Ice Crystal Icing Detection Avoidance Research Jeanne Mason

Ice Crystal Icing –Detection & Avoidance Research Jeanne Mason, Associate Technical Fellow, Boeing Propulsion Engine Operability Melissa Bravin, Boeing Propulsion Engine Operability Erick Hall, Boeing Flight Operations Engineering Copyright © 2015 Boeing. All rights reserved. Author, 3/3/2021, Filename. ppt | 1

Ice Crystal Icing – Ice on the Engine Core Flowpath Physics: • Crystals can form ice on engine surfaces warmer than freezing • These warm engine surfaces are in the compressor (aft of the fan) Fan Core air travels downstream to the combustor Potential ice crystal accretion areas Ice shed from compressor surfaces can cause engine instability Copyright © 2015 Boeing. All rights reserved.

Ice Crystals Characteristics § 80% of all events - few or no returns at flight level § Amber and returns below the aircraft § Low updraft velocity, light to moderate turbulence § Aircraft in deepest (tallest) part of the cloud § Clouds can spread out hundreds of miles Tropopause Freezing level Copyright © 2011 Boeing. All rights reserved. Copyright © 2015 Boeing. All rights reserved. Many engine events occur without flight-level weather radar returns

Ice Crystal Icing Reports from the Flight Deck Conditions Reported: • Speckled green on weather radar • Rain on the windscreen • Small collection of ice particles on wiper post • “Shhh” sound • TAT near zero • Humid cockpit • Ozone smell • St. Elmo’s fire • Engine vibration Copyright © 2015 Boeing. All rights reserved. IMC (always) Rain effect

General Statistics From Current Database § Latitude 32°S - 52°N, with 92% contained within the tropical and subtropical latitude band § Altitudes from FL 110 to FL 450 with a median of FL 350 § Generally warmer than ISA standard day (ISA + 5 to +25) Subtropical Tropical Equator Tropical Subtropical Copyright © 2015 Boeing. All rights reserved.

HIWC Cloud Identification § Engine events typically occur when the airplane traverses greater than 100 km of significant convective anvil cloud associated with a mesoscale convective systems (MCS). • These MCS have a large, often oval region of convective anvil cloud above cruise flight levels • Within that region, there often are embedded convective core/updraft areas which push moisture up above the tropopause, resulting in overshooting cloud Copyright © 2015 Boeing. All rights reserved. Oval Enhanced Region. Analyzed on IR satellite IR Enhanced distance

Engine ICI Event Analysis no turbulence very light turbulence Moderate turbulence Pilot’s responses from questionnaires Using the turbulence characterization of NASA/TM-2012 -217337 Flight Tests of Turbulence Prediction and Warning System (TPAWS) Hamilton et al, Langley Research Center Copyright © 2015 Boeing. All rights reserved. Author, 3/3/2021, Filename. ppt |7

Commercial Flights of Flight in HIWC Analysis of Frequency Null = No Engine Event To provide insight into frequency that commercial flights potentially encounter HIWC cloud, 10, 025 flight tracks in Asia-Pacific region analyzed for large MCS encounters • None had ice crystal icing engine event • Found flights when enhanced distance greater than 100 km • Cruise level (FL 390 -FL 410) • Results: • 60 MCS with enhanced distance greater than 100 km with no adverse effect • Oval Enhanced Region. Analyzed on IR satellite IR Enhanced distance Not all big MCS cause events – Boeing is researching what is unique about clouds that do Copyright © 2015 Boeing. All rights reserved. Grzych et al |8

Convective Maximum ICI Event Characteristics § Such regions are hypothesized to be associated with local maximum in ice water content § Some convective clouds peak in intensity overnight, diminishing during daytime hours § Commercial aircraft fly during all hours of the day – Flight crews have more challenges / fewer cues to identify weather at night, since ice crystals are poor reflectors of radar energy § Flight test measurements will be taken only in day-time hours for safety reasons Copyright © 2015 Boeing. All rights reserved. SAE 2015 -01 -2130

Detailed Investigation of Japan and South China Sea Regions § 50 cases in these two regions § 11 cases have been chosen for detailed meteorological analysis § General Properties of 11 cases analyzed § From April to November (monsoon and tropical storm season) § All oceanic MCS § 9 in cruise, 2 in descent § -56°C to -7°C SAT, median -43°C § Evenly distributed across solar-time § FL 185 -400, median FL 370 Copyright © 2015 Boeing. All rights reserved. Cases chosen for detailed analysis Remaining cases SAE 2015 -01 -2130

Mesoscale Convective Systems are Like a Pot of Boiling Water Enhanced Level Overshoot -79 C High IWC MCS High IWC Tropopause or Enhanced Level -69 C Freezing level The area of cloud reaching cold temperatures can be a surrogate for high altitude ice concentrations Copyright © 2015 Boeing. All rights reserved.

Analysis Method Copyright © 2015 Boeing. All rights reserved. § Convective cloud in which event occurred was tracked from 24 hours before event to 6 hours after event § MTSAT IR images used: approximate cloud top temperatures § Moving box defined to encompass cloud – statistics inside box computed versus time § Prime statistics: § Total area of cloud colder than equilibrium ‘enhanced’ level (top altitude a rising air parcel reaches before becoming negatively buoyant) – brightwhite + red cloud at left § Total area of cloud 10°C colder; enhanced level overshoot (ELO) – red cloud at left § Red areas are indicators of center of updraft in cloud SAE 2015 -01 -2130

When Does the Engine Event Occur Relative to the Peak of the Convection Lifecycle § Study of 11 cases in Japan – South China Sea region: engine events occur in close proximity in time and space to the peak cloud area These results suggest that the airplane must be near a local convective core for an engine event Copyright © 2015 Boeing. All rights reserved. SAE 2015 -01 -2130 An Investigation into Location and Convective Lifecycle Trends in an Ice Crystal Icing Engine Event Database – Bravin et al

Results of 11 Cases Parameter Range Average Comment Enhanced Distance Traversed 72 -272 km 194 km Similar to findings of Grzych and Mason (2010) Event time – time of max enhanced (black line) -1. 2 to + 8. 6 hrs. 2. 1 hrs. Events typically after peak in enhanced level area Event time – time of max ELO (red line) -0. 8 to 3. 5 hrs. 0. 88 hrs. Events closely aligned in time to peak in area of local ELO– average < 1 hour Aircraft minimum distance to center of ELO 22 -54 km 41 km Aircraft always passed fairly close to center of an ELO, close to the time of its maximum area Aircraft penetrated ELO (red) area? 8 of 11 cases (2 of 3 that did not were in descent) Events tend to occur at or close to the peak in ELO area Copyright © 2015 Boeing. All rights reserved. SAE 2015 -01 -2130

Summary § Boeing database has been updated from Mason et al. (2006) with 46 events to current 162 events § Types of clouds, temperature levels, and length of cloud traversed similar to previous studies § Regional influences hard to identify due to low frequencies in each region § No evidence overall of a time-of-day bias in events, even though research literature reports time-of-day influences on deep convective cloud frequencies § Most significant new finding is that events occur in MCS close to a local active cell near the time of its peak intensity, identified herein as an "Enhanced Level Overshoot" (ELO). Time of the event ELO may be quite different from the time of the MCS system overall peak intensity 15 ICE-0113 Copyright © 2015 Boeing. All rights reserved.

Copyright © 2015 Boeing. All rights reserved.