<p> Accurate unsteady measurements are required for studying the flows in high speed turbomachines, which rely on the interaction between rotating and stationary components. Using statistics of phase locked ensembles simplifies the problem, but accurate frequency response in the 10-100 kHz range significantly limits the applicable techniques. This research advances the state of the art for phase resolved measurement techniques using for high speed turbomachinery flows focusing on the following areas: development, validation, and uncertainty quantification. Four methods were developed and implemented: an unsteady total pressure probe, the multiple overheat hot-wire method, the slanted hot-wire method, and the phase peak yaw hot-wire method. These methods allow for the entire phase locked average flow field to be measured (temperature, pressure, and velocity components, swirl angle, etc.). No trusted reference measurement or representative canonical flow exists for comparison of the phase resolved quantities, making validation challenging. Five different validation exercises were performed to increase the confidence and explore the range of applicability. These exercises relied on checking for consistency with expected flow features, comparing independent measurements, and cross validation with CFD. The combined uncertainties for the measurements were quantified using uncertainty estimates from investigations into the elemental error sources. The frequency response uncertainty of constant temperature hot-wire system was investigated using a novel method of illuminating the wire with a laser pulse. The uncertainty analysis provided estimates for the uncertainty in the measurements as well as showing the sensitivity to various sources of error.</p>

<p> A robust proportional-integral (PI) controller was synthesized for the F-16 VISTA (Variable stability In-flight Simulator Test Aircraft) using a linear matrix inequality (LMI) approach, with the goal of eventually designing and implementing a linear parameter-varying PI controller on high performance aircraft. The combination of classical and modern control theory provides theoretically guaranteed stability and performance throughout the flight envelope and ease of implementation due to the simplicity of the PI controller structure. The controller is designed by solving a set of LMIs with pole placement constraints. This closed-loop system was simulated in MATLAB/Simulink to analyze the performance of the controller. A robust <i> H</i>_&infin; controller was also developed to compare performance with PI controller. The simulation results showed stability, albeit with poor performance compared to the <i>H</i>_&infin; controlle </p>

<p> An experiment consists of a remote-controlled T-REX 600E Pro helicopter piloted above an underway 108-foot Patrol Craft&rsquo;s (YP) flight deck to analyze the effect of <i>Re_H</i> &ap; 3.4&times;10⁵ turbulent ship air wake on helicopter angular motion. The motion is measured by an inertial measurement unit (IMU) sensor mounted on the helicopter transmitting variably-spaced data in a 3-D Cartesian reference frame. Data is collected with the helicopter in locations above the flight deck which closely match actual takeoff/landing positions of US Navy H-60 helicopters above flight decks of air-capable ships (CG/DDG/FFG). This method is both an indirect way of qualifying the flow field and a practical way of measuring the actual effects of ship air wake turbulence on rotary-wing aircraft. Fourier analysis is performed on helicopter angular velocities to determine predominant frequencies of motion. Frequencies of this motion are compared with both pilot input frequencies and vortex shedding frequencies of incompressible, subsonic flow around 2-D and 3-D backward-facing step (BFS) geometries. These frequencies of motion are discussed in relation to the pilot&rsquo;s perception of motion and the phenomena of spatial disorientation. The spectra of pilot inputs in roll and pitch were found to have cutoff frequencies of 0.5 Hz and 0.7 Hz, respectively, which agree well with full-size helicopter pilot autospectra. A 3 Hz non-pilot initiated disturbance is noted; the source of this disturbance is thought to be due to interactions with airflow over the flight deck hangar or dynamic effects of the flybar on the helicopter.</p>

<p> Modern unmanned aerial vehicles (UAV) are made of lightweight structures, owing to the demand for longer ranges and heavier payloads. These lightweight aircraft are more susceptible to vibrations caused by atmospheric turbulence transmitted to the fuselage from the wings. These vibrations, which can cause damage to the payload or on board avionics present a serious problem, since air turbulence is expected to increase over the next few decades, due to climate change. </p><p> The objective of this thesis is to predict the vibration of an aircraft fuselage by establishing a relationship between wing and fuselage vibration. A combination of ANSYS^&reg; and MATLAB^&reg; modeling are used to simulate aircraft vibrations. First, the displacement of a lumped mass aircraft model to step and sinusoidal forces acting on the wings are compared to displacements calculated using modal superposition equations. Next, a state space representation of this system is found using system identification techniques, which uses wing displacement as input, and provides fuselage displacement as output. This state space model is compared to a derived state space model for validation. Finally, a three dimensional aircraft with distributed displacement sensors on its wings is modeled. A state space representation is established using the wing displacement output from the sensors as its input and the motion and rotation of the fuselage along the X, Y and Z axes as the output. </p><p> It is seen that the displacement results of the lumped mass system match with those calculated using modal superposition equations. The state space model can also accurately predict the fuselage vibration of the lumped mass system, when provided with wing displacement as input. More importantly, results have shown that the distributed vibration sensors on the three dimensional plane model are able to measure the wing displacements. Using the output from these distributed sensors, the motion and rotation of the fuselage about all three axes can be effectively predicted.</p>

<p> A low-speed wind tunnel experiment was performed to study the effect of a backward-facing step on transition in a swept-wing flow. Detailed hot-wire measurements were used to assess the flow field characteristics on a swept flat plate with and without a backward-facing step. A pressure body was installed on the ceiling to induce a pressure field simulating that of an infinite swept wing. The step height was approximately 50% of the boundary-layer thickness at the step. Measurements without the step confirmed the dominance of the stationary crossflow instabilities leading to a high-frequency secondary-instability breakdown. The backward-facing step had a local destabilizing effect on the growth of the dominant stationary crossflow mode and the harmonic of the dominant mode. The stationary crossflow disturbances reached small amplitudes (3 to 5% U_e) before breakdown occurred. The transition front moved forward as the initial amplitude of the stationary crossflow disturbance was increased. The step introduced a flow field rich with unsteady disturbances. Three different families of unsteady disturbances were identified corresponding to three distinct frequency bands in the 80 to 1500 Hz range. Wave angles and phase speeds were measured for each type of disturbance. The disturbances are believed to correspond to a traveling crossflow-type disturbance, a TS-type disturbance, and a free shear layer instability. Each of the disturbances were modulated through interaction with the stationary crossflow modes. The spanwise modulation was different for each family and was seen in the distortion of the amplitude and phase. Larger stationary crossflow vortices resulted in larger peak amplitudes of the unsteady disturbances at similar streamwise locations. The mean-flow modulation appears to affect the local stability of the unsteady disturbances even at low stationary-crossflow amplitudes. The local destabilization of the unsteady disturbances is believed to be responsible for the sensitivity of transition location to stationary crossflow amplitude. Breakdown was initiated despite the low amplitude of the unsteady disturbances (2 to 4% U_e). Nonlinear interactions were observed between the different unsteady disturbances and may be ultimately responsible for breakdown to turbulence.</p>

<p> An analytical technique for investigating transport aircraft handling qualities is exercised in a study using models of two such vehicles, a Boeing 747 and Lockheed C-5A. Two flight conditions are employed for climb and directional tasks, and a third included for a flare task. The analysis technique is based upon a &ldquo;structural model&rdquo; of the human pilot developed by Hess. The associated analysis procedure has been discussed previously in the literature, but centered almost exclusively on the characteristics of high-performance fighter aircraft. The handling qualities rating level (HQRL) and pilot induced oscillation tendencies rating level (PIORL) are predicted for nominal configurations of the aircraft and for &ldquo;damaged&rdquo; configurations where actuator rate limits are introduced as nonlinearites. It is demonstrated that the analysis can accommodate nonlinear pilot/vehicle behavior and do so in the context of specific flight tasks, yielding estimates of handling qualities, pilot-induced oscillation tendencies and upper limits of task performance. A brief human-in-the-loop tracking study was performed to provide a limited validation of the pilot model employed.</p>

<p> The problem of controlling the behavior of a spacecraft in an optimal manner is one that has been studied since the beginning of the space era in the late 1950s. Recently, the complexity of such optimization problems has been increased by the introduction of spacecraft that are comparatively small in size and capable of being reconfigured, either in between missions or on-orbit. While such spacecraft have the potential to greatly expand the variety of missions that can be undertaken, they also increase the basic number of design variables that must be optimized. This dissertation presents a novel approach for the design of spacecraft control systems that optimizes both controller gain parameters and physical attributes of the spacecraft <i> in parallel</i>. The central design tool for this new strategy is a genetic algorithm, which applies concepts from evolutionary biology to search a complex design space in an efficient manner. Results are presented for spacecraft of various sizes, and the genetic algorithm design results are compared to a number of more traditional design approaches. </p><p> In the first part of this dissertation the genetic algorithm is applied to the problem of tuning the gain parameters of a nonlinear control law. This controller is used within a small spacecraft performing an attitude tracking maneuver, and must compensate for multiple environmental disturbance moments and the imposition of actuator saturation limits. While the control law under study has proven stability properties, no work has yet been done on optimizing the gain parameters for a specific application. The combined complexity of the controller itself and the spacecraft system make gain tuning via traditional approaches very difficult, and as such a genetic algorithm is utilized. The genetic algorithm can search a broad swath of the overall design space, and does so more efficiently than a human engineer applying their intuition in an "informed" trial-and-error approach to the same problem. </p><p> With the basic efficacy of the genetic algorithm established, in the next phase of the dissertation a novel controller optimization approach known as location-scheduled control is introduced. Under location-scheduled control, the use of the genetic algorithm is extended to not only optimize the gain parameters of a given control law but also the physical location of the control actuators within the spacecraft. This <i>dual</i> optimization of both controller gain parameters and physical properties of the spacecraft yields a catalog of design solutions that can be called upon from mission to mission, which significantly reduces the time required for control system design. The ability to easily relocate the hardware components of a spacecraft control system is enabled by a class of spacecraft known as CubeSats, which will be described in detail. </p><p> In the final portion of this dissertation the location-scheduled control methodology is applied to a real-word testbed system and hardware results are compared to those obtained via simulation. This system makes use of a unique property of superconducting physics known as flux-pinned interfaces that allows CubeSat-scale test articles to be easily reconfigured. The location-scheduled control approach is used to simultaneously determine the optimal configuration of the reconfigurable spacecraft system and the appropriate gains for a single-axis reorientation maneuver. It is shown through hardware testing that the genetic algorithm once again yields a combination of system configuration and controller gain values that outperforms those found by a control systems engineer.</p>

<p> The communications and electronic systems that comprise a distributed control architecture for a robotic manipulator tie the high level control and motion planning to the electromechanical components. Custom solutions to this problem can be expensive in terms of time, cost, and maintenance. The integration of commercial off the shelf (COTS) motion controllers, combined with a robust communication standard, offers the potential to reduce the costs and development times for new robots. This thesis demonstrates an implementation of this architecture using commercial controllers and the CANopen communications bus on two existing dexterous robots. Testing is conducted to quantify the single joint performance of these modules. Additionally, the implementation of the system on a second robot arm was conducted in order to test the flexibility of the system for use with different actuators and feedback.</p>

<p> A robust test bed is needed to facilitate future development of guidance, navigation, and control software for future vehicles capable of vertical takeoff and landings. Specifically, this work aims to develop both a hardware and software simulator that can be used for future flight software development for extra-planetary vehicles. To achieve the program requirements of a high thrust to weight ratio with large payload capability, the vehicle is designed to have a novel combination of electric motors and a micro jet engine is used to act as the propulsion elements.</p><p> The spacecraft simulator underwent several iterations of hardware development using different materials and fabrication methods. The final design used a combination of carbon fiber and fiberglass that was cured under vacuum to serve as the frame of the vehicle which provided a strong, lightweight platform for all flight components and future payloads.</p><p> The vehicle also uses an open source software development platform, Arduino, to serve as the initial flight computer and has onboard accelerometers, gyroscopes, and magnetometers to sense the vehicles attitude. To prevent instability due to noise, a polynomial kalman filter was designed and this fed the sensed angles and rates into a robust attitude controller which autonomously control the vehicle' s yaw, pitch, and roll angles.</p><p> In addition to the hardware development of the vehicle itself, both a software simulation and a real time data acquisition interface was written in MATLAB/SIMULINK so that real flight data could be taken and then correlated to the simulation to prove the accuracy of the analytical model.</p><p> In result, the full scale vehicle was designed and own outside of the lab environment and data showed that the software model accurately predicted the flight dynamics of the vehicle.</p>

<p>Reynolds averaged Navier Stokes (RANS) solvers have become the workhorse for simulating turbulent flows for most practical purposes. While the incompressible turbulence models used with RANS equations have improved considerably in their predictive capability, significant breakthrough has not been achieved for their compressible counterparts. With the advancement in computing power, high resolution direct numerical simulation (DNS) of low Reynolds number turbulent flows has become feasible. DNS of simple turbulent flows provides a detailed database which can be used for developing and testing turbulence models. In this work, we perform DNS of compressible homogeneous turbulence&mdash;decaying isotropic turbulence and homogeneous shear flow&mdash;for a range of initial turbulent Mach numbers, (<i>M_t</i>₀ = 0.05&ndash;0.4) using the more natural initial conditions. Simulations were performed on grids with 128³ and 256³ points. Compressibility effects on the evolution of turbulent kinetic energy were studied. We found negligible compressibility effects for decaying isotropic turbulence, while homogeneous shear flow demonstrated compressibility effects in the growth rate of turbulent kinetic energy. Compressibility corrections to turbulence models in the form of the ratio <i>&epsis;_d/&epsis;_s</i>, have been tested with the results from the simulations. For decaying isotropic turbulence a [special characters omitted] scaling was found to be better than [special characters omitted] while for homogeneous shear flow it was the opposite. The small value of the ratio <i>&epsis;_d/&epsis;_s</i> in decaying isotropic turbulence makes the [special characters omitted] scaling less relevant. Based on the DNS results of homogeneous shear flow, a new correction parameterized by the gradient Mach number, <i> M_g,</i> is proposed. The parameter <i>C_&mu;</i>, which is assumed constant for incompressible two equation eddy viscosity models, is computed explicitly from the DNS data. An <i>M_g,</i> dependence of the parameter, <i>C_&mu;</i>, is proposed. </p>