Gregory H. Parker

Flight Test Measurement and Assessment of a Flapping Micro Air Vehicle

Flight test of flapping micro air vehicles (FMAVs) is carried out using an instrumented measurement system to obtain various engineering parameters and hence to assess the flight performance of the vehicles through the data investigation. An indoor flight test facility equipped with a motion capture system and tracking cameras is used for the work presented in this paper. Maneuvers including straight-level flight, ground flapping, takeoff and landing are tested. Spatial position and orientation data are obtained from the retro-reflective tracking markers attached to the vehicles. Subsequent test analysis is carried out by generating performance parameters from raw data and then assessing the flight performance by comparison of the vehicles. The main findings of this work confirm that the test method and procedures presented here enable the systematic numerical data measurement and assessment of the flying performances of these vehicles, and show the applicability for the test and evaluation of general flapping MAVs.

Computational Aeroelastic Analysis of Micro Air Vehicle With Experimentally Determined Modes

This paper presents the results of a combined computational and experimental aeroelastic analysis of a Micro Air Vehicle. This MAV has a 24 inch wing span, and is designed for local area reconnaissance. The fuselage and leading edge of the wings are made of carbon flber, with nylon parachute material spanned by carbon flber ribs making up the rest of the wing. Wind tunnel data for the MAV with a rigid carbon flber wing and a ∞exible carbon flber ribbed nylon wing are used to validate the computational approach. The general aeroelastic algorithm consists of coupling a structural model based on experimentally determined mode shapes and a conventional deforming grid CFD solution. The mode shapes used in the structural model are measured using laser vibrometry on the harmonically excited vehicle. The aeroelastic algorithm is used to calculate static structural deformations coupled with a steady ∞ow solution, as well as the dynamic structural response coupled with a time accurate ∞ow solution. Quantitative comparisons of the statically deformed results are made with wind-tunnel testing, and a qualitative analysis of the dynamic response is also presented.

Computational Aeroelastic Analysis of Store-Induced Limit-Cycle Oscillation

Limit-cycle oscillation was simulated for a rectangular wing referred to as the Goland + wing. It was found that the aerodynamic nonlinearity responsible for limit-cycle oscillation in the Goland + wing was shock motion and the periodic appearance/disappearance of shocks. The Goland + structural model was such that in the transonic flutter dip region, the primary bending and twisting modes were in phase and coupled to produce a single-degree-of-freedom, torsional flutter mode about a point located ahead of the leading edge of the wing. It was determined that the combination of strong trailing-edge and lambda shocks which periodically appear/disappear, limited the energy flow into the structure. This mechanism quenched the growth of the flutter, resulting in a steady limit-cycle oscillation. Underwing and tip stores were added to the Goland + wing to determine how they affected limit-cycle oscillation. It was found that the aerodynamic forces on the store transferred additional energy into the structure increasing the amplitude of the limit-cycle oscillation. However, it was also found that the underwing store interfered with the airflow on the bottom of the wing, which limited the amplitude of the limit-cycle oscillation.

The Physics of an Optimized Flapping Wing Micro Air Vehicle

An optimized stroke path for a flapping wing micro air vehicle in hover is studied. A flapping plate in quiescent flow is used as the basis for the study. The body is modeled as a thin two-dimensional flat plate and is represented as a collection of discrete vortices. Unsteady effects, including leading-edge separation, are captured by vortex shedding. Singleparameter studies suggest an angle of incidence of approximately 40 is optimal for simple flapping motion. A more complex figure-eight stroke path with a fixed translational stroke and a variable angle of incidence based on B-splines is imposed. Using three B-splines, an optimal solution for the angle of incedence as a function of stroke path is presented. Small changes in design variables can yield local maxima which can challenge traditional gradient-based optimization techniques.

A Design Optimization Strategy for Micro Air Vehicles

This paper presents progress on the development of a computational framework for synthesizing and optimizing the design of micro air vehicles with flapping wings. It is anticipated that aeroelastic tailoring can be used to improve the performance of mechanical flapping-wing micro air vehicles relative to designs with rigid structure. Due to the physics associated with micro air vehicles, the design framework must be able to capture nonlinear physical behavior with high fidelity, while at the same time maintaining the computational efficiency required for effective design space exploration. In the present paper, the authors present three newly developed components of the framework: a spectral element method for computing the system dynamics, an adjoint method for calculating sensitivities of the system with respect to changing structural design variables and an optimization component built upon existing gradient-based methods. These new components are integrated with existing components and the entire framework is used to compute flapping cycles for both rigid and flexible micro air vehicle models and to demonstrate consistency between the direct and adjoint forms for sensitivity computation.

The Role of Viscosity in Store-Induced Limit-Cycle Oscillation

Aircraft with stores can exhibit an aeroelastic phenomenon characterized by limited amplitude, self-sustaining oscillations produced by aerodynamic-structure interactions known as limit-cycle oscillation. In order to study this phenomenon, a first order accurate code has been developed to interface a modal structural model with a commercial, parallel, Navier-Stokes fluid solver with a deforming grid capability. The commercial solver chosen was FLUENT 6.1 by FLUENT Inc. Initial testing of this code has been completed on the Goland+ and AGARD 445.6 wings. Limit-cycle oscillation was successfully obtained with the Goland+ wing. Grid refinement was found to slightly decrease the oscillating frequency. The Spalart-Allmaras turbulence model was used in one Goland+ test case. However, due to the length of the run times, conclusions can not yet be drawn. A tip store and underwing stores were successfully added to the Goland+ wing. These stores did not affect the limit-cycle oscillation. The code was also used to get a preliminary look at a documented F-16C limit-cycle oscillation flight test scenario.

Flapping Wing Micro Air Vehicle Bench Test Setup

The purpose of this research was to develop testing methods that can be used to determine the forces, moments, and deflections involved in flapping wing aerodynamics. To pursue the research, a flapping wing mechanism and wings with spans ranging from 9.1 inches to 12.1 inches were built. A variety of mechanisms, capable of, alternatively, purely flapping, flapping with pitch, and flapping with pitch and out-of-plane motion were conceptualized and drawn using solid modeling software. Two of the simpler designs, a single degree-of-freedom flapping mechanism and the two-degree of freedom flapping mechanism were fabricated using a rapid prototype 3-D printer, and sustained operation was demonstrated. A thrust stand and a six-component force balance were used to gather force data from the flapping-only mechanism, combined with a variety of wing shapes. Four high-speed cameras were used to capture the motion of the wings. To minimize intrusiveness an array of laser dots was projected onto the wing during flapping and photogrammetry software was used to analyze the images and determine a shape profile of the wing composed of a frame and membrane during flapping. While the focus of this research was on the bench test setup development, some insight into the influence of wing design on the forces acting on the mechanism was gained.

A Computational Design Framework for Flapping Micro Air Vehicles

The authors have embarked on the development of a Computational Design Framework as the means to synthesize and optimize highly integrated Micro Air Vehicle designs with flapping wings. This paper describes both the computational physics and the framework requirements that contribute to the overall design. Computational physics address inherently non-linear flapping aerodynamics and non-linear structures. Mathematical models of symbiotic non-linear behavior will feed design optimization in ways that bring out new and beneficial aeroelastic interactions. A preliminary design was initiated with simplified physics in order to begin design connectivity within the design framework. The framework builds on earlier reported developments with software classes (objects) that facilitate coordinated graphical animation using either a single or multiple windows.

The Complicating Effect of Uncertain Flapping Wing Kinematics on Model Validation

In order to experimentally validate an aeroelastic modeling tool, a flapping mechanism has been built to flap a flexible wing structure about a single axis. A comparison of the flexible tip deformation during the flapping stroke is of particular interest. Due to the small size of fully-scaled mechanisms, the “commanded” kinematics may differ substantially from what is observed. A crucial input into any numerical flapping wing model, the temporal derivatives of the flapping kinematics, is then not known with certainty. For the current work, the flap rotation is measured with a non-contact image correlation technique, and a Fourier series fit is used to obtain flapping velocities and accelerations. The resulting match between experimental and numerical tip displacements is satisfactory for a very small range of harmonic number. A recommended strategy for future numerical modeling efforts is to include the entire flapping system (power source, actuation, wings) into the framework in order to improve the validation process, with less dependence upon experimental data to “tune” the computational models input kinematics.

Multidisciplinary Optimization of a Hovering Wing with a Service-Oriented Framework and Experimental Model Validation

An aeroelastic flapping analysis is integrated with a commercial, gradient-based optimizer within a computational framework. The aeroelastic analysis couples a geometrically-nonlinear beam formulation with a quasi-steady blade element aerodynamics tool and trigonometric flapping kinematics. Analytic gradient information is produced for peak power required, cycle-averaged lift, and maximum von Mises stress with respect to element chord and thickness and nine kinematic parameters, alleviating the burden of finitedifference gradients. The chord and thickness distributions and kinematics were simultaneously optimized to provide a wing requiring minimum flapping power under constraints on lift and stress. Three optimized designs are presented, yielding more than 70% reduction in peak power requirement from baseline designs, and 28% reduction from a design produced by another optimization method. This work concludes with an experimental validation of the aeroelastic tool through the comparison of various static, dynamic, and flapping metrics.