Post on 15-Dec-2015
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
1M
Loss of Control of Light Aircraft:Loss of Control of Light Aircraft:Balancing Cost and Safety in Flight Balancing Cost and Safety in Flight
TestTest
Mike BromfieldMike BromfieldGuy GrattonGuy Gratton
SETP/SFTE 4th European Flight Test Safety Workshop London 28~29, September 2010
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
3G
What’s the question?• GASCo 28 year review of GA fatal accidents
– 35-50% stall/spin– Significant type variations
• Why type variations– C150 rate >> C152 rate (~17:1)– PA28 “Hershey Bar” wing ~population mean, tapered
wing no fatalities
• Understand the reasons– For flying training– For design
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
4M
Spot the difference…?
Cessna 150L
Cessna 152
0.71 fatals/ 100,000 hrs0.71 fatals/ 100,000 hrs
0.04 fatals/ 100,000 hrs0.04 fatals/ 100,000 hrs
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Events
0 2 4 6 8 10
Task: Baulked Ldg
Pilot: Instructor
Pilot: 2 POB
Enviroment: Wind 30 kts+
Aircraft: Cessna 152
Aircraft: Cessna 150 L/M
10 events: C150/C152 stall/spin
fatalities
5M
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Methods & EquipmentFlight Test basics
Go-Pro Cockpit mounted camera
Appareo GAU1000A Flight Data Recorder
Garmin 296
6M
J31-Calibration
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Flight Test Programme Build-up
•8 aircraft•17 test sorties + 3 checkouts + 1 decline•25hrs 35mins flight test
7M
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Flight Test using Rental Aircraft
• Aircraft Variables (ageing?)– maintenance, performance, W&CG
• Organisation– Different priorities
• Flight test .v. Flight School– Aft CG stalling at 3000ft anybody?
• Local environment– Area & procedures, controlled airspace, ATC, Wx
8G
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
-7.00-6.00-5.00-4.00-3.00-2.00-1.000.001.002.003.004.005.006.007.00
40 50 60 70 80 90 100 110 120
Sti
ck
Fo
rce
Pu
ll (d
aN
) >
>>
Speed KCAS >>>
Polynomial Curve Fit for Apparent LSS - Cruise
F150L G-GBLR CRF150M G-BCRT CRC152 G-BOFL CRStall BoundaryPull Force to Stall Boundary
BTP/Sortie: 2008-06-02/03/04A/C= Cessna F150L,F150M,C152CG = Mid-CGVtrim = 84~88 KCAS
Stick Force to Change AirspeedCessna 150L,M & 152 with Flaps a) UP b) DOWN
(L30)
9M
Flaps UP
0
3
6
9
Start Climb
Mid ClimbEnd Climb
HQR Climb & Point Track
G-BCUH F150M
G-BFLU F152
Desirable: +/- 2 MIAS/KIAS Adequate: +/-5 MIAS/KIAS
BTP/Sortie: 2008-06-12A/C= Cessna F150M G-BCUHDate: 05/03/10Gross Wt.= 1599 lbsCG = 35.03" AoD (25.87%MAC/58.76% CG/Mid -CG)Vtrim = 69 KCAS (81 MIAS)
BTP/Sortie: 2008-06-10A/C= Cessna F152 G-BFLUDate: 12/12/09Gross Wt.= 1655 lbsCG = 33.8" AoD (23.78%MAC/50.71% CG/Mid -CG)Vtrim = 69 KCAS
-7.00
-6.00
-5.00
-4.00
-3.00
-2.00
-1.00
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
40 50 60 70 80 90 100 110 120
Sti
ck
Fo
rce
Pu
ll (
da
N)
>>
>
Speed KCAS >>>
Polynomial Curve Fit for Apparent LSS - Landing (30)
F150L L30F150M L30C152 L30Stall BoundaryPull Force to Stall Boundary
BTP/Sortie: 2008-06-02/03/04A/C= Cessna F150L,F150M,C152CG = Mid-CGVtrim = 66~68 KCAS
Flaps DOWN
Desirable Pull Force to Stall (PFtS – 10 lbf/4.4475daN)
Desirable Pull Force to Stall (PFtS – 10 lbf/4.4475daN)
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
10M
Radio Calls >> Increased Workload??
Climbing Flight: Cessna 150M
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Flight Test Results• Variability between aircraft, C150 consistently lower pitch forces
– Mean gradient factors ~2-3– C150 sometimes neutral (e.g. L40)
• Stick force dependent on:- – (1) Flaps, (2) Power, (3) Trim, (4) Weight, (5) CG
• Low stick force >> poor airspeed management
• Stall warning– C150 non-compliant with (current) part 23 in certain configs.– Simultaneous stall warning + aerodynamic stall
11M
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Simulation Experiments- Scenario-based Testing
Simulator• PC7 Fixed-base research simulator
– Controllable force feedback– 150 x 40 deg. Visuals
Scenarios• 20 pilots x 5 scenarios x 3 stick force gradients
– Circuits, EFATO, Climb-out, Go-around, Base to Finals Turn
Data /Analysis– Workload (Heart Rate, NASA TLX)– Flight dynamics + RT/intercom
12M
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
13M
Typical simulator test results:Typical simulator test results: margin of safety & effect of stick margin of safety & effect of stick
forcesforces
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Example: Pilot 24 Go-around, Full Pwr / Full Flap
Key:Stick Force Gradients
HighMediumLow
Pilot 24 (high hrs):• ATPL• 12,000+ hrs PIC• Age 60~64• Avge 350+ hrs per yr
Non-dimensionalised AoA .v. Elevator
Pilot 23 (medium hrs):• PPL (no IMC) • 1200+ hrs PIC• Age 65~69• Avge 33+ hrs per yr
14M
Elevator
Alpha
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
15
Conclusions Type training - C150 and C152 are different
Design factors - Certification standards are too subjective, consider use of Human Factors tools
Aircraft design, loading & configuration influence airspeed management & stall avoidance via pilot cues
Low stick force gradients can result in:-– Poor airspeed management cues & stall avoidance
cues– More mentally demanding tasks– Increasing workload– Decreasing safety margins
………which all makes loss of control due to distraction or pilot inattention more likely
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Lessons Learned• Independence
– Manage FTI, check maintenance– Always check W&CG reports
• Workup– Essential but spread across sites and aircraft– Use learning curve
• Efficiency– Critical test points– Concentrate flying periods
• But always review data between sorties
• Best practice– Qualified pilot opinion– Redundancy in data collection
16G
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Any Questions?Any Questions?
17M/G
Acknowledgements:-Acknowledgements:- Dr Mark YoungDr Mark Young Thomas Gerald Gray Thomas Gerald Gray
Charitable TrustCharitable Trust GASCo, UK-CAAGASCo, UK-CAA Cranfield & Sheffield Cranfield & Sheffield
UniversitiesUniversities
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
18
Type of Accident (UK) - 1980 to 2006 (GASCo)
Medical/Suicide3%
Airframe Failure4%
Collision with Grnd Object
5%
Mid-air Collision6%
Forced Landing12%
CFIT12%
Low Flying/Aeros16%
Loss of Control IMC8%
Loss of Control VMC25%
Undetermined5%
Low Approach4%
Fixed Wing <5,700kg (non-microlight) fatal
accident causal factors: 1980 to 2006 (UK)
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
PilotAircraft
Dynamics
ConditionError
(ConditionCues)
ControlOutputs
Gust/Turbulence
ActualCondition
TrackingPoint
(Optimal)
ControlSystem
ControlInputs
Boundary(Unsafe
Condition)Actual
Condition
SafetyMargin
(Safety Cues)
+ -
+
-
+
+
Planned Margin
19
BFSL Safety Model
Aircraft
Environ.
Pilot
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Cessna 152
CG Moves aft =Lower stick forces =
More severe stall & spin
20
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
HQR - Climb & Point Track C152, F150M
0
3
6
9
Start Climb
Mid ClimbEnd Climb
HQR Climb & Point Track
G-BCUH F150M
G-BFLU F152
Desirable: +/- 2 MIAS/KIAS Adequate: +/-5 MIAS/KIAS
BTP/Sortie: 2008-06-12A/C= Cessna F150M G-BCUHDate: 05/03/10Gross Wt.= 1599 lbsCG = 35.03" AoD (25.87%MAC/58.76% CG/Mid -CG)Vtrim = 69 KCAS (81 MIAS)
BTP/Sortie: 2008-06-10A/C= Cessna F152 G -BFLUDate: 12/12/09Gross Wt.= 1655 lbsCG = 33.8" AoD (23.78%MAC/50.71% CG/Mid -CG)Vtrim = 69 KCAS
21
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Sim. Experiment 3:Workload Results
22
‘High’ Stick Force :- - Lower Mental Workload- Higher Physical Workload
0
5
10
15
20
25
30
***Practice Circuit***BRS Rwy 27
CircuitBRS Rwy 27
Approach & FULL Flap Landing @65 kts, Rwy 27
2nm700' AGL
(with Go-around)
Base to Finals Turn w/landing
BRS Rwy 27 R/HMid-base750' AGL
Total Workload - All Stick Force Gradients
Total Workload - High
Total Workload - Medium
Total Workoad - Low
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
***Practice Circuit***BRS Rwy 27
CircuitBRS Rwy 27
Approach & FULL Flap Landing @65 kts, Rwy 27
2nm700' AGL
(with Go-around)
Base to Finals Turn w/landing
BRS Rwy 27 R/HMid-base750' AGL
Mental Demand - All Stick Force Gradients
Mental Demand - High
Mental Demand - Medium
Mental Demand - Low
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
***Practice Circuit***BRS Rwy 27
CircuitBRS Rwy 27
Approach & FULL Flap Landing @65 kts, Rwy 27
2nm700' AGL
(with Go-around)
Base to Finals Turn w/landing
BRS Rwy 27 R/HMid-base750' AGL
Physical Demand - All Stick Force Gradient
Physical Demand - High
Physical Demand - Medium
Physical Demand - Low
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
Example: Pilot 24Climb-out, Full Pwr / Zero Flap
Key:Stick Force Gradients
HighMediumLow
Pilot 24 (high hrs):• ATPL• 12,000+ hrs PIC• Age 60~64• Avge 350+ hrs per yr
Non-dimensionalised AoA .v. elevator
Pilot 23 (medium hrs):• PPL (no IMC) • 1200+ hrs PIC• Age 65~69• Avge 33+ hrs per yr
23
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
24
Example: Handling assessment in climb-out
1.3 VStall
1.2 VStall
1.1 VStall
Point Tracking
(Maintain)
Boundary Avoidance
(Avoid)
VVStallStall
13 (Sat.)
Speed
Time
46 (Unsat.)
79 (Unaccept.)
10+ Control lost10+ Control lost
Cooper-Harper for boundary avoidance – can be automated
Workload
‘Top-down’
‘Bottom-up’@crossover switch from ‘top-down’
(point tracking) to ‘bottom-up’
(boundary avoidance)
www.brunel.ac.uk/about/acad/sed/sedres/cem/bfsl
School of Engineering and Design
25M
Conclusions – Causes of LoC
• Stall Boundary crossing (not point tracking)– Handling characteristics & workload
• Type training– A C150 is not a C152
• Design factors– Certification standards are too subjective– Consider Human Factors tools
• Further work needed– Historical certification standards for stalling– Apparent LSS results and models (FCMC in longstab calcs)