1  20
Next
Number of results to display per page
Online 1. 3D particle concentration measurement in turbulent flows using magnetic resonance imaging [2018]
 Borup, Daniel Duffy, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Turbulent, dispersed multiphase flows are of critical importance in a wide range of application areas. Experimental investigations are indispensable in the study of such flows, as experiments can provide reliable domainspecific knowledge and/or validation for computational fluid dynamics (CFD) tools. However, only a limited set of experimental techniques currently are available for studying particleladen flows: pointwise measurements provide high temporal resolution but poor spatial coverage, while laserbased techniques can allow for 2D or 3D measurements, but only in geometrically simple flows. Magnetic Resonance Imaging (MRI) is a powerful tool that can provide fully quantitative, 3D experimental data without the need for optical access. Currently, MRI can provide the timeaveraged, 3component velocity and/or scalar concentration fields in turbulent singlephase flows of arbitrary geometric complexity. In recent years MRI has been applied to the study of singlephase flows across a broad range of problems from the engineering, environmental, and medical arenas. MRI data sets are particularly well suited for validating CFD simulations of complex 3D flows because comprehensive data coverage can be obtained in a relatively short time. The present work describes development, validation, and application of a new diagnostic, wherein MRI is used to obtain the 3D mean volume fraction field for solid microparticles dispersed in a turbulent water flow. The new method is referred to as Magnetic Resonance Particle concentration, or MRP. This technique was designed to maintain the same advantages of existing MRIbased techniques: quantitative data can be obtained in 3D for fully turbulent flow in arbitrarily complex geometries. MRP is based on a linear relationship between the MRI signal decay rate and particle volume fraction (Yablonskiy and Haacke, 1994). The MRP method and underlying physics were validated through several studies, increasing in complexity from a single particle suspended in a gel to a fully turbulent channel flow seeded uniformly with particles. The channel flow case showed that the signal decay rate varied linearly with particle volume fraction, and that the measured proportionality constant was within 5% of the value predicted by the theory of Yablonskiy and Haacke (1994). This good agreement was observed for two fully turbulent Reynolds numbers, 6300 and 12,200, and over most of the measurement domain. However, the measured proportionality constant was lower than expected in the the furthest upstream portion of the channel; several potential reasons for this discrepancy were identified, but none could be proven conclusively at this stage. Following the validation experiments, MRP was applied to three application cases drawn from realworld flows of interest. First, the dispersion of two particle streaks in a model human nasal passage was studied. The results showed that almost all particles reaching the upper portions of the nasal passage (e.g., the olfactory region) entered the nose near the nostril tip, even at high breathing rates where the flow was not laminar. The second case involved MRP concentration measurements for a particle streak in a generic gas turbine blade internal cooling passage. Results in this case provided evidence that small dustlike particles ingested into a cooling passage may behave inertially in the presence of fine flow features, such as the recirculation regions behind ribbed flow turbulators. In the final case, the performance of a particle separator device proposed by Musgrove et al. (2009) was quantified using both MRP and a samplebased analysis performed outside the MRI environment. The two techniques were in agreement regarding the poor overall effectiveness of the separator, and the 3D MRP data were used to examine the particle transport physics and suggest potential design improvements. Taken together, results from the three test cases showed that MRP can provide quantitative, 3D particle concentration data in applicationrelevant flows, leading to unique insights that would not be possible with existing measurement techniques.
 Also online at
Online 2. Aerodynamic interactions in arrays of verticalaxis wind turbines [2018]
 Brownstein, Ian David, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Understanding the aerodynamic interactions between turbines in a wind farm is essential for maximizing power generation. In contrast to horizontalaxis wind turbines (HAWTs), for which wake interactions between turbines in arrays must be minimized to prevent performance losses, verticalaxis wind turbines (VAWTs) in arrays have demonstrated beneficial interactions that can result in net power output greater than that of turbines in isolation. These synergistic interactions have been observed in previous numerical simulations, laboratory experiments, and field work. This dissertation builds on previous work by identifying the aerodynamic mechanisms that result in beneficial turbineturbine interactions and providing insights into potential wind farm optimization. The experimental data presented indicates increased power production of downstream VAWTs when positioned offset from the wake of upstream turbines. Comparison with threedimensional, threecomponent flow measurements demonstrates that this enhancement is due to flow acceleration adjacent to the upstream turbine, which increase the incident freestream velocity on appropriately positioned downstream turbines. A loworder model combining potential flow and actuator disk theory accurately captures this effect. Laboratory and field experiments were used to validate the model's predictive capabilities, and an evolutionary algorithm was deployed to investigate array optimization. Furthermore, changes in upstream turbine performance are related to variations in the surrounding flow field due to the presence of the downstream rotor. Finally, threedimensional vortex interactions behind pairs of VAWTs are observed to replenish momentum in the array's wake. These effects are described along with their implications for wind farm design.
 Also online at

Online 3. Algorithms for realtime timedependent density functional theory and calculation of phase diagrams for twodimensional phasechange materials [2018]
 Rehn, Daniel Adam, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Twodimensional materials have found use in a wide range of engineering applications, owing in part to their unique properties not found in the bulk and their potential to enable fabrication of devices that approach atomic sizes. The inherently quantum mechanical nature of 2D materials can be treated using density functional theory (DFT)based methods, with some limitations that can in principle be addressed with the timedependent formulation of DFT, known as timedependent density functional theory (TDDFT). Here, we explore the use of DFT and TDDFT for applications of 2D materials, with two primary focuses: (1) the implementation and analysis of numerical integration schemes for timedomain TDDFT and associated applications to 2D and bulk materials properties and (2) the application of groundstate DFT to the study of thermodynamic properties of 2D materials that undergo structural phase transformations under electrostatic gating. Towards focus 1, we have implemented a variety of implicit and explicit integration schemes in the context of planewave realtime TDDFT and assessed the accuracy, stability, and computational cost of these methods. We find that for planewaves, highorder explicit multistep methods, including AdamsBashforth and AdamsBashforthMoulton methods, outperform commonlyused methods including CrankNicolson and RungeKutta. We have developed a code, called rttddft, as an extension to the popular Quantum ESPRESSO code, which will enable researchers performing groundstate planewave DFT calculations to immediately benefit from our timedomain integration schemes. The code is written in objectoriented Fortran and is designed using objectoriented design patterns and builtin testing. We also apply the code to a new timedomain formulation of the adiabatic connection fluctuationdissipation theorem (ACFDT) to compute electron correlation energies, an application especially relevant to van der Waals interactions in 2D materials systems. Towards focus 2, we have developed modeling methods to predict the phase transformation properties and phase diagrams of electrostaticallygated 2D materials. Using DFTbased calculations, we are able to predict the critical charge and critical voltage required to induce a structural phase transformation in a variety of 2D materials, and also determine how these critical values change with temperature. Using these methods, we are able to map out detailed phase diagrams for different 2D materials systems. In addition, we study the potential to use the electrostatic gating mechanism to absorb heat from the surroundings, an effect we term the electrostaticaloric effect. The methods we develop are useful for a variety of realistic engineering applications, including the use of 2D materials for applications in phasechange memory, controllable atomicscale heat absorbers, and atomic sensors.
 Also online at
Online 4. Applications of modern medical imaging techniques for quantitative engineering flow measurements [2018]
 Vasquez Guzman, Pablo Adolfo, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Conventional engineering flow measurement techniques are usually limited to providing measurements at discrete points or planes and are either intrusive or require direct optical access. Obtaining fullfield engineering flow measurements using existing techniques is difficult and seldom done. However, an accurate and detailed fullfield understanding of relevant engineering flow processes is needed to optimize performance and/or test computational models. Modern medical imaging techniques offer the potential to provide noninvasive detailed, threedimensional, fullfield, information of relevant engineering flows. The work presented herein describes the methodology and application of modern medical imaging techniques, Xray Computed Tomography (CT) and Magnetic Resonance Imagining (MRI), as quantitative measurement techniques to noninvasively acquire detailed fullfield information of engineering flows with direct relevance to gas turbines. Results reveal threedimensional characteristics and provide quantitative validation data for highfidelity simulations. Xray CT was used to acquire liquid mass concentration measurements of pressure swirl atomizers operating in ambient conditions. While modern medical Xray CT systems have the capability of providing accurate detailed liquidmass distribution measurements of sprays, such systems are not optimized for acquiring liquidmass distribution measurements of sprays. Numerous parameters that influence the performance of an Xray CT system were investigated to optimize a tabletop Xray CT system for imaging sprays. Result provide quantitative information in the nearfield spray region of pressureswirl atomizers and qualitative information of the features within the atomizers. Measurements were acquired up to a distance of approximately ten orifice diameters downstream of the pressure swirl atomizers. Magnetic Resonance Thermometry (MRT) coupled with Magnetic Resonance Velocimetry (MRV) was used to acquire temperature and velocity measurements of a conjugate jetincrossflow configuration. MRT is a recent development that has extended the capability of established MRIbased measurement techniques for investigating complex thermofluid flows with conjugate conditions. The 3D MRT measurements provide an overall film cooling effectiveness estimate that allows for the investigation the film cooling performance and conjugate heat transfer effects. Higher resolution 2D MRT results provide an estimate of the heat transfer coefficient. The experimental results were compared to ANSYS Fluent RANS simulations of the conjugate jetincrossflow configuration. Consistent with previous studies, the RANS simulations overpredicted the temperature distribution and did not properly capture the kidneyshaped structure due to the CVP. In general, Xray CT provides the means to acquire accurate detailed threedimensional liquid mass concentration measurements of twophase flows and MRI provides the means to acquire detailed threedimensional temperature and velocity measurements of liquid flows. Furthermore, these modern medical imaging techniques offer the potential of allowing for a more rapid design and development cycle, not only saving time and cost but providing a more detailed understanding of relevant engineering flow processes.
 Also online at

Online 5. Atomic layer deposition for energy and semiconductor applications [2018]
 Kim, Yongmin, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

In this thesis, I contribute to two instrumental sectors in our current society  semiconductor and energy  by developing novel materials through advanced nanoscale engineering through atomic layer deposition (ALD). ALD is a deposition technique in vacuum, which can make highly uniform films with a thickness range of 1100 nm on both 2D and 3D structured substrates. ALD chemistries enable the deposition of a wide diversity of materials. Furthermore, performing diverse ALD chemistries in single deposition is possible, leading to fabrication of both homogenous mixtures and heterogeneous structures (e.g. nanolaminate and metal/metaloxide composite). First, I will discuss about the use of ALD to fabricate barium titanates and silicon nitrides for the applications in key semiconductor components in Chapters 2 and 3. The continuous downscaling of integrated circuits (ICs) necessitates the development of ultrathin oxides and nitrides, mainly used for dielectric materials and sidewall spacers in transistors, respectively. Oxide films with high dielectric constants (i.e. highk) and low electrical leakage currents are essential for the applications of information storage devices such as dynamic randomaccess memory (DRAM). Using ALD, ultrathin (< 7 nm) barium titanates (BaxTiyOz) with different cation stoichiometries are fabricated and the tunability on electrical properties by changing BatoTi cation ratio is successfully demonstrated. Also, aluminum incorporation into BaxTiyOz shows it can further decrease the electrical leakage current with a marginal sacrifice of dielectric constants. Another contribution on semiconductor of this thesis is ALD of Nanolaminates of silicon nitridealuminum nitride (SiNAlN) for sidewall spacer. The development of superior wetetching resistant and electrically insulating films is very crucial for modern transistor fabrication processes with a process temperature limitation below 350 °C. The developed SiNAlN nanocomposites meet industry requirements by sufficiently lowering the wet etch rates as well as the electrical leakage currents. This dissertation also describes the use of ALD on energy materials in Chapter 4. In the past, ALD has been a key manufacturing process in the semiconductor industry. Transferring this mature technology to energy applications has recently gathered much attention since ALD is capable of precisely controlling loadings and morphologies of both metal and metal oxides, already demonstrated in several semiconductor devices. These capabilities can bring about technical innovation in a wide range of energy devices with nanoscale components such as fuel cells, solar cell, and batteries. Specifically, I will showcase a novel synthesis route for PtTi alloy nanomaterials through ALD of metal/metaloxide composite in conjunction with post thermal annealing in hydrogen. The developed PtTi alloys successfully prove their capability as cathode electrocatalysts in Polymer electrolyte fuel cells (PEFCs), i.e. oxygen reduction reaction (ORR). PEFCs, regarded the most promising implementation of fuel cells in the automotive sector, still require the significant reduction of platinum used in electrocatalysts. One of the most effective ways to minimize platinum is the utilization of platinum alloys. Among several Pt3M alloys (M = Y, Ni, Co, and Fe), Pt3Ti alloy is one of the most promising candidates. In this chapter, a simple and scalable synthesis route of Pt3Ti through ALDs of Pt and TiO2, followed by thermal annealing at 800 °C in hydrogen for intermixing of Pt and TiO2 layers, is successfully demonstrated. The best performing PtTi alloy catalysts show excellent stability and more than 2fold improvement in ORR Ptbased mass activity in comparison to commercial Pt powder catalyst. I envision that these aforementioned ALD materials introduced in this thesis will be actually utilized in semiconductor and energy applications, which will eventually accelerate the innovations of ICs and contribute to change the landscape of electrical energy production toward a more efficient and ecofriendly direction.
 Also online at
Online 6. Autonomous vehicle motion planning with ethical considerations [2018]
 Thornton, Sarah Marie, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Human drivers navigate the roadways by balancing values such as safety, legality, and mobility. An autonomous vehicle driving on the same roadways as humans likely needs to navigate based on similar values. For engineers of autonomous vehicle technology, the challenge is then to connect these human values to the algorithm design. To address this challenge, a mapping of philosophical frameworks to mathematical frameworks is used in order to motivate various design choices in a motion planning algorithm. Deontological ethics parallels rulebased mathematical concepts while consequentialism parallels costbased mathematical concepts. The philosophical theory of virtue ethics is also used to help motivate the relative weightings between the design objectives of path tracking, obstacle avoidance, and adherence to traffic laws. Experimental results of an autonomous vehicle navigating an obstructed twolane roadway with a double yellow line demonstrate the implications of the various design choices in a model predictive steering controller. In order to determine the success of the human values captured in an algorithm, the iterative methodology of value sensitive design (VSD) is used to formalize the connection of human values to engineering specifications. A modified VSD methodology is used to develop an autonomous vehicle speed control algorithm to safely navigate a pedestrian crosswalk. Two VSD iterations are presented that model the problem as a partially observable Markov decision process and use dynamic programming to compute an optimal policy to control the longitudinal acceleration of the vehicle based on the belief of a pedestrian crossing. The speed control algorithms are also tested in realtime on an experimental vehicle on a closedroad course.
 Also online at

Online 7. Biomechanical differences between rearfoot striking and forefoot striking in running [2018]
 Yong, Jennifer Rachel, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Running is a popular and beneficial form of exercise, but has notoriously high injury rates. One modification that has been suggested for reducing injury is changing foot strike pattern. The majority of recreational runners land on their heels when running, which is known as rearfoot striking. However, converting to forefoot striking and landing on the ball of the foot rather than the heel may be useful for reducing overuse injuries. Understanding the differences between rearfoot striking and forefoot striking during running and how these differences affect injury risk is necessary before promoting one foot strike pattern as protective. The goal of this dissertation was to improve the understanding of biomechanical differences between rearfoot striking and forefoot striking. Initially, I gained an understanding of how these running patterns differ by comparing muscle activity between habitual rearfoot striking runners and habitual forefoot striking runners in ten major lower limb muscles. Next, I used acute gait retraining on habitual rearfoot striking runners, asking them to perform two different running adaptations: converting to a forefoot striking pattern and increasing their cadence by 10%. I compared how these adaptations affected biomechanical parameters associated with a history of tibial stress fractures. From this study, I found that converting to forefoot striking and increasing cadence reduced different injury risk parameters. Both of these adaptations appear to potentially reduce the risk of tibial stress fractures, but they do so in different ways. Finally, I analyzed the effect of converting to forefoot striking on plantar flexor muscletendon dynamics using simulations driven by measured kinematics and electromyography. With these simulations, I evaluated acute adaptation to forefoot striking on elastic energy storage in the Achilles tendon and plantar flexor muscle mechanics. Based on increased activation, increased negative work, and eccentric contraction during early stance, runners interested in converting to forefoot striking should be mindful of the increased demands on the gastrocnemius. The experimental results and simulations from this thesis are publicly available, allowing others to build upon this work and continue to probe the differences between habitual rearfoot striking and converted forefoot striking.
 Also online at
Online 8. Biomechanics of hovering vertebrates [2018]
 Ingersoll, Robert Rivers, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Insects, hummingbirds, and nectar bats evolved the ability to hover in front of flowers to get access to energyrich nectar. It has been established that insects generate up to half of the lift needed to support their body weight during the upstroke. Estimates show that hummingbirds generate at least a quarter of their weight support during the upstroke by inverting their feathered wings more than generalist birds. In contrast, bats evolved membrane wings that they partially fold during the upstroke. While it has been hypothesized that hovering nectar bats generate vertical lift force during the upstroke, the complex structure of their wakes makes it hard to quantify this through flow measurement. To understand how hummingbirds and bats manipulate aerodynamic forces with their wings to perform these feats, we developed a new instrument that accurately measures these aerodynamic forces in vivo. This Aerodynamic Force Platform and an array of 3Dcalibrated highspeed cameras simultaneously recorded the vertical aerodynamic forces and wing kinematics. The pressure field generated by the animal travels to the boundaries of the flight volume at the speed of sound. The top and bottom plate mechanically integrate the pressure distribution, which is measured by three force sensors on each plate. By using the Aerodynamic Force Platform to measure these wingbeatresolved aerodynamic lift forces in vivo, we highlight similarities and differences across species and taxa. While it is known that insects improve efficiency by using elastic recoil for stroke reversal, it is unclear if hummingbirds converged on a similar solution, due to asymmetries in their lift generation and specialized flight muscle apparatus. We measured the aerodynamic force and kinematics of Anna's hummingbirds to resolve wing torque and power within the wingbeat. Comparing wingbeat resolved aerodynamic weight support measurements across species, we find that hummingbirds have low induced power losses similar to flies, much lower than typical for a generalist bird in slow hovering flight. We also show how hummingbirds' early muscle activation furnishes elastic recoil through stroke reversal to stay within the physiological limits of the pectoralis and supracoracoideus flight muscles. Expanding our species of interest, we traveled to Costa Rica with our new measurement device. We resolved the aerodynamic force and wing kinematics of 104 slow hovering hummingbirds and bats across 20 total species in vivo. While all hummingbirds we studied converged on an efficient horizontal wingbeat with an active upstroke to generate lift, the bats did notthey relied on drag to fully support their body weight. Remarkably, the nectar bats generate a significantly elevated vertical force during the upstroke compared to fruit batsby inverting their wing more like hummingbirdssuggesting convergent evolution.
 Also online at

Online 9. Building a nonMarkovian coarsegrained model [2018]
 Lee, Hee Sun, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Many scientific problems deal with timeevolution equations that are computationally too costly to solve in full resolution of a system because of the system size and the timescale involved in the problems. One strategy to reduce the computational cost is to describe the system in reduced dimensions, which can be divided into two tasks: 1) identification of the coordinates that effectively describe the characteristics that we are interested in, 2) finding a governing equation for those selected coordinates, which should account for the effect of ignored degrees of freedom. Following the common naming convention in the context of biomolecular systems, the description in reduced dimensions will be referred to as a coarsegrained (CG) model; and the selected coordinates will be referred to as CG coordinates, and the governing equation for CG coordinates that only involves the CG coordinates will be referred to as a CG equation. In this dissertation, we build CG equations based on the MoriZwanzig (MZ) formalism. The MZ formalism gives an exact governing equation for CG coordinates, which we will call the MZ equation. The MZ equation consists of the three terms: the mean term, the memory term and the fluctuation term. The memory term depends on current and past values of CG coordinates. The fluctuation term is a function of finegrain coordinates. So, to build a CG equation using the MZ equation, we need to 1) compute the memory term and 2) model the fluctuation term. The exact memory term given by the MZ formalism is prohibitively costly to compute since it includes a solution of a high dimensional partial differential equation. In the first part of the dissertation, we approximate the memory term so that we can evaluate the memory term without prohibitive computational cost. In the second part of the dissertation, we propose CG equations that are different from the standard MZ equation and thus have different forms of the memory term. Computing the memory term in our CG equations only requires trajectories of a full system without having to compute the high dimensional partial differential equation. With a certain type of approximation on the memory term, the MZ equation reduces to the the generalized Langevin equation (GLE). In the first part of the dissertation, we present a GLE approach when CG coordinates are multiple generalized coordinates, defined in general as nonlinear functions of microscopic coordinates. The CG equation for multiple generalized coordinates is described by the multidimensional GLE, which include the full memory matrix with nonzero offdiagonal entries. We first present a method to compute the memory matrix in a multidimensional GLE using trajectories of a full system. Then, in order to reduce the computational cost of computing the multidimensional friction with memory, we introduce a method that maps the GLE to an extended Markovian system. In addition, we study the effect of using a nonconstant mass matrix in the CG model. In particular, we include massdependent terms in the mean force. We used the proposed CG model to describe the conformational motion of a solvated alanine dipeptide system, with two dihedral angles as the CG coordinates. We showed that the CG model can accurately reproduce two important kinetic quantities: the velocity autocorrelation function and the distribution of first passage times. In the second part of the dissertation, we propose two CG equations that include more complicated memory terms than the GLE. The first CG equation has the memory term that can be nonlinear in CG coordinates. We tested the CG equation using a simple heatbath system. The second CG equation includes a coordinate dependent memory model. Only preliminary numerical results are presented for those two CG equations.
 Also online at

Online 10. Calculations of power flow in a cochlear model and measurements of innerhaircell stereocilia motion [2018]
 Wang, Yanli, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

The cochlea is an intricate biomechanical apparatus that converts the mechanical energy of sound into electrochemical energy in the auditory nerve fibers. The ability of mammals to distinguish different frequencies is attributed to the cochlea's ability to map specific sound frequencies onto different locations on the basilar membrane, a collagendense membrane extending along the length of the cochlea, like a mechanical Fourier transformer. The ability to hear an exceptionally large range of soundpressure levels is attributed to the nonlinearcompressive amplification granted by the active processes in a live cochlea, which also sharpens the frequency resolution of hearing. The exact mechanisms by which these active processes give rise to amplification and the sharpening of frequency tuning remain elusive. It is known that the electromotile outer hair cells, located on top of the basilar membrane alongside the sensory inner hair cells, are an important part of the mechanical apparatus and the source of the active processes. Both types of hair cell feature specialized apical modifications called stereocilia, which are closely related to microvilli. The two cell types have distinct tasks. Upon pushing and pulling on the stereocilia of the inner hair cells, action potentials of the nerve fibers innervating the inner hair cells are triggered sending neural signals to the auditory cortex. On the other hand, for the outer hair cells, deflection of the stereocilia instead causes a change in the length of the cell body via a unique form of electromotility, which helps to accomplish cochlear amplification. The organ that houses the two types of hair cell, the organ of Corti, has a sophisticated structure that facilitates the transfer of motion between the basilar membrane, the stereocilia, and the outer hair cells. The exact mechanisms for the transfer of motion within the organ of Corti also remain unknown. It is hoped that this thesis will bring us closer to explaining the mechanism of cochlear amplification and stereocilia stimulation for the inner hair cells. Chapter 1 of this thesis introduces cochlear mechanics, the morphology and structural implications of the organ of Corti, the physiology of the hair cells, and what is known and unknown about cochlear amplification and stereocilia stimulation. Ever since the discovery of the outerhaircell electromotility, whether or not the outer hair cells actually generate energy to accomplish cochlear amplification has been heatedly debated. Chapter 2 addresses this question with a theoretical approach using a mouse cochlea model. The power output of the outer hair cells is directly calculated, and the common misconceptions around energy dissipation are clarified. The results show that the outer hair cells do provide power into the cochlea, on the order of a few fW, and contrary to what have been assumed, the power increases with soundpressure level, so does the energy dissipated within the cochlea. Chapter 3 of this thesis aims to shine light on the mechanisms of stereocilia stimulation and their physiological relevance for the inner hair cells, by directly observing the innerhaircell stereocilia motion in situ using opened cochleae specimens from adult (P2021) mice. The stereocilia motions were recorded using a highspeed camera at 12,500 frames per second, in response to a 2 or 3 kHz stimulation applied to the stapes, which is the middleear bone that transmits sound into the cochlea. Subpixel motions of individual stereocilia were extracted using a customdesigned algorithm. It is observed that the individual stereocilia within one bundle exhibit semiindependent motion. The magnitude and phase of the stereocilia motion relative to the organ of Corti are also valuable for understanding the stimulation mechanism and validating computational models.
 Also online at

 Everhart, Camille Latrice Marie, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

The thermal properties of the materials used in microelectromechanical systems (MEMS) are of increasing concern as devices progressively miniaturize with improved fabrication technologies. Temperature sensitivity (or insensitivity) is often a major consideration for MEMS sensors, as many components of a sensor's output drift with and depend on temperature. Thus, thermal characterization should be a necessary step for any sensor material. This thesis begins with thermal conductivity measurements for sizelimited highlydoped silicon and ALD platinum. Though these results were mainly internal to labgroup needs, the methods used are widely applicable to other studies. As MEMS manufacturing techniques and technologies advance and mature, previously encountered limitations should no longer be assumed. The thermal accelerometer is a device that has found limited use within the MEMS community, thanks to its high power consumption and notably low bandwidth. However, with new fabrication techniques the thermal accelerometer's performance can be pushed in new and exciting ways, potentially expanding its usability for a larger variety of applications. Here we present the results from the first thermal accelerometer fabricated using Plasma Enhanced Atomic Layer Deposition (PEALD). PEALD allows for ultrathin, low defect, highdensity platinum films that deliver excellent stability and accuracy. We offer a 100× crosssection reduction relative to previous thermal accelerometers, thereby increasing the heating efficiency and decreasing thermal time constants. The suspended resistance thermometers and heaters capitalize on the properties provided by PEALD and Pt: pronounced stability, high resistivity, and linear temperature coefficient of resistance (TcR), giving our device tractable and consistent results. For a heater temperature rise of 150 degrees C and acceleration between +/ 1g, this device has a raw sensitivity of 54e3 C/g, excellent crossaxis isolation, highlinearity, and remarkably low drift.
 Also online at
Online 12. Coherent pattern identification in fluid flows [2018]
 SchlueterKuch, Kristy Lynn, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

The motion of complex fluid flows, and material advection in them, can be understood though the identification and analysis of persistent transport features in the flow. These features, often referred to as coherent structures, act as a robust framework of material surfaces that separate the flow into distinct regions which resist mixing with surrounding regions. Understanding the evolution of these structures is a critical step in identifying the mechanisms underlying fluid transport. Material transport is central to many geophysical flow phenomena, such as the development of marine ecosystems, and the spread of pollutants including volcanic ash and oil. However, many of the techniques currently available for measuring such flows, including tracking arrays of Lagrangian drifters (e.g. ocean surface drifters and weather balloons), result in sparse and spatially irregular velocity data. This is insufficient for the use of many coherent structure detection algorithms, which rely on an assumption of initially closelyspaced fluid tracers. Additionally, current methods often focus on determination of the full boundaries of coherent sets, whereas in practice, it is often more valuable and practical to identify the complete set of trajectories that are coherent with an individual trajectory of interest. Motivated by the successes in using coherent structure analysis to study transport, and considering the limitations of measuring largescale geophysical flows, this work details the development of an algorithm for detecting coherent patterns from potentially sparse Lagrangian flow trajectories. The method, based on principles from graph coloring and unsupervised machine learning, groups trajectories based on a measure of pairwise dissimilarity between their displacement patterns. Through the use of several analytical and experimental validation cases, this method detects coherent flow patterns using significantly less data than is required by existing techniques. Consideration of less optimal groupings of trajectories also allows for the detection of all trajectories coherent with a chosen trajectory of interest. Due to the robustness of the coherent patterns to effects such as chaotic advection, the flow structures identified prove useful in the assimilation of Lagrangian trajectory data into models of geophysical flows. Importantly, although the method is demonstrated here in the context of fluid flow kinematics, the generality of the approach allows for its potential application to other unsupervised clustering problems in dynamical systems.
 Also online at

Online 13. Computational analysis of canonical problems arising in the interaction of heated particles and a fluid [2018]
 Ganguli, Swetava, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Numerical simulations of turbulent solidparticleladen flows are challenging due to the simultaneous occurring of multiple phenomena spanning widely different scales. Presence of heat transfer, which occurs when the particulate phase is irradiated with radiation, complicates this problem further. An important application of multiphase particleingas flow is that of a solidparticlebased, solarthermal receiver to convert solar energy to electrical energy. To this end, the Predictive Science Academic Alliance Program (PSAAP) II at Stanford University, focuses on advancing the stateoftheart in largescale, predictive simulations of irradiated particleladen turbulence relevant to concentrated solar power (CSP) systems. The aim of this thesis is to design and analyze a hierarchical set of increasingly complex and physically representative canonical problems that involve the interaction of subsets of the aforementioned physical phenomena. This thesis is divided into two parts  the first dealing with the interaction of a single heated particle with an initially quiescent flow; the second dealing with the interaction of heated particle(s) with an initially uniform flow. In part I, the temporal evolution of the initial shock front and the low Mach flow field produced behind the front due to the presence of a spherical, finitesize heat energy source in a gas is investigated. The source of energy could either be some internal heat source within the sphere or radiation heating the sphere assuming that the fluid is optically thin. This leads to either a Dirichlet or a Robin boundary condition on the surface of the sphere. When the source is internal, the assumption is that the timescale of heating is much smaller than the diffusion time scale of the fluid. While the study of the sphere is of physical interest, the analogous problems of a uniformly heated infinitely long cylindrical wire and an infinite plate are also considered. These problems serve as model problems to study and quantify finitesize effects, nonBoussinesq effects and compressibility effects without making any assumptions on the amount of heat addition from the sphere into the fluid. The study encompasses heating regimes where the Boussinesq approximation holds and regimes where it breaks down. At small times after the boundary conditions are imposed, compressibility effects are significant and a strong shock wave forms. This shock wave weakens as it moves away from the source eventually leading to an acoustic wave as the pressure settles everywhere to the ambient value for sufficiently large times. The features of this shock wave are analyzed in detail. Following the shock wave, the fluid motion occurs at a much lower speed which allows for a low Mach number formulation of the NavierStokes equations. In this low Mach regime, the resulting nonlinear energy equation is solved analytically using the method of Homotopy Perturbation Expansion. This leads to a weak decoupling of finitesize effects and nonBoussinesq effects thereby allowing the quantification of the individual impacts of both phenomena on the total solution. In part II, fully resolved simulations are first used to quantify the effects of heat transfer in the absence of buoyancy on the drag of a spatially fixed heated spherical particle at low particle Reynolds numbers (Re) in the range 0.001 to 10, in a variable property fluid. This analysis is carried out without making any assumptions on the amount of heat addition from the sphere and thus encompasses both, the heating regime where the Boussinesq approximation holds and the regime where it breaks down. The particle is assumed to have a low Biot number, which means that the particle is uniformly at the same temperature and has no internal temperature gradients. Large deviations in the value of the drag coefficient as the temperature of the sphere increases are observed. When Re is less than 0.01, these deviations are explained using a low Mach perturbation analysis as irrotational corrections to a StokesOseen base flow. Correlations for the drag and Nusselt number of a heated sphere are proposed for the range of Reynolds numbers from 0.001 to 10, which fit the computationally obtained values with less than 1% and 3% errors respectively. These drag and Nusselt number correlations can be used in simulations of gassolid flows where the accuracy of the drag law affects the prediction of the overall flow behavior. Finally, an analogy to incompressible flow over a modified sphere is demonstrated. Then, fully resolved simulations are used to quantify the effects of heat transfer in the presence of buoyancy on the drag of a spatially fixed heated spherical particle at low Reynolds numbers in the range 0.001 to 10 in a variable property fluid with the same assumptions as outlined before. A nondimensional number called Buoyancy Induced Viscous Reynolds Number, which is analogous to the Grashof number, is derived using scaling analysis. No assumptions are made on the magnitude of this number either. The effects of the orientation of gravity relative to the freestream velocity are also examined. Under appropriate constraints on the Buoyancy Induced Viscous Reynolds Number and the Reynolds number, the total drag on a heated sphere in the presence of buoyancy is shown to be, within 10\% error, the linear superposition of the drag computed in two canonical setups: one being the setup studied in chapter 9 and the other being natural convection. To demonstrate the effect of the drag modification due to heat transfer and buoyancy, the settling time and the settling velocity of a falling particle is shown using the proposed correlations. Then, heated spherical particles in infinite periodic arrays are considered. Drag enhancement is observed as the distance between particles decreases. However, the ratio of the viscous drag to the pressure drag on the sphere remains very close to 2:1. Finally, the individual effect of inertial, thermal, and clustering corrections to Stokes drag is evaluated on three sets of particle clusters extracted from the stateoftheart pointparticle simulations of the PSAAPII channel. When compared to the true drag experienced by these particles in these clusters (as evaluated from particleresolved simulations), just using Stokes' drag formula results in up to 18\% error which can be reduced to less than 2\% if the corrections proposed in this thesis are accounted for.
 Also online at

Online 14. Computational modelling and uncertainty quantification of blood flow in the coronary arteries [2018]
 Tran, Justin Sheldon, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Atherosclerotic coronary artery disease continues to negatively impact the lives of millions worldwide. Computational fluid dynamics modeling of coronary blood flow has the potential to help improve clinical outcomes and aid in treatment planning. Significant advancements in coronary blood flow modeling in recent years have opened a wide range of applications such as assessing risk for disease progression or providing a platform for virtual surgery and treatment planning. To encourage the growth of this field and promote adoption of computational results in the clinic, it is crucial that these tools be made as automated as possible so they can be applied to large patient cohorts. In addition, the variability of computational results with respect to uncertainties in the inputs and model must be better understood and systematically quantified. Addressing these concerns is the subject of this thesis. In the first part, a framework for automatically tuning the lumped parameter boundary conditions in simulations of coronary blood flow is developed and demonstrated. Specifying boundary conditions in complex computational models is not a trivial task, especially when the dimensionality of the input space is high and multiple constraints on the outputs need to be satisfied simultaneously. Specifically in the context of patientspecific coronary simulations, clinical data such as the blood pressure, cardiac output, and coronary flow waveforms must be simultaneously satisfied with a large set of input parameters that include lumped resistances, capacitances, and heart model parameters. A typical user can eventually gain expertise to modify the input parameters to satisfy targets, but this manual tuning is timeconsuming and not easily reproduced. We thus formulate the automated tuning process as a Bayesian inverse problem in which the model parameters are treated as random variables, and optimal parameters are determined by finding the maximum of the posterior distribution of input parameters. We also perform sensitivity analysis on the input parameters to determine a subset of thirteen parameters that most influence the clinical targets. In the second part, we perform uncertainty quantification on patientspecific simulations of coronary artery bypass graft hemodynamics. Vein graft failure in patients with coronary bypass continues to be a major clinical issue with relatively little knowledge about the mechanisms for failure. Simulations have shown that predicted quantities such as wall shear stress or wall strain can be useful in predicting vein graft failure, but adoption of such results in clinical practice is hindered due to the fact simulations can only produce deterministic results with no range of confidence. Uncertainty quantification provides a framework for quantifying the uncertainty in computational results, and we applied it to assess the variability in computed predictions of timeaverage wall shear stress and wall strain under uncertainty in the lumped parameter boundary conditions and vessel wall material properties. To achieve this aim efficiently, we develop a novel submodeling strategy for reducing the computational cost of the analysis. We also, for the first time, consider spatial variability in the graft wall material properties by using a random field description. We finally propagate these uncertainties forward using a newly developed multiresolution approach. The results show that the timeaveraged wall shear stress is relatively well estimated with confidence intervals about 35\% of the mean value, but the wall strain exhibited significantly more variability due to the large uncertainty in the material properties. In the third part, we perform a comparison of methods for modeling wall deformability in vascular blood flow simulations. Though sometimes neglected, wall deformability can have significant impacts on the computational results, affecting predictions of wall shear stress and precluding calculation of stresses and strains in the vessel wall. There are several methods proposed in the literature for modeling wall deformability, two of the most popular being the Arbitrary Lagrangian Eularian (ALE) and Couple Momentum Methods (CMM). Although both methods capture the essential characteristics of wall deformability, they can produce different results and computational performance. This provides a rigorous comparison which will aid in the choice of deformable wall model. Additionally, we consider the concept of prestress. Because the geometry for a patientspecific simulation is extracted from medical image data of the \textit{in vivo} cardiovascular system, the vessel walls carry an internal stress which holds the geometry in equilibrium with hemodynamic pressures and viscous stresses. We implement prestress in both ALE and CMM contexts and confirm that it is necessary to avoid overinflation of the anatomic domain. Although studied mostly within the context of coronary flow simulations, the methods and approaches outlined in this thesis are designed to be generally applicable across other domains in computational modeling, fluid dynamics, and biomechanics. Automated tuning is a general framework for assimilating multiple sources of target data to inform optimal input parameter values, and can broadly be applied in multiscale modeling. The methods for uncertainty quantification can be adapted to assess variability of simulations in other computational fluid mechanics and biomechanics contexts. The results from the wall deformability comparison can also be extended to apply to other contexts including other cardiovascular diseases, respiratory flow, and medical devices. In addition to providing insights into coronary flow simulations, this thesis aims to motivate the importance of tuning, uncertainty quantification, and model comparisons for other cardiovascular simulations and multiscale biological modeling more broadly.
 Also online at
Online 15. A consistent reaction mechanism and model for the combustion and gasification of coal and biomass, and the cofiring thereof [2018]
 Tilghman, Matthew B., author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

This work seeks to provide a complete and thermodynamicallyconsistent reaction mechanism for the combustion and gasification of coal and biomass chars. In pursuit of this, mass loss data obtained via thermogravimetric analysis as well as data from temperature programmed desorptions were used to determine kinetic and thermochemical parameters. These parameters were determined via their adjustment in an intrinsic chemical reactivity model, which was developed to predict char conversion rates and offgas compositions when pulverized coal and biomass char particles are exposed to reactive gases within our pressurized thermogravimetric analyzer (PTGA). Char reactivity tests were performed in H2O/H2/N2 environments in order to obtain data to determine kinetic parameters for char reactivity to H2O, in CO2/CO/N2 environments in order to obtain data to determine kinetic parameters for char reactivity to CO2, and in O2/N2 environments in order to obtain data to determine kinetic parameters for char reactivity to O2. This work then seeks to assess whether the reactive properties of the two individual fuels are altered when they are cofired together. The two fuels were cofired together at various mixture fractions, and subjected to similar reactivity tests as the pure fuels. Results indicate that kinetic parameters of the pure fuels alone cannot be used to predict reactive properties when cofired, as properties of the coal char are indeed affected when devolatilizing in close proximity to biomass. To explain and account for these changes, this work proposes two models, a mixedchar surface area model, and a mixedchar reactivity model. The underlying principle is the same: observed changes can largely be predicted by employing a combination of the properties if three types of chars instead of the originally assumed two types. A coal particle that devolatilizes in an area that had just immediately been affected by a prior biomass particle devolatilization possesses different properties than a coal particle that devolatilizes in an unaffected area. Thus, the mixedchar models include properties of the original coal char, the original biomass char, and a third "affected" coal char. Finally, this work then seeks to extend the results beyond the kineticallycontrolled regimes in which the data were obtained. In many conditions of interest, diffusion phenomena begins to impact the particle conversion rates, so predicting intrinsic reactivity alone is not enough to predict particle conversion rates. A fullyresolved DNS model was created that jointly simulates the internal pore diffusion of the particle as well as the locationresolved char reactivity in order to assess methods for accounting for the effects of diffusion. The main method this work investigates is Thiele's method, which makes use of an analytical solution that relates the Effectiveness Factor (η) to the Thiele Modulus (φ). This work reveals that the analytical η − φ relationship set forth by Thiele does not accurately apply to CH2O and CCO2 reactions, due primarily to the fact that the elementary reactions are no longer irreversible and 1storder (but rather are reversible and inhibited by their products). However, Thiele's method can still be used with slight changes proposed in this work.
 Also online at

Online 16. Control of electric motors and drives via convex optimization [2018]
 Moehle, Nicholas, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Electric motors and drives have been in common use for more than a century. Traditionally, alternatingcurrent electric motors are driven by applying symmetric, multiphase sinusoidal voltage waveforms to their terminals. However, due to the wide availability of switching power converters as well as microcontrollers for controlling them, it is now possible to drive motors using specialized, nonsinusoidal waveforms, allowing us to design the drive voltage waveform along with the motor. This is one of our main points in this work: techniques and technologies now exist for controlling unconventional electric motors. We first consider optimal voltage and current waveform design for electric motors, then considers the design of feedback controllers for switchedmode power converters that could be used to implement these waveforms. First we give energyoptimal current waveforms for a permanent magnet synchronous motor that result in a desired average torque. Our formulation generalizes previous work by including a general backEMF wave shape, voltage and current limits, an arbitrary phase winding connection, a simple eddy current loss model, and a tradeoff between power loss and torque ripple. Determining the optimal current waveforms requires solving a convex optimization problem. We show how to use the alternating direction method of multipliers to find the optimal current in milliseconds or hundreds of microseconds, depending on the processor used, which allows the possibility of generating optimal waveforms in real time. We next present a method for finding current waveforms for induction motors that minimize resistive loss while achieving a desired average torque output. Our method is not based on referenceframe theory for electric machines, and therefore directly handles induction motors with asymmetric winding patterns, nonsinusoidally distributed windings, and a general winding connection. We do not explicitly handle torque ripple or voltage constraints. Our method is based on converting the torque control problem to a nonconvex linearquadratic control problem, which can be solved by using a (tight) semidefinite programming relaxation. We then address the problem of finding current waveforms for a switched reluctance motor that minimize a userdefined combination of torque ripple and RMS current. The motor model we use is fairly general, and includes magnetic saturation, voltage and current limits, and highly coupled magnetics (and therefore, unconventional geometries and winding patterns). We solve this problem by approximating it as a mixedinteger convex program, which we solve globally using branch and bound. We demonstrate our approach on an experimentally verified model of a fully pitched switched reluctance motor, for which we find the globally optimal waveforms, even for high rotor speeds. After considering the motor design problem, we consider control of switchedmode power electronic converters. First we consider the theoretical problem of controlling general switchedaffine dynamical systems i.e., the problem of selecting a sequence of affine dynamics from a finite set in order to minimize a sum of convex functions of the system state. We develop a new reduction of this problem to a mixedinteger convex program (MICP), based on perspective functions. Relaxing the integer constraints of this MICP results in a convex optimization problem, whose optimal value is a lower bound on the original problem value. We show that this bound is at least as tight as similar bounds obtained from two other wellknown MICP reductions (via conversion to a mixed logical dynamical system, and by generalized disjunctive programming), and our numerical study indicates it is often substantially tighter. Finally, we consider the problem of controlling switchedmode power converters using model predictive control. Model predictive control requires solving optimization problems in real time, limiting its application to systems with small numbers of switches and a short horizon. We propose a technique for using offline computation to approximate the model predictive controller. This is done by dividing the planning horizon into two segments, and using a quadratic function to approximate the optimal cost over the second segment. The approximate model predictive algorithm minimizes the true cost over the first segment, and the approximate cost over the second segment, allowing the user to adjust the computational requirements by changing the length of the first segment. We conclude with two simulated examples.
 Also online at

Online 17. Design and fabrication of high efficiency metal oxide and siliconbased photoelectrodes for solar water splitting [2018]
 Zhao, Jiheng, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Renewable energies have attracted significant attention in recent years due to the increasing energy demand and critical environmental problems. Among all the renewable energy sources, solar energy is the most promising one with the largest capacity (12,000 TW), minimal environmental impact, and inexhaustible usage. Photoelectrochemical (PEC) water splitting, which can directly convert the solar energy into chemical bonds (hydrogen and oxygen) through splitting water molecules, is one of the most promising ways to solve those issues. This technique has several advantages: First of all, since it can directly convert the solar energy to chemical bonds (e.g. H2), so the solar energy can be appropriately stored, which can the solve the intermittent supply of sunlight. Second, the stored hydrogen can be used freely based on the demands. Third, the whole solar to hydrogen converting process is completely carbon free, which can ultimately solve the environmental issues by reducing the greenhouse gas emissions. Thus, researchers have spent tremendous efforts on exploring and developing efficient and stable photoelectrodes for effective photocatalytic/electrocatalytic solar conversion. Metal oxides and silicon are two promising candidates due to their unique material properties. However, both metal oxides and silicon have their own drawbacks that limiting their performance for solar fuel generation. This thesis presents some novel methods developed towards improving metal oxide and silicon based photoelectrodes performance. The mismatch between the light absorption depth and charge carrier diffusion length is the biggest challenge for metal oxide photoanodes. Fabricating ultrathin lightabsorbers on textured substrates offers an effective way to solve this issue. Such ultrathin films with the thickness smaller than the minority charge carrier diffusion length, can effectively enhance the charge collection efficiency. Additionally, textured substrates would greatly enhance both light absorption and surface reactions of photoanodes, and thus reduce the total amount of lightabsorbers needed. Furthermore, the overall PEC performance can be further enhanced by properly control the interfacial layer morphologies. An amphiphilic graft copolymer selfassembly method was developed to create a mesoporous metal oxide interfacial layer to significantly reduce the interfacial charge recombination. The biggest challenges about Si are the instability when operating under photoanodic conditions in aqueous electrolyte and surface kinetics for water splitting reaction. To date, the most effective protection layer for Si photocathodes in alkaline solutions is the atomic layer deposited (ALD) dense TiO2 layer combined with noble metalbased catalysts on top. However, those high vacuum based techniques are expensive and hard to scale up for practical applications. Herein, two novel solutionbased deposition methods were developed to replace the highvacuum based techniques for efficient Si photoelectrode protection. The first method used a modified hydrothermal method to deposit earthabundant NiFelayered double hydroxide (LDH) to simultaneously protect and catalyze Si photocathodes in alkaline solutions. The second method used a modified electroless deposition (ELD) method to uniformly deposit protective and catalytic Ni films on Si wafers, resulting in an efficient and stable Si photoanode for solar water oxidation.
 Also online at

Online 18. Effect of constraint on mechanical behavior of polymer layers and membranes in energy applications [2018]
 Yuen, Pak Yan, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

The role of mechanical constraint is well known in the mechanical behavior of engineering materials such as adhesive joints and alloys. However, in the case of polymer layers and membranes, the effect of mechanical constraint is not always understood. In particular, in energy technologies, device geometry is becoming increasingly complex but the role of mechanical constraint is not explored. In proton exchange membranes (PEM) fuel cells, the PEM is constrained by the adjacent current collector plates such that the membrane can only deform in the flow channel areas. I showed that the tearing energy decreases with increasing constraint associated with a smaller plastic zone and is much lower than the tearing energy with no constraint. In photovoltaics, the backsheet structure is a layered film where the tearing behavior of the individual layers does not necessarily represent the tearing behavior of the entire backsheet. I showed that such characteristic arises from the interaction between the individual layers during the tearing process, where one layer of the backsheet is mechanically constrained by its neighboring layers and the layers may debond from each other. Finally, I present the mechanism through which mechanical constraint enhances the stiffness of a novel polymer composite electrolyte for batteries. The composite electrolyte is composed of a silica aerogel monolith filled with a polymer electrolyte. When the voids of the aerogel are filled with the polymer, the polymer constrained the bending of the cell wall and effectively enhanced the stiffness of the composite. This work highlights the importance of considering the effects of mechanical constraint as it can lead to surprising effects on the mechanical and fracture properties of materials.
 Also online at

Online 19. Enabling multimodal robots via controllable adhesives [2018]
 Estrada, Matthew Alfonso, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

This thesis is about the design and analysis of robots that use adhesives to combine multiple modes of operation. Examples of multimodal operation include ballistic or powered flight combined with perching and crawling. In each case, the robots are made possible by a proliferation of components  from microprocessors to sensors and motors  that have accompanied the growth of drones or quadrotors in consumer markets. The components are compact and light enough that it is possible to support multiple modes of operation on a small platform. The thesis takes a cue from small creatures such as insects, most of which have multiple modes of operation (e.g. flying and crawling) and which can often move objects many times their weight through the use of attachment mechanisms at the tarsus of each limb. Examining the strategies of insects such as wasps leads to insights for designing and modeling the multimodal robots outfitted with geckoinspired adhesives. The resulting platforms are capable of tasks that no single mode of operation can support. For example, one of these platforms, named FlyCroTugs, can fly rapidly to remote sites, attach a tether to a heavy object, land and then pull that object with a force many times the robot's weight. The FlyCroTugs use geckoinspired adhesives to anchor themselves when applying large forces. Another example involves a small robot, named KlingOn, that can transition from ballistic flight to crawling on a vertical surface, using the same geckoinspired adhesives. A third example involves a gripper that can capture freefloating objects with a flexiblebacked adhesives. Further insights arise when considering scaling laws applicable to small multimodal robots that use adhesion. At the scale of these robots, contact forces typically dominate the dynamics when the robots are interacting with objects. Therefore, modeling the force constraints associated with the adhesives leads to corresponding dynamic constraints on the robots, in terms of their trajectories and velocities. The adhesive force constraints also have implications for the dimensions, geometry, stiffness and damping of the robot attachment pads and grippers. Ultimately this thesis posits that increased attention to robotic endeffectors and attachment mechanisms can promote the efficacy of small, multimodal robotic systems interacting with their environment.
 Also online at

Online 20. Energy consumption and salt adsorption in capacitive deionization [2018]
 Hemmatifar, Ali, author.
 [Stanford, California] : [Stanford University], 2018.
 Description
 Book — 1 online resource.
 Summary

Challenges for clean water are global and diverse, and flexible water treatment approaches are needed. To date, over 4 billion people, more than half of the world's population, live in water scarce areas where withdrawn water surpasses the amount the region can sustainably support for at least one month per year. Slightly under 4 billion live in areas of severe scarcity conditions. Aggravating the situation further, naturally occurring toxins, chemical contamination introduced by human activities, and high salinity make many already limited water sources unsafe for consumption. Water desalination and disinfection provides viable solution to deficit of clean water. Seawater desalination is currently the main source of clean water production, however, availability of seawater is geographically uneven and thus not an available option targeted towards inland regions. Brackish water (water with a low to moderate salt content), however, is relatively common and can provide an appealing source of potable water with appropriate treatment technologies. Capacitive desalination or capacitive deionization (CDI) is an electrosorptive desalination method that leverages porous and conductive electrodes for electrostatic ion adsorption. Upon application of a small voltage (order 1 V) across each electrode pair, salt ions are removed from feed water and electrostatically held within the electrode pores. The CDI cell is then regenerated by removing or reversing the voltage, which spontaneously releases the ions and forms brine solution. CDI has a number of advantages over common desalination techniques. Most importantly, it does not require high pressure or temperature to operate, is widely scalable, and thus relevant for distributed applications (as investment and infrastructure cost is low and is directly proportional to plant capacity). CDI is potentially energy efficient and cost effective for brackish water desalination, since the energy cost per volume of treated water roughly scales with the amount of removed salt (rather than volume of treated water). CDI is thus the most advantageous in brackish water desalination as well as water recycling and reuse where salt content is far below that of seawater. In addition, CDI has the great potential for selective removal of ionic species based on ion valence, hydrated ion size and pore size, surface chemistry, and pH environments. We first focus on electrosorptive desalination energy, in both theory and practice. We present a general topdown approach to show minimum energy of ion separation is indeed Gibbs free energy of separation for most known EDLs irrespective of EDL geometry and thickness. We fabricate a low series resistance CDI cell, operate the cell at various current and flow rates, and demonstrate lowenergy desalination with unprecedented 9% thermodynamic efficiency and only 4.6 kT energy requirement per removed ion. We further experimentally quantify individual loss mechanisms and show resistive and Faradaic losses as two main loss mechanisms. We show the two loss mechanisms favor different charging rates: resistive losses are dominant at high charging currents, but Faradaic losses are dominant at low charging rates, as the cell spends longer time at high voltage. Our results provide a powerful tool for optimizing CDI operation. In addition to study of desalination energy, we study charge and species transport of electrosorption process. We formulate and solved the first twodimensional model of a CDI cell coupling external electrical network, charge conservation, and mass conservation in bimodal pore structure electrodes. We fabricate a labscale CDI cell, experimentally calibrate the model, and show a good agreement between model results and experimental data. Our results show CDI process has two distinct phases: a fast adsorption step at the beginning of charging followed by a slow salt removal step. Finally, we study the effect of surface functional groups on pH dependent salt adsorption and ion selectivity by developing theory and performing controlled experiments. To this end, we expanded the current surface charge models by coupling a double layer model with acidbase equilibria theory and further validate the model by wellcontrolled titration experiments. The fitted model with one acidic and one basic surface group showed a very good agreement with the experiments. Our results show (1) specific adsorption of cations and expulsion of anions at electrolyte pH values higher than pK of acidic groups, and (2) specific adsorption of anion and expulsion of cations at pH values lower than pK of basic groups.
 Also online at
