Scott Schreck

Real-time prediction of unsteady aerodynamics: Application for aircraft control and manoeuvrability enhancement

The capability to control unsteady separated flow fields could dramatically enhance aircraft agility. To enable control, however, real-time prediction of these flow fields over a broad parameter range must be realized. The present work describes real-time predictions of three-dimensional unsteady separated flow fields and aerodynamic coefficients using neural networks. Unsteady surface-pressure readings were obtained from an airfoil pitched at a constant rate through the static stall angle. All data sets were comprised of 15 simultaneously acquired pressure records and one pitch angle record. Five such records and the associated pitch angle histories were used to train the neural network using a time-series algorithm. Post-training, the input to the network was the pitch angle (alpha), the angular velocity (dalpha/dt), and the initial 15 recorded surface pressures at time (t (0)). Subsequently, the time (t+Deltat) network predictions, for each of the surface pressures, were fed back as the input to the network throughout the pitch history. The results indicated that the neural network accurately predicted the unsteady separated flow fields as well as the aerodynamic coefficients to within 5% of the experimental data. Consistent results were obtained both for the training set as well as for generalization to both other constant pitch rates and to sinusoidal pitch motions. The results clearly indicated that the neural-network model could predict the unsteady surface-pressure distributions and aerodynamic coefficients based solely on angle of attack information. The capability for real-time prediction of both unsteady separated flow fields and aerodynamic coefficients across a wide range of parameters in turn provides a critical step towards the development of control systems targeted at exploiting unsteady aerodynamics for aircraft manoeuvrability enhancement.

Unsteady Fluid Mechanics Applications of Neural Networks

The capability to harness or alleviate unsteady aerodynamic forces and moments could dramatically enhance aircraft control during severe maneuvers as well as signie cantly extend the life span of both helicopter and wind turbine blade/rotor assemblies. Using recursive neural networks, time-dependent models that predict unsteady boundary-layer development, separation, dynamic stall, and dynamic reattachment have been developed. Further, these models of the e ow› wing interactions can be used as the foundation upon which to develop adaptive control systems. The present work describes these capabilities for three-dimensional unsteady surface pressures and two-dimensional unsteady shear-stress measurements obtained for harmonic and constant-rate pitch motions. In the near future, it is predicted that such techniques will provide a viable approach for the development of six degree-of-freedom motion simulators for severe vehicle maneuvers as well as a foundation for the active control of unsteady e uid mechanics in a variety of systems.

Neural networks: Applications and opportunities in aeronautics

Abstract Technologies based on neural networks are currently being developed which may assist in addressing a wide range of complex problems in aeronautics. The review indicates that the proper utilization of this technology offers a feasible approach to help meet current and future technological needs. This might include, but is not limited to, the following: the implementation of active control devices to harness or suppress unsteady aerodynamic effects such as dynamic stall on helicopter rotor blades; the parallel data processing of 10s to 1000s of sensors either for actuation and control or for system and component ‘health’ monitoring and fault detection; the development of simulators and control algorithms for severe, ‘unsteady’, six degree-of-freedom vehicle maneuvers where the linearized equations of motion do not adequately describe the vehicle dynamics; and the requirement to monitor these sensors, simulate the maneuvers and instigate control all on a time-scale which must be faster than the real-time phenomena. Clearly, neural networks alone will not solve these problems. However, neural networks provide a practical approach for determining solutions of complex nonlinear problems such as equations of motion, for the parallel processing of 1000s of sensors, and this can be achieved with the required computational speed.

Horizontal Axis Wind Turbine Blade Aerodynamics in Experiments and Modeling

Turbine aerodynamics remains a challenging and crucial research area for wind energy. Blade aerodynamic forces responsible for power production must be augmented to maximize energy capture. At the same time, adverse aerodynamic loads that fatigue turbine components need to be mitigated to extend machine service life. Successful resolution of these conflicting demands and continued cost of energy reduction require accurate blade aerodynamic models. This, in turn, depends on clear physical understanding and reliable numerical modeling of rotational augmentation and dynamic stall, the two phenomena principally responsible for amplified turbine blade aerodynamic loads. The current work examines full-scale turbine blade aerodynamic measurements and current modeling methodologies to better understand the physical and numerical attributes that determine model performance

Blade Dynamic Stall Vortex Kinematics for a Horizontal Axis Wind Turbine in Yawed Conditions

Horizontal axis wind turbines routinely suffer significant time varying aerodynamic loads that adversely impact structures, mechanical components, and power production. As lighter and more flexible wind turbines are designed to reduce overall cost of energy, greater accuracy and reliability will become even more crucial in future aerodynamics models. However to render calculations tractable, current modeling approaches admit various approximations that can degrade model predictive accuracy. To help understand the impact of these modeling approximations and improve future models, the current effort seeks to document and comprehend the vortex kinematics for three-dimensional, unsteady, vortex dominated flows occurring on horizontal axis wind turbine blades during non-zero yaw conditions. To experimentally characterize these flows, the National Renewable Energy Laboratory Unsteady Aerodynamics Experiment turbine was erected in the NASA Ames 80 ft×120 ft wind tunnel. Then, under strictly-controlled inflow conditions, turbine blade surface pressures and local inflow velocities were acquired at multiple radial locations. Surface pressure histories and normal force records were used to characterize dynamic stall vortex kinematics and normal forces. Stall vortices occupied approximately two-thirds of the aerodynamically active blade span and persisted for nearly one-fourth of the blade rotation cycle. Stall vortex convection varied dramatically along the blade radius, yielding pronounced dynamic stall vortex deformation. Analysis of these data revealed systematic alterations to vortex kinematics due to changes in test section speed, yaw error, and blade span location.

Unsteady vortex dynamics and surface pressure topologies on a finite pitching wing

Abstract : A straight wing having an NACA 0015 cross section and rectangular planform was attached to a circular splitter plate. This configuration was pitched at a constant rate to angles exceeding the static stall angel. The unsteady, vortex-dominated flow that developed over the wing and splitter plate was characterized in detail using surface pressure measurements and flow visualization. Both types of data showed that the leading-edge vortex underwent profound three-dimensional alterations to cross section and convection over the entire wing span. These changes in leading-edge vortex structure and kinematics were correlated with prominent spanwise variations in force coefficients. When appropriately dissected, visualization results and pressure data suggested physical mechanisms to account for these three-dimensional variations in unsteady forces and surface pressures.

Yaw aerodynamics analyzed with three codes in comparison with experiment

Yaw aerodynamics were computed with three codes of different complexity; 1) The 3D Navier Stokes solver Ellipsys3D using 5–8 million grid points; 2) HAWC3D which is a 3D actuator disc model coupled to a blade element model and using 20–30.000 grid points and 3) HAWC, a finite element based aeroelastic code using The Blade Element Momentum (BEM) model for the aerodynamics. Simulations were performed for two experiments. The first is the field rotor measurements on a 100 kW turbine at Risoe where local flow angle (LFA) and local relative velocity (LRV) at one radial station have been measured in a yaw angle interval of ±60°. The other experiment is the NREL measurements on a 10 m rotor in the NASA Ames 80 ft × 120 ft wind tunnel. LFA were measured at five radial stations and data for the 45° yaw case were analyzed. The measured changes in LFA caused by the yawing were used as the main parameter in the comparison with the models. In general a good correlation was found comparing the Ellipsys3D results with the LFA measured on the NREL rotor whereas a systematic underestimation of the amplitude in LFA as function of azimuth was observed for the two other models. This could possibly be ascribed to upwash influence on the measured LFA.Copyright

Pitch Rate and Reynolds Number Effects on Unsteady Boundary-Layer Transition and Separation

A NACA 0015 airfoil was pitched at a constant rate through static stall to elevated angles of attack. Shear stress measurements of high spatial and temporal resolution were performed near the airfoil leading edge, in the vicinity of subsequent dynamic stall vortex initiation. Using these data, unsteady boundary-layer reversal and transition were characterized for a range of nondimensional pitch rates and Reynolds numbers. Analyses revealed the independent influences of nondimensional pitch rate and Reynolds number upon unsteady boundary-layer reversal and transition. Temporal and spatial relationships between unsteady boundary-layer reversal and transition imply that unsteady boundary-layer reversal is a precursor and principal determinant in unsteady boundary-layer transition. Comprehension of these and other fundamental unsteady flow physics are crucial for the control of dynamically separated flows generated by maneuvering aircraft, rotorcraft, and wind energy machines

Turbine Inflow Characterization at the National Wind Technology Center

Utility-scale wind turbines operate in dynamic flows that can vary significantly over timescales from less than a second to several years. To better understand the inflow to utility-scale turbines, two inflow towers were installed and commissioned at the National Renewable Energy Laboratorys (NREL) National Wind Technology Center near Boulder, Colorado, in 2011. These towers are 135 m tall and instrumented with a combination of sonic anemometers, cup anemometers, wind vanes, and temperature measurements to characterize the inflow wind speed and direction, turbulence, stability and thermal stratification to two utility-scale turbines. Herein, we present variations in mean and turbulent wind parameters with height, atmospheric stability, and as a function of wind direction that could be important for turbine operation as well as persistence of turbine wakes. Wind speed, turbulence intensity, and dissipation are all factors that affect turbine performance. Our results show that these all vary with height across the rotor disk, demonstrating the importance of measuring atmospheric conditions that influence wind turbine performance at multiple heights in the rotor disk, rather than relying on extrapolation from lower levels.

Neural network prediction of three-dimensional unsteady separated flowfields

Unsteady surface pressures were measured on a wing pitching beyond static stall. Surface pressure measurements confirmed that the pitching wing generated a rapidly evolving, three-dimensional unsteady surface pressure field. Using these data, both linear and nonlinear neural networks were developed. A novel quasilinear activation function enabled extraction of a linear equation system from the weight matrices of the linear network. This equation set was used to predict unsteady surface pressures and unsteady aerodynamic loads. Neural network predictions were compared directly to measured surface pressures and aerodynamic loads. The neural network accurately predicted both temporal and spatial variations for the unsteady separated flowfield as well as for the aerodynamic loads. Consistent results were obtained using either the linear or nonlinear neural network. In addition, fluid mechanics modeled by the linear equation set were consistent with established vorticity dynamics principles.