Document
National Aeronautics and Space Administration www.nasa.gov
May 21, 2014
Aircraft Loss of Control: Research and Technology Directions
University of Missouri Seminar Series
Assuring Safe and Effective Control under Hazardous Conditions Christine Belcastro, Senior Research Engineer & Project Scientist NASA Langley Research Center
LOC Hazards Analysis Vehicle Dynamics Modeling Guidance, Control, & Systems Validation
– – – –
Introduction: Aircraft Loss of Control (LOC) Research Approach Selected Research Results Future Research Directions Summary & Concluding Remarks
• • • • •
Outline
60,000 lbs.
≥ (LOC-I) Loss of Control Inflight Other Categories can Contribute Note: These statistics apply to jet transport aircraft CAST – Commercial Aviation Safety Team / ICAO – International Civil Aviation Organization
LOC Overview: Relevance to Accident Statistics
What is Aircraft LOC and what Causes it?
Note: This Aircraft (and many others involved in LOC accidents) < 60,000 lbs Crashed During Approach to Landing (instrument landing system approach) Aircraft: Bombardier DHC-8-400, N200WQ; Date: February 12, 2009 Clarence Center, New York (Near Buffalo-Niagara International Airport)
• • •
LOC Example: Colgan Air Flight 3407
flight state vehicle : dive) : response system, rates, system improper component abnormal disturbances or loading, upsets thunderstorms, spiral flight sensor / : control control angular ‐ inaction impairment trajectory errors recovery, & engines / / energy, rain night) maneuver, configuration, non improper vehicle engine, and and disturbances atmospheric flight vortices (including & controlled haze, & action fatigue or airspeed, / attitude, conditions vehicle airfoil, moving) dynamics ineffective and wake turbulence, aggressive from crew or airframe failures, crew (fog descent to forces, component, input, aircraft weather icing, vehicle attitude, / shear, dynamics instrumentation, of (fixed impairment faults, hazards onboard visibility Inappropriate contaminated damage Control deck Loss awareness, control procedure, wind snow » » » » inclement vehicle system inappropriate poor obstacle abnormal abnormal asymmetric uncontrolled stall/departure – – – – – – – – – – Adverse External Abnormal • • •
Causal & Contributing Factors
LOC Usually Results from Multiple Causal & Contributing Factors of / loss 3) ‐ vehicle or (1 vehicle / to response condition qualities into Stall handling above control
Causes
Primary
of effectiveness dynamic and flying (including asymmetric effects) the upset (e.g., Departure) 2. Reduction 3. Changes 4. Combinations 1. Entry and Exhibit, Providence, Rhode Island, 16-19 August / to of be (or rates to variable and effects is (adjusted behavior pilot more aircraft cannot by coupling acceptable large state qualities by predictable angular that or unrecoverable inability of and (i.e, altitude, inputs be small nonlinear high the one divergent 2 inertial envelopes altered longer / to in / not 1, by by by handling motion outside
LOC inputs
no flight controlled is is heading, phase) need result tolerances normal level to ‐ path control degrade aircraft LOC predictably pilot) kinematic disproportionately responses changes, oscillatory flight following:
Characteristics
− − − for pilot response the that displacements, maintain wings tracking predictably autoflight system) AIAA Atmospheric Flight Mechanics Conference 2004.
LOC Problem Definition
LOC: characterized the • outside • not • characterized • likely • characterized • flight Note: Wilborn, J. E. and Foster, J. V., “Defining Commercial Aircraft Loss-of-Control: a Quantitative Approach,” Lambregts, A. A., Nesemeier, G., Wilborn, J.E., Newman, R. L., “Airplane Upsets: Old Problem, New Issues,” AIAA Modeling and Simulation Technologies Confernece and Exhibit, Honolulu, Hawaii, 18-21 August 2008 1. 2.
Flight Safe Systems Onboard Enhanced Transition Technologies Technology under LOC Hazards Attitudes Trajectory under LOC Hazards & Models Enhanced Improved Crew Training Upset Vehicle Representative Recover Simulation Methods Technology Products Abnormal Abnormal Stall/Departure • • • Improved Situational Awareness, Guidance & Control Distraction Visibility) / Complexity) (Poor Testing (System Awareness Crew Response Mitigate Inappropriate Disorientation Confusion Situational Guidance, Control & Systems Poor Spatial Mode Evaluation Requirements • • • Simulation Fault, Disturbance / Unsafe Comprehensive Technology Evaluations Hazard Detect Damage Technology Development & Evaluation Impairment, LOC RWG (Industry, Government, Academia) Safe Vehicle Problem Analysis / External Hazard LOC Test Scenarios Vehicle Failure, External Vehicle Dynamics Modeling Unsafe • • Holistic Approach to Breaking LOC Precursor Sequences: Flight Normal Training Analysis Precursor Crew LOC Hazards Prevent Risks Improved under Conditions Problem Definition LOC • Crew Action / Inaction Vehicle / System Environmental / Atmospheric Vehicle Upsets New / Emergent Upsets • • • •
LOC Research Approach
Abnormal Flight / Upset Hazards Extreme Attitudes Abnormal Energy States Abnormal Control Response Stall / Departure Objectives: individual precursors worst-case precursor combinations precursor sequencing » » » Events that involved loss of control regardless of official classification Wide spectrum of commercial transport aircraft (at or above 12,500 lbs, jets & props) Team consensus process Evaluation of causal & contributing factors (or precursors) should include Based on current trends Determined by CAST Based on & correlated to accident / incident analysis & future risks Any additional conditions needed for resilience testing
– – – – – – – –
Establish Analysis Team (NASA, NTSB, NIA, STI, MIT)* Define an extensive accident (and incident) set over a recent 15-year time period Perform a thorough analysis of this accident / incident set Identify future LOC risks Develop a comprehensive set of LOC test scenarios * LOC Analysis Team: C. Belcastro & J. Foster (NASA), L. Groff & D. Crider (NTSB), R. Newman (NIA), D. Klyde (STI), A. Huston (MIT) 1. 2. 3. 4. 5.
LOC Hazards Analysis (1)
LOC Accident / Incident Data Set
Failure by crew to maintain control, Weather encounters, Abrupt maneuvers, and Reduced control capability due to equipment malfunction or failure • • • • Aircraft Accident Reports on DVD (R. Dorsett, 2006) Australian Transport Safety Bureau (ATSB) Aviation Safety Network (ASN) Canadian Transportation Safety Board (TSB) Flightglobal (Ascend Database) French Bureau d'Enquêtes et d'Analyses pour la sécurité de l'aviation civile (BEA) German Bundesstelle für Flugunfalluntersuchung (BFU) International Civil Aviation Organization (ICAO) Irish Air Accident Investigation Unit (AAIU) National Transportation Safety Board (NTSB) “loss-of-control” “upset” “unusual attitude” “stall” “uncontrolled” – – – – – – – – – – – – – – – Data Sources Search Criteria Resulted in Broader LOC Accident Set than LOC-I, Including Accidents & Incidents Involving:
• •
LOC Hazards Analysis (2)
78 70 5803 1234 7185 615 339 2224 3858 7185 Fatalities On-Board Fatalities On-Board 87 28 17 143 275 Events 38 96 42 44 50 Events Total Total Operation Aircraft Accident Set is Provided in Appendix A LOC Events by Type of Operation: Classification LOC Events by Aircraft Classification: Wide-body Turbojets Narrow-body Turbojets Business Jets Turboprop Transports Piston Transports Commuter Airplanes of the 2014 SciTech Paper (see Refs in Backup) Scheduled Airlines Non-Scheduled Non-Revenue Operations Executive Transportation 0 2 69 77 37 15 270 156 152 374 1241 1697 2008 1087 7185 2938 2143 2104 7185 On-Board Fatalities On-Board Fatalities
7185 Onboard Fatalities, 235 Ground Fatalities
6 8 3 9 8 5 2 275 Accidents and Incidents (1996 – 2010)
79 43 41 17 22 34 18 99 74 275 102 275 Events Events LOC Events by Phase of Flight: LOC Events by 5-Year Intervals: Total Timeframe 1996 to 2000 2001 to 2005 2006 to 2010 Flight Regime Takeoff Initial Climb Climb Cruise Descent Holding Approach VFR Pattern Circling Final Approach Landing Go-around Missed Approach Maneuvering Total
LOC Hazards Analysis (3)
Stall / Departure Abnormal Attitude Abnormal Airspeed Uncontrolled Descent Uncommanded Motions Abnormal Angular Rates Abnormal Flight Trajectory Oscillatory Vehicle Response Abnormal Control for Trim / Flight Undesired Abrupt Dynamic Response Abnormal / Counterintuitive Control Response Abnormal Vehicle Dynamics & Upsets Abnormal Vehicle Dynamics Vehicle Upset Conditions Night Fixed Moving Fog / Haze Turbulence Wind Shear Snow / Icing Wake Vortex Precursors / Hazards Precursor Categories Precursor Sub-Categories Thunderstorms / Rain External Hazards & Disturbances Inclement Weather & Atmospheric Disturbances Poor Visibility Obstacle Engine Sensor Engine Damage Control Component Improper Procedure Contaminated Airfoil Aggressive Maneuver System Operational Error Improper Loading (Cargo) Crew Fatigue / Impairment Flight Deck Instrumentation Airframe Structural Damage Improper / Ineffective Recovery Improper Loading (Weight / CG) Loss of Energy State Awareness Loss of Attitude State Awareness System / Sub-System (non-control) Inappropriate Vehicle Configuration Abnormal / Inadvertent Control Input Inadequate Crew Resource Monitoring Lack of Aircraft / System State Awareness Adverse Onboard Conditions Vehicle Impairment System & Component Failure / Malfunction Ineffective Crew Action / Inaction
LOC Hazards Analysis (4)
% 8.4 4.3 94.0 35.8 43.8 61.8 42.2 24.2 18.4 75.4 74.0 601 312 6750 2576 3150 4444 3036 1741 1324 5416 5315 Fatalities % 5.8 87.3 31.3 42.6 58.2 36.7 23.6 10.9 80.0 17.1 68.4 86 65 30 16 47 240 117 160 101 220 188 Incidents Accidents / Obstacle Malfunctions Disturbances Poor Visibility
Category & Sub-Category Level
Individual LOC Hazards Statistics: Vehicle Impairment Vehicle Upset Conditions Abnormal Vehicle Dynamics System & Component Failures / Inclement Weather & Atmospheric Individual Precursor Contributions are Provided in the 2014 SciTech Paper Inappropriate Crew Action / Inaction Hazard Category / Sub-Category Adverse Onboard Conditions External Hazards & Disturbances Abnormal Dynamics & Vehicle Upset Conditions
LOC Hazards Analysis (5)
19 Accidents Disturbances External Hazards & – More – 199 – 299 – 499 – 999 – 99 Fatalities 0 1 100 200 300 500 59 Accidents 58 Accidents
Worst-Case Hazards Combinations (1)
to is Size Accidents Preliminary Worst-Case Analysis Performed at Sub-Category Level 32 Accidents of Proportional Sphere Number Directly Adverse Onboard Conditions None / Abnormal Unknown Dynamics & Vehicle Upsets
LOC Hazards Analysis (6)
Vehicle Upset Abnormal Dynamics Explored at Combination Hazards Level Disturbances External Hazards & – More – 199 – 299 – 499 – 999 – 99 Fatalities 0 1 100 200 300 500 to is
Worst-Case Hazards Combinations (2) Size
Accidents of Proportional Sphere Number Directly Adverse Onboard Conditions None / Abnormal Unknown Dynamics & Vehicle Upsets
LOC Hazards Analysis (7)
Vehicle Upset Abnormal Dynamics – 199 – 299 – 499 – 999 – More – 99 Fatalities 0 1 100 200 300 500 Poor Visibility
Precursor Combinations for
to is
Crew Action / Inaction – Poor Visibility – Vehicle Upset
Size Accidents of Proportional Sphere Number Directly Inappropriate Crew Action / Inaction
LOC Hazards Analysis (8)
Vehicle Upsets Stall / Departure Uncont. Descent Abnorm. Traject. Undes. Abrpt. Resp.. Abnorm. Ang. Rates Abnorm. Airspeed Abnorm. Attitude - 0 0 0 0 0 0 0 0 3 0 3 3
7th
- 3 0 0 3 1 0 0 1 1 11 10 15
6th
- 3 1 6 0 0 0 0 4 10 33 29 43
5th
- 4 5 2 1 0 1 8 39 30 55 47
4th
- 4 1 2 1 88 11 10 67 78 14 64
3rd
- 6 6 4 32 35 86 16 89 23 66
2nd
9 0 0 0 41 84 42 86 58 19 22
1st 167 275
TOTALS Obstacle Malfunctions Disturbances Poor Visibility Vehicle Impairment Vehicle Upset Conditions
Temporal Sequencing: Category & Sub-Category Totals
Precursor
Precursor Sequence Information is Provided in the 2014 SciTech Paper Abnormal Vehicle Dynamics System & Component Failures / Inclement Weather & Atmospheric Inappropriate Crew Action / Inaction Adverse Onboard Conditions External Hazards & Disturbances Abnormal Dynamics & Vehicle Upset Conditions Unknown Precipitating Events
LOC Hazards Analysis (9)
3 3 1 1 0 529 143 152 Fatalities 1 1 2 1 1 1 1 Events LOC
→
Upset Conditions Vehicle LOC LOC
→ → →
Crew Upset Upset Inaction / Conditions Conditions Vehicle Vehicle Action LOC Inappropriate LOC LOC LOC
→ → → → → → →
Crew Crew Crew Vehicle Upset Upset & Inaction Inaction Inaction Abnormal / / / Conditions Conditions Conditions Vehicle Vehicle Action Action Action Upset Inappropriate Inappropriate Inappropriate Unknown Dynamics LOC
→ → → → → → →
Crew & Failures Upset Vehicle Weather Inaction / Dynamics System Conditions Atmospheric Malfunctions Disturbances Vehicle & / Action Abnormal Inclement Inappropriate Component
→ → → → →
Sequence Diagrams – Example: Initiated by Poor Visibility
Total Poor Visibility Visibility → FLIGHT Poor NORMAL Sequence Diagrams at Category & Sub-category Level are Provided in Appendix B of the 2014 SciTech Paper
LOC Hazards Analysis (10)
LOC Recovery Improper / Ineffective Stall / Departure State Energy Awareness / Inappropriate Management Loss of Energy
Colgan Air 3407 (2/12/2009)
Example Sequence at Precursor Level: Crew Fatigue / Impairment Normal Flight
LOC Hazards Analysis (11)
execute they were full thrust; procedures Recovery too late that and applied and altitude, the autopilot for recovery Improper / losing speed Ineffective disconnected crew failed to Crew realized Stall / activated Departure Stick shaker Energy Loss of Pilots got co-pilot ASI the ASI mis- pilot and co- warning, and Inadequate Awareness / match between by stick shaker while getting an Energy State Management confused due to pilot, decreasing excessive speed excessive speed warning followed advisory Airspeed" Failure / and "Mach decreasing; warning was while Co-pilot "Rudder ratio" imminent stall Instrument warnings were issued to crew Flight Deck Malfunction ASI read 200 kts excessive speed followed by stick shaker indicating kts) Error / System Autopilot / Autothrottle and reduced 350 kts (when Operational Inadequacy inappropriately erroneous ASI of increased pitch-up airspeed based on actual ASI was 220 2/6/1996: Birgenair 301 (B-757) En Route, Example Detailed Precursor Sequence: flight) Sensor / Failure / Malfunction
Near Puerto Plata, Dom. Republic (189 Fatalities)
Sensor System (ASI) was not working working; Incorrect ASI by a blocked pitot tube for 3-4 days prior to this falsely high ASI reading; Pilot's air speed indicator properly and resulted in a readings possibly caused (which was left uncovered Co-pilot's ASI seemed to be Night factor were a Unclear conditions contributing whether night Action / Inaction Improper for 3-4 days
LOC Hazards Analysis (12)
left uncovered Maintenance Pitot tubes were prior to the flight
LOC test scenarios also provide engineering simulation requirements
10 International Databases Searched for LOC Commercial transports at or above 12,500 lbs 275 accidents & incidents identified resulting in 7185 fatalities Based on Six Accident / Incident Subsets (45-46 Events) Individual Precursor Statistics Worst-Case Precursor Combinations Temporal Sequencing Re-Evaluation Based on Team Consensus Approach (In Progress) Definition of Future LOC Risks (To be Coordinated with CAST / ATLAS) Development of LOC Test Scenarios Final Results to be Submitted for NASA TP and Journal Publication
−
– – – – – – – – – – – Analysis results and test scenarios can be used in the development and evaluation of technology solutions for LOC prevention and recovery (e.g., Onboard Systems) Potential for wider application of this research to broader LOC solutions Comprehensive Set of Accidents / Incidents Compiled for 1996 – 2010 Preliminary Analysis Results Obtained Ongoing Research
• •
• • •
LOC Hazards Analysis Summary
AIAA , National Harbor, AIAA Guidance, Navigation and SciTech Forum , , Toronto, 2010.
, Toronto, August 2-5, 2010.
, Minneapolis, Minnesota, August 2012.
, Minneapolis, Minnesota, August 2012.
http://www.boeing.com/news/techissues/pdf/statsum.pdf AIAA Conference on Guidance, Navigation, and Control . (To be submitted in 2014) AIAA Guidance, Navigation and Control Conference , Toronto, August 2-5, 2010 AIAA Guidance, Navigation and Control Conference AIAA Conference on Guidance, Navigation and Control “Statistical Summary of Commercial Jet Airplane Accidents, Worldwide Operations, 1959-2011”, Boeing Commercial Airplanes, July 2012. URL: Evans, Joni K., “An Examination of In Flight Loss of Control Events During 1988–2004”, Alliant Techsystems, Inc., NASA Langley Research Center, Contract No.: TEAMS:NNL07AM99T/R1C0, Task No. 5.2, 2007. Wilborn, J. E. and Foster, J. V., “Defining Commercial Aircraft Loss-of-Control: a Quantitative Approach,” Atmospheric Flight Mechanics Conference and Exhibit, AIAA, Providence, Rhode Island, 16-19 August 2004 Belcastro, Christine M. and Foster, John V.: Aircraft Loss-of-Control Accident Analysis; Control Conference Belcastro, Christine M., Groff, Loren, Newman, Richard L., Foster, John V., Crider, Dennis A., Klyde, David H., and Huston, A. McCall, “Preliminary Analysis of Aircraft Loss of Control Accidents: Worst Case Precursor Combinations and Temporal Sequencing”, Maryland, January 2014. Belcastro, Christine M. and Jacobson, Steven: Future Integrated Systems Concept for Preventing Aircraft Loss-of- Control Accidents; Belcastro, Christine M., “Loss of Control Prevention and Recovery: Onboard Guidance, Control, and Systems Technologies,” Belcastro, Christine M.: Validation and Verification of Future Integrated Safety-Critical Systems Operating under Off- Nominal Conditions; Belcastro, Christine M., “Validation of Safety-Critical Systems for Aircraft Loss-of-Control Prevention and Recovery,” AIAA Guidance, Navigation, and Control Conference Belcastro, Christine M., Groff, Loren, Newman, Richard L., Foster, John V., Crider, Dennis A., Klyde, David H., and Huston, A. McCall, “Aircraft Loss of Control Analysis and Test Scenarios for Technology Development and Validation”, NASA Technical Paper • • • • • • • • • •
LOC Problem Publications
to
/
is Size Accidents of
More Proportional External
– 199 – 299 ‐ Hazards
Fatalities – 99 Sphere
Disturbances
0 1 100 200 300 Number Directly & Crew Actions / Inactions +/- Upsets (Terminal Area) Future High-Density Operations Control Component Failures Icing Effects Wakes / Wind Shear
Vehicle/Environment-Related Hazards • • •
Conditions
Onboard
Analysis of 64 Accidents with 2821 Fatalities from 2000 – 2009 (10 Years) Attitude / Energy
Adverse Asym
/ Descent
Upset
− Attitude Trajectory Departure Unknown Loss of Aircraft State Awareness Spatial Disorientation Rates / / /
Multiple Hazards Guidance, Mitigation, & Upset Prevention / Recovery
Crew-Related Hazards • •
Conditions Va
Stall
Initial LOC Hazards Prioritization
None
Vehicle
Abnormal Ab. Abnormal Uncontrolled
Simulation
Upsets Impairment Disturbances
LOC
Vehicle Vehicle External
Integrated
Models
Propulsion
&
Structure Angle of Attack, deg Open Loop Fan Speed Control EPR Control Aerodynamics Airframe x 10
Database
2.4 2.5 2.6 2.7 2.8 2.9 Net Thrust, lbf NRA Partner: Boeing Flight
Representative Modeling Research Approach
Testing
Acquisition CFD
tunnel Subscale / Testing
Data
Wind Laboratory Experimental
Integrated
Class-Representative Integrated LOC Simulations (Upsets, Impairment, External Disturbances)
Vehicle Dynamics Modeling Technologies (VDMT) for Characterizing Effects of LOC Hazards
External Disturbances Propulsion Atmospheric Disturbance CFD Vehicle Impairment LOC Hazards Aerodynamics Integrated Real-Time LOC Simulation Structural / Aeroelastic Vehicle Upsets Experimental Multidisciplinary Modeling of Flight Dynamics Effects Airframe Engine Airframe (Partnered with MVS) Engine Control Components Collateral Damage Effects Engine » » » » » » » Aerodynamic Effects Engine Effects Airframe Structure Effects Icing Effects (Partnered with AEST) Damage System Failures Wind Shear Wake Vortices Turbulence Multidisciplinary Hazard Effects Multiple Hazards that can Lead to LOC Integrated Multidisciplinary Real-Time Simulation Provide Means of Capturing Vehicle-Level Effects – – – – – – – – – – – Vehicle Upset Modeling Vehicle Impairment Modeling Integration of Existing Atmospheric Disturbance Models Integrated Real-Time Simulation Development
VDMT Accomplishments Summary
• • • •
Research Objectives
» Hazards detection and flight safety impacts assessment » Upset prevention, detection and recovery » Multiple hazards mitigation (system failures, icing, wakes / wind shear) » Improved situational awareness specific to LOC » Anticipatory guidance for LOC prevention » Control cueing for recovery » Analysis » Simulation » Experimental Testing Develop an integrated system architecture and technologies that provide Develop preliminary crew interface concepts that provide Evaluate GCST technologies with support by CTE technologies
– – –
Goal: Develop and evaluate onboard systems technologies that provide improved real-time situational awareness, guidance, and control under hazards that can lead to LOC Research Objectives: Guidance, Control, and Systems Technologies (GCST) for Safe & Effective Control under LOC Hazards • • SSCI & Load MVS: UM AEST: Effects Engine) Barron for Icing NASA / Structural (Airframe & Real-Time NASA / Estimation Detection / UIUC, UM Icing Models Identification NRA – UTSI, NRA - UIUC / SBIR – NRA – UIUC, Assuring Safe Control (ASC) NASA / NRA – UIUC & UM Maintaining Vehicle Safety (MVS) Enables Upset Prevention / Detection / Recovery under Multiple Hazards SBIR - Barron NASA / Crew Decision Making (CDM) CDM (SE 207) SBIR – Barron UM (Upset Detection) NASA / NRA – UIUC & NRA – UTSI / UIUC Guidance, Control, and Systems Technologies (GCST) for Safe & Effective Control under LOC Hazards NRA: University of Illinois (UIUC), University of Michigan (UM), University of Tennessee (UTSI); SBIR: Scientific Systems (SSCI), Barron Associates (Barron) NASA / NRA – UIUC LOC Prediction NASA / NRA – UIUC Resilient Flight Control NASA / NASA / Protection NRA – UIUC NRA – UIUC
Multiple Hazards Mitigation, LOC Prediction & Dynamic Envelope Protection
Pilot Interfaces Dynamic Envelope
UIUC Technical Approach: iReCoVeR
aircraft
impaired
violated
is
accretion
it
if
ice
possibly
under
Prediction:
or
augmentation
the
of
LOC
envelope
margins
control
and
impairment
envelope
envelope
adaptive
operational
(DFEP)
vehicle
L1
safe
maneuverability
e.g.
an
(FED):
operational
the
operational
(FDI):
and
to
DFEP
the
RFC
safe
to
of
Protection
improved
its
(RFC): conditions,
aircraft
conditions
with
controller
Isolation
conditions
the
within
estimate
envelope
adverse
and
reconfiguring Determination
Envelope of
Control
flight
faults/failures
stays
baseline
for
return
a
impairment
stabilization isolate
accurate
of
Flight
Flight
signal estimate
aircraft
and
term
Envelope
‐
Detection
Challenging Moderate Vehicle
− − −
Ensure Automatically Short Consisting Detect Provide Determine Provide
Dynamic Fault Flight
Resilient
− − − − − − − −
• • •
•
UIUC Research Objectives
of of
envelopes envelopes
3 3
AIAA Atmospheric
Excursion Excursion
least) least)
LoC = LoC =
(at (at
Control: a Quantitative Approach,” ‐ of ‐ , Providence, RI, August 2004.
Flight Mechanics Conference and Exhibit [1] Wilborn, J. E. and Foster, J. V., “Defining Commercial Aircraft Loss
LOC Prediction & Flight Envelope Protection (1)
laws; highest ϴ‐ : specific modifying limits; the for protection by adjusted with limiting control ‐ total has FEP be & , protection speed β can , factor s p limits Command protection C*, energy Dynamic Hierarchical architecture longitudinal system; Speed integrated protection. Protect ϴ‐ Limits online Load priority
(simplified)
› › › ›
protection
envelope
flight
dynamic
for
architecture
Control
LOC Prediction & Flight Envelope Protection (2)
#2 350 350 scheme) 300 300 Foster) FEP limits in 250 250 LoC envelopes EAS [kts] EAS [kts] LoC prediction) FEP EAS [kts] used 200 200 LoC envelopes (for (Wilborn & 150 150 Extended (limits 0 1 2 3 0 1 2 3 -1 -1 0.5 1.5 2.5 0.5 1.5 2.5 -0.5 -0.5 z z z n [g] n [g] n [g] #1 60 60 40 40 [%] [%] 20 20 lat lat δ 0 0 0 [deg] [deg] [deg]
φ Example
-20 -20 Lateral control -50 -40 -40 Roll Control -60 -100 -60 0 0 0 10 20 30 50 10 20 30 -20 -10 -50 -20 -10 -100 [deg] [deg] θ [deg] Dynamic Roll φ ’ [deg] Dynamic Roll ' [deg] 10 10
Illustrative
[%] 5 5 [%] lon δ lon 0 0 [deg] 0 [deg] [deg] β Pitch control -5 -5 -50 Pitch Control -10 -10 -100 0 2 4 6 8 0 2 4 6 8 -6 -4 -2 -6 -4 -2 10 12 14 0 10 12 14 10 20 30 40 50 60 [deg] -40 -30 -20 -10 [deg] Dynamic Pitch ' [deg] α [deg] Dynamic Pitch θ ’ [deg] Limit Adjustment Logic:
‐
Keep count of the number of LoC and extended LoC envelope excursions. Adjust FEP limits depending on the number of exceeded LoC envelopes.
› ›
LoC Prediction: FEP
LOC Prediction & Flight Envelope Protection (3)
up;
‐
sideslip build
‐
of
‐
pilot sets directly EPR; pilot commands altitude changes; pilot commands speed changes.
− − −
term speed stability
‐
rate & AoSS flight control law
‐
axis roll
‐
path control with long
‐
vector roll maneuvers without angle
‐
Baseline Robust Flight Control Law (Non-Adaptive)
precise flight commensurate with industry standard (e.g., Boeing) Directional Control Augmentation System: velocity coordinated turns at zero sideslip and crosswind flight. Mode 0: Normal operational mode Mode 1: Specific potential energy Mode 2: Specific kinetic energy
‐
− − − − − − −
C*U flight control law Integrated stability Total energy control law
Longitudinal Control Augmentation System: Lateral Automatic Throttle Control System:
› › ›
Resilient Flight Control (1)
and turns
pilot
climbing
certificated
ft/min ‐ deg bank turns ‐ Straight and level flight 25 1,000 descents Increase/Decrease airspeed Alt: 9,000 – 12,000 ft. Bank angle: 0 to 25 deg Airspeed: 250 to 280 kts › › › › › › ›
category
Execution of standard maneuvers in normal flight operation: Range of operation:
transport
a
by
performed
being
are
Piloted Simulation Results: Nominal Operations
tests
the
of
All
Resilient Flight Control (9)
excursions excursion excursions 1 2 >2
active
LoC prediction & prevention
Piloted Simulation Results: Nominal Operations
Proposed System Does Not Appear to Affect Pilot Performance under Normal Operations
Resilient Flight Control (10)
‐ with fixed shaped” time ‐ fields cos) − velocity Tailwind, downdraft, and lateral gust modeled as “(1 varying respect to the NED frame.
Piloted Simulation Results: Microburst
Maintain wings level at 200 KCAS Minimize altitude loss Maintain initial heading
› › ›
Test instructions:
Resilient Flight Control (12)
excursions excursion excursions 1 2 >2
not active
LoC prediction & prevention
Piloted Simulation Results: Microburst
LOC Conditions Occur for Substantial Amount of Time; Pilot Recovered after 2500 ft Altitude Loss
Resilient Flight Control (12)
excursions excursion excursions 1 2 >2
active
prevention
&
LoC prediction
Piloted Simulation Results: Microburst
Significant Reduction in LOC Conditions; Pilot Recovered after < 1000 ft Altitude Loss
Resilient Flight Control (13)
Safe envelope Source: NASA
Safe
,
in
::
tested
environment
controllable here?
conditions be
to
A/C Is
Testable
flight
::
unable
risk
‐
application
High
of
some
target
wind
to help
Repeatable
presence
::
the
Adaptive Control Augmentation
1 in
:
L
envelope
Predictable
objectives
predictable aircraft response
data
law
Provide the pilot avoid excursions outside the tunnel aircraft impairment.
• 10 publications directly related to work with AirSTAR GTM (2009-2012):
conference proceedings, journal articles, magazine articles, book chapter
Control
Resilient Flight Control (14)
advanced Control System CHR 3 (FQ L1) CHR 5 ( FQ L2) CHR 7 (FQ L3)
L1 Adaptive Control System
50% reduction in pitch control effectiveness Ft. Pickett, VA September 2010 Deployment,
Open-Loop Aircraft CHR 4 (FQ L2) CHR 10 (Uncontrollable)
Open-Loop Aircraft
Note: Subscale Test Vehicle Response is 4.25X Faster than Full-Scale Aircraft Example Result: Offset Landing with Emulated Destabilizing Failure: Pitch Stability degraded by 2 inboard elevator segments Roll Damping Stability degraded by spoilers Initial offset: 90 ft. lateral, 1800 ft. downrange, 100 ft. above the runway Flying qualities ratings taken for nominal, neutrally stable, unstable airplane
AirSTAR Testing under Vehicle Impairment
Nominal Neutrally Stable Unstable 35 35 30 30 25 25 20 20 time, sec 15 15 Open-loop roll departure 10 10 5 5 0 0 5 0 Example Result: T2 FLT 58 C14 WT02a 10 15 20 25 through departure and recovery. 0.2 -0.6 -0.4 -0.2 p l ˆ C Ft. Pickett, VA (deg) characterization during approach to stall, Demonstrated real-time stability and control May 2011 Deployment,
slow transition through stall boundary and improved stall/departure recovery
Flights 54, 55, 58
Applied L1 adaptive control to lengthen time on condition with stabilization that allowed
Example Result: Upset Test Condition - Stall / Departure (Pilot + Advanced Control System)
AirSTAR Testing under Upset Conditions
.
who must control the
provide operators with
, but rather its inappropriate
situation awareness
quantitative human performance modeling techniques
automation interaction interfaces
‐
automation failures (in aviation):
‐
human
Objective
inadequate feedback and interaction with the humans
the problem is not the presence of automation
“As automation increasingly takes its place in industry, especially high risk industry, it is often blamed for causing harm and increasing the chance of human error when failures do occur… design… operations under normal operating conditions are performed appropriately, but there is overall conduct of the task…” American Airlines 4184 (Roselawn, IN): Automation returns control to pilots with no indication of current envelope for safe control input or recovery; Scandinavian Airlines 751 (Gottröra, Sweden): Automation engages in actions not desired nor understood by the pilot; AirFrance 447 (Atlantic Ocean): Automation fails to clearly disclose sensor inconsistencies as detected by the automation until autopilot disconnects.
D.A. Norman (1990): Develop and validate ensuring that the information for achieving high levels of › Examples of human
› › ›
UIUC Pilot Interfaces & Situational Awareness (1)
path)
ownship
limit
limit
+
state
as
flight
Absolute (scales energy changes guides
Ownship
Desirable
β
Pilot
display)
α
Guides
Tunnel
state
Display
energy
Energy
Tunnel
after
Adaptive
dimensional
‐
(3
Adaptive
Tunnel
before
Conventional
UIUC Pilot Interfaces & Situational Awareness (2)
Based on the quantitative definition of loss of control proposed by Wilborn and Foster; Critical flight parameters are maintained within desired limits; Loss of control mitigated in the (reduced) set of test maneuvers considered; The system does not seem to limit the ability of the pilot to perform aggressive, evasive maneuvers. Integrate FDI solutions into the control architecture; Integrate ice accretion models; Design and integrate pilot/aircraft/automation interfaces; Expand the set of test scenarios; Validate the developed technologies; Investigate ‘optimal’ recovery procedures.
Preliminary control architecture for LOC prevention: Ongoing efforts:
− − − − − − − − − −
• •
Multiple Hazards Mitigation, LOC Prediction & Dynamic Envelope Protection: Summary
, ,” in ,” Master’s ,” Master’s Thesis , Technical University of Loop Testing ‐ the ‐ control prediction and prevention for , National Harbor, MD, January 2014.
in ‐ ‐ of ‐ ,” Semester’s Thesis Loss Flight Enveloped Protection for NASA’s GTM Integration of a Simulink Dynamics Model into a Full , National Harbor, MD, January 2014.
Champaign, May 2014. (to be submitted) Control Prediction and Prevention System for NASA's ‐ ‐ of ‐ Loop testing using NASA’s Transport Class Model ‐ the , National Harbor, MD, January 2014.
‐ in ‐ AIAA Guidance, Navigation and Control Estimation of Airspeed Using Continuous Polynomial Adaptive Estimator , National Harbor, MD, January 2014.
Control Prevention, and Upset Recovery Systems for NASA’s Transport Class ‐ ,” in of ‐ , University of Illinois at Urbana , National Harbor, MD, January 2014.
Champaign, August 2014. (to be submitted) ‐ An Adaptive Unkown Input Observer for Fault Detection and Isolation of Aircraft Actuator AIAA Guidance, Navigation and Control Loop Testing ‐ Champaign, May 2013.
,” in the ‐ ‐ loop simulated flight tests of a Loss in ‐ ‐ the , Technical University of Munich, March 2014.
‐ ,” Master’s Thesis AIAA Guidance, Navigation and Control in ‐ Axis Flight Envelope Protection for NASA’s GTM Research Aircraft ,” in Flight Simulator Development for Pilot ‐ Pilot Integration of the GTM T2 Model into a Full Size Simulator for Human Flight Envelope Protection, Loss Pitch AIAA Guidance, Navigation and Control ,” Master’s Thesis ,” in
Theses AIAA GNC 2014
N. Tekles, “ Munich, April 2013. S. T. Pelech, “ University of Illinois at Urbana N. Tekles, “ Model J. Chongvisal, “ Transport Class Model K. Ackermann, “ Thesis, University of Illinois at Urbana N. Tekles, E. Xargay, R. Choe, N. Hovakimyan, I. M. Gregory, and F. Holzapfel, “ Research Aircraft J. Chongvisal, N. Tekles, D. Talleur, A. Kirlik, N. Hovakimyan, and C. M. Belcastro, “ NASA’s Transport Class Model H. Lee, S. Snyder, & N. Hovakimyan, “ Faults K. A. Ackerman, S. T. Pelech, R. S. Carbonari, N. Hovakimyan, and A. Kirlik, “ Sized Simulator for Human H. Felemban, J. Che, C. Cao, and I. M. Gregory, “ AIAA Guidance, Navigation and Control At Least 10 Publications on L1 Adaptive Control Development with AirSTAR Conference papers, Journal articles, Magazine articles, Book chapter − − − − − − − − − − −
• •
•
Multiple Hazards Mitigation, LOC Prediction & Dynamic Envelope Protection: Recent Publications SBIR: Scientific Systems (SSCI), Barron Associates (Barron) NRA: University of Illinois (UIUC), University of Michigan (UM), University of Tennessee (UTSI); Enables Upset Prevention / Detection / Recovery & Safe Landing under Multiple Hazards Planner Landing Emergency
GCST for Safe & Effective Control under LOC Hazards: Status
Airframe Engine Airframe Structure (Current Activity with MVS, Longer Term) Propulsion System – – – – Sensors (Focus on Dynamics & Control) Control Actuator Failures Propulsion System Icing Damage Robustness under Turbulence Wind Shear Wake Vortices Inappropriate / Ineffective Control Inputs Ineffective Recovery » » » » » » » » » » System Failures Vehicle Impairment External Hazards Inappropriate Crew Actions / Inaction LOC Prediction Upset Detection Safe Flight Envelope Estimation Flight Safety (Longer Term) – – – – – – – – Hazards Effects Detection, Identification, & Mitigation (Dynamics and Control Effects) Vehicle Level Effects Prediction / Detection
• •
GCST Accomplishments Summary
Real-Time Monitoring Monitoring • Fault Detection • Real-Time Multiple Hazards Integrated System Evaluations under • Ground-Based • In-Flight Risk Evaluations Integrated & High- Experimental Testing (Bristol University and Drexel University) Simulation Evaluations Guided Monte Carlo & Piloted • Vehicle Upsets/Impairment • Vehicle Impairment Unsafe Analysis Potential Safe Problems Technology
Comprehensive Technology Evaluation using Realistic Hazards Test Scenarios
Assess Effective Hazards Coverage, and Identify System Limitations & Weaknesses
Identification of NRA Partners: University of Minnesota, University of West Virginia, Georgia Institute of Technology Assessment • Stability / Robustness • Confidence Level
Comprehensive Technology Evaluation (CTE)
Unsafe Deterministic Stochastic Conventional Transports within Visual Range T-Tail Transports (Developing research aircraft via SBIR I & II) Beyond Visual Range (Developing capability in-house) Test Vehicle with SHM Infrastructure (SBIR IIe ?)
» » » » » » Nonlinear Analysis Methods (Bifurcation, etc.) Robustness Analysis for Nonlinear Uncertain Systems Analysis for Fusion-Based Stochastic Estimation Systems Analysis of Pilot-Vehicle Systems Analysis of Pilot-Automation Systems Analysis of Complex Integrated Systems Batch / Monte Carlo Real-Time Piloted Evaluations AirSTAR Testbed (Potential for Integrated System Vehicle-Level Testing) SAFETI Lab (Potential for Integrated System Vehicle-Level Testing) Preliminary Test Scenarios Developed (2012) Final Set to be Developed (2014) – – – – – – – – – – – – Analysis Methods Simulation Methods (Based on Enhanced LOC Hazards Simulation) Experimental Test Methods LOC Test Scenarios Development
• • • •
CTE Accomplishments Summary
10% 10% 10% 10% 10% 10% 20% 20% 20% 30% 40% 50% Risks Future Conditions Vehicle Upset 3. Decreased Airspeed, Asymmetric Forces / Moments, Stall / Departure Coverage Sets 1 0 0 0 0 0 1 0 0 1 1 1 Risk Risks Future Covered Additional External Cumulative Hazards & Disturbances % Potential 0.79% 2.38% 6.35% 9.52% 10.32% 11.11% 11.90% 12.70% 14.29% 14.29% 15.08% 15.08% Accidents Future & of 10% 10% 10% 10% 10% 10% 10% 20% 20% 10% 10% 10% Risks Future Set Data Response Data Inappropriate Crew Coverage 2. Crew Distraction Resulting in Delayed Response Followed by Excessive Response % 0.79% 1.59% 3.97% 3.17% 0.79% 0.79% 0.79% 0.79% 1.59% 0.00% 0.79% 0.00% Accidents Historical on 1 1 1 1 1 1 1 2 2 1 1 1 of by Risks Future Number Covered Scenario Based Conditions 10 10 3 3 3 3 3 3 7 8 4 by 10 Adverse Onboard Risks 3, 3, Future Covered Scenario Hazards 1. Single Engine Failure (100% Thrust Loss); 4. Various Levels of Structural Damage with and without Loss of Control Effector of Data 1 2 5 4 1 1 1 1 2 0 1 0 of by Set Flight Cruise Number Covered Scenario Accidents from Condition Coverage 8, 18, 63 79 Data 7 3 2 110 20, 56 13 16 113 N/A N/A Covered 15, Scenario 62, 41, 2, Crew 1, 17, Accidents from Vehicle Damage Set by Scenario Leading to Upset and Distraction Description Followed by Engine Failure E D D D D D D D D D D D D, Realistic Test Scenarios with Traceability to the Hazards Sets Four Precursor LOC Scenarios: Vehicle Failure –> Inappropriate Crew Response –> Upset –> Vehicle Damage Sequence Generalized Batch Piloted Methods Analysis, Evaluation Simulation Simulation, Recommended 1 2 3 4 5 6 7 8 9 10 11 12 Set Scenario Number Set Number Scenario Upset Vehicle Crew Response Inappropriate Hazards Analysis Hazards Sequences / Unique & Generalized Sequences Loss of Aircraft State Awareness Spatial Disorientation Extreme Attitudes Abnormal Energy States Abnormal Control Response Stall / Departure Control Component Failures Icing Effects Wake Vortices Upsets Accident Data / Future Risks Abnormal Flight / Upset Hazards • • • • Crew Hazards • • Vehicle and Environmental Hazards • • • Worst-Case Hazards Combinations Hazard Vehicle External Problem
Approach for Developing LOC Test Scenarios
Partners: NTSB & NIA (CAST / ATLAS) & , Hazards Disturbances – More – 99 – 199 – 299 – 499 – 999 Fatalities 0 1 External 100 200 300 500 to is Size Accidents Conditions of Proportional Sphere Worst Case Precursor Combinations Upsets Number Onboard , National Harbor, MD, January 2014.
Dynamics Directly AIAA Conference on Guidance, Navigation, and Control 275 LOC Accidents & Incidents (1996-2010) Vehicle Upset Example Preliminary Analysis Results 237 Unique Sequences Identified at the Precursor Level Adverse 108 Unique Sequences Identified at Sub-Category Level; & Dynamics Unknown Abnormal / Vehicle None Ref: Belcastro, C. M., Groff, L., Newman, R. L., Foster, J. V., Crider, D. A., Klyde, D. H., and Huston, A. M., “Preliminary Analysis of Aircraft Loss of Control Accidents: Worst Case Precursor Combinations and Temporal Sequencing”, SciTech Forum Abnormal 10 International Databases Searched for LOC Commercial transports at or above 12,500 lbs 275 accidents & incidents identified resulting in 7185 fatalities Based on Six Accident / Incident Subsets (45-46 Events) Individual Precursor Statistics Worst-Case Precursor Combinations Temporal Sequencing Re-Evaluation Based on Team Consensus Approach (In Progress) Definition of Future LOC Risks (To be Coordinated with CAST / ATLAS) Development of LOC Test Scenarios Final Results to be Submitted for NASA TP and Journal Publication – – – – – – – – – – – Comprehensive Set of Accidents / Incidents Compiled for 1996 – 2010 Preliminary Analysis Results Obtained Ongoing Research
Status of LOC Hazards Analysis & Test Scenarios Development
• • •
Evaluation & Testbeds Methods, Tools, Technology Constraints Limitations & under LOC Hazards Degree of Technology Effectiveness Level of Enhanced Onboard Systems Technologies for Improved Hazards Coverage Situational Awareness, Guidance & Control Comprehensive Technology Evaluation (CTE) Upset Vehicle Crew Response Inappropriate Combinations / Results Level of Incident Hazards Sequences Based on Worst-Case Hazards Accident / Analysis & Vehicle Hazard External Problem Future Risks Confidence in the Evaluation Test Scenarios LOC Hazards Analysis
LOC Research Integration
Safety-Assured Autonomy Real-Time Safety Assurance Resilient Control & Mission Management • • Resilient Control under LOC Hazards LOC Prediction, Prevention & Recovery Resilient Mission Planning • • • Pilot-Optional Aircraft Dynamic Envelope Protection Resilient Control under Off-Nominal Conditions Upset Detection & Recovery Automatic Collision Avoidance Emergency Landing Planning • • • • • Baseline: Altitude Hold, Autoland, Nominal Envelope Protection, TCAS, EGPWS, No Significant Warnings or Guidance under LOC Hazards Single-Pilot Operations UAS Remotely Piloted Current Operations Future Research Directions: Safety-Assured Autonomy for Advanced Manned / Unmanned Aircraft Baseline: Technology Used to Automate Routine Operations under Nominal Conditions and Provide Information & Alerts Resilient Systems Provide Safety Augmentation, Guidance & Emergency Intervention to Support Baseline Systems and Human Operator Fully Autonomous Systems Enable Safety-Assured Operations at All NAS Levels (Vehicles, Infrastructure, and Operations) Variable Autonomy Systems Enable Synergistic Dynamic Teaming Between Human and Intelligent Systems Largest Fatal Accident Category Most Complex Accident Category (Many Causal & Contributing Factors) Large Accident / Incident Set Worst-Case Hazards Combinations Detailed Hazards Sequences Comprehensive LOC Test Scenarios Vehicle Dynamics Modeling & Simulation of LOC Precursor Conditions Guidance, Control, & Systems for LOC Prevention / Recovery Improved Crew Situation Awareness, Guidance, & Cueing under LOC Hazards Validation of Safety-Critical Technologies Developed for LOC Prevention / Recovery Resilient Systems Variable Autonomy Systems Fully Autonomous Systems
– – – – – – – – – – – – –
Aircraft LOC Contributes Significantly to Accidents & Fatalities NASA is Conducting Unique Hazards Analysis of LOC Problem to Enable Holistic Technology Solution NASA is Conducting Research and Developing Technologies for LOC Prevention & Recovery Future Research Direction towards Resilient Autonomous Aircraft
• • • •
Summary & Concluding Remarks
757-864-4035
Hampton, VA 23681
Dr. Christine M. Belcastro
8 Langley Boulevard, MS 308
NASA Langley Research Center christine.m.belcastro@nasa.gov
NASA Aviation Safety Program
Dynamic Systems and Control Branch
Assuring Safe & Effective Control under Hazardous Conditions
Contact Information:
Airframe Engine Fast Engine Response Integrated Flight / Propulsion Control
– – – –
Abstract Vehicle Impairment under Icing Multiple Hazards Mitigation
• • •
Backup
The
Hence, there is no single
Aircraft loss-of-control (LOC) accidents are
The approach includes onboard systems technologies for LOC
Aircraft loss of control is a leading cause of fatal accidents across all transport airplane and operational classes. highly complex in that they can result from numerous causal and contributing factors acting alone or (more often) in combination. intervention strategy to prevent these accidents. This presentation will define LOC as a dynamics and control problem, summarize LOC accident analysis results, and discuss recent NASA research that supports a holistic approach for significantly reducing aircraft LOC accidents for current and future aircraft. prevention and recovery, as well as a parallel effort for their validation. onboard systems technologies provide improved situational awareness, guidance, and control under realistic LOC hazards, and include: hazards effects detection and mitigation; upset detection, prevention and recovery; and multiple hazards mitigation. Validation technologies include analytical, simulation, and experimental methods, tools and testbeds as well as the development of a comprehensive set of realistic LOC test scenarios. Future research directions will also be discussed.
Abstract
Icing NASA / Airframe NRA – UTSI, Icing NASA / Airframe NRA – UTSI,
Vehicle Impairment Identification
Break NR PR NR NR -- UK NR -- NR Pusher ES ES ES ES -- UK -- -- NR Shaker ES ES ES ES -- UK ES -- NR ES = Early stall Albury Eildon Weir American Eagle 3008 No ice lift curve (deg) Stick Pusher Saab 340 NR = Not responded to Icing Event Comparison Angle of Attack - Stick Shaker Accident/Incident L Lift Coefficient - C PR = Proper response Saab 340A, VH-LPI, Eilden Weir, Victoria, Nov 11, 1998 Saab 340A, VH-KEQ, Albury, New South Wales Australia, June 18, 2004 Saab 340B, VH-OLM, Bathurst, New South Wales Australia, June 28 2002 American Eagle 3008, Saab 340B+, January 2, 2006 Air Canada Flight 646, Canadair CL-600-2B219, December 16, 1997 Cessna Citation 560, Pueblo, Colorado, February 16, 2005 Comair Flight 3272, Embraer EMB-120RT, January 9, 1997 Skywest 3855, Bombardier CL-600-2B19, January 17, 2004 ComAir 5054, EMB-120, March 19, 2001 UK = Unknown
Airframe Icing Problem Relevance
IO = Inoperative Ref: Dennis Crider, NTSB, LOC RWG Kickoff Meeting, August 16,2012 » Loss of roll control, upset » fatal accident » Roll oscillations » Recovered with flap change » May occur after stall, if at all » Stall break (roll off and roll control) » Post stall flight characteristics 72-212, October 31, 1994 Canada Flight 1130; A321-211, December 7, 2002 low and banked.
– American Eagle Flight 4184 ATR – Air Canada Flight 457 & Air – Stick shaker and pusher: – Crews may not respond to: – Lack of stall recognition with nose
• Control • Icing Stall
Vehicle Impairment Identification – Airframe Icing (1)
Airframe Icing Problem Relevance
is reduced right now), and
» Stall occurring before normal stall angles of attack » Changes in control effectiveness » detecting icing and its effects on the aircraft stability and control, » giving the pilot specific warnings in advance (e.g., elevator control effectiveness » showing flight envelope restrictions based on the current aircraft condition.
Pilots aren’t familiar with what adverse effects can come from them, such as Pilots don’t realize they are in icing conditions, and do not receive advance warning of impending stall Icing problem can build up slowly, then the aircraft can quickly get into an upset from a trigger event, such as maneuvering or a configuration change (e.g., flaps or power changes) At that point, there is little time to think or react. ICEPRO addresses this problem directly by
– – – – –
Summary of Airframe Icing Problem: Potential Technology Solution:
• •
Vehicle Impairment Identification – Airframe Icing (2)
Utilizes Dynamic Inversion Control Evaluation System (D-ICES) to compare current aircraft state to a nominal baseline state. ID mode triggered when control differences or drag degradation thresholds met Executes a series of optimally designed control excitations during periods of low pilot control activity Utilizes Real-Time Parameter Identification (RTPID) to compute an icing severity parameter (ISP) based on comparisons of current stability and control to baseline Alert messages and cues displayed to the pilot to remain within safe flight envelope Low speed cues are dynamically set to maintain the 5% stall margin based on icing severity • • • • • • Monitor Mode: ID Mode: Reporting Mode:
ICEPRO System Description
1) 2) 3)
Pilot seat, controls, and turbo-prop throttle quadrant Column force system and elevator trim switch 4 flat panel screens for out-the-window graphics Multiple PC’s hosting D-Six simulation model, drive graphics, and an electro- mechanical control loader Sets up initial conditions for each training block Video recording & monitoring devices Intercom for communications between pilot & instructor Supplement FTD training with multi- media training
– – – – – – – –
Portable, fixed-base FTD Instructor’s Workstation
• •
ICEPRO System Evaluation (1)
Statistically Significant Results
Reference: Contractor Report Submitted to NASA GRC by UTSI » Fewer incipient upsets » Better Situation Awareness » No increase in workload with ICEPRO Raised situational awareness Assisted in flying the task Reduced the risk of upsets and eventual LOC from a hazardous icing encounter
NASA GRC’s ICEFTD Simulated on-board excitation system (OBES)
– –
ICEPRO was Evaluated by 30 Pilots During the 2007-2011 NASA NRA using All pilots were constrained to fly a hazardous configuration. The ICEPRO pilot group showed By providing the pilot reliable cuing and messages, ICEPRO 1. 2. 3.
• • •
ICEPRO System Evaluation (2)
, Minneapolis, MN, August 2012
AIAA Atmospheric Flight Mechanics Conference
B. Martos, “Identifying and Correcting First Order Effects in Explanatory Variables for Longitudinal Real Time Parameter Identification Methods in Atmospheric Turbulence,” University of Tennessee, 2013. D. R. Gingras, R. Ranaudo, B. Barnhart, T. Ratvasky and E. Morelli, “Envelope Protection for In-Flight Ice Contamination," AIAA-2009-1458, 2009. D. R. Gingras, B. Barnhart, R. Ranaudo, B. Martos, T. Ratvasky and E. Morelli, "Development and Implementation of a Model-Driven Envelope Protection System for In- Flight Ice Contamination," AIAA-2010-8141, 2010. R. Ranaudo, B. Martos, B. Norton, D. Gingras, B. Barnhart, T. Ratvasky and E. Morelli, "Piloted Simulation to Evaluate the Utility of a Real Time Envelope System for Mitigating In-Flight Icing Hazards," AIAA-2010-7987, 2010. B. Martos and E. A. Morelli, "Using Indirect Turbulence Measurements for Real-Time Parameter Estimation in Turbulent Air," AIAA-2012-4651, 2012. Morelli, E.A. and Cunningham, K. “Aircraft Dynamic Modeling in Turbulence,” AIAA-2012- 4650, B. Martos, R. Ranaudo, B. Norton, D. Gingras, B. Barnhart, T. Ratvasky and E. Morelli, "Final Report NASA Grant NNH06ZEA001N: Development, Implementation and Pilot Evaluation of a Model-Driven Envelope Protection System to Mitigate the Hazard of In- Flight Ice Contamination," NASA CR-2013-0000, (To be published in 2014).
Airframe Icing Publications
Dissertation: • Conference Publications: • • • • • •
Icing NASA Engine
Vehicle Impairment Identification: Engine Icing Effects
, 2011 th Ice Accumulation Engine Malfunction Normal Engine Performance Occurs under high- altitude storm clouds with massive quantities of small ice crystals Not currently detectable on pilot radar Ice crystals are drawn into engine where some are ingested with air flowing into the compressor Under core flow compression, ice crystal accumulation can occur Ice can break off from the compressor components causing the engine to surge, stall, flame out, or experience other malfunctions • • • • •
Engine Icing Problem Relevance 14 total power loss (all engines)
153 Power-Loss events 1988-2010 *
Temporary or sustained power loss, uncontrollability, engine shutdown Compressor surge Flame-out due to combustor ice ingestion Sensor Icing Engine Rollback
– − − − −
* Fisher, John, “Aircraft Turbine Engine Icing: Current Issues and Future Vision,” presentation at SAE International Aircraft Icing Conference, June 16 Ice has been found to accrete in the compressors of commercial aircraft engines during operations under High Ice Water Content (HIWC) conditions More than 150 power loss events reported in last 20 years Many possible causes for power losses
• • •
Vehicle Impairment Identification – Engine Icing
50 50 50 50 45 45 45 45 40 40 40 40 20% pressure drop 23% pressure drop 26% pressure drop /s] m 35 35 35 35 30 30 30 30 100 Mass Flow Rate [lb time [s] 0% Blockage 10% 20% 27% 25 25 25 25 1.2 1.4 1.6 1.8 2.2 2.4 2.6 2.8 20 20 20 20 Pressure Ratio 15 15 15 15 4 4 x 10 x 10 10 10 10 10 1.2 0.98 1.02 1.04 1.06 1.08 1.14 1.16 1.18 1.22 3780 3800 3820 3840 3860 3880 2800 3000 3200 net f c f TGT [ R] F [lb ] N [rpm] N [rpm]
: Successful simulation of Engine
:
Engine Icing Effects Modeling & Simulation
: Use NASA C-MAPSS40k simulation
: Engine Rollback Phenomena
Currently available for U.S. citizens Open-source thermodynamic modeling platform under development Occurs once blockage become severe enough Results in decreased thrust and fan speed, and increase in Turbine Gas Temperature (TGT) Unresponsive to throttle commands Realistic controller, sensor noise, modular » » Modify Low Pressure Compressor (LPC) maps to include the effect of discrete levels of ice blockage Caused by the engine controller limiting fuel when certain safety limits are reached Enables technology evaluation under engine icing Enables development of detection/mitigation methods Ref: May, R.D., Guo, T-H., Veres J.P., Jorgenson, P.C.E., “Engine Icing Modeling and Simulation (Part 2): Performance Simulation of Engine Rollback Phenomena,” 2011-38-0026, SAE International Conference on Aircraft and Engine Icing and Ground Deicing, Chicago, IL, Jun 13-17, 2011. doi:10.4271/2011-38-0026 Problem – – – Approach − − Accomplishment Rollback using LPC blockage map and dynamic controller − Significance – –
Engine Icing Effects Detection (1)
• • • •
Detection
False-Positive = 0.2% True-Positive = 99.6% Average Blockage Level at Detection = 4.55%
• • • Early Results: Threshold • Systems-Level Perspective of Engine Ice Accretion
Engine Icing Impairment Detection
: : Presented to the Engine Icing Working Group, Derby, England, September 19-20, 2012 from expected sensor values FADEC » Balance false positive rate and detection time » Detect a decrease in the flow capacity » Distance Measure (Dm) to determine deviation » Low memory usage » Should be capable of operating real-time in typical
efficiency and flow capacity) based on available sensors sensor
– Estimate the change in LPC “health” (LPC – Threshold Selected to – Linear estimator approach – Early results promising – Further improvements achievable using HIWC – Development of mitigation strategies next step
Approach Accomplishments
Engine Icing Effects Detection (2)
• •
» Complete development of model of Honeywell engine using T-MATS software » Implement & test detection algorithms on simulated engine » Verify against experimental data No other component faults This ensures that all changes in engine operation are due to ice accretion Detection algorithm based on combination of LPC efficiency and LPC flow capacity Developing a technique for full envelope detection and transient operations Complete dynamic detection algorithm for full envelop and transient operations Follow up on work done in NASA PSL Develop mitigation strategies – iterate with the NASA GRC icing code to determine how the change in operating point impacts the accretion of ice & possible testing in PSL
– – – – – – –
Assumptions: Ongoing Research Future Plans Engine Icing Effects Detection Technology Status: TRL 1
• • • •
Engine Icing Effects Detection (3)
Publications
May, R.D., Simon, D.L., Guo, T-H., “Modeling and Detection of Ice Particle Accretion in Aircraft Engine Compression Systems,” AIAA Atmospheric Flight Mechanics Conference, Minneapolis, MN, Aug 13-16, 2012. May, R.D., Guo, T.H., Simon, D.L., “An Approach to Detect and Mitigate Ice Particle Accretion in Aircraft Engine Compression Systems,” ASME-GT2013- 95049, ASME TurboExpo 2013, San Antonio, TX, June 3-7, 2013. May, R. D., Guo, T-.H., Simon, D. L., “Detection of the Impact of Ice Crystal Accretion in an Aircraft Engine Compression System During Dynamic Operation,” AIAA 2014-0270, AIAA Guidance, Navigation, and Control Conference, National Harbor, MD, January 13-17, 2014 R. May, J.W. Chapman, J.P. Veres, T. Guo, M.J. Oliver, “Development and Validation of an Aircraft Engine Simulation Including the Impact of Engine Ice Accretion, “Propulsion and Energy 2014 Forum, Cleveland, OH, July 28-30, 2014. (To Appear)
• • • •
Engine Icing Effects Detection (4)
NRA – P&W NASA GRC / LaRC Fast Engine Response / Integrated Flight-Propulsion Control
Fast Engine Response & Integrated Flight-Propulsion Control (IFPC)
limits and by using existing actuators in novel ways
Engine behavior is highly non-linear depending on operation conditions Involving risk of engine failures Increase controller bandwidth Relax engine limits Off-nominal operation C-MAPSS40k simulation Engine response time can be improved with only changes to gains and Any change will increase the risk of engine failure Pilot based evaluations of the fast response engine are very favorable
– – – – – – – – –
Fast response engine can increase the likelihood of recovering the aircraft from loss of control scenarios Challenges: Approaches: Results:
• • • •
Fast Engine Response Research (1)
10 10 Nominal MAS time [s] 5 5 Time, s Hot Day 14 4 16 18 20 22 24 26 28 HPC Surge Margin [%] x 10 0 0 0 1 2 3 0 10 20 30 40 Thrust, lbf Surge Margin, % 10 10 = nom = 1.4 * nom p p K K in Rise Time 5 5 22% Reduction time [s] Time, s Standard Day x 10 0 0 0 1 2 3 0 10 20 30 40 x 10 Thrust, lbf Surge Margin, % 2.2 2.3 2.4 2.5 2.6 2.25 2.35 2.45 2.55 2.65 net f F [lb ] Changing the setpoint controller’s gain will improve response during small transients Increase in gain reduces stability margins Marginal impact on HPC Surge Margin Relax acceleration limits Reduced compressor surge margin Engine condition considered to minimize the risk Rise time reduction depending on the engine health condition
Control gain modification • • • Risk-based limit modification • • • •
Fast Engine Response Research (2)
25 25 25 25 20 20 20 time [s] 15 15 0 2 4 0 5 f m 10 20 40 60 -60 -40 -20 W [lb /s] cust m HPC SM [%] VSV [deg] W [lb /s] 25 25 25 25 Nominal HSI HSI+ 20 20 time [s] 15 15 x 10 15 x 10 0 2 4 gross f 0.9 1.1 1.2 20 40 60 F [lb ] c 2000 3000 4000 LPC SM [%] N [rpm] f N [rpm]
High Speed Idle Operation
2.28 s 1.81 s 1.36 s
Time Constant
HSI
HSI+
Move compressor vanes off- nominal Bleed air from engine components Extract power from shafts
Nominal
Control Law
– – –
Designed to improve engine response during approach Operate at higher shaft speeds Reduces the potential fan/core mismatch Higher speeds lead to more thrust During approach/landing it is critical to balance the aircraft’s energy Engine can be operated in an off- schedule manner to reduce excess thrust
• • • • • •
Fast Engine Response Research (3)
Prevent unintended/premature ground contact during approach phase of flight Automatic aggressive recovery maneuver of collision is otherwise inevitable Define protected envelop of the flight landing path Continuous prediction of altitude loss during maneuver Define recovery control commands Determine trigger point for necessary recovery Can not interfere with normal landing procedure Model-predictive automatic recovery system to prevent unintended ground contact during landing Automatic application of recovery maneuver through flight and propulsion control override Very successful pilot evaluations for approach/landing in different scenarios
– – – – – – – – – –
Model-Predictive Automatic Recovery System (M-PARS)
Proof of concept as an example of the Integrated Flight and Propulsion Control Objective: Approaches: Results:
• • • •
Integrated Flight-Propulsion Control (1)
Automatic Recovery System
Define aggressive recovery control commands (e.g., autopilot GA mode, full power and pitch up, etc.) Continuous prediction of altitude loss due to maneuver If prediction violates threshold, initiate recovery maneuver
• • •
Integrated Flight-Propulsion Control (2)
Models and control systems for aircraft and engines (TCM + C-MAPSS40k) Flight path predictor MPARS flight/propulsion control override algorithms
– – –
Full cockpit with standard pilot/copilot controls and instrumentation PC 1: X-Plane PC 2: Displays PC 3: Everything else
• • • •
Spare Aircraft X-Plane Transport (TCM & C- MAPSS40K Master_PC1 Visuals_PC2
NASA GRC Simulation Testbed
Integrated Flight-Propulsion Control (3)
Flight path is from right to left Altitude threshold set to 50 feet AGL 1000 feet from runway: altitude threshold set to 0 feet 50 feet AGL: protection system deactivates Simple rules work well unless trying to touchdown at runway edge
• • • • •
3 3 3 2.5 2.5 2.5 AGL AGL predicted 2 2 2
Successful Landing
1.5 1.5 1.5 1 1 1 Distance to Runway, nm 0.5 0.5 0.5 0 0 0 0 0 0 -0.5 -0.5 -0.5 10 20 30 -80 -60 -40 -20 Throttle, % Predicted Alt. Change, ft 1000 AGL, ft
Integrated Flight-Propulsion Control (4)
3 3 2.5 2.5 2 2 1.5 1.5 1 1 Distance to Runway, nm 0.5 0.5 Variety of failed approaches to stress test recovery system All cases would have touched down short of runway Circular point denotes MPARS activation (notice throttle change) 0 0 0 0 10 20 50 -20 -10 Pitch Angle, deg Throttle, %
• • •
3 3 3
Failed Approaches
2.5 2.5 2.5 2 2 2 1.5 1.5 1.5 1 1 1 0.5 0.5 0.5 0 0 0 0 0 500 100 120 140 160 1000 2000 4000 Air Speed, kts -4000 -2000 AGL, ft Vert. Speed, fpm
Integrated Flight-Propulsion Control (5)
prevent unintended ground contact (flight/propulsion controls override) activation modes
– Model-predictive automatic recovery system developed to – Primary application: low-speed, low-altitude – Automatic application of aggressive recovery maneuver – Continuous flight path prediction to ensure “last-second” – Forward distance prediction & sloping threshold – Uncertainty analysis of flight path predictor – Incorporation of risk-based enhanced engine performance control Status Ongoing & future work:
• •
Fast Engine Response & Integrated Flight-Propulsion Control Summary (1)
Publications (1)
McGlynn, G. E., Litt, J.S., Lemon, K.A., and Csank, J.T., “A Risk Management Architecture for Emergency Integrated Aircraft Control,” AIAA-2011-1568, Infotech@Aerospace 2011, St. Louis, Missouri, Mar. 29-31, 2011. also NASA/TM—2011-217143, December 2011. Csank, J.T., Chin, J.C., May, R.D., Litt, J.S., and Guo, T.-H., “Implementation of Enhanced Propulsion Control Modes for Emergency Flight Operation,” AIAA-2011-1590, Infotech@Aerospace 2011, St. Louis, Missouri, Mar. 29-31, 2011. Avishai Weiss, Ilya Kolmanovsky, Walter Merrill, "Incorporating Risk into Control Design for Emergency Operation of Turbo-Fan Engines," AIAA-2011-1591, Infotech@Aerospace 2011, St. Louis, Missouri, Mar. 29-31, 2011. Chuan Wang, Reza Sharifi, Chengyu Cao, "L1 Adaptive Control of Uncertain Nonlinear Systems in the Presence of Output Limits," AIAA-2011-1400, Infotech@Aerospace 2011, St. Louis, Missouri, Mar. 29- 31, 2011 James Urnes, Timothy Smith, "Use of Propulsion Commands to Prevent Loss-of-Control Aircraft Accidents," AIAA-2011-1567, Infotech@Aerospace 2011, St. Louis, Missouri, Mar. 29-31, 2011. Csank, J.T., May, R.D., Litt, J.S., and Guo, T.-H., “A Sensitivity Study of Commercial Aircraft Engine Response for Emergency Situations,” NASA/TM—2011-217004, April 2011. Richter, Hanz, and Litt, Jonathan S. , "A Novel Controller for Gas Turbine Engines with Aggressive Limit Management," AIAA 2011-5857, 47th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, July 31-August 3, 2011, San Diego, CA. Csank, J. T., May, R. D., Guo, T-H., Litt, J.S., “The Effect of Modified Control Limits on the Performance of a Generic Commercial Aircraft Engine,” AIAA 2011-5972, 47th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, July 31-August 3, 2011, San Diego, CA. Also NASA/TM--2012- 217261, January 2012.
• • • • • • • •
Fast Engine Response & Integrated Flight-Propulsion Control Summary (2)
Publications (2)
May, R. D., Csank, J. T., Guo, T-H., Litt, J.S., “Improving Engine Responsiveness during Approach through High Speed Idle Control,” AIAA 2011-5973, 47th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, July 31-August 3, 2011, San Diego, CA. Lemon, Kimberly A., Litt, Jonathan S., and May, Ryan D., “An Emergency Engine Response Requirement Analysis Tool for Lateral-Directional Dynamic Aircraft Stability,” AIAA 2011-6308, AIAA Guidance, Navigation & Control Conference, Portland, OR, August 8-11, 2011. May, Ryan D., Lemon, Kimberly A., Csank, Jeffrey T., Litt, Jonathan S., and Guo, Ten-Huei, “The Effect of Faster Engine Response on the Lateral Directional Control of a Damaged Aircraft,” AIAA 2011-6307, AIAA Guidance, Navigation & Control Conference, Portland, OR, August 8-11, 2011. Also NASA/TM--2012- 217216. March 2012. Litt, Jonathan S., Sowers, T. Shane, Owen, A. Karl, Fulton, Chris, Chicatelli, Amy, “Flight Simulator Evaluation of Enhanced Propulsion Control Modes for Emergency Operation,” AIAA 2012-2604, Infotech@Aerospace 2012, Garden Grove, CA, June 19-21, 2012. Litt, Jonathan S., Guo, Ten-Huei, Sowers, T. Shane, Chicatelli, Amy K., Fulton, Christopher E., May, Ryan D., and Owen, A. Karl, "Pilot-in-the-Loop Evaluation of a Yaw Rate to Throttle Feedback Control with Enhanced Engine Response," AIAA-2012-5027, AIAA Guidance, Navigation, and Control Conference, Minneapolis, Minnesota, Aug. 13-16, 2012. Zaretsky, Erwin V., Litt, Jonathan S., Hendricks, Robert C., Soditus, Sherry M., "Determination of Turbine Blade Life From Engine Field Data," JOURNAL OF PROPULSION AND POWER, Vol. 28, No. 6, November- December 2012, pp. 1156-1167, also NASA/TP—2013-217030, April 2013. Liu, Y., Litt, J. S., Guo, T.-H., “Design and Demonstration of Emergency Control Modes for Enhanced Engine Performance,” 49th AIAA/ASME/SAE/ASEE Joint Propulsion Conference, San Jose, CA, 14-17 July 2013, also NASA/TM—2013-216552, September 2013.
Fast Engine Response & Integrated Flight-Propulsion Control Summary (3)
• • • • • • •
Publications (3)
Invited Presentations
New Technology Disclosures
“High Speed Idle Engine Control Mode” non-provisional utility patent application filed on 12/13/2012. “Model-Predictive Automatic Recovery System,” Co-inventors are Jonathan Litt, Yuan Liu, T. Shane Ten-Huei Guo, “Enhanced Engine Control Overview,” 4th NASA GRC Aeronautics Propulsion Control and Diagnostics Research Workshop, Dec. 11-12, 2013, Cleveland, OH James Liu, “Controller Design for Enhanced Engine Response,” 4th NASA GRC Aeronautics Propulsion Control and Diagnostics Research Workshop, Dec. 11-12, 2013, Cleveland, OH Jonathan Litt, “Using Propulsion System for Loss of Control Prevention and Mitigation,” 4th NASA GRC Aeronautics Propulsion Control and Diagnostics Research Workshop, Dec. 11-12, 2013, Cleveland, OH James Liu, “Piloted Simulation of a Model-Predictive Automated Recovery System,” at Aerospace Control and Guidance Systems Committee Meeting #113, March 12-14, 2014, Englewood, CO.
Litt, Jonathan S., Liu, Yuan, Sowers, T. Shane, Owen, A. Karl, Guo, Ten-Huei, “Piloted Simulation Evaluation of a Model-Predictive Automatic Recovery System to Prevent Vehicle Loss of Control on Approach,” AIAA 2014-0036, AIAA ATMOSPHERIC FLIGHT MECHANICS CONFERENCE, National Harbor, MD, January 13-17, 2014, also NASA/TM—2014-216644, March 2014. Liu, Yuan, Litt, Jonathan S., Sowers, T. Shane, Owen, A. Karl, Guo, Ten-Huei, “Application and Evaluation of Risk-Based Performance Enhancing Engine Control Modes,” Propulsion and Energy 2014 Forum, Cleveland, OH, July 28-30, 2014. (To Appear) Co-inventors are Jeff Csank, Ryan May, Jonathan Litt, and Ten-Huei Guo Sowers, and A. Karl Owen.
• • • • • • • •