Presentation is loading. Please wait.

Presentation is loading. Please wait.

Ice Crystal Icing –Detection & Avoidance Research

Similar presentations


Presentation on theme: "Ice Crystal Icing –Detection & Avoidance Research"— Presentation transcript:

1 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

2 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 What is ICI? As AP flies through cloud, engine encounters ice crystals if it is in HIWC environment Physics - Crystals can form ice on engine surfaces warmer than freezing These warm engine surfaces are in the compressor (aft of the fan) Ice crystal accretions which shed can cause engine events such as powerloss, compressor damage TRANSITION: as part of the industry work being done for ICI Potential ice crystal accretion areas Ice shed from compressor surfaces can cause engine instability

3 Ice Crystals Characteristics
80% of all events - few or no returns at flight level Amber and red 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 Many engine events occur without flight-level weather radar returns Tropopause Freezing level Copyright © 2011 Boeing. All rights reserved.

4 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 IMC (always) Rain effect

5 General Statistics From Current Database
Latitude 32°S - 52°N, with 92% contained within the tropical and subtropical latitude band Altitudes from FL110 to FL450 with a median of FL350 Generally warmer than ISA standard day (ISA + 5 to +25) Subtropical Tropical Equator Tropical Subtropical

6 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). Oval Enhanced Region- Analyzed on IR satellite 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 IR Enhanced distance

7 Engine ICI Event Analysis
no turbulence very light turbulence light turbulence Moderate turbulence Pilot’s responses from questionnaires Using the turbulence characterization of NASA/TM Flight Tests of Turbulence Prediction and Warning System (TPAWS) - Hamilton et al, Langley Research Center

8 Oval Enhanced Region- Analyzed on IR satellite
Analysis of Frequency of Flight in HIWC Commercial Flights 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 (FL390-FL410) Results: 60 MCS with enhanced distance greater than 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

9 ICI Event Characteristics
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 Convective maximum One question industry has asked is whether engine events are occurring near the peak of convective activity Some convective clouds peak in intensity overnight, diminishing during daytime hours Have seen local overnight ocean convection peak –e.g. Darwin (regional effect) – enhancement may be specific to region I’ll delve more into detail about Japan in a minute, but I want to mention that while the region around Japan doesn’t seem to have an overnight peak, the events are still occurring near the peak of the storms 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 – a question that one could ask is: Will flight test day-time-only measurements be biased because of potential nighttime convective maxima? Based on the previous criteria, we next evaluated whether it is important to gather data for the engineering standard near the peak of convective activity SAE

10 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 , median FL 370 Cases chosen for detailed analysis Remaining cases Detailed investigation of Japan and South China Sea Regions for event location and timing 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 – similar to other events in database -7°C to -56°C SAT, median -43°C – vs median 36 C in database Evenly distributed across solar-time FL , median FL vs median FL350 in database TRANSITION: A large part of the detailed analyses for these 11 cases involved IR satellite imagery SAE

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

12 Analysis Method 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) – bright-white + 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 Analysis method ANIMATION- let run twice 11 cases chosen for detailed investigation of the lifecycle of event cloud Engine events were found to occur in fairly large MCS, with substantial cloud reaching the equilibrium level (i.e. very high cloud reaching the tropopause or near to it). The average distance traversed of such cloud was 195 km. 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) – bright-white + 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

13 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 SAE An Investigation into Location and Convective Lifecycle Trends in an Ice Crystal Icing Engine Event Database – Bravin et al

14 Results of 11 Cases Parameter Range Average Comment Enhanced Distance Traversed 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 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) Results of 11 cases Enhanced distances were similar to previous work done by Grzych and Mason published in 2010 Events tend to occur near the peak of the local ELO average, and the airplane tends to pass very close to the center of the ELO close to the time of the ELO max area Engine events most closely related in time and space to an equilibrium level overshoot. Overshoots are a reflection of the updraft regions where high IWC is expected. events were -0.8 to 3.5 after the time of maximum of a local overshoot area, averaging 0.88 hours after. Events were km from the center of a local overshoot, averaging only 41 km In 8 of 11 cases, the event aircraft passed through the area defined as the overshoot Events tend to occur at or close to the peak in ELO area SAE

15 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 Boeing database of 162 engine event properties have been updated from Mason et al. (2009) with 46 events Similar properties to 2009 except now events in Africa and South America Examined for regional differences in seasonal frequency and time of day frequency As before, events linked to regions and times with high deep convection frequency – typically local summer months Regions subject to tropical storms have extended seasons into the fall African mostly continental events have overnight maximum, possible early morning maximum in South American events Japan/South China sea area found to continue to be event-rich regions; 50 of 162 events were located there. 11 were thoroughly analyzed Long season due to spring/summer monsoon and summer/fall tropical storms Almost all tropical oceanic cases No indication of bias to higher frequency of nighttime events in spite of global observation of weak overnight maximum on oceanic convection 15ICE-0113

16


Download ppt "Ice Crystal Icing –Detection & Avoidance Research"

Similar presentations


Ads by Google