Mechanical Engineering - Ph.D. / Sc.D.
http://hdl.handle.net/1721.1/7683
Mon, 26 Sep 2016 20:51:17 GMT2016-09-26T20:51:17ZDetermining thermal stratification in rooms under mixing and displacement ventilation
http://hdl.handle.net/1721.1/104255
Determining thermal stratification in rooms under mixing and displacement ventilation
DomÃnguez Espinosa, Francisco Alonso
Computational Fluid Dynamics (CFD) simulations of a typical office under both mixing and displacement ventilation were performed to study the effects of room geometry (height and area of the supply), ventilation parameters (supply momentum and heat gain intensity) and radiation heat transfer on the thermal stratification of the air and the temperatures of the surfaces in the space. The air stratification and the temperatures of the surfaces are two important parameters when determining thermal comfort conditions in the room. Different room configurations were characterized in terms of their Archimedes number, which compares the effects of buoyancy and supply momentum, and dimensionless geometric variables. A high Archimedes space was found to be divided into a warm region of uniform temperature above the occupants and a zone where the temperature increases approximately linearly with height. In a low Archimedes space the air is mixed by the supply jet in the lower part of the room, especially near the outlet, resulting in this area having uniform temperature. However, the supply jet was found to be less efficient at mixing the air near the ceiling, resulting in higher temperatures in this zone than with higher Archimedes numbers. For a given Archimedes number, as the supply area increased, the air temperature was found to decrease in the lower part of the room but increase near the ceiling. The supply height was found to increase the vertical mixing in the room. Correlations were proposed to establish the temperature profile within 5% of the temperature rise of the room, which include the effects of the Archimedes number and room geometry. Correlations were developed to estimate the temperatures of the surfaces in a room, based on a dimensionless parameter that characterizes the amount of free area to convect heat to the air. The temperatures of the surfaces were found to be a function of this convective area, regardless of the view factors and convective heat transfer coefficients of the surfaces. A larger amount of convective area was found to result in lower surfaces temperatures but higher air temperatures. A simple methodology to estimate all of the radiative view factors in an occupied office for use in multizone models was proposed. It was shown that the commonly ignored view factor among occupants can be of importance, not only because occupants exchange radiation among themselves, but also because they block radiation that would otherwise reach other surfaces in the room. In addition, techniques to estimate the view factors between other surfaces, such as partitions and furniture, were also developed. Estimated view factors between surfaces encountered in practical situations were found to be within 10% of the results from ray tracing software. The estimated view factors were then incorporated into a thermal resistor network akin to the thermal circuits used to model heat transfer in multizone software. Results from the resistor network showed good agreement with CFD results, although the accuracy depends on the convective heat transfer coefficients used. Finally, it was demonstrated that scale models that use water as the working fluid are not capable of replicating the air thermal stratification, the temperatures of the surfaces or the mass flow rate of a full-sized space, because they neglect the effects of thermal radiation transfer.
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2016.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 322-331).
Fri, 01 Jan 2016 00:00:00 GMThttp://hdl.handle.net/1721.1/1042552016-01-01T00:00:00ZComputation of nonlinear hydrodynamic loads on floating wind turbines using fluid-impulse theory
http://hdl.handle.net/1721.1/104254
Computation of nonlinear hydrodynamic loads on floating wind turbines using fluid-impulse theory
Chan, Godine Kok Yan
Wind energy is one of the more viable sources of renewable energy and offshore wind turbines represent a promising technology for the cost effective harvesting of this abundant source of energy. To capture wind energy offshore, horizontal-axis wind turbines can be installed on offshore platforms and the study of hydrodynamic loads on these offshore platforms becomes a critical issue for the design of offshore wind turbine systems. A versatile and efficient hydrodynamics module was developed to evaluate the linear and nonlinear loads on floating wind turbines using a new fluid-impulse formulation - the Fluid Impulse Theory(FIT). The new formulation allows linear and nonlinear loads on floating bodies to be computed in the time domain, and avoids the computationally intensive evaluation of temporal and spatial gradients of the velocity potential in the Bernoulli equation and the discretization of the nonlinear free surface. The module computes linear and nonlinear loads - including hydrostatic, Froude-Krylov, radiation and diffraction, as well as nonlinear effects known to cause ringing, springing and slow-drift loads - directly in the time domain and a stochastic seastate. The accurate evaluation of nonlinear loads by FIT provides an excellent alternative to existing methods for the safe and cost-effective design of offshore floating wind turbines. The time-domain Green function is used to solve the linear and nonlinear free-surface problems and efficient methods are derived for its computation. The body instantaneous wetted surface is approximated by a panel mesh and the discretization of the free surface is circumvented by using the Green function. The evaluation of the nonlinear loads is based on explicit expressions derived by the fluid-impulse theory, which can be computed efficiently.
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2016.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 199-202).
Fri, 01 Jan 2016 00:00:00 GMThttp://hdl.handle.net/1721.1/1042542016-01-01T00:00:00ZBayesian inference of chemical reaction networks
http://hdl.handle.net/1721.1/104253
Bayesian inference of chemical reaction networks
Galagali, Nikhil
The development of chemical reaction models aids system design and optimization, along with fundamental understanding, in areas including combustion, catalysis, electrochemistry, and biology. A systematic approach to building reaction network models uses available data not only to estimate unknown parameters, but to also learn the model structure. Bayesian inference provides a natural approach for this data-driven construction of models. Traditional Bayesian model inference methodology is based on evaluating a multidimensional integral for each model. This approach is often infeasible for reaction network inference, as the number of plausible models can be very large. An alternative approach based on model-space sampling can enable large-scale network inference, but its efficient implementation presents many challenges. In this thesis, we present new computational methods that make large-scale nonlinear network inference tractable. Firstly, we exploit the network-based interactions of species to design improved "between-model" proposals for Markov chain Monte Carlo (MCMC). We then introduce a sensitivity-based determination of move types which, when combined with the network-aware proposals, yields further sampling efficiency. These algorithms are tested on example problems with up to 1000 plausible models. We find that our new algorithms yield significant gains in sampling performance, with almost two orders of magnitude reduction in the variance of posterior estimates. We also show that by casting network inference as a fixed-dimensional problem with point-mass priors, we can adapt existing adaptive MCMC methods for network inference. We apply this novel framework to the inference of reaction models for catalytic reforming of methane from a set of ~/~ 32000 possible models and real experimental data. We find that the use of adaptive MCMC makes large-scale inference of reaction networks feasible without the often extensive manual tuning that is required with conventional approaches. Finally, we present an approximation-based method that allows sampling over very large model spaces whose exploration remains prohibitively expensive with ex-act sampling methods. We run an MCMC algorithm over model indicators and for each visited model approximate the model evidence via Laplace's method. Limited and sparse available data tend to produce multi-modal posteriors over the model indicators. To perform inference in this setting, we develop a population-based approximate model inference MCMC algorithm. Numerical tests on problems with around 109 models demonstrate the superiority of our population-based algorithm over single-chain MCMC approaches.
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2016.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 189-198).
Fri, 01 Jan 2016 00:00:00 GMThttp://hdl.handle.net/1721.1/1042532016-01-01T00:00:00ZHarvesting photon energy : ultra-thin crystalline silicon solar cell & near-field thermoradiative cells
http://hdl.handle.net/1721.1/104252
Harvesting photon energy : ultra-thin crystalline silicon solar cell & near-field thermoradiative cells
Hsu, Wei-Chun
Photons from the sun and terrestrial sources have great potential to satisfy the energy demand of humans. This thesis studies two types of energy conversion technologies, photovoltaic solar cells based on crystalline silicon thin films and thermal-radiative cells using terrestrial heat sources, focusing on managing photons but also concurrently considering electron transport and entropy generation. Photovoltaic technology has been widely adopted to convert solar energy into electricity. Crystalline silicon material occupies ~90% of the photovoltaic market. However, the silicon material in a photovoltaic module with ~180-pm-thick silicon material contributes more than 30% of the overall cost, giving rise to an obstacle to compete with fossil fuel energy. One promising solution to break this barrier is the technology of thin-film crystalline silicon solar cells if the weak absorption of silicon can be overcome. To maintain its high energy conversion efficiency, nanostructure is designed considering both light trapping and electron collection. This design guided the fabrication of 10-pm-thick crystalline silicon photovoltaic cells with efficiencies as high as 15.7%. To reach efficiency >20% in industry, multiple strategies have been investigated to further improve the performance including the least-common-multiple rule for the double gratings structure, external optical cavity, high quality silicon in bulk material and interfaces, and optimal contact spacing and doping. For the energy conversion of terrestrial heat source, a direct bandgap solar cell can work in the reverse bias mode to convert energy into electricity companied by emission of photons as entropy carriers. Photon spectral entropy and fluxes are used to develop strategies for improving the heat to electricity conversion efficiency. Near-field radiative transfer, especially using phonon polariton material to couple out emitted photons from electron-hole recombination, is proposed to enhance energy conversion efficiency as well as the power density. We predict that the InSb thermoradiative cell can achieve the efficiency and power density up to 20.4 % and 327 Wm-2, respectively, between a hot source at 500K and a cold sink at 300K, if the sub-bandgap and non-radiative losses could be avoided.
Thesis: Ph. D., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2016.; Cataloged from PDF version of thesis.; Includes bibliographical references (pages 134-148).
Fri, 01 Jan 2016 00:00:00 GMThttp://hdl.handle.net/1721.1/1042522016-01-01T00:00:00Z