performing a repetitive, intentional, ice shedding engine run-up procedure. You should demonstrate stabilized ice accretion, either naturally or with run-ups, to achieve a ground icing taxi time which is appropriate for long taxi times in icing conditions.
(2) Engine ice shed characteristics (for example, period, extent, location(s)) during the duration of the § 33.68(d) test demonstration, should be examined (that is, by means of visual and/or suitable engine flowpath instrumentation) to determine the engine’s natural (unassisted by operator input) tendency to shed compressor system rotating and static surface ice accumulations as well as the effectiveness of any intentional (that is, prescribed, operator initiated) ice shedding
engine run-up procedure(s). Understanding these natural and, if applicable, assisted ice shedding characteristics will aid in determining the necessary ground operational limitations in icing requirements (for example, total and periodic exposure durations, prescribed procedures) for the engine.
(3) The resulting substantiated ground operating procedure in icing must be included in the engine’s operating instructions. If required, the applicable aircraft flight manual should specify the maximum total time allowed for ground icing operation and, if appropriate, the maximum time period between run-ups consistent with the compliance test demonstration.
Alternate run-up procedures are also acceptable if you can show these will result in acceptable engine acceleration to takeoff power.
i. Ice Crystal Icing (part 33 turbofan and turboprop, only).
(1) Ice crystal icing conditions that are hazardous to turbine engines are found in and around cloud formations containing deep convection, in scales ranging from isolated and organized convective cells and complexes, to tropical storms (for example, monsoons, hurricanes, typhoons). Information on the mixed phase icing condition is located in the reference shown in paragraph 5.i. of this AC. Total water content (TWC) up to 5 gm/m3 can exist at these high altitudes, composed mostly of ice crystals. The maximum size of ice particles within convective weather clouds can be up to 10 mm or larger, but mass is typically concentrated at a much smaller median mass diameter of approximately 200 microns.
(2) Appendix D of part 33 defines compliance requirements in terms of ice crystal and mixed phase icing environments. This mixed phase environment extends beyond the icing envelopes depicted in appendices C and O of part 25 or appendix C of part 29. TWC concentration levels in appendix D of part 33 are based on predicted “adiabatic” condensation rates in convective storm updrafts. This has been shown to be a reasonable upper estimate based on available flight test data, mostly from the 1950s. However, the accuracy of these flight test data is no longer adequately traceable, and the flight patterns adopted were not optimum for deriving TWC extent (cloud length) statistics. Therefore, the TWC extent probability of 99% put forth by appendix D is likely increasingly conservative with increasing cloud length. Until a more comprehensive mixed phase and glaciated atmospheric envelope is defined from flight test data using modern cloud instrumentation to determine TWC and ice crystal size measurement, appendix D of part 33 should be treated as the critical ice crystal icing condition when demonstrating compliance for § 33.68. Likewise, applicants should evaluate engine inlet probes for the conditions depicted in appendix D of part 33.
j. Test Setup Considerations. The LWC levels depicted in appendices C and O of part 25 and appendix C of part 29 are intended as supercooled droplet ambient icing conditions. Tests may be conducted with a simulated cloud which is produced outside of the inlet and ingested into the engine. Under such a test environment, the LWC within the inlet ducting should replicate the engine operating conditions of an airplane in icing conditions at actual airspeeds. The inlet icing concentration or dilution effect is dependent on droplet size, engine fan speed (for turbofan engines), and simulated forward airspeed. For example, engine operation at idle descent power with simulated forward airspeed that is less than flight speed, due to facility limitations, might require a compensating increase in the test level of LWC concentration above what is depicted in appendices C and O of part 25. This increase would be greater for larger supercooled droplet diameters. The engine size is also a variable that affects the difference in the LWC inlet concentration between flight conditions and the engine test environment, with small engines potentially needing the greatest compensation. Rotorcraft engine applications may have different criteria and applicants should propose a test that addresses those specific turboshaft installation issues.
k. Direct Connect Inlet Engine Test Facilities. Icing tests that provide a simulated icing cloud by direct connection of facility piping to the front flange of the engine, where no inlet air spillage is allowed, may cause test facility effects to alter the test parameters (for example, LWC and MVD). You must provide the FAA with data that demonstrates the simulated test conditions are representative of an installed engine operating in the icing environments as depicted in appendices C and O of part 25 and appendix C of part 29, as applicable. Suitable data may be direct measurement data concerning LWC within the inlet, or an acceptable validated analysis of water droplet trajectories for the test setup. In some cases, you may need to adjust the LWC to address any effects of the test setup (for example, non-uniformity across the engine face).
l. Testing in Mixed Phase or Glaciated Atmospheric Conditions (part 33 turboprop and turbofan engines only).
(1) Flight testing is a method of demonstrating engine operation in icing conditions, particularly mixed phase or glaciated atmospheric conditions. Under these conditions, two important flight test considerations are: the measurement of ambient meteorological data, and the ability to correlate the measured engine performance to a more severe icing point. To address the correlation, it is necessary that you have a fully instrumented engine with temperature sensors strategically located in the core flow passage. This instrumentation will collect data during the icing environment encounter. Since the environment encountered may not be as severe as the levels needed, scaling of the measured data to show satisfactory engine operation at the more severe point should be attempted to the extent possible. Any proposals for scaling should be supported by data from experiments, and the scaling approach should be conservative.
(a) An accurate measurement of the ambient meteorological condition is essential. Past engine flight tests in this environment demonstrated that a combination of LWC probe, TWC probe, particle sizing and imaging probe, ice accretion detector, and TAT probe are necessary to fully characterize the ambient environment.
(b) An onboard real-time meteorological data display with GPS positioning capability is also helpful to find high ice crystal concentration areas. This allows the pilot to position the aircraft for the test. Since testing around a thunderstorm elevates the risk level of a flight test, an on-site rapid data reduction capability would give a timely indication if the test objective is met, and minimizes the number of flights into this hazardous environment.
(2) Simulation of the Critical Mixed Phase and Glaciated Atmospheric Condition in a Ground Icing Facility. As indicated in the reference shown in paragraph 5.i. of this AC, facility simulation of ice crystal icing conditions is difficult and not currently done routinely. It is not known how well any facility simulation method actually replicates the natural environment.
Consequently, no testing standard for this icing condition currently exists. Therefore, ground simulation to demonstrate compliance to the ice crystal icing certification requirements will be evaluated on a case-by-case basis until uniform industry standards are developed.
m. Test Results and Compliance Issues. During all icing tests (that is, supercooled droplet, or ice crystal icing), the engine should operate without accumulating ice that would adversely affect engine operation. Accumulating ice can adversely affect engine operation (for example,
flameout, surge, stall, run-down, high vibrations, slow acceleration, or lack of throttle or power lever response), or can cause a sustained loss of power or thrust. Additionally, the applicant should accurately monitor icing test point conditions. Video surveillance or other instrumentation may provide the means to identify any source of ice damage, especially in those instances where test apparatus may also shed ice (for example, icing nozzles, special test instrumentation, or icing tunnel walls).
(1) Sustained Loss of Power or Thrust and Power Loss Instabilities. The engine should not experience a sustained power loss while operating at approved ratings in icing conditions.
(2) Temporary Power Loss. Temporary power losses below engine power and thrust ratings selected under § 33.8 are acceptable if you show that the engine has a sufficient margin to avert any power loss instability, such as rollback, surge, stall, high vibrations or flameout.
(3) Momentary Power Loss. Momentary power loss caused by pumping or processing of ice debris through the fan module and compressor during the ice shed ingestion process is usually acceptable. Any accepted temporary or momentary power loss or temporary high vibrations must be documented in the engine installation manual.
n. Mechanical Damage. The engine should not exhibit more than limited engine damage from any cause because of § 33.68 icing testing. Some limited damage can be acceptable if the resultant power loss is minor. Additionally you must fully account for cumulative damage from repeat encounters. Limited engine damage will be considered acceptable provided you satisfy the following criteria:
(1) Continued In-Service Use. You should evaluate any resultant damage and demonstrate that it does not affect the engine performance for continued in-service use. This includes both continued safe operation, with no imminent failures expected, and no significant power loss.
(2) Sustained Power Losses. The engine should not experience any sustained power loss beyond 1.5% (that is, the nominal accepted level considered to be within measurement capability).
(3) Temporary or Momentary Power Loss. Temporary or momentary power loss should be reviewed by both the airframe and engine manufacturer to assess installed operational impact.
Any resultant temporary or momentary power loss, or high vibrations found acceptable by the FAA during the compliance demonstration, should be recorded in the installation manual by the engine manufacturer.
(4) Validation Basis. Analytical tools used to substantiate the criteria for determining acceptable damage should demonstrate an acceptable validation basis. For example, validation could utilize engine tests or rig tests to substantiate the accuracy of results. An acceptable analytical tool validation basis includes test data which yield conservative results.
(5) Engine Damage. Damage to the engine or engine components as a result of icing compliance testing should not exceed the maintenance manual limits. Cumulative damage for