AIA/EIWG Subcommittee on Engine Probe Icing:

Transcription

1

2 : A Process for Evaluating the Performance of Temperature Probes, Combined Temperature and Pressure Probes, and Static Pressure Probes in Icing Conditions 6 October 2017

3 Introduction In response to tasking by the FAA as defined in a letter dated July 3, 2013, titled, Request Formation of Advisory Group to Address Specific Engine and Installation Icing Issues, the Engine Icing Working Group (EIWG) has studied the issue of engine probe icing and its effect on engines and airframe systems. Guidance to evaluate the performance of Pitot (total pressure) probes and Pitot static (combined total and static pressure) probes in icing conditions exists in the form of SAE Aerospace Standard AS5562 and FAA TSO C16. In order to address temperature probes, combined temperature and pressure probes, or static pressure probes, the EIWG has developed a process which can be used to evaluate probe behavior in icing conditions and the resulting impact on engines and airframe systems. The original FAA letter addressed only engine inlet probes, however the process that has been developed can be used on any temperature probe, combined pressure/temperature probe, or static pressure probe regardless of its installation location. This report will be provided to the FAA in response to their tasking request for use in developing policy and guidance for engine probe icing. The FAA has participated on this AIA committee; however, conclusions stated within this report do not necessarily represent the views of the FAA. Once this report is submitted to the FAA, the FAA has stated that they will review the final conclusions, respond to the recommendations and make a decision as to how to proceed. Executive Summary The EIWG has developed a structured methodology for showing compliance with the requirements related to the performance of temperature probes, combined temperature and pressure probes, or static pressure probes in icing conditions. The process details areas for applicants to target in their analysis and describes how applicants can develop appropriate test points tailored to their application to show compliance with the engine and airframe requirements related to the performance of the probes in icing conditions. The EIWG recommends the FAA incorporate the process into policy and guidance on showing compliance to the regulations related to probe icing. Background The industry has experienced multiple cases of ice crystals blocking turbine engine air data (temperature and pressure) probes on small to medium-sized turbine powered airplanes in the recent past. These ice blockages have resulted in engine power rollbacks, lack of throttle response, and flight crew warnings for out-of-range probe data. Some airplane manufacturers have eliminated certain airframe air data probes and, instead use the engine probe data for airframe systems. Eliminating dedicated airframe probes and using engine inlet probe data for key airframe systems may result in the need for greater emphasis on icing compliance and hazard safety analysis to support overall aircraft safety. Eliminating the airframe probes also removes the ability of the engine control system to use the airframe signal as the classical referee, that is, the engine uses the airframe signal to validate data from the engine probes. Without the use of an airframe probe, the engine sensor is substantially more critical in measuring air data throughout the operating envelope of the aircraft. Page 2 of 45

4 Some engine manufacturers use heated inlet probes to prevent ice from accreting on or in the probe, while others use unheated probes. Heated probes can be effective in preventing ice from accreting on the probe while in supercooled droplet icing conditions that are typically found at low to mid-altitudes. Heated probes, however, can be susceptible to ice accretion from ice crystal icing conditions. These conditions are typically found at high altitude but can exist at lower altitudes. Applicants have redesigned certain heated temperature probes to reduce potential for ice crystal blockage but have not been able to completely eliminate it. Unheated probes are less susceptible to ice crystal icing conditions, but are more susceptible to lower and mid-altitude supercooled droplet icing conditions. Therefore, whether manufacturers use heated or unheated probes, problems with ice accretion can still occur. Engine manufacturers have incorporated Full Authority Digital Engine Control (FADEC) logic to detect and annunciate probe blockage and compensate for the probe error to maintain acceptable engine operation, but this logic has not been able to completely eliminate all issues. Discussion No specific rule targets the turbine engine air data probes, however they are indirectly addressed within 14 CFR sections 33.28, 33.68, and when part of the engine type design and within 14 CFR sections (b), (b), (b) and (b) when part of the aircraft installation. Historically, engine air data probes were tested during the engine s induction system icing compliance testing for supercooled water droplets, as defined by Appendix C of part 25. They have also been subjected to DO-160 standards and test procedures during testing, but they have not addressed the effects of ice crystals on inlet probes. Note that testing completed at the probe level does not relieve the engine manufacturer from meeting the requirements of at the engine level. Amendment to 14 CFR part 33 and Amendment to 14 CFR part 25 introduced requirements for the engine and airframe systems to operate in mixed phase and ice crystal icing conditions (defined by Appendix D to part 33) and supercooled large drop (SLD) icing, (defined by Appendix O to part 25). Similar regulations are included in EASA CS-E and CS-25, including identical icing envelopes for ice crystal icing (defined in CS-25 Appendix P) and SLD icing (defined in CS-25 Appendix O). The FAA provides guidance in advisory circular (AC) A for demonstrating compliance with the engine induction system icing and engine installation icing requirements of 14 CFR parts 23, 25, 27, 29, and 33. While AC A discusses where probes could be susceptible to icing and the need to evaluate them, it does not provide specific guidance on how to do so. SAE Aerospace Standard AS5562 and FAA TSO C16b provide guidance on means to evaluate electrically heated Pitot and Pitot-static probes for in-flight icing conditions, however there is no guidance available on evaluating temperature probes. EASA regulation CS-E 780 includes similar requirements to the icing requirements of 14 CFR part 33. The related guidance included in AMC E 780 directs the applicant to determine the critical probe icing conditions using the guidance of AMC of CS-25. EASA regulation CS requires each flight instrument external probes system, including, but not limited to temperature probes, to be heated or have an equivalent means of preventing malfunction in the heavy rain conditions of table 1 of CS and in icing conditions. AMC provides some guidance related to the test setup and the conditions to be tested to show compliance with the regulation. Page 3 of 45

5 The EIWG has worked to develop a structured methodology for showing compliance with the referenced FAA requirements. The process details areas for an applicant to target in their analysis and describes how an applicant can develop appropriate test points tailored to their application to show compliance with the requirements related to probe icing. The process is applicable to any temperature or pressure probe that provides data to the electronic engine control system or other airframe systems, whether the probe is installed in the engine or on the airframe. The process is shown graphically in Figure 1. A detailed explanation of each step in the process follows. 1 Start Compare the Engine & Aircraft Operating Envelopes with the Icing Envelopes Consider 1) the Atmospheric condition and extent aircraft will operate in; and 2) the possible failure modes of probes, including the loss of heater function. 2 Determine Ambient Conditions Calculate water flux as a function of total temperature 6 Determine How to Address Candidate Test Points Determine: 1) which points can be covered by other test conditions; 2) which points can actually be tested (and modify conditions to allow testing if facility limitation is why test can t be completed); and 3) which points can be covered by analysis/test 7 Complete Testing/Analysis Provide Engine Response to Probe Icing to Airframe OEM Change in thrust 2. Changes in control schedules 3. Failure indications/fault accommodation 4. Changes in displayed parameters 5. probe signal error 12 Determine Effect on Aircraft 1.Effect of signal error on airframe systems 2. Effect of engine response (change in thrust / fault accommodation/etc.) 3 Determine Local Conditions at Probe Based on Ambient Conditions Scoop Factor Calculations for inlet mounted probe, concentrations effect, local velocities, etc. for airframe mounted probe. 8 Compile Test/Analysis Results Quantify effect of test conditions on probe: 1. signal error (temperature, pressure, etc.) including transient response rate including any build/shed cycle effects and when icing conditions are introduced or removed 2. accretion/shedding size and shape of ice build up 13 Aircraft Effect Acceptable? Yes No 4 Identify Candidate Test Points Based on local conditions at probe High IWC/short duration Lower IWC/longer duration Include AC A Table Points for completeness Etc. 9 Provide Results to Engine OEM OEM determine acceptability of probe response on engine control 1) corruption due to icing 2) any other failure due to icing (including Turbo machinery damage from probe icing build/shed cycle) Include in safety assessment for 33.28: 1) effect of loss of heater function 14 Completion Engine OEM incorporate engine response into final installation manual Airframe OEM document aircraft response and compliance with Installation Manual requirements as required 5 Conduct System Level Analysis Probe/Engine/Airframe OEMs develop pass/fail criteria 10 Engine effect acceptable? No Yes Figure 1 Process Flow Chart Page 4 of 45

6 Step 1: Compare the Engine & Aircraft Operating Envelopes with the Icing Envelopes The initial step requires defining the engine and aircraft operating envelopes (altitude, airspeed, temperature). The part 25 Appendix C and part 33 Appendix D icing envelopes need to be compared to the aircraft operating envelope to determine applicability of those envelopes. The part 25 Appendix O icing envelope should also be evaluated as required to support engine and/or aircraft certification. Due to the tendency of probes to react in much shorter time frames than ETOPS conditions, no unique evaluation is required for ETOPS airplanes. In support of engine and airframe safety analyses, possible failure modes of the proposed probe should be considered at this stage as well. For example, if the probe is a heated probe, failure of the heater element should be considered. These failure modes do not necessarily need to be evaluated as part of the test program, but they do need to be addressed so the installer of the probe understands the impacts of failure modes of the probe. The integrity of the data provided from the probe should also be evaluated so the airframe and engine manufacturers can determine what systems the probe is adequate for use in (i.e. systems with catastrophic safety effects will require a more robust probe or multiple data sources, whereas systems with only minor safety effects can use a less robust probe). Step 2: Determine Ambient Conditions The envelopes of part 25 Appendices C and O and part 33 Appendix D define ambient or free stream concentrations of liquid water and ice crystals. Based on these concentrations and the aircraft operating speeds, water flux as a function of total temperature can be calculated. For glaciated testing the highest water flux cases tend to be the most severe. This is not necessarily the case for super-cooled liquid water or mixed phase. Other parameters to consider in selecting critical points include highest total cooling load, minimum predicted surface temperature, maximum water to air mass flux ratio, etc. Refer to Appendix 1 for an example of the process of defining ambient conditions to be evaluated. Within the ice crystal envelope defined by part 33 Appendix D, total water content (TWC) in g/m 3 has been determined based upon the adiabatic lapse defined by the convective rise of 90% relative humidity air from sea level to higher altitudes and scaled by a factor of 0.65 to a standard cloud length of 17.4 nautical miles. In-service occurrences show that several air data probe (both temperature and pressure probe) icing events in glaciated conditions occurred outside of the Appendix D envelope in terms of altitude and outside air temperature. In that context, the environment to be considered should be Appendix D enlarged to encompass ISA +30 C conditions and should be extended to a minimum temperature of -70 C. In addition, the Appendix D envelope should be expanded to cover ISA -5 C conditions above 25,000 ft. This expanded Appendix D envelope and a sampling of conditions where air data probes have experienced in-service ice crystal icing events is shown in Figure 2. Page 5 of 45

7 Ambient Temperature ( C) Appendix D Envelope Appendix D Expansion Industry Reported Probe Icing Events ISA+30 ISA+20 ISA Altitude (ft) Figure 2 Expanded Convective Cloud Ice Crystal Envelope In addition, based on data available from multiple sources, EASA AMC suggests that the standard cloud of 17.4 NM and the associated average TWC concentration values provided by part 33 Appendix D may not provide the most conservative conditions for air data probe testing. The max or peak TWC concentration values should be considered instead of the 17.4 NM values provided by part 33 Appendix D. These max or peak values are available in FAA document DOT/FAA/AR-09/13 and correspond to the 17.4 NM values multiplied by a factor of (1/0.65). Step 3: Determine Local Conditions at the Probe Based on Ambient Conditions Based on the ambient conditions determined in Step 2, the local conditions at the probe should be determined. Probes are typically mounted a sufficient distance from the mounting surface (e.g. the fuselage skin or engine inlet) to accurately sense the freestream parameter of interest (total temperature, total pressure, etc.). However, when flying through particles such as supercooled water droplets, ice crystals or rain, there can be a concentration effect near the mounting surface. This concentration effect is primarily due to inertia and drag effects, but can also be affected by large particles which splash and/or break up yet remain in proximity to the boundary layer. This effect is highly installation dependent and it can vary significantly depending on probe location and probe design. Page 6 of 45

8 The determination of local conditions involves reviewing the installed position of the probe (e.g. in the engine inlet or on the airframe) and determining local flow velocities and concentration effects for airframe mounted probes or scoop factor effects for inlet mounted probes. AC 20-73A presents some of the methods that have been used to calculate drop impingement and water catch at the location of interest for supercooled liquid water cases. For ice crystal icing conditions, there are currently no accepted methods for calculating ice crystal trajectories or the effects of ice crystal bouncing and break up on the local concentration levels at the point of interest; however, preliminary research indicates the ice crystal concentration can be significantly higher than ambient conditions (reference SAE papers and ), so conservative assumptions should be made in regard to local ice water content for ice crystal icing conditions in accordance with the state of the art analytical tools. The assumptions made regarding concentration factors should be documented. Similar analysis would need to be completed to determine local conditions for probes installed in other places, such as behind the fan on a turbofan engine or in the core of the engine. The local conditions may be significantly different than the free stream or ambient conditions, including local probe angle of attack, Mach number, pressure, temperature, etc. Step 4: Identify Candidate Test Points Based on the local conditions at the probe, the candidate test points can be determined. An example of one method to determine the candidate test points for ice crystal icing conditions is included in Appendix 1. A range of temperatures representative of the icing envelopes should be tested for the probes. The probe design should be reviewed to determine whether the probe is more susceptible to higher water concentrations for short durations, to lower water concentrations for longer durations, or to cyclic conditions, such as those defined in Test Point 3 of Table 1 (Amendment 33-34). As noted in Step 2, the maximum ambient water concentrations should account for peak TWC conditions corresponding to the values of part 33 Appendix D multiplied by Testing at the peak TWC values is consistent with the guidelines of EASA AMC In addition to the peak TWC conditions, the maximum ambient water concentrations can be reduced and the duration extended to address conditions where lower water concentrations for longer durations are critical. Testing at different water concentrations also provides evidence that the installed probe will work throughout the operating envelope and not just at the maximum level. The peak concentration can be scaled to a standard (17.4 nm) cloud by dividing the peak TWC values by It is also recommended that a concentration of one half of the peak value be evaluated. From Appendix D, Figure D3 the scale factor for the standard cloud is 1, and the scale factor for one half of the peak value is ½*1.538 = Based on Appendix D Figure D3, this scale factor corresponds to a cloud extent of approximately 215 nm. Review of in-service engine ice crystal icing events (reference SAE paper ) shows that the 99 th percentile cloud length for events where engine damage occurred is 354 nm (657 km), with the large majority of engine events occurring in clouds of less than 215 nm (400 km) in length. Page 7 of 45

9 Experience has shown that in general, probes do not take as long to respond to ice crystal icing as engines do, and therefore it is reasonable to reduce the maximum cloud length for a probe to 215 nm and use the corresponding TWC of one half the peak value. The duration of each test point also needs to be considered. AMC and AS5562 both require test conditions in ice crystal icing conditions be run for a minimum of 2 minutes. In general, this is a sufficient length of time to evaluate the probe s behavior. However, for cases where the probe builds ice, the duration should be extended as required to quantify the size of ice which may accrete and to ensure any build/shed cycles are completely evaluated. For the reduced concentration points, the tests should be run for a sufficient length of time to traverse a 17.4 nm or 215 nm cloud as appropriate to the test condition. For typical cruise airspeeds, traversing a 17.4 nm cloud takes on the order of 2 to 3 minutes. For these cases, the conditions completed with the peak TWC for 2 minutes is a more severe test than 2 to 3 minutes at a reduced TWC, and the lower TWC case can therefore be eliminated from consideration. For typical cruise airspeeds, traversing a 215 nm cloud takes approximately 25 minutes. To ensure the test adequately addresses the concerns, the one half of peak TWC conditions should be run for 30 minutes (17.4 nm / 2 minutes = 8.7 nm/minute, therefore 215/8.7 = 24.7 minutes rounded up to 30 minutes). In regards to liquid water icing testing, for engine mounted components test/analysis points should be identified per the conditions stated in Table 1 of This requires 10-minute duration glaze ice and rime ice tests at various power settings (airflows) as well as 45-minute glaze and rime ice holding conditions. For airframe mounted temperature probes, test/analysis points should be identified per the conditions stated in Tables 1 and 2 of EASA AMC of CS-25. This requires 15 minute durations per continuous maximum cloud concentrations, 5 minute durations per intermittent maximum cloud concentrations and either 30 minute or 10 minute duration cyclical liquid water concentrations. For both engine mounted and airframe mounted probes, in addition to the above requirements a critical point analysis (CPA) should be performed based on the Appendix C requirements to determine if there are additional critical points within the operating envelope that should be considered for test/analysis. The critical points which the engine is tested against to show compliance with may or may not be critical points for the probe. For example, high airflow conditions like maximum continuous power may be more critical for an inlet mounted probe than for the engine as a whole. It is therefore important for the applicant to determine the critical conditions for the probe as part of this step and not rely solely on the points for demonstrating proper operation of the probe. The points defined in and AMC represent conditions at the engine inlet for the installation. For a component (probe) test, the appropriate concentration factor needs to be determined so that the local conditions at the probe match what it will see as installed. For unheated probes in liquid water environments, the applicant should evaluate the impact of ice accumulation and shedding in regards to altitude. Analysis and/or testing at altitude should be considered. For heated probes, it is accepted to test liquid water conditions in a nonpressure controlled wind tunnel since this is considered more conservative than testing at higher altitudes. Page 8 of 45

10 In regards to mixed liquid and ice crystal conditions, test/analysis points should be selected with the guidance stated in section 12.2 of EASA AMC of CS-25. This specifies the total water content (TWC) should be based on a 2.6 nautical mile cloud. One clarification needed is that the AMC does not specify how to designate the amount of liquid water vs ice crystal contents. Appendix D defines the liquid water portion of mixed phase conditions to be 1.0 g/m 3 for clouds of less than 50 nautical mile extents for temperatures above -20 C and zero for temperatures below -20 C. SAE Aerospace Standard AS5562 assumes the liquid water content is per the Appendix C intermittent maximum cloud and the remainder of the TWC is ice crystals. The EIWG s recommendation is to align with SAE Aerospace Standard AS5562 and assume a liquid water content (LWC) per the Appendix C intermittent maximum cloud and the balance of the TWC to be ice crystals for all mixed phase conditions. This recommendation results in consistency with AS5562 and conservatively results in mixed phase conditions in colder conditions with more liquid water present than the Appendix D requirements. An example of one method to determine the candidate test points for mixed phase conditions is included in Appendix 1. The duration for mixed phase icing conditions also needs to be considered. For the maximum TWC in mixed phase conditions, it is appropriate to run test conditions for 2 minutes as this is sufficient time for an air data probe to reach a steady state and stabilized condition. In addition, flight testing in well developed, large diameter mesoscale convective systems completed as part of the HAIC/HIWC flight test campaigns showed a low frequency of mixed phase regions. At -10 C, mixed phase regions amount to less than about 5% of the total in-cloud distance traversed and maximum average distances across mixed phase zones were about 8 nautical miles. Frequency of mixed phase zones decreased with decreasing temperature. At -30 C, the spatial fraction of mixed phase zones was less than 0.2%, and maximum average distance across mixed phase zones was about 1.9 nautical miles. It is believed that these welldeveloped storm cells produce large amounts of falling and recirculating ice that would tend to glaciate any new updraft, thus resulting in lower liquid water contents, and smaller regions of mixed phase conditions. For smaller or still developing cells with less glaciation and less circulation, regions of mixed phase conditions with higher liquid water content could exist for longer times. These areas of mixed phase conditions are still likely to exist as separate regions within the storm, resulting in alternating between mixed phase and fully glaciated conditions as the storm is traversed. A conservative approach to represent these smaller or fresh convective cells is to perform a cyclic test alternating between mixed phase and fully glaciated conditions. Therefore, in addition to the test conditions for maximum TWC conditions for two minutes, tests should be completed cycling between mixed phase and fully glaciated conditions. For these cyclic conditions, the TWC should be set to one half of the 2.6 nautical mile scaled TWC. This lower TWC is justifiable as the HAIC/HIWC flight test results indicate that extended regions of liquid water seem less likely in high IWC conditions. Therefore, lower IWC, and resulting lower TWC, values are required for mixed phase conditions to be sustained without transitioning to fully glaciated conditions. The LWC for the mixed phase conditions in a developing storm cell is not yet well understood. In order to define a conservative test, the LWC should be assumed to be the Appendix C Intermittent Maximum value for the conditions to be tested. As discussed in Steps 10 and 13, other engine or airframe level mitigation may be necessary to ensure Page 9 of 45

11 acceptable aircraft operation depending on the test results in these conservative conditions. Each cycle should alternate between 2 minutes in mixed phase conditions and 2 minutes in fully glaciated conditions. To simulate flying through a developing storm system or holding in such conditions, the cycles should be continued until repetitive, stabilized operation has been shown, or for a maximum of 30 minutes. For the case of a heated probe, test conditions with the heater turned off should be considered, whether to address an inadvertent type encounter where the heater isn t turned on immediately or to address a transient power interruption. The timing should be coordinated with the engine and airframe manufacturers to ensure appropriate conditions are defined. Step 5: Conduct a System Level Analysis A system level analysis should be jointly completed by the probe, engine and airframe manufacturers to define the pass/fail criteria for the probe when exposed to the candidate test points determined in Step 4. The probe remaining free of ice is not necessarily the pass criteria. The acceptance criteria need to consider the criticality of the probe and the air data it provides to the engine and the airframe. A simplified acceptance criterion is for the engine and airframe systems that utilize the data from the probe to continue to operate within some acceptable range in icing conditions. The probe manufacturer can typically only quantify what the effect is at the probe level (e.g. the probe output/accuracy is ±x when exposed to a specific condition). The engine manufacturer needs to translate that probe effect into the resulting system impact at the engine level (e.g. ±x of measured temperature error equates to ±y thrust or to some effect on engine operability) The airframe manufacturer needs to know the probe level effect (e.g. ±x measured temperature error) and the installation level effect on airframe systems that use data from the probe along with the engine effect and the impact the engine effect has at the aircraft level. Shedding of any ice that accretes on the probe should also be evaluated as part of the pass/fail criteria to ensure no damage to downstream components occurs. The acceptance criteria need to consider the types of conditions and the possible failure modes of probes. Classic failure mode of a temperature probe in ice crystal icing conditions is to clog with partially melted crystals driving the probe reading to 0 C (the temperature of the slush). Other failure modes exist, such as smaller errors for some short period of time or fluctuations in the probe output with build and shed cycles. Damage to downstream components as a result of ice shedding. Step 6: Determine How to Address Candidate Test Points An analysis should be completed of the candidate points identified in Step 4. The points should be evaluated to determine which points are covered by other conditions, or can be covered by analysis supported by testing at other conditions. Page 10 of 45

12 The points should also be evaluated to determine if they can be tested at the proposed test facility. When a particular test condition cannot be achieved due to a test facility limitation, the scaling methods of AS5562 sections through may be used to vary temperature, airspeed and water flux to define test conditions which can be achieved. If a significant number of test points need to be modified in this manner, an alternate test facility capable of testing more of the proposed envelope, should be considered so as to minimize the number of candidate points which are not directly tested. As noted in AS5562 section there is no acceptable method to scale the altitude for ice crystal icing conditions at this time. Therefore, ice crystal icing conditions must be tested at the altitude determined in Step 4. It may be possible to change the probe design such that any points that cannot be tested are no longer critical. Alternatively, it may be possible to impose limitations on the probe s operating envelope. However, this may require engine or aircraft level design changes or limitations. Any limitations on the probe installation must be clearly identified and communicated to the engine and aircraft manufacturers so those limitations can be adhered to as installed. As stated in Step 1 above, the system level effects of the loss of heater function should also be considered. This may be addressed either in a test matrix or by analysis. This activity is intended to document the effect on the probe with the heater inoperative in the worst case icing conditions being evaluated. Step 7: Complete Testing/Analysis Testing of all conditions identified as needing to be tested in Step 6 should be completed in a facility capable of meeting the test conditions. The configuration of the test article should match the intended installation including orientation as closely as possible, and the probe itself should match the type design configuration, except as necessary to install instrumentation or other test equipment. The following requirements should be met for each test condition. Where noted and emphasized in italics, the requirements were copied from SAE AS5562, the requirements are reprinted with permission SAE AS5562 Copyright 2017 SAE International. Further distribution of this material is not permitted without prior permission from SAE. 1. Probe Mounting Location (AS ) The location of the probe within the calibrated test section shall be selected such that the LWC and/or IWC at the surfaces of interest meet or exceed the required test condition. 2. Probe Mount Heating Requirements (AS ) If it is necessary to heat the probe mounting arrangement in the icing wind tunnel, the mount heating system shall be designed so as not to invalidate the test results. 3. Installation Heat Sink Effects (AS ) The probe shall be mounted to an aluminum heat sink of 0.10 inches (2.5 mm) in thickness and 100 square inches (645 cm2), or any required configuration of equivalent thermal capacity, which will be exposed to the ambient environment of the icing wind tunnel. Page 11 of 45

13 Note: if specific installation data is available, the as-tested heat sink effects should reflect the conditions that would exist in the installation. 4. Probe Power for Electrically Heated Probes (AS ) A probe heater voltage shall be specified for use in all test conditions. The value of the probe voltage utilized shall be recorded in the test results and form part of the test record. NOTE: It is commonly accepted to test the probe at 10% below the nominal rated voltage to cover voltage variation at the probe location on aircraft during normal aircraft operation. NOTE: The probe test voltage value shall be provided to the installer to support installation approval. 5. Non-Electrically Heated Probes For probes which are heated using some method other than an electrical heater, the heat source should be set to the minimum allowable value expected for the installation. For example, a probe heated using bleed air should be tested with the bleed air supply at the minimum expected regulation pressure and temperature. The specific conditions tested regarding the air supply shall be provided to the installer to support installation approval. It is important to fully characterize this air supply including: in-line pressure drop, piping clearances, local heat transfer characteristics and any control orifices in the supply line and/or sensor. 6. Tunnel Blockage (AS ) The percent blockage, using the projected frontal area of the UUT [Unit Under Test], of the icing tunnel test section shall be assessed to verify it does not invalidate the test results. 7. Angle of Attack The effect of angle of attack (AOA) in the intended installation should be considered. If it is determined that the angle of attack will have no significant impact, all test conditions can be run at a single angle of attack. If it is determined the angle of attack could have an impact, or if it cannot be determined if it would, the probe shall be tested at angles of attack of -15, 0 and +15 for each icing test condition, or if nominal angle of attack is known, nominal - 15, nominal, and nominal + 15 for each icing condition. 8. Data Collection Sample Rate (AS ) The data collection sample rate for all tests shall be a minimum of 20 Hz. 9. Electrically Heated Probe Test Unit Selection (AS ) Probe qualification tests shall be performed on a unit fitted with a heater circuit having the lowest electrical performance acceptable on a production article as defined by the acceptance test procedure. The test voltage may be adjusted to simulate the lowest performing probe. 10. Non- Electrically Heated Probe Test Unit Selection Page 12 of 45

14 Probe qualification tests shall be performed on a unit having the lowest performance acceptable on a production article as defined by the acceptance test procedure. The inputs may be adjusted to simulate the lowest performing probe (e.g. a bleed air heated probe may have the air pressure and/or air temperature adjusted to simulate the lowest performing probe heater). 11. Test Duration Ice crystal conditions tests completed at the peak TWC values shall be run for 2 minutes. Tests completed at one half of the peak TWC value shall be run for 30 minutes. For mixed phase condition tests, tests completed at the maximumtwc values shall be run for 2 minutes, and tests completed at the reduced TWC values shall be run as cyclic tests. Each cycle should alternate between 2 minutes in mixed phase conditions and 2 minutes in fully glaciated conditions. The cycles should be continued until repetitive, stabilized operation has been shown, or for a maximum of 30 minutes For liquid water icing conditions, test/analysis points should be identified per the conditions stated in Tables 1 and 2 of EASA AMC of CS-25. This requires 15 minute durations per continuous maximum cloud concentrations, 5 minute durations per intermittent maximum cloud concentrations and either 30 minute or 10 minute duration cyclical liquid water concentrations as defined in the AMC. 12. Test Particle Size Distribution The particle size MMD (Median Mass Dimension) for the test conditions shall match that defined by 14 CFR part 33 Appendix D, unless it can be justified that a different size will not have a significant effect on the test. Step 8: Compile Test/Analysis Results The test and analysis results should quantify the effect of the conditions on the probe including: 1. Signal error (temperature, pressure, etc.) including transient response rate and including any build/shed cycle effects and when icing conditions are introduced or removed. 2. Measurement accuracy and any changes when icing conditions are introduced or removed, and any changes when the heater is turned on. 3. Accretion/shedding - size and shape of ice buildup. 4. Frequency of build/shed cycles and impact on signal error and the transient probe response to shedding Step 9: Provide Results to Engine OEM The results of the testing and analysis should be provided to the engine manufacturer so that the acceptability of the probe response can be determined. This evaluation should be based on Page 13 of 45

15 corruption of the probe output signal due to icing and any other failure mode seen during the testing. Step 10: Determine Acceptability of Engine Effects Based on the results of the testing and analysis, the engine manufacturer should determine whether the effect of the probe response to the icing conditions is acceptable or not. If the response is acceptable, proceed to Step 11, if not, the pass/fail criteria should be reviewed to ensure an acceptable engine response is defined and the process should be repeated from Step 1. The revised system analysis may require review of the candidate test points and/or design changes to the probe or installation. It is important to note that the pass/fail criteria should not simply be changed to match the results, rather the system analysis step should be repeated to ensure correctness and the design reviewed or changed as necessary. Uncertainty regarding test data validity: For an engine inlet probe, static testing outside of the engine inlet system may not exhibit the same build and shed behavior due to variations in airflow and vibration levels. Results from an integrated inlet/engine test are likely to be quite different due to changes in vibration and local airflow, and likely to result in differences in shedding behavior compared to an isolated probe in an icing tunnel. Step 11: Provide Engine Response to Probe Icing to Airframe OEM The engine manufacturer should document the response of the engine based on the probe response to the icing conditions tested and provide that response to the airframe manufacturer. The engine response evaluation should address: 1. Probe Signal error 2. Failure indications/fault accommodation 3. Changes in engine operating characteristics (surge/stall, flameout, etc.) due to signal error 4. Change in thrust or power setting 5. Change in displayed parameters Step 12: Determine Effect on Aircraft The airframer should review the data provided by the engine manufacturer in Step 11 to determine the impact to the aircraft. Step 13: Determine Acceptability of Aircraft Effects Based on the results of the testing and analysis, the airframe manufacturer should determine whether the effect of the probe response to the icing conditions is acceptable at an aircraft level or not. If the response is acceptable, proceed to Step 14, if not, the pass/fail criteria should be reviewed to ensure an acceptable aircraft response is defined and the process should be repeated from Step 1. The revised system analysis may require review of the candidate test points and/or design changes to the probe or installation, and/or re-evaluation of the aircraft response. Page 14 of 45

16 Step 14: Document the Acceptability of Aircraft Effects Once the aircraft effect has been determined to be acceptable, final documentation of the probe, engine and aircraft response to the icing environments should be completed. This includes the engine manufacturer incorporating information about the engine response into the engine installation instructions provided under This information should include the data provided to the airframe OEM in Step 11, a description of the icing environments evaluated and any other pertinent data from the safety assessments required by 33.28(e) and 33.75(a) describing the probe and engine response. The airframe manufacturer should document the response and compliance with any requirements included in the engine installation instructions as required. This can include information included in the Airplane Flight Manual, the system safety documentation for the aircraft, compliance documents, etc. Conclusions and Recommendations The EIWG has developed a process to evaluate probe behavior in icing conditions and the resulting impact on engine and airframe systems. The EIWG proposes that this process can be used to show compliance with the engine and airplane-level requirements of temperature probes, combined temperature and pressure probes or static pressure probes in the icing conditions of supercooled water, mixed phase and ice crystal icing, and SLD icing, as defined in the respective envelopes of 14 CFR part 25 Appendix C, 14 CFR part 33 Appendix D, and 14 CFR part 25 Appendix O. This process is applicable to any temperature probe, combined temperature and pressure probe, or static pressure probe that provides data to the engine control system or any airframe system, whether the probe is installed in the engine or on the airframe. The 14 steps of this process are, in summary: 1. Compare the Engine & Aircraft Operating Envelopes with the Icing Envelopes 2. Determine Ambient Conditions 3. Determine Local Conditions at the Probe 4. Identify Candidate Test Points 5. Conduct a System Level Analysis 6. Determine How to Address Candidate Test Points 7. Complete Testing/Analysis 8. Compile Test/Analysis Results 9. Provide Results to Engine OEM 10. Determine Acceptability of Engine Effects 11. Provide Engine Response to Airframe OEM 12. Determine Effect on Aircraft 13. Determine Acceptability of Aircraft Effects 14. Document the Acceptability of Aircraft Effects The EIWG proposes that the FAA review this process for incorporation into policy and guidance on showing compliance to the regulations applicable to temperature probes, combined temperature and pressure probes, or static pressure probes when operating in all types of icing Page 15 of 45

17 conditions. The EIWG further proposes that the FAA review this process with EASA to achieve harmonization with EASA policy and guidance. In addition, the EIWG proposes expanding the Appendix D envelope to ISA +30 C and ISA -5 C above 25,000 ft with both the ISA +30 C and ISA -5 C extended to a minimum temperature of -70 C as shown in Figure 2. Any changes made to the Appendix D envelope regarding altitude/temperature envelope or total water content levels need to be incorporated into this document and any FAA guidance developed from this document. It is the EIWG s opinion that the mixed phase test conditions defined in Step 4 are conservative in terms of water concentrations and durations of the conditions, however no data exists to justify using any lower concentration or duration. The EIWG recommends additional research be conducted to understand the prevalence of mixed phase conditions, and the total and liquid water concentrations in mixed phase conditions to ensure these conditions are not overly conservative. Additionally, there are currently no accepted methods for calculating ice crystal trajectories or the effects of ice crystal bouncing and break up on the local concentration levels at the point of interest in ice crystal icing conditions. The EIWG recommends additional research be conducted to provide additional data in order to determine acceptable methodologies and/or assumptions with regards to local ice water content for ice crystal icing conditions. References and Related Material 1. FAA Letter to Aerospace Industries Association, Request Formation of Advisory Group to Address Specific Engine and Installation Icing Issues dated 3 July SAE AS5562, Ice and Rain Minimum Qualification Standards for Pitot and Pitot-static Probes, issued 08/ FAA TSO C16b, ELECTRICALLY HEATED PITOT AND PITOT-STATIC TUBES, dated January 27, FAA Advisory Circular AC A Turbojet, Turboprop, Turboshaft and Turbofan Engine Induction System Icing and Ice Ingestion, dated October 22, EASA CS-E Amendment 4, AMC E EASA AMC , Flight instrument external probes, Amendment No: 25/16 7. DOT/FAA/AR-09/13 Technical Compendium from Meetings of the Engine Harmonization Working Group, March FAA Advisory Circular AC 20-73A, Aircraft Ice Protection, dated August 16, SAE International An Analysis of Turbofan Inlet Water and Ice Concentration Effects in Icing Conditions, S. Liao, X. Liu and M. Feulner 10. SAE International Ice Crystal Ingestion In a Turbofan Engine, M. Feulner, S. Liao, B. Rose and X. Liu 11. SAE International Studies of Cloud Characteristics Related to Jet Engine Ice Crystal Icing Utilizing Infrared Satellite Imagery, T. Tritz, J. Mason, M. Bravin, and A. Sharpsten Page 16 of 45

18 Appendix 1: Sample Candidate Test Point Identification Using Water Mass Flux as the Critical Parameter The following is an example application to illustrate the process of a critical point analysis, assuming that water mass flux is the parameter of criticality. A CPA using water mass flux as the critical parameter will naturally select the highest airspeeds within the flight envelope. However, these types of conditions may not be the most critical for all probe designs and/or types of icing conditions. Therefore, it is recommended that other parameters besides water mass flux be considered. These may include, for example, maximum total cooling load, minimum dynamic pressure, maximum water-to-air mass flux ratio and minimum anti-icing heater power available. Consideration of other parameters of criticality will tend to result in a wide range of airspeeds considered. In addition, the CPA may need to consider a range of additional influences that are not detailed herein, for example: - Inlet scoop factor - Engine power setting - Droplet size In order to determine the water flux a probe will be exposed to in service, the ambient conditions to be evaluated need to be defined. The following process defines a method of determining the ambient conditions for ice crystal and mixed phased icing based on the aircraft operating conditions. First, the ice crystal conditions are identified and then the process is repeated for mixed phase conditions. The ice crystal icing envelope is defined as a function of altitude and temperature in Appendix D of 14 CFR part 33 as shown in Figure 3. Page 17 of 45

19 Ambient Temperature ( C) Altitude (ft) Figure 3 Part 33 Appendix D Convective Cloud Ice Crystal Envelope Within the ice crystal envelope, total water content (TWC) in g/m 3 has been determined based upon the adiabatic lapse defined by the convective rise of 90% relative humidity air from sea level to higher altitudes and scaled by a factor of 0.65 to a standard cloud length of 17.4 nautical miles. To determine the peak TWC, the values from part 33 Appendix D are multiplied by (1/0.65). Figure 4 displays peak TWC values over a range of ambient temperature within the boundaries of the ice crystal envelope. Note that Figure 4 also includes an isothermal line of total water content for -70 C which is not included in part 33 Appendix D. The -70 C curve was obtained by extrapolating the other curves to a -70 C condition. The TWC data of Figure 4 is also presented in tabular form in Table 1. Page 18 of 45

20 TWC (g/m3) 10 Peak TWC Levels: Adiabatic Lapse from Sea 90% Relative Humidity C -10 C -20 C -30 C -40 C -50 C -60 C -70 C Altitude (ft) Figure 4 Total Water Content Peak Values Table 1 Total Water Content Peak Values Ambient Temperature Altitude (ft) 0 C -10 C -20 C -30 C -40 C -50 C -60 C -70 C Page 19 of 45

21 Ambient Temperature ( C) Ambient Temperature Altitude (ft) 0 C -10 C -20 C -30 C -40 C -50 C -60 C -70 C In service events show that several probe icing events (both temperature and pressure probes) in glaciated conditions have occurred outside of the 14 CFR part 33 Appendix D domain in terms of altitude and outside air temperature. Furthermore, a reported event occurred at a temperature of -70 C. In that context, the convective cloud ice crystal envelope should be enlarged to encompass ISA +30 C conditions and to extend to -70 C. In addition, the Appendix D envelope should be expanded to cover ISA -5 C conditions above 25,000 ft. This expanded envelope is in Figure Appendix D Envelope -10 Appendix D Expansion Altitude (ft) Figure 5 Expanded Convective Cloud Ice Crystal Envelope Page 20 of 45

22 TWC (g/m3) Figure 5 can be overlaid on Figure 4 to show the relationship of maximum water content as a function of altitude and temperature, including the expansion of the Appendix D envelope, as shown in Figure 6 10 Peak TWC Levels As a Function of Altitude and Temperature Altitude (ft) 0 C -10 C -20 C -30 C -40 C -50 C -60 C -70 C Appendix D Temperature/Altitude Envelope Appendix D Expansion Figure 6 Peak Total Water Content Levels as a Function of Altitude and Temperature The peak values of each TWC curve can also be plotted to show the peak TWC as a function of temperature, as shown in Figure 7. Page 21 of 45

23 Appendix D Peak TWC Appendix D Peak TWC Extrapolated to -70 C 7.00 TWC (g/m 3 ) Ambient Temperature ( C) Figure 7 Maximum Total Water Content as a Function of Temperature Points of interest for the analysis can be plotted on the altitude/temperature envelope. The points of interest should include the most critical conditions where icing is likely to occur. For this example, it is assumed that those critical conditions occur at the corners of the envelope within the design operating envelope of the aircraft. In addition to the corner points of the envelope, several points in the middle of the envelope are also selected for completeness. The points selected for this example are plotted in Figure 8. Note that the highlighted points at higher altitude conditions have been selected to correspond to the peak total water content for the selected ambient temperature rather than following the edge of the temperature/altitude envelope. When selecting these points, the analysis may need to address other considerations in addition to the peak total water content, for example, system failure modes that impact engine core airflow. The completed analysis should provide an evaluation of the total threat relative to the target operating envelope and determine the most threatening conditions based on that evaluation. Page 22 of 45

24 Ambient Temperature ( C) 10 0 Appendix D Envelope Appendix D Expansion Points for Analysis Points correspond to peak TWC for given ambient temperature Altitude (ft) Figure 8 Selected Ice Crystal Points for Analysis The points can then be plotted on the total water content versus altitude envelope of Figure 6 to determine the total water content at each point; this is shown in Figure 9. Page 23 of 45

25 TWC (g/m3) Altitude (ft) 0 C -10 C -20 C -30 C -40 C -50 C -60 C -70 C Appendix D Temperature/Altitude Envelope Appendix D Expansion Selected Points for Analysis Figure 9 Total Water Content for Selected Points The points represented on Figure 9 give the altitude (ambient pressure), temperature and total water content for the points of interest. These parameters are summarized in Table 2. Table 2 Selected Ice Crystal Points for Analysis (Altitude, Temperature, TWC) Altitude Ts Peak TWC Point # Ft C g/m Page 24 of 45