Feedback Driven 6-DoF Modeling of a Tilt-Duct Aircraft STAR Global - - PowerPoint PPT Presentation
Feedback Driven 6-DoF Modeling of a Tilt-Duct Aircraft STAR Global - - PowerPoint PPT Presentation
Feedback Driven 6-DoF Modeling of a Tilt-Duct Aircraft STAR Global Forum May 12, 2010 Paul Vasilescu AD-150 High Speed VTOL UAV - Designed around USMC Tier III requirements (Now Group IV) - VTOL from air capable ships - Exceed 200 kts airspeed
2
AD-150 High Speed VTOL UAV
- Designed around USMC Tier III
requirements (Now Group IV)
- VTOL from air capable ships
- Exceed 200 kts airspeed
Length 14.5 ft Wing Span 17.5 ft Height 4.75 ft Takeoff Weight 2,200 lbs - 2,800 lbs Max Payload 500 lbs - 1,000 lbs Max Speed 300 kts Max Range 736 Nmi
Our Business Model
Create a multi-mission, economical, high-speed VTOL UAV to fill a market void
Team with industry leaders for payload, subsystem and contracting support
Secure production contracts for airframes and lift and propulsion systems
Designed & Manufactured by ADFS in Jessup, MD Designed by ADFS, Manufactured by ADFS and/or Subcontractors Designed & Manufactured by Subcontractors
Turboshaft Powerplant HTAL System Drive System Main Landing Gear Vectoring Nozzle Flight Control Actuators Flaperon Ruddervator Mission Payload Avionics Navigation & Control Power Management Nose Landing Gear HTAL Nacelle Airframe HTAL Fan
AD-150 Development Timeline
2006 2007 2008 2009 2010 2011
Tilt-Duct VTOL ~ Key Challenges
Transition between hover and forward flight
Dynamic conditions
Large pitching moments
Potential duct stall
Low control authority
Static / steady state limitations
Transient 6-DoF solution A full understanding of flight conditions experienced in transition requires a comprehensive transient simulation
6 Use is limited to Proprietary Restriction on the cover 6
AD-150 6-DoF CFD Model
Create a robust 6-DoF CFD simulation to accurately model transition between hover and forward flight
Support changes in geometric configuration between time steps
Duct angle relative to the fuselage
Angle-of-Attack
Support basic autopilot user code
Run on top of the simulation
Read inertial data at each time step
Implement geometric and power changes
Must be accurate but also feasible
Split the propulsion model from the aircraft model
Verify data using multiple CFD methods
Validate CFD models with experimental tests
7 Use is limited to Proprietary Restriction on the cover 7
Steady State Propulsion Model
Star-CCM+ steady state segregated flow model with moving reference frame
Automated workflow solved every case in the test matrix
Test matrix includes variations
Duct Angle
Feestream Velocity
Fan RPM
Measured data includes
6 Component Forces and Moments
Fan Torque
Airflow Velocity (above and below fan disk)
Data was fit to mathematical curves to enable integration into any flight condition modeled by the unsteady model
8 Use is limited to Proprietary Restriction on the cover 8
Aircraft 6-DoF Model
Star-CCM+ unsteady 6-DoF
Nacelles modeled in cylindrical volumes
Enables rotation of the ducts without remeshing
In-place interfaces used between all volumes
Boundary conditions replace rotating fans
Computed dynamically for each time step based on data from the propulsion system CFD
Validated by comparing 6 component forces and moments of both models
Angle-of-Attack changes are achieved by rotating all volumes about the aircraft’s CG
Basic autopilot code sets power, duct angle, and angle-of-attack at each time step
Initial runs enabled translation in all directions and rotation in the yaw axis
9 Use is limited to Proprietary Restriction on the cover 9
Feedback Driven CFD 6-DoF Workflow
10 Use is limited to Proprietary Restriction on the cover 10
AD-150 6-DoF Initial Cases
11 Use is limited to Proprietary Restriction on the cover 11
AD-150 CFD 6-DoF Team
Qlogic NETrack Development Center (NDC)
Test drive HPC systems
16 Dell PowerEdge R610 Compute Nodes w/ Qlogic TrueScale InfiniBand
128 Total CPU Cores and 384 GB RAM
Performance Improvement
Reduced 3000 hours of compute time to 48 hours for the stead state propulsion simulation test matrix
Reduced 93.4 hours of compute time per second of flight time of the unsteady / transient aircraft simulation to 2.4 hours of compute time per second of flight time
Experimental Validation Aircraft Wind Tunnel Test
Built and tested a 1/3 scale model of the
AD-150 aircraft
Tests conducted at the Glen L. Martin Wind
Tunnel in Q1 2009
Performed static and dynamic tests over a wide
range of combinations of angle of attack, sideslip and forward speed
Confirmed CFD performance data and
validated airframe design
Funding provided by the state of Maryland
through the Maryland Industrial Partnerships program (MIPS)
“The American Dynamics Flight Systems team came to the wind tunnel test with a more complete set of CFD results directly comparable to the tunnel conditions than any other group we have worked with up to now.” -- Dr. Jewel Barlow, Director of the Glenn L. Martin Wind Tunnel
Experimental Validation Propulsion Wind Tunnel Test
Build an test a scaled AD-150 propulsion
system in a wind tunnel
Simulate conditions experienced during
transition between hover and forward flight
Test will be conducted and the UMD Glenn L.
Martin Wind
Tests will include variances in freestream
velocity, fan velocity, and duct angle
Measurements will be taken and compared
to equivalent data taken from the computational models
Funding provided by the state of Maryland
through the Maryland Industrial Partnerships program (MIPS)
Experimental Validation Propulsion Static Test
Build and integrate a fully instrumented test rig
for the AD-150 lift and propulsion system
Tests to be conducted at Aberdeen Proving
Grounds later this year
Mechanically validate propulsion and drive
components
Confirm CFD performance data for hover in
ground effect (HIGE) and hover out of ground effect (HOGE)
Provide a comprehensive data set of system
performance and reliability
Funding provided by the state of Maryland
through the Maryland Technology Transfer Fund (MTTF) program
15 Use is limited to Proprietary Restriction on the cover 15
Next Steps
Movable control surfaces
Stability augmentation code
Turboshaft engine code
Replace AP code with software and hardware interfaces to a physical autopilot