Thesis: Ph. D. in Computational Science and Engineering, Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, 2018 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 255-266). / The polar ice sheets have enormous potential impact on future global mean sea level rise. Recent observations suggest they are losing mass to the ocean at an accelerated rate. Skillful prediction of the ice sheets' future mass loss remains difficult, however; observations of key variables are insufficient and physical processes are poorly understood. Even when a relatively accurate dynamical model is available, computational limitations make it difficult to characterize uncertainties associated with the model's predictions. To address this prediction challenge, this thesis presents complementary developments in glaciology and in Bayesian computation. / In particular, (i) we develop new models of marine-terminating glaciers whose dynamics are controlled by an extended set of physical processes and geometric constraints; and (ii) we develop new sampling algorithms to efficiently characterize selected marginals of a high-dimensional probability distribution describing uncertain parameters. The latter algorithms have broader utility in Bayesian modeling and inference with computationally intensive models. We begin by studying laterally confined ice streams that terminate in the ocean, where they may form floating ice shelves. Such marine-terminating outlet glaciers are the main conduits by which Greenland and Antarctica drain their ice mass into the ocean. Ice shelves play an important role in buttressing the grounded inland ice. The seaward ice flow is typically accompanied by acceleration and thinning. Increased thinning eventually leads to flotation of the ice supported by buoyant forces from the ocean. / The transition region from grounded to floating ice is referred to as the grounding line (or zone), and the mass transport across the grounding line as the output flux. Previous work by Weertman (1974) and Schoof (2007) considers laterally unconfined ice streams, showing that their output flux is a monotonically increasing function of the bedrock rock depth at the grounding line. This scenario leads to the marine ice sheet instability (MISI): retreating into deeper water increases the output flux, and retreat accelerates. Therefore, stable steady states cannot exist on downward sloping beds. We extend this analysis to laterally confined glaciers and investigate when side-wall drag is sufficient to stabilize glaciers on downward sloping beds. Additionally, we include a parameterization of sub-shelf melt. We find that, whereas lateral drag can stabilize glaciers that would otherwise be subject to the MISI, sub-shelf melt can destabilize them. / Our ultimate goal is to predict future ice sheet volume and to quantify its uncertainty. We do so in the Bayesian statistical setting, conditioning our prediction on available observations. Yet characterizing a posterior distribution-using, for example, Markov chain Monte Carlo (MCMC)-involves repeated evaluations of an ice stream model, which are prohibitively expensive. Furthermore, the model parameters that need to be inferred are high dimensional, even though we are primarily interested in a low dimensional quantity: the future ice volume. We address this computational challenge by developing new structure-exploiting Monte Carlo methods that combine marginalization with surrogate modeling. Given a high-dimensional (posterior) distribution on the model parameters, whose density evaluations are computationally intensive, we construct an MCMC chain that directly targets a particular low-dimensional marginal of interest. In general, the marginal density is not available analytically. / Instead, we can compute unbiased noisy estimates of this density. Our MCMC algorithm incrementally constructs a local regression approximation of the target marginal density using these estimates. Continual refinement of the approximation, as MCMC sampling proceeds, leads to an asymptotically exact characterization of the desired marginal distribution. Analysis of the bias-variance tradeoff guides an ideal refinement strategy that balances the decay rates of different components of the error. Our approach exploits regularity in the marginal density to significantly reduce computational expense relative to both full-dimensional and pseudo-marginal MCMC. / by Andrew D. Davis. / Ph. D. in Computational Science and Engineering / Ph.D.inComputationalScienceandEngineering Massachusetts Institute of Technology, Department of Aeronautics and Astronautics

This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Thesis: Ph. D. in Computational Science and Engineering, Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, 2018 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 187-195). / Design optimization methods are a popular tool in computational fluid dynamics for designing components or finalizing the flow parameters of a system. The adjoint method accelerates the design process by providing gradients of the design objective with respect to the system parameters. But, typically, adjoint-based design optimization methods have used low fidelity simulations like Reynolds Averaged Navier-Stokes (RANS). To reliably capture the complex flow phenomena like turbulent boundary layers, turbulent wakes and fluid separation involved in high Reynolds number flows, high fidelity simulations like large eddy simulation (LES) are required. Unfortunately, due to the chaotic dynamics of turbulence, the adjoint method for LES diverges and produces incorrect gradients. In this thesis, the adjoint method for unsteady flow equations is modified by adding artificial viscosity to the adjoint equations. The additional viscosity stabilizes the adjoint solution and maintains reasonable accuracy of the gradients obtained from it. The accuracy of the method is assessed on multiple turbulent flow problems, including subsonic flow over a cylinder and transonic flow over a gas turbine vane. The utility of the method is then tested in performing shape optimization of the trailing edge of a transonic turbine vane. The optimal design, found using a modified gradient-based Bayesian optimization algorithm, shows approximately 15% better aero-thermal performance than the baseline design. Such design optimizations are possible due to the availability of massively parallel supercomputers. Designing high performance fluid flow solvers for the next generation supercomputers is a challenging task. In this thesis, a two-level computational graph method for writing optimized distributed flow solvers on heterogeneous architectures is presented. A checkpoint-based automatic differentiation method is used to derive the corresponding adjoint flow solver in this framework. / by Chaitanya Anil Talnikar. / Ph. D. in Computational Science and Engineering / Ph.D.inComputationalScienceandEngineering Massachusetts Institute of Technology, Department of Aeronautics and Astronautics

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2002. / Includes bibliographical references (p. 188-189). / During the boost phase of ascent, winds have a significant impact on a launch vehicle's angle of attack, and can induce large structural loads on the vehicle. Traditional methods for mitigating these loads involve measuring the winds prior to launch and designing trajectories to minimize the vehicle angle of attack (a). The current balloon-based method of collecting wind field information produces wind profiles with significant uncertainty due to the inherent time delays associated with balloon measurement procedures. Managing the mission risk caused by these uncertain wind measurements has always been important to control system designers. This thesis will describe a novel approach to managing structural loads through the combination of a Light Detection and Ranging (LIDAR) wind sensor, and Model Predictive Control (MPC). LIDAR wind sensors can provide near real-time wind measurements, significantly reducing wind uncertainty at launch. MPC takes full advantage of this current wind information through a unique combination of proactive control, constraint integration and tuning flexibility. This thesis describes the development of two types of MPC controllers, as well as a baseline controller representative of current control methods used by industry. A complete description of Model Predictive Control theory and derivation of the necessary control matrices is included. The performance of each MPC controller is compared to that of the baseline controller for a wide range of wind profiles from both the Eastern and Western U.S. Test Ranges. Both MPC controllers are shown to provide reductions of greater than 50% in a, Qa and structural bending moments. In addition, the effects of wind measurement delays and uncertainty on the performance of each controller are investigated. / by Andrew Allen Martin. / S.M.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1990. / Vita. / Includes bibliographical references (leaves 243-252). / by Daniel Michael Merfeld. / Ph.D.

Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1996. / Includes bibliographical references (leaves 52-54). / by Martin Bowyer Graf. / M.S.

Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1991. / Includes bibliographical references (leaves 119-121). / by Alexandra Maria Landsberg. / M.S.

Thesis (M.S.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 1990. / Includes bibliographical references (leaves 149-150). / by George Pappas. / M.S.

Thesis (M.S.)--Massachusetts Institute of Technology. Dept. of Aeronautics and Astronautics, 1986. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND AERO. / Bibliography: leaf 135. / by Craig Howard Smith. / M.S.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2008. / Includes bibliographical references (leaves 145-153). / As unmanned aerial vehicles (UAVs) take on more prominent roles in aerial missions, it becomes necessary to increase the level of autonomy available to them within the mission planner. In order to complete realistic mission scenarios, the UAV must be capable of operating within a complex environment, which may include obstacles and other no-fly zones. Additionally, the UAV must be able to overcome environmental uncertainties such as modeling errors, external disturbances, and an incomplete situational awareness. By utilizing planners which can autonomously navigate within such environments, the cost-effectiveness of UAV missions can be dramatically improved.This thesis develops a UAV trajectory planner to efficiently identify and execute trajectories which are robust to a complex, uncertain environment. This planner, named Efficient RSBK, integrates previous mixed-integer linear programming (MILP) path planning algorithms with several implementation innovations to achieve provably robust on-line trajectory optimization. Using the proposed innovations, the planner is able to design intelligent long-term plans using a minimal number of decision variables. The effectiveness of this planner is demonstrated with both simulation results and flight experiments on a quadrotor testbed.Two major components of the Efficient RSBK framework are the robust model predictive control (RMPC) scheme and the low-level planner. This thesis develops a generalized framework to investigate RMPC affine feedback policies on the disturbance, identify relative strengths and weaknesses, and assess suitability for the UAV trajectory planning problem. A simple example demonstrates that even with a conventional problem setup, the closed-loop performance may not always improve with additional decision variables, despite the resulting increase in computational complexity. A compatible low-level troller is also introduced which significantly improves trajectory-following accuracy, as demonstrated by additional flight experiments. / by Brandon Luders. / S.M.

Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2009. / Includes bibliographical references (p. 91-94). / As unmanned aerial vehicles (UAVs) become more involved in challenging mission objectives, the need for agility controlled flight becomes more of a necessity. The ability to navigate through constrained environments as well as quickly maneuver to each mission target is essential. Currently, individual vehicles are developed with a particular mission objective, whether it be persistent surveillance or fly-by reconnaissance. Fixed-wing vehicles with a high thrust-to-weight ratio are capable of performing maneuvers such as take-off or perch style landing and switch between hover and conventional flight modes. Agile flight controllers enable a single vehicle to achieve multiple mission objectives. By utilizing the knowledge of the flight dynamics through all flight regimes, nonlinear controllers can be developed that control the aircraft in a single design. This thesis develops a full six-degree-of-freedom model for a fixed-wing propeller-driven aircraft along with methods of control through non conventional flight regimes. In particular, these controllers focus on transitioning into and out of hover to level flight modes. This maneuver poses hardships for conventional linear control architectures because these flights involve regions of the post-stall regime, which is highly nonlinear due to separation of flow over the lifting surfaces. Using Lyapunov back stepping control stability theory as well as quaternion-based control methods, control strategies are developed that stabilize the aircraft through these flight regimes without the need to switch control schemes. The effectiveness of each control strategy is demonstrated in both simulation and flight experiments. / by Frantisek Michal Sobolic. / S.M.