Beatrice Roget

Computational Study of Flexible Wing Ornithopter Flight

This paper presents the development and evaluation of a computational fluid dynamics based methodology to predict the aerodynamic forces produced by a flexible flapping wing. The computational fluid dynamics analysis code solves the compressible Reynolds-averaged form of the Navier-Stokes equations on structured curvilinear grids. A grid deformation algorithm is devised that deforms the body-conforming volume grid at each time step consistent with the measured wing motions. This algorithm is based on geometric considerations and is both computationally efficient and capable of handling very large deformations. This methodology is validated using experimental data obtained from a test on an ornithopter with flexible wings. Test data include measurements of the wing surface deformations as well as the generated forces in the horizontal and vertical directions. Correlation with test data shows good agreement with measured vertical force and satisfactory agreement with measured horizontal force at low flapping frequencies. However, the prediction accuracy degrades with an increase in Happing frequency. Evidence of resonance in the vehicle system was detected from the analysis of the experimental data. Unmodeled inertial effects from the vehicle body and support mounts may be one of the contributors to disagreement between the data and analysis.

Experimental Determination of Ornithopter Membrane Wing Shapes Used For Simple Aerodynamic Modeling

This paper addresses a method to analyze flapping membrane wing aerodynamics using kinema tic data. An alternative to computational fluid dynamics, this experimental method places small reflective markers on a wing and tracks them using a Vicon motion measurement system. The dynamic membrane shapes of the ornithopter wing are then quantified as moving points in a three dimensional coordinate system. This motion tracking process was completed for two ornithopters with approximately one meter span rip -stop nylon wings over a series of flapping frequencies. Aerodynamic loads were recorded in additi on to wing shape data to confirm the aerodynamic modeling results. The wing shape data was used to form an analytical aerodynamic model that uses blade element theory and quasi -steady aerodynamics to account for the local twist, stroke angle, membrane shap e, wing velocity and acceleration. Results from the aerodynamic model show good correlation between the magnitude of lift and thrust produced but some phase errors exist between the measured and calculated force curves .

Prediction of Helicopter Maneuver Loads Using a Fluid-Structure Analysis

A fluid-structure analysis framework that couples computational fluid dynamics and computational structural dynamics is constructed to study the aeromechanics of a helicopter rotor system under maneuvering-flight conditions. The computational fluid dynamics approach consists of the solution of unsteady Reynolds-averaged Navier-Stokes equations for the near field of the rotor coupled with the dynamics of trailed vortex wake that is computed using a free-vortex method. The computational structural dynamics approach uses a multibody finite element method to model the rotor hub and blades. The analysis framework is used to study the utility tactical transport aerial system pull-up maneuver of the UH-60A helicopter. Results shown illustrate the correlation of predicted performance, aerodynamic and structural dynamic loading, with measured flight-test data. The normal load factor and the peak-to-peak structural and aerodynamic loading show good correlation with flight-test data, indicating that the analysis framework is suitable for preliminary design purposes. Important phenomena such as advancing-blade transonic effects and retreating blade flow separation are predicted satisfactorily. However, deficiencies are noted in the accurate prediction of stall onset, reattachment, and shock-induced separation.

Wind-Tunnel Testing of Rotor with Individually Controlled Trailing-Edge Flaps for Vibration Reduction

A control method is proposed to reduce vibrations in helicopters using active trailing-edge flaps on the rotor blades. Each blade is controlled independently, taking into account possible blade dissimilarities. The method consists of performing simultaneous system identification and closed-loop control at each time step. For the system identification, different inputs are applied to each blade, and the relationship between the individual blade inputs and the resulting loads in the fixed frame is estimated on-line, assuming a linear-time-periodic model of the helicopter. Closed-loop tests are conducted using a four-bladed Mach-scaled rotor with piezobender trailing-edge flaps. The rotor model is fitted on a bearingless model-scale hub and tested in the Glenn L. Martin wind tunnel. These tests demonstrate the controllers ability to account for blade dissimilarities and generate different optimal inputs for each blade. The 1 and 4/rev components of fixed frame loads are reduced individually by 50 and 60%. Simultaneous reduction of 1 and 4/rev components is also demonstrated (43% reduction). However, vibration increases are noted for some nontarget hub loads.

Robust Individual Blade Control Algorithm for a Dissimilar Rotor

A new control methodology is formulated for vibration reduction at the rotor hub by controlling trailingedge e aps. The novelty of the proposed methodology lies in its ability to control each rotor blade separately and optimally, taking into account blade-to-blade dissimilarities, while using exclusively e xed-frame measurements. The controller is formulated in the time domain and adaptively generates in real time individual control inputs to the trailing-edge e aps to achieve vibration reduction. Numerical simulations using a hingeless rotor model show that thecontrollergeneratescontrolinputsto each blade,taking bladedissimilaritiesinto account, andsuccessfully minimizes hub vibrations. Nomenclature CT=ae = blade loading F = e xed-frame load F0 = uncontrolled e xed-frame load IK = identity matrix of size, K kN=rev = integer multiple of N times the rotor frequency N = number of rotor blades Ns = number of samples per revolution R = rotor radius S = permutation matrix ® = sectional angle of attack ± = e ap dee ection angle @®=@± = e ap effectiveness π = advance ratio A = rotor revolutions per minute Subscripts k = blade number n = iteration number

Design of a Martian Autonomous Rotary-Wing Vehicle

The design of a Martian autonomous rotary wing vehicle (MARV) is described. MARV is a 50-kg gross takeoff mass, coaxial helicopter designed for Mars exploration. Powered by a fuel cell system, it carries a payload of 10.8 kg over a range of 25 km with an endurance of 39 min including hover capability for 1 min. MARV is designed in response to the Request For Proposal from NASA/Sikorsky for the Year 2000 American Helicopter Society student design competition. The design covers aerodynamic and structural design of rotor blades, vehicle power plant, fuselage and landing gear, control system, transmission, and vehicle lander communications. A detailed mechanism for autonomous deployment of the vehicle from the lander is also described. This preliminary design study indicates that controlled vertical flight on Mars is feasible with existing technology

TRAILING EDGE FLAP CONTROL METHODOLOGY FOR VIBRATION REDUCTION OF HELICOPTER WITH DISSIMILAR BLADES

This paper formulates a new control methodology for vibration reduction at the rotor hub by controlling trailing edge flaps. The novelty of the proposed methodology lies in its ability to control each rotor blade separately and optimally taking into account blade-to-blade dissimilarities, while using exclusively fixed frame measurements. The controller is formulated in the time domain, and adaptively generates in real time individual control inputs to the trailing edge flaps to achieve vibration reduction. Numerical simulations using a hingeless rotor model show that the controller generates control inputs to each blade, taking blade dissimilarities into account, and successfully minimizes vertical hub vibrations.