Document
Engine Yaw Augmentation for Hybrid-Wing-Body Aircraft
via Optimal Control Allocation Techniques !
NASA Dryden Flight Research Center !
Brian R. Taylor !
Seung Y. Yoo !
AIAA GNC 2011 !
Agenda !
• Motivation !
• Prior Research !
• Aircraft Background !
• Approach !
• System Architecture !
– Diagram !
– Optimal Allocator Objectives !
• Results !
– Nominal !
• Frequency Response !
• Attainable Moment Set !
– Robustness !
• Aerodynamic Modeling Errors !
• Sensor Noise !
• Conclusions !
AIAA GNC 2011 !
Motivation !
• Environmentally Responsible Aviation project goal: !
– Improve noise and efficiency of future aircraft !
• Hybrid-Wing-Body (HWB) aircraft have potential to reduce fuel burn
and noise compared with current aircraft !
• Due to lack of a large vertical tail with a large moment arm aft of the
center of gravity, HWB aircraft tend to have reduced bare airframe
yaw stability and control !
– Some aircraft augment with a closed-loop flight control system which uses split ailerons to create yaw moment with asymmetric drag !
Use asymmetric engine thrust to reduce control surface deflection !
AIAA GNC 2011 !
Prior Research !
• Propulsion controlled aircraft research began following the complete
loss of hydraulic power on United Airlines Flight 232 !
– Pilots manually operated engines in order to control the aircraft and attempt a landing in Sioux City, Iowa !
• Led to extensive research in the use of propulsion control to replace
or augment the control authority of the baseline aircraft in the event
of failures !
• Research on thrust vectoring to reduce trim drag reductions on a
NASA F-15 aircraft !
– 3.5% drag reduction for pitch thrust vectoring !
– 1.5% drag reduction for yaw thrust vectoring !
Controller to reduce surface activity during trim and low frequency
inputs on HWB aircraft !
Implemented as an add-on to the baseline control laws !
AIAA GNC 2011 !
X-48B Background !
• Research partnership of Boeing, NASA, and AFRL !
– Design and fabrication contracted to Cranfield Aerospace !
• Airframe !
– Remotely piloted from ground control station !
– 8.5% dynamically scaled (rigid body) !
• Wingspan: 20.4 ft !
• Weight: 525 lbf !
• Thrust: 54 lbf each (3 JetCat turbojet engines) !
– Closed-loop flight control system !
– 20 control surfaces !
• 4 split ailerons !
• 2 winglet rudders !
AIAA GNC 2011 !
Approach !
• Engine yaw control implemented as an add-on to the baseline
control laws !
– Objective: to reduce the amount of control surface deflection while not degrading performance of baseline control laws !
– Baseline control laws have no “knowledge” of the add-on !
• Optimal control allocation techniques used to determine the optimal
thrust for each engine !
– Track the total thrust command from the power lever angles (PLA) !
– Generate yaw moment to drive the split ailerons to zero !
– Keep the individual engines as close to their individual PLA commands as possible !
AIAA GNC 2011 !
System Diagram !
• Rate limiting: !
– +/- 3% of the total engine thrust !
– Step size 1% of the total engine thrust !
– 343 computations per frame !
AIAA GNC 2011 !
Optimal Allocator Objectives !
• Optimal control allocation techniques used to determine the optimal
thrust for each engine !
– Track the total thrust command from the power lever angles (PLA) !
– Generate yaw moment to drive the split ailerons to zero !
– Keep the individual engines as close to their individual PLA commands as possible !
Yaw Moment Individual PLA Total Thrust 2 2 2
J = T − u + ε v − Bu + γ T − u + T − u + T − u
( ) ∑ ( ) ( ) ( )
i i 1 1 2 2 3 3 i = 1 u ≤ u ≤ u Subject to: i i i min max
€
AIAA GNC 2011 !
€
Implementation !
• Implemented in X-48B non-linear simulation !
– Ensure engine yaw add-on did not degrade performance of baseline control laws !
– Measure the benefits of the engine yaw controller !
– Ensure engine yaw controller is robust to modeling errors and instrumentation noise !
• Simulation tests: !
– Lateral-Directional frequency response !
– Benefits and performance around the attainable moment set boundary !
– Aerodynamic modeling errors !
– Instrumentation noise !
AIAA GNC 2011 !
Frequency Response !
• Frequency sweep of rudder to sideslip angle with engine add-on
turned on and off !
• Matches well below 6 rad/s !
– Engine response has 6 dB attenuation at 6 rad/s !
ï 1 0 Amplitude Ratio (dB) 10 10 EYC Off EYC On 0 Phase Angle (deg) ï 1 0 10 10 AIAA GNC 2011 !
Freq (rad/sec)
Attainable Moment Set Testing !
• 2% rudder step !
• PLA ramp at 80 seconds !
• Total thrust tracked before and after the PLA ramp !
0.95 0.9 0.85 0.8 0.75 0.7 0.65 Normalized Total Thrust 0.6 Command Response 0.55 0 20 40 60 80 100 120 140 Time (sec) AIAA GNC 2011 !
Attainable Moment Set Testing !
• Asymmetric thrust within attainable moment set !
• Convergence time approximately 50 seconds !
Engine 1 ramped to maintain asymmetric thrust Engine 1 Engine 2 0.9 0.9 Engine 3 Total Thrust Command 0.8 0.8 0.7 0.7 0.6 0.6 0.5 0.5 Normalized Thrust 0.4 Normalized Thrust 0.4 0.3 0.3 0.2 0 20 40 60 80 100 120 140 AIAA GNC 2011 ! Time (sec) 0.2 0 20 40 60 80 100 120 140 Time (sec)
Attainable Moment Set Testing !
• Split aileron deflection is driven to zero !
– Due to baseline control laws, winglet rudder deflection is also driven to zero !
• As the asymmetric thrust is reduced to track total thrust, surface
deflection is increased to meet the yaw command !
Left Upper Surface Right Lower Surface Right Split Aileron (deg) Winglet Rudder Deflection (deg) 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Time (sec) Time (sec) AIAA GNC 2011 !
Aerodynamic Modeling Errors !
• 2% rudder step and +/-50% error used on the aerodynamic model
affecting the aero achieved yawing moment !
• Convergence times change with error, but similar steady state is
reached !
Engine 1 No Mdl Error Engine 2 No Mdl Error 0.9 0.9 Engine 3 No Mdl Error Engine 1 +50% Mdl Error 0.8 0.8 Engine 2 +50% Mdl Error Engine 3 +50% Mdl Error 0.7 0.7 Engine 1 ï 50% Mdl Error Engine 2 ï 50% Mdl Error 0.6 Engine 3 ï 50% Mdl Error 0.6 Total Thrust Command 0.5 0.5 Normalized Thrust Normalized Thrust 0.4 0.4 0.3 0.3 0.2 0 20 40 60 80 100 120 140 0.2 Time (sec) 0 20 40 60 80 100 120 140 AIAA GNC 2011 !
Time (sec)
Aerodynamic Modeling Errors !
• Split ailerons are still driven to zero deflection !
• Convergence time is about 10 sec shorter for +50% error and
approximately 30 sec longer for -50% error compared to nominal !
Upper Surface No Mdl Error Lower Surface No Mdl Error Right Split Aileron (deg) Upper Surface +50% Mdl Error No Mdl Error Lower Surface +50% Mdl Error Winglet Rudder Deflection (deg) +50% Mdl Error Upper Surface ï 50% Mdl Error ï 50% Mdl Error Lower Surface ï 50% Mdl Error 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Time (sec) Time (sec) AIAA GNC 2011 !
Instrumentation Noise !
• Noise added to measured RPM signals equal to 5% of the maximum !
• Low pass filters (0.5 rad/s ) on all of the signals from the sensors !
• Increased noise in the thrust commands, but similar steady state
values !
Engine 1 no Noise 0.9 Engine 2 no Noise 0.9 Engine 3 no Noise 0.8 Engine 1 with Noise 0.8 Engine 2 with Noise 0.7 Engine 3 with Noise 0.7 Total Thrust Command 0.6 0.6 0.5 0.4 0.5 Normalized Thrust Normalized Thrust 0.3 0.4 0.2 0.3 0.1 0 20 40 60 80 100 120 140 0.2 Time (sec) 0 20 40 60 80 100 120 140 AIAA GNC 2011 !
Time (sec)
Instrumentation Noise !
• Convergence time is approximately 25 seconds slower with noise on
the measured RPM signals !
• Right split aileron is still driven to zero deflection !
No Noise With Noise Right Split Aileron (deg) Upper Surface no Noise Lower Surface no Noise Winglet Rudder Deflection (deg) Upper Surface with Noise Lower Surface with Noise 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Time (sec) Time (sec) AIAA GNC 2011 !
Drag Reduction !
• Potential to reduce drag by 2 – 4% depending on flight condition !
• Estimated from the maximum yaw capability of the engine add-on
and the amount of split aileron required to create the same moment !
• 2.8% drag reduction for this case seen in flight: !
ï 40 Upper Surface Lower Surface ï 30 ï 20 ï 10 Split Aileron Surface Position (deg) AIAA GNC 2011 !
0 50 100 150 200 250 300 350 400 Time (sec)
Design Considerations !
• Potential drag reduction from engine yaw add-on !
• Operates engines off-nominal condition which may increase the
amount of fuel used even though drag is reduced !
– Application dependent !
• Amount of closed-loop yaw stability and control needed and the type of surfaces used !
• Thrust and location of engines !
• Specific fuel consumption for off-nominal operation !
– Can implement fuel flow as the second objective in the cost function instead of yaw moment to take into account the effect of operating engines off-nominal conditions !
• Additional tool and trade-off for aircraft designers !
AIAA GNC 2011 !
Conclusions !
• Add-on to baseline control laws !
• Asymmetric engine thrust was used to reduce
deployment of aerodynamic surfaces !
– Objectives: !
• Preserve baseline aircraft control characteristics !
• Reduce drag !
• Optimal control allocation techniques: !
– Track total thrust command !
– Generate yaw moment to drive split ailerons to zero !
– Keep individual engines close to PLA !
• Robust to aerodynamic modeling errors
and RPM noise !
– Convergence time differences, but similar steady state !
• Drag reduction of 2 – 4% !
• Planned for flight research on X-48C in 2012 !
AIAA GNC 2011 !
AIAA GNC 2011 !
Your Title Here