A Mars Tumbleweed Rover is a spherical, wind driven, planetary rover. Compared with conventional rovers, a tumbleweed rover can travel further faster and gain access to areas such as valleys and chasms that previously were inaccessible. This paper will present results of the design, testing, mathematical modeling, and computer simulation of a tumbleweed rover propelled by a constant wind loading over a rough surface representing the planet's surface. Results will show different trajectories for specific cases where initial conditions such as mass distribution and angular velocity are varied.

An uninhabited aero vehicle platform design is presented. This encompasses the landing gear system and the structures of the vehicle. The landing gear system consisted of the design, construction and testing of the main and nose gears. The testing of the landing gear resulted in a valid system that could then be integrated into the vehicle. The vehicle structures are composed of various configurations of composite sandwiches. Extensive material testing was conducted to experimentally produce the physical properties of the materials. These properties and techniques can be utilized by other vehicle designs. The structural design was refined and ultimately verified within a finite element analysis program, ANSYS. This analysis implemented a composite shell element that utilized all of the material properties gained from the material testing. This work resulted in an analyzed and constructed vehicle. Ultimately the vehicle was load tested to verify the analytical results.

<p>SRIVASTAVA, RAVI K. An Adaptive Grid Algorithm for Air Quality Modeling. (Under the direction of Dr. D. Scott McRae.)The physical and chemical processes responsible for air pollution span a wide range of spatial scales. For example, there may be point sources, such as power plants that are characterized by relatively small spatial scales compared to the size of the region that may be impacted by such sources. To obtain accurate distributions of pollutants in an air quality simulation, the pertinent spatial scales can be resolved by varying the physical grid node spacing.A new dynamic adaptive grid algorithm, the Dynamic Solution Adaptive Grid Algorithm - PPM (DSAGA-PPM), is developed for use in air quality modeling. Given a fixed number of grid nodes, DSAGA-PPM distributes these nodes in response to spatial resolution requirements of the solution field and then updates the solution field based on the resulting distribution of nodes. DSAGA-PPM is implemented dynamically to resolve any evolving solution features. Tests with model problems demonstrate that DSAGA-PPM calculates advection much more accurately than the corresponding static grid algorithm (SGA-PPM) and, therefore, would assure more accurate starting concentrations for chemistry calculations. For example, after one revolution of four rotating cones, 87% of each of the cone peaks is retained using DSAGA-PPM while only 63% is retained using SGA-PPM. The root-mean-square errors in DSAGA-PPM results are about 4-5 times lower than those in the corresponding SGA-PPM results. Tests with reacting species and sources demonstrate that DSAGA-PPM provides the needed solution resolution. In simulations of a rotating and reacting conical puff, the root-mean-square errors in DSAGA-PPM results are about 4-6 times lower than those in the corresponding SGA-PPM results. In simulations of a power plant plume, the DSAGA-PPM solution reflects the early, the intermediate, and the mature stages of plume development; these stages are not seen in the corresponding SGA-PPM solution. Finally, it is demonstrated that DSAGA-PPM provides an accurate description of the ozone production resulting due to dynamic interactions between emissions from two power plants and an urban area. In general, these results reflect that DSAGA-PPM is able to provide accurate spatial and temporal resolution of rapidly changing and complex concentration fields. Performance achieved by DSAGA-PPM in model problem simulations indicates that it can provide accurate air quality modeling solutions at costs 10 times less than those incurred in obtaining equivalent static grid solutions. <P>

<p>McDaniel, Keith S. Three Dimensional Simulation of Time-DependentScramjet Isolator / Combustor Flowfields Implemented onParallel Architectures, ( Under the directions of Dr. J. R. Edwards). The development of a parallel Navier-Stokes solver for computing time-dependent,three-dimensional reacting flowfields within scramjet (supersonic combusting ramjet)engines is presented in this work. The algorithm combines low-diffusion upwinding methods, timeaccurate implicit integration techniques, and domain decomposition strategies to yield an effectiveapproach for large-scale simulations. The algorithm is mapped to a distributed memoryIBM SP-2 architecture and a shared memory Compaq ES-40 architecture using the MPI-1 message-passingstandard. Two and three-dimensional simulations of time-dependent hydrogen fuel injection into a modelscramjet isolator / combustor configuration at two equivalence ratios are performed. Thesesimulations are used to gain knowledge of engine operability, inlet performance, isolatorperformance, fuel air mixing, flame holding, mode transition, and engine unstart.Results for an injection at a ratio of 0.29 show qualitative agreement withexperiment for the two-dimensional case, but revealed a slow progression towardengine unstart for the three-dimensional case. Injection at an equivalence ratio of 0.61resulted in engine unstart for both two-dimensional and three-dimensional cases.Engine unstart for the three-dimensional case occurs as a response to the formation and growthof large pockets of reversed flow along the combustor side wall. These structuresdevelop at an incipient pressure above 154 kPa and result in significant blockage of the core flow,additional compression, and chemical reaction within the boundary layer. All of these factors promotea much more rapid unstart as compared with the two-dimensional case.<P>

<p>Manning, Melissa Lynn. Computational Evaluation of Quiet Tunnel Hypersonic Boundary Layer Stability Experiments. (Under the direction of Dr. Ndaona Chokani.) A computational evaluation of two stability experiments conducted in the NASA Langley Mach 6 axisymmetric quiet nozzle test chamber facility is conducted. Navier-Stokes analysis of the mean flow and linear stability theory analysis of boundary layer disturbances is performed in the computations. The effects of adverse pressure gradient and wall cooling are examined. Calculated pressure, temperature and boundary layer thickness distributions show very good overall agreement with experimental measurements. Computed mass flux and total temperature profiles show very good quantitative agreement with uncalibrated hot-wire measurements obtained with the hot-wire operated in high and low overheat modes respectively. Comparisons between calibrated hot-wire data and mean flow computations show excellent agreement in the early stages of the transitional flow. However, examination of the wire Reynolds number and mass flux and total temperature eigenfunction profiles suggest that when operated in high overheat mode the sensitivity of the hot-wire to total temperature is significant. Thus, while uncalibrated hot-wire measurements are useful to characterize the overall features of the flow, calibrated hot-wire measurements are necessary for quantitative comparison with stability theory. Computations show that adverse pressure gradient and wall cooling decrease the boundary layer thickness and increase the frequency and amplification rate of the unstable second mode disturbances; these findings are consistent with the experimental observations. <P>

<p> Lamb waves based ultrasonic testing has been studiedfor many years. However, conventional methods of generatingand collecting of Lamb waves usually require bulky instruments and manual interference, thus can not be applieddirectly for in-situ or in-service monitoring of thestructural health. Especially, the method of interpretingthe Lamb waves in an active structural health monitoring(SHM) system with built-in piezoelectric sensors/actuatorsis not available yet. The objective of this study was to propose and validate, through numerical simulation and experimental studies, the feasibility of adopting the geophysical migration method to interpret the ultrasonic Lamb wave signals for the purpose of realizing quantitative damage identification. A homogeneous isotropic plate with a surface-mountedlinear piezoelectric ceramic (PZT) disk array is studied as an example. The piezoelectric disks act as actuators to excite Lamb waves and also as sensors to receive the waves reflected from the structural anomaly in the plate. The migration technique, which is an advanced technique in geophysics to reverse the reflection wave field and to image the Earth interior, is then used to back-propagate the recorded wave signals and to visually image the damage in the plate. Mindlin plate theory is adopted to model the propagating waves, and a two-dimensional 2-6 order explicit finite difference algorithm is used to synthesize the reflection waves and to implement the migration process. The stability and accuracy criteria of the finite differencealgorithm when used in plate problems is discussed. An analytical solution is derived for the transient Lamb waves of an infinite plate subject to a point loading. This solution is used to verify the accuracy of the finite difference calculation. Both poststack and prestack migration are studied to propagate the reflection energy back to the damages. For the poststack migration, a one-way version of flexural wave equation is derived and the data pre-processing procedures before migration, such as muting direct arrival, deconvolution and stacking, are discussed. For prestack migration, an excitation-time imaging condition specifically for the migration of waves in a plate is introduced based on ray-tracing concepts and the asymptotic properties of flexural wave velocities and the migration is proceeded through the full-way wave equation. The results of numerical simulation show that the migration method possesses the capability of identifying multiple discrete damages without a priori assumption on the distribution pattern of the damages. Thus not only the existence but also the shape and the dimensions of the damages can be visually identified. An experimental apparatus is then set up to validate the conclusions drawn from the synthetic data. For calibration of the system, an analytical model of the waves in a plate incorporated with PZT sensors/actuators is developed. The agreement between the model calculated data and the measured data in the experiment shows that A0 mode Lamb waves are accurately generated and collected. Finally, the migration results from the reflection waves of an artificial damage in an arc shape recorded in the experiment are presented. It is shown that the existence of the damage could be correctly imaged through the migration process as it was shown in the numerical simulation.<P>

<p>The numerical investigation into the dynamics of unsteady inlet flowfields is applied to a three-dimensional scramjet inlet-isolator-diffuser geometry designed for hypersonic type applications. The Reynolds-Averaged Navier-Stokes equations are integrated in time using a subiterating, time-accurate implicit algorithm. Inviscid fluxes are calculated using the Low Diffusion Flux Splitting Scheme of Edwards. A modified version of the dynamic solution-adaptive point movement algorithm of Benson and McRae is used in a coupled mode to dynamically resolve the features of the flow by enhancing the spatial accuracy of the simulations. The unsteady mesh terms are incorporated into the flow solver via the inviscid fluxes. The dynamic solution-adaptive grid algorithm of Benson and McRae is modified to improve orthogonality at the boundaries to ensure accurate application of boundary conditions and properly resolve turbulent boundary layers. Shock tube simulations are performed to ascertain the effectiveness of the algorithm for unsteady flow situations on fixed and moving grids. Unstarts due to a combustor and freestream angle of attack perturbations are simulated in a three-dimensional inlet-isolator-diffuser configuration.<P>

<p>Low Reynolds number experiments have been conducted in the North Carolina State University Subsonic Wind Tunnel Facility to determine the feasibility of using an adaptive mechanical turbulator to control in a predetermined manner a laminar separation bubble. The experimental set-up consists of a flat plate with an inverted airfoil mounted above its surface. This set-up permits laminar separation bubbles to be examined over a range of Reynolds numbers and pressure distributions. An SMA-actuated mechanical turbulator is developed in the present work. A multi-element hot-film array is used to detect the laminar separation, transition to turbulence, and turbulent reattachment in the laminar separation bubble. The effects of turbulator height and Reynolds number on the mean, the standard deviation, the skewness, and the kurtosis of the hot-film signals are examined. The experimental results agree well with the theoretical predictions of laminar separation, transition, and turbulent reattachment obtained using a viscous/inviscid analysis code.For the range of trip heights examined, that is up to 30 percent of boundary layer thickness, the adaptive boundary layer trip is not suited to provide proportional control of a laminar separation bubble, as no monotonic variation of location in separation, transition, or reattachment is observed with varying trip height. However, the hot-film measurements show promise for providing in real time the identification of the primary features of the laminar separation bubble.<P>

<p>Experiments have been conducted using vane-type vortex generators to control flow separation and exit flow distortion in a diffusing duct. The primary purpose is to examine the feasibility of using surface-mounted hot-film sensors to determine the extent of exit flow distortion. The experimental set-up consists of a two-dimensional blow down type wind tunnel with a variable diffuser exit. One diffuser wall is curved to produce a Stratford-like pressure gradient. The wall's placement is adjustable such that the adverse pressure gradient can be adjusted to promote separation. An active vortex generator array that can be placed at three streamwise locations is used to reduce the extent of flow separation and exit distortion.Diffuser surface pressure and exit total pressure measurements are obtained and compared to the hot-film data. The time-averaged mean and rms voltages from the hot-film data are used as indicators of flow separation and exit flow distortion.Results show that, with the use of the vortex generators, high mean voltages and low levels of rms voltage correlate well with improved pressure recovery. Conversely, poorer pressure recovery is associated with lower mean voltages and higher rms values compared to the baseline cases. Increased total pressures at the diffuser exit are accompanied by increases in hot-film mean voltages. These indicate higher shear stresses, which also correspond to increased flow uniformity. Lower variations in the rms voltages compared to the baseline cases also correlated well with improved total pressures at the diffuser exit.<P>

<p>The work presented in this thesis involved the analysis of adhesive bonded joints for composite bridge decks and was divided into three phases. The first phase involved a parametric study on a single lap joint using ANSYS finite element analysis software. The purpose of the parametric study was to alter the geometry and material properties of the joint and study their effects on the stress distribution in both the adherends and adhesive. The four different cases studied included adding a taper to the adherends, different edge shapes on the adhesive layer, a material stiffness imbalance and a geometric stiffness imbalance. It was found that for the taper case and the edge shape case, the stress field in the joint was affected slightly. The material and geometric stiffness imbalance cases had the most drastic affect on the stress field of both the adhesive and adherend. Phase two of this study involved physical tests on single lap joints pulled in uniaxial tension. Tests were performed on three different types of laminates in order to study the interfacial effects these laminates had on the adhesive bond. It was found that by changing the surface of the composite, the mode of failure changed significantly.Phase three of this research involved a study on surface preparation. Three different surface preparations were conducted on the adherends of a butt-strap joint. The first included an acetone wipe. The second involved sanding the adherends. The third surface preparation involved adding APRIME-2, a secondary bonding agent, to the adherends before adding the strap. By simply sanding the adherends, the load to failure was increased by 350% compared to an acetone wipe. The ATPRIME-2 improved the load to failure by an additional 60% as well as improved the failure mode to a fiber tear. It was concluded that surface preparation has a major impact on the behavior of adhesively bonded joints.<P>