Dissertations and Theses - Aerospace Engineering
http://hdl.handle.net/2142/14800
Wed, 23 Aug 2017 12:05:52 GMT2017-08-23T12:05:52ZAerodynamics of lifting surfaces in propeller slipstreams
http://hdl.handle.net/2142/95413
Aerodynamics of lifting surfaces in propeller slipstreams
Chadha, Sparsh Adhir
The high-speed flow in the wake of the propeller also known as propeller wash, or simply propwash, can severely affect the aerodynamic forces on a lifting surface. Steady-state computational results for a symmetric SD8020 airfoil with chord length of 1 ft in a propeller slipstream at a freestream Reynolds number of 100,000 are presented in this study. ANSYS FLUENT was used to solve the flow equations inside the control volume. For the two-dimensional analysis, the propeller was modeled as an actuator line across which the nondimensionalized pressure jump was varied from 3.4 to 13.6. The aerodynamic performance was obtained for the airfoil in configurations with varying horizontal and vertical distance between the actuator line center and airfoil leading edge as well as different diameter-to-chord ratios. As compared with the clean configuration, the lift coefficient and drag coefficient increased by a factor of 5 and 25, respectively, for the slipstream with the highest pressure jump case. The two-dimensional lift curve remained linear throughout the angle of attack range from 0 to 12 deg, and aerodynamic stall was not observed for the computed cases. In general, reducing the diameter-to-chord ratio and shifting the airfoil downstream improved aerodynamic performance. Vertical offset in the airfoil location affected the local flow due to the slipstream and resulted in a low wing configuration producing high lift, and low drag relative to the baseline configuration. Three-dimensional simulations were performed with a circular actuator disk and a rectangular span lifting surface with a semi-span of 1 ft. Due to the wall mirroring effect, the setup simulated a system with infinite propellers upstream of a lifting surface with infinite span. A high spanwise variation of lift in the slipstream shear layer resulted in induced trailing vortices. The trailing vortices caused downwash on the sections within the slipstream flow and upwash on the sections located outside the slipstream which led to an early onset of stall on the outboard sections.
Aerodynamics; Computational fluid dynamics (CFD); Airfoils; Wings; Propeller Slipstreams; Distributed Propulsion; Propeller wing interaction; Slipstream effects
Fri, 09 Dec 2016 00:00:00 GMThttp://hdl.handle.net/2142/954132016-12-09T00:00:00ZChadha, Sparsh AdhirDynamic modeling of a robotic spacecraft for NASA’s asteroid redirect robotic mission
http://hdl.handle.net/2142/95399
Dynamic modeling of a robotic spacecraft for NASA’s asteroid redirect robotic mission
Saxena, Ayush ASaxena
NASA’s Asteroid Redirect Robotic Mission (ARRM) aims to pick up a boulder from the surface of a large asteroid and transport it to a distant retrograde orbit around the Moon for future exploration by a manned mission. This thesis presents a detailed analysis of the dynamic modeling of the ARRM spacecraft grasping the boulder. This model is used for three-axis attitude control design and simulation of the system.
This thesis presents a 30 degree-of-freedom nonlinear lumped-mass model for the structural dynamics of the spacecraft-boulder system. This model is derived using the Euler-Lagrange formulation and simulated in the Matlab-Simulink environment. Another model is derived using Kane’s formulation and SD/FAST, a software package dedicated to deriving dynamic models. Both models are linearized numerically about an equilibrium point. The frequency domain analysis of these linearized models is presented to understand the system behavior and dominant modes. Both models are compared to each other and to an independently developed finite-element model to validate the modeling approach.
Dynamics; Modeling; National aeronautics and space administration (NASA); Asteroid redirect robotic mission (ARRM); Asteroid; Euler; Lagrange
Wed, 07 Dec 2016 00:00:00 GMThttp://hdl.handle.net/2142/953992016-12-07T00:00:00ZSaxena, Ayush ASaxenaAdjoint-based optimization for hyperbolic balance laws in the presence of discontinuities
http://hdl.handle.net/2142/95395
Adjoint-based optimization for hyperbolic balance laws in the presence of discontinuities
Fikl, Alexandru
In this thesis, we are interested in optimization in multiphase flows using discrete adjoint-based methods. The main issues we will endeavor to study are the impact on the linearized and adjoint equations of discontinuous solutions, the behavior of the THINC-family of numerical schemes under linearization and the rigorous formulation of an optimization problem when the interface is represented by a Heaviside marker function.
We will study the two main discontinuous wave patterns that appear in hyperbolic balance laws: contact discontinuities and shocks. These two have very different characteristics in the forward problem and we will see that they require different treatments for the adjoint formulations as well. Specifically, shocks require additional information, in the form of the shock location, to properly define first-order variations necessary for gradient methods. Another important aspect in the numerical treatment of multiphase flows is the use of an anti-diffusive numerical scheme for the interface advection equation. The scheme we will investigate is the fairly recent THINC scheme, because it is one of the few numerical schemes in the field that is differentiable (with respect to the volume fraction). However, we will need to extend the classic formulation of the THINC scheme to force it to behave correctly in the linearized and adjoint regime. Even with these extensions, we will see that the adjoint converges everywhere except at the interface, due to the discontinuity. Finally, we define a simplified optimization problem for the volume fraction, where the velocity is given by a velocity potential instead of the Navier-Stokes equations. This allows us to specifically study the formulation of the cost functional and the viability of using a Heaviside function as a representation of the interface (as opposed to e.g. level set methods). We will show that such a formulation is indeed possible and leads to well-posed optimization problems.
Adjoint; Optimization; Hyperbolic; Conservation laws; Interface; Thinc
Tue, 06 Dec 2016 00:00:00 GMThttp://hdl.handle.net/2142/953952016-12-06T00:00:00ZFikl, AlexandruPiecewise Bézier curve trajectory generation and control for quadrotors
http://hdl.handle.net/2142/95352
Piecewise Bézier curve trajectory generation and control for quadrotors
Lakshmanan, Arun
Quadrotors have the capability of being immensely useful vehicles to aid humans in labor intensive tasks. The critical challenge of using quadrotors inside homes is the efficient navigation of these vehicles in tight spaces while avoiding obstacles. Although methods exists to generate collision free trajectories, they often do not account for the dynamics of quadrotor.
This thesis presents an approach for trajectory generation and control that can harness the complete dynamics of the quadrotor to achieve efficient navigation in cluttered spaces. First, the equations of motion for a quadrotor model is derived. It is also shown that the quadrotor system is differentially flat, which allows the analytical conversion of a time parameterized trajectory to states and outputs of the vehicle. Next, the thesis describes a control design approach in the non-Euclidean state space of quadrotors to achieve improved tracking performance for complex trajectories.
A novel trajectory generation method is presented to achieve smooth and graceful paths for quadrotors. The trajectory generation is formulated as an optimization problem that generates piecewise Bézier curves which minimize snap over the complete trajectory. The optimization method generates these trajectories from collision-free waypoints indicative of the constrained environment. Further, the Bézier curves are time parameterized to satisfy the dynamic constraints and ensure feasibility.
quadrotors; trajectory generation; optimization; nonlinear control systems; geometric controller; minimum snap; bezier curve; differential flatness
Fri, 09 Dec 2016 00:00:00 GMThttp://hdl.handle.net/2142/953522016-12-09T00:00:00ZLakshmanan, ArunNonlinear attitude control of spacecraft with strain-actuated solar arrays
http://hdl.handle.net/2142/95273
Nonlinear attitude control of spacecraft with strain-actuated solar arrays
Nakka, Yashwanth Kumar
This thesis presents a mathematical framework for precision attitude control of a spacecraft using the inertial coupling between the spacecraft and solar arrays. The spacecraft with solar arrays is modeled as a one degree of freedom cylinder (rigid body rotation) with flexible appendages (infinite-dimensional system). The equations of motion that describe system evolution are derived using the extend generalizations of the Lagrangian for infinite dimension systems. Precision attitude control is achieved by bending the flexible appendage using strain actuators. Global asymptotic convergence of the controller’s is proved using the Lyapunov direct method, which ensures that the control objectives of trajectory tracking and slewing are achieved. The Input-to-State stability of these controllers is used to generalize the control laws in terms of a variable that scales the stiffness term. The closed-loop system is simulated numerically for different values of the variable to verify stability.
An experimental setup, that mimics a spacecraft with solar arrays is designed as a cylinder that is secured to a flexible beam using an interference fit. The strain actuation of the beam is achieved using piezoelectric actuators. The rotation of the cylinder and bending in beam are estimated using measurements from a Vicon motion capture system. The closed-loop system is tested in real-time to achieve controlled rotation of the cylinder.
Attitude control; Nonlinear control; Strain-actuated solar array; Lyapunov direct method; Piezoelectric material
Fri, 02 Sep 2016 00:00:00 GMThttp://hdl.handle.net/2142/952732016-09-02T00:00:00ZNakka, Yashwanth KumarImproving the robustness of GPS direct position estimation
http://hdl.handle.net/2142/95250
Improving the robustness of GPS direct position estimation
Ng, Yuting
Global Positioning System (GPS) receivers are increasingly used for positioning in urban environments and precise timing in critical infrastructures. These scenarios are challenging for GPS receivers because building reflection and obstruction contribute to GPS signal degradation in urban environments, while potential jamming and spoofing attacks disrupt GPS time synchronization in critical infrastructures.
We propose using Direct Position Estimation (DPE), augmented with additional navigation information, to enable robust GPS receiver operation in challenging scenarios. Unlike conventional methods, such as scalar and vector tracking, DPE performs Maximum Likelihood Estimation (MLE) of the navigation solution on the raw GPS signal. DPE initializes multiple navigation candidates and searches for the candidate that maximizes the cross-correlation between the expected GPS signal reception and the received GPS signal. The direct search and inherent joint optimization across multiple satellite signals make DPE more robust than scalar and vector tracking. In addition, since the parameter of interest is the navigation solution, DPE provides a natural framework for directly incorporating additional navigation information.
The contribution of this thesis is to design and experimentally validate algorithms for deeply integrating additional navigation information into DPE.
To improve the robustness of DPE in multipath caused by building reflection, we propose transforming non-line-of-sight (NLOS) GPS signals from being unwanted interferences to useful navigation signals. We include NLOS GPS signals into the expected GPS signal reception as additional line-of-sight (LOS) GPS signals to virtual satellites at mirror-image positions. The satellite mirror-image positions are calculated using information of building reflection surfaces, estimated from available three-dimensional (3D) maps. We conducted experiments in front of the 50~m by 40~m wind tunnel located at NASA's Ames Research Center in Mountain View, California, utilizing the surface of the wind tunnel as a reflector of GPS signals. We demonstrated through our experiment, improved robustness in terms of horizontal positioning accuracy, due to the constructive integration of NLOS GPS signals.
In urban environments where GPS sensing is hindered by building obstruction, we propose addressing buildings as additional navigation features instead of undesirable obstacles. We deeply integrate DPE with image map-matching of images captured by an onboard camera against geo-referenced images. The navigation solutions directly estimated from both DPE and image map-matching are fused and used in close-loop GPS signal and camera image tracking. We conducted experiments with joint collections of GPS signals and camera images on our university campus in Urbana, Illinois. We demonstrated, through our experiment, improved robustness in terms of positioning availability, due to the additional vision information.
In addition to positioning, GPS receivers are used for time synchronization in critical infrastructures, where they are vulnerable to malicious attacks. For robust GPS time estimation, we propose using the known, static GPS receiver location as prior information. Estimation of the 3D position, clock bias, 3D velocity and clock drift parameters is reduced to estimation of only the clock bias and clock drift parameters. We conducted experiments on the rooftop of our laboratory in Urbana, Illinois, using the collected signals in simulated jamming and spoofing attacks. We demonstrated improved robustness in terms of anti-jamming and anti-spoofing, due to the information redundancy gained from parameter reduction.
Global Positioning System (GPS); GPS signal; robustness; positioning; position estimation; navigation; urban environment; timing; time synchronization; time estimation; critical infrastructures; jamming; spoofing; Direct Position Estimation (DPE); Maximum Likelihood Estimation (MLE); non-line-of-sight (NLOS); three-dimensional maps; 3D maps; accuracy; map-matching; image; vision; camera; availability; phasor measurement unit (PMU); anti-jamming; anti-spoofing
Wed, 23 Nov 2016 00:00:00 GMThttp://hdl.handle.net/2142/952502016-11-23T00:00:00ZNg, YutingMechanical response of spring films under compression and shear loading
http://hdl.handle.net/2142/89241
Mechanical response of spring films under compression and shear loading
Mott, Ryan Nicholas
Thin films of discrete micro and nanostructures fabricated by the method of Glancing Angle Deposition (GLAD) provide a means to build compliant interfaces that maintain important properties of the constituent materials, such as thermal and electrical conductivity while enabling interfacial resilience. In this dissertation research, the normal compressive and transverse stiffnesses of a variety of Si and Cu spring films were determined experimentally in order to assess the effect of geometric parameters on the mechanical response. Si springs of either 4 or 10 coil turns were deposited on unseeded and seeded Si wafers with seed spacing of 900 nm or 1500 nm. The Cu springs had 10 coil turns and were deposited on silicon wafers with 2000 nm, 2400 nm, 2800 nm, or 3200 nm seed spacing or unseeded Si substrates. Larger seed spacing resulted in Cu springs with larger coil diameters and larger wire thickness compared to seeded Si springs.
Compression tests were conducted at stress amplitudes between 0.5 MPa and 50 MPa on Si films and between 5 MPa and 50 MPa on Cu films. The test samples were circular areas of ~90 μm diameter, subjected to compression with a flat punch. The force vs. displacement curves were used to compute the film stiffness while scanning electron microscope (SEM) images were collected to measure the residual compression. The stiffness of the Si films at the lowest applied stress of 500 kPa varied between 24 ± 0.2 and 66 ± 3.4 MPa for different spring configurations. At the common test condition of 5 MPa applied stress, the stiffness of Si films was between 44 ± 0.2 and 165 ± 1 MPa, while Cu spring films had stiffnesses between 184 ± 2 and 353 ± 15 MPa. At the other extreme of an applied stress of 50 MPa, the stiffness of Si films ranged between 291 ± 0.5 and 810 ± 6 MPa and of Cu films between 611 ± 5 and 1308 ± 28 MPa. Notably, the Si films experienced more permanent deformation at lower stresses compared to Cu, reaching 6.5% at 5 MPa, while Cu films showed no permanent strain until 20 MPa, at which point they experienced only 2% permanent strain. The maximum permanent strains occurring at 50 MPa were 38% for Si and only 12% for Cu.
Shear tests were performed with both types of films using a custom apparatus. The shear stiffness was between 7 ± 0.6 and 27 ± 4 MPa for Si, and between 218 ± 37 and 322 ± 85 MPa for Cu. The higher stiffness of Cu films originated in their significantly larger coil and wire diameter compared to Si. However, the shear strength of seeded Cu springs, between 2 ± 0.4 and 4 ± 1.2 MPa, was approximately the same as that of Si, which had a range of 1 ± 0.05 to 4 ± 0.7 MPa. Cu springs failed at the seeding post which was the most slender point in the structure. Unseeded Cu springs failed within the spring layer with shear strength of 16 ± 0.9 MPa. The largest residual compression was measured for unseeded Cu films reaching a value of 19 ± 2.2 %. In shear, the Si films experienced failure at different locations between the capping layer and just above the seed post, whereas the seeded Cu springs experienced failure directly at the seed post, which prevented the determination of the true shear strength of seeded Cu spring films.
Nanosprings; Shear; Compression; Cyclic; Loading; Response; Springs; Microsprings
Fri, 11 Dec 2015 00:00:00 GMThttp://hdl.handle.net/2142/892412015-12-11T00:00:00ZMott, Ryan NicholasAutonomous visual-inertial navigation and absolute visual scale estimation
http://hdl.handle.net/2142/89154
Autonomous visual-inertial navigation and absolute visual scale estimation
Deng, Xinke
In this thesis, we present a system that uses a single camera and an inertial
measurement unit (IMU) to navigate an Unmanned Aerial Vehicle (UAV) in a
previously unknown environment. The approach consists of two parts. First,
we apply a state-of-the-art simultaneous localization and mapping (SLAM)
method to the video stream of a onboard camera. From the SLAM system,
an up-to-a-scale pose of the camera is estimated, because the absolute size
of the environment cannot be estimated with a single camera. Second, the
estimated pose is fused with the data from IMU to resolve the scale ambiguity.
While analyzing the performance of the system, we find that the conver-
gence rate of scale decreases when the magnitude of scale increases. This
relationship has not been demonstrated and explained before. In this thesis,
we present an analysis and explanation of this phenomenon.
Visual-inertial Navigation; Scale Estimation
Tue, 08 Dec 2015 00:00:00 GMThttp://hdl.handle.net/2142/891542015-12-08T00:00:00ZDeng, XinkeShape optimization of cambered airfoils using a genetic algorithm and a multipoint inverse method
http://hdl.handle.net/2142/89100
Shape optimization of cambered airfoils using a genetic algorithm and a multipoint inverse method
Tsai, Kyle
This thesis presents a process to optimize a cambered airfoil with the MATLAB genetic algorithm (GA) and a multipoint inverse method called PROFOIL. XFOIL was used to evaluate the aerodynamic performance of each airfoil. Data processing techniques and a custom penalty function were developed in order to overcome challenges in integrating these tools. The viability of this approach was assessed in three airfoil optimization studies. First, the optimizer was tuned using a parameterized study of various GA configurations for optimizing the (Cl/Cd)max for an 18% thick airfoil at the design conditions: Cm = -0.063 at Re = 6.88 * 10^6 and Cm = -0.030 at Re = 2.00 * 10^6. The first is a typical flow condition for a wind turbine at the r/R = 0.75 blade section location, and the second is identical to the requirements used in designing the Liebeck L1003. This tuned GA was used for the rest of the thesis. In the second study, the optimization of (Cl/Cd)max for an 18% thick airfoil with design Cm = -0.060 was conducted at Re = 6.00 * 10^6, which is a typical condition for general aviation aircraft. It was observed that the optimized airfoil resembles the Liebeck L1003 airfoil, which was designed with a Stratford pressure recovery distribution. In the third study, a series of Clmax optimization runs was performed for varying pitching moments at Re = 6.00 * 10^6, revealing final solutions that segregated into two types of airfoils that differ in camber. It was shown that the optimizer converged on reflexed airfoils for design coefficients of moment Cm = 0.000 through Cm = -0.050 and on aft-loaded airfoils for Cm = -0.075 through Cm = -0.200. In addition, both groupings of airfoils exhibited an increase in Clmax concomitant with increasing nose-down pitching moment. The results indicate that this approach can reproduce airfoils designed with central design philosophies using only a limited number of design inputs.
Genetic Algorithm (GA); Evolutionary Algorithms (EA); airfoil shape optimization; inverse design; multipoint inverse design; PROFOIL; MATLAB; XFOIL
Fri, 11 Dec 2015 00:00:00 GMThttp://hdl.handle.net/2142/891002015-12-11T00:00:00ZTsai, KyleDevelopment of an AMR Octree DSMC approach for shock dominated flows
http://hdl.handle.net/2142/89046
Development of an AMR Octree DSMC approach for shock dominated flows
Sawant, Saurabh S
Key strategies used in the development of a scalable, three-dimensional direct simulation Monte Carlo (DSMC) program are described.
The code employs an Octree based adaptive mesh refinement (AMR) that gives flexibility in capturing multi-scale physics.
It is coupled with a robust cut-cell algorithm to incorporate complex triangulated geometries.
With the use of distributed memory systems and Message-Passing-Interface (MPI) for communication, the code is potentially scalable.
However, to simulate continuum-like conditions involving multi-scale physics, better scalability that has yet been achieved is desirable.
The thesis identifies two main performance bottlenecks in simulating at continuum-like conditions, first, improving the scalability of the code for more than 128 processors by reducing the communication and evenly balancing the computational load, and second, improve the algorithmic performance of the code by eliminating the expensive recursive tree traversal inherent in Octree based mesh structure.
In order to resolve the first issue sophisticated graph-partitioners have been used, however, without success.
The thesis also explains the special considerations required for embedded geometries in a parallel computational environment.
An efficient algorithm is discussed that allows for the checking of particle-surface interaction only if they are close enough to the geometry.
The code calculates various surface coefficients and employs the Borgnakke-Larsen continuous relaxation model to simulate inelastic collisions of diatomic molecules.
Finally, these strategies and models are validated by simulating hypersonic flows of argon and nitrogen over a hemisphere and double-wedge configuration and the solutions are compared with the results obtained from an older DSMC code known as SMILE.
Direct simulation monte carlo; Hypersonic; Octree; Adaptive Mesh Refinement (AMR); Scalable Unstructured Gasdynamic Adaptive mesh Refinement (SUGAR); Shock dominated flows
Fri, 11 Dec 2015 00:00:00 GMThttp://hdl.handle.net/2142/890462015-12-11T00:00:00ZSawant, Saurabh S