Typescript (photocopy). / Digitized by Kansas Correctional Industries

Typescript (photocopy). / Digitized by Kansas Correctional Industries

No description available.

The microelectronics industry has been managing the RC delay problem arising from aggressive line scaling, by replacing aluminum (Al) by copper (Cu) and oxide dielectric by low-k dielectric. Electromigration (EM) turned out to be a serious reliability problem for Cu interconnects due to the implementation of mechanically weaker low-k dielectrics. In addition, line width and via size scaling resulted in the need of a novel diffusion barrier, which should be uniform and thin. The objective of this dissertation is to investigate the impacts of Ta barrier process, such as barrier-first and pre-clean first, and scaling of barrier and line/via on EM reliability of Cu/low-k interconnects. For this purpose, EM statistical test structures, having different number of line segments, line width, and via width, were designed. The EM test structures were fabricated by a dualdamascene process with two metal layers (M1/Via/M2), which were then packaged for EM tests. The package-level EM tests were performed in a specially designed vacuum chamber with pure nitrogen environment. The novel barrier deposition process, called barrier-first, showed a higher (jL)[subscript c] product and prolonged EM lifetime, compared with the conventional Ta barrier deposition process, known as pre-clean first. This can be attributed to the improved uniformity and thickness of the Ta layer on the via and trench, as confirmed by TEM. As for the barrier thickness effect, the (jL)c product decreased with decreasing thickness, due to reduced Cu confinement. A direct correlation between via size and EM reliability was found; namely, EM lifetime and statistics degraded with via size. This can be attributed to the fact that critical void length to cause open circuit is about the size of via width. To investigate further line scaling effect on EM reliability, SiON (siliconoxynitride) trenchfilling process was introduced to fabricate 60-nm lines, corresponding to 45-nm technology, using a conventional, wider line lithograph technology. The EM lifetime of 60-nm fine lines with SiON filling was longer than that of a standard damascene structure, which can be attributed to a distinct via/metal-1 configuration in reducing process-induced defects at the via/metal-1 interface. / text

Not available

Electron transport in nanoparticle single-electron transistors (SETs) is a fruitful method to explore a wide range of physical phenomena at the nanometer scale. In this thesis, we investigate electron transport in SETs incorporating various nanoparticles, including gold nanoparticles in both classical and quantum regimes and Pb nanoparticle in both superconducting and normal states. SETs have been successfully fabricated by incorporating individual gold nanoparticles into the gaps between two electrodes. Although single-electron tunneling behavior is prominent, quantized energy levels cannot be resolved in these SETs due to their relatively large particle sizes. A novel method has been developed to achieve SETs incorporating gold nanoparticles whose sizes are small enough to resolve discrete quantum energy levels. The devices consist of spontaneously-formed ultrasmall gold nanoparticles linked by alkanedithiols to gold electrodes. The devices reproducibly exhibit addition energies of a few hundred meV, which enables the observation of single electron tunneling at room temperature. At low temperatures, resonant tunneling through discrete energy levels in the Au nanoparticles is observed, which is accompanied by the excitations of molecular vibrations at large bias voltage. Having explored the SETs in normal state, we have extended the experiments to superconducting single-electron transistors (SSETs). We first fabricated and characterized Pb superconducting electrodes with nanometer-sized separation. Our observation clearly shows that conventional Barden-Cooper-Schrieffer theory remains valid to interpret the tunneling behavior between two nanometer-spaced Pb electrodes. Furthermore, by incorporating Pb nanoparticles between the two Pb electrodes, we have fabricated SSETs and investigated the transport properties of these devices. In the superconducting state, the conductance is suppressed by a combination of the single electron tunneling effect and the absence of density of states within the superconducting gap. In the suppression regime, the tunneling spectroscopy shows current features that arise from quasiparticle tunneling caused by singularity matching. At low temperature, the features can only be observed for odd charge states in SSETs. At high temperature, the odd-even parity effect is smeared out. Upon application of a magnetic field, the superconducting state is suppressed and single-electron tunneling behavior for normal metallic nanoparticles is recovered.

Not available / text

M.A. / This dissertation comprises an in-depth investigation into the domain of computer vision, with specific reference to the textile industry. The study consists of three main sections. In the first section, the computer-vision process is scrutinised in its entirety. Attention is given in this regard to what computer vision is, where it originated from, how it compares with human vision and what the motivations are for its implementation. Following, the computer-vision process is divided into four main components, namely image acquisition, image processing, image analysis and image interpretation. Subsequently, each component is discussed in greater detail, as well as aspects such as the hardware used in the course of the process, the algorithms that are implemented and the specific applications used for the process of analysis. In the second section, the focus is shifted to the textile industry, where our main focus lies. In this regard, examples are examined of the successful implementation of computer-vision technologies in the textile industry. In the third section, an investigation is launched into the specific problem for which a solution needs to be found in the present study, namely to determine whether computer vision constitutes a cost-effective way in which to locate broken thread during the spinning process. A wide range of algorithms has been applied for this purpose, whereafter the results of these experiments are announced.

Advancements in robotics have the potential to aid humans in many realms of exploration as well as daily life: from search and rescue work, to space and deep sea exploration, to in-home assistance to improve the quality of life for those with limited mobility. One of the main milestones that needs to be met for robotics to achieve these ends is a robust ability to manipulate objects and locomote in cluttered and changing environments. A prerequisite to these skills is the ability to understand the current state of the world as well as how actions result in changes to the environment; in short, a robot needs a way to model itself and the world around it. With recent advances in machine learning and access to cheap and fast computation, one of the most promising avenues for creating robust models is to learn a neural network to approximate the dynamics of the system. Learning a data-driven model that accurately replicates the dynamics of a robot and its environment is an active area of robotics research. This model needs to be accurate, it needs to operate using sensors that are often high dimensional, and it needs to be robust to changes within the system and the surrounding environment. In this thesis, we investigate ways to improve the processof learning data-driven dynamics models as well as ways to reduce the dimensionality of a robot’s state space. We start by trying to improve the long-term accuracy of neural network based forward models. Learning forward models is more complicated than it appears on the surface. While it is easy to learn a model to predict the change of a system over a short horizon, it is challenging to assure this performance over a long horizon. We investigate the concept of adding temporal information into the loss function of the forward model during training; we demonstrate that this improves the accuracy of a model when it is used to predict over long horizons. While we are currently working with low dimensional systems, we eventually want to apply our learned models to robots with high dimensional state spaces. To make learning feasible, we need to find ways to learn a lower dimensional representation of the state space (also known as a latent space) to make learning models in the real world computationally feasible. We present a method to improve the usefulness of a learned latent space using a method we call context training: we learn a latent space alongside a forward model to encourage the learned latent space to retain the variables critical to learning the dynamics of the system. In all of our experiments, we spend significant time in analysis and evaluation. A large portion of literature demonstrating the effectiveness of data-driven forward models in robot control settings often only presents the final controller performance. We were often left curious about what the model was learning independent of the control scenario. We set out to do our own deep dive into exactly what data-driven forward models are predicting. We evaluate all of our models over long horizons. We also look deeper than just the mean and median loss values. We plot the full distribution of loss values over the entire horizon. The literature on data-driven models that do evaluate model prediction accuracy often focuses on the mean and median prediction errors; while these are important metrics, we found that looking at these metrics alone can sometimes obscure subtle but important effects. A high mean loss is often a result of poor performance on only a subset of the test dataset; one model can outperform other models with lower mean error values on a majority of the test set, but it can be skewed to look like the worst performer by having a few highly inaccurate outliers. We observe that models often have a subset of a test dataset on which they perform best; we seek to limit the use of a model to regions of the test dataset where it has high accuracy by using an ensemble of models. We find that if we train an ensemble of forward models, the accuracy of the models is higher when they all agree on a prediction. Conversely, when the ensemble of models disagrees, the prediction is often poor. We explore this relationship and propose future ways to apply it. Finally, we look into the application of improved model accuracy and context trained latent spaces. We start by testing the performance of our context training architecture as a method to reduce the state space dimensionality in a model-free reinforcement learning (MFRL) reaching task. We hypothesize that a policy trained with a latent space observation derived using our context trained encoder will outperform a policy trained with a latent space observation derived from a standard autoencoder. Unfortunately, we found no difference in task performance between the policies learned using either method. We end on a bright note by looking at the power of model-based control when we have access to an accurate model. We successfully use model predictive control (MPC) to generate robust locomotion for a simulated snake robot. With access to an accurate model, we are able to generate realistic snake gaits in a variety of environments with very little parameter tuning that are robust to changes in the environment.

The purpose of this study was to develop a simple one- dimensional finite element for the nonlinear analysis of symmetrically and unsymmetrically laminated composite beams including shear deformation. There is a need for a simple and efficient method for analyzing unsymmetrically laminated beams since no other study on this topic is currently available. The beam element has ten degrees of p freedom at each of the two nodes: the axial displacement, the transverse deflection due to bending and shear, the twisting angle, the inplane shear rotation, and their derivatives along the axial direction. The formulation, solution procedure, and the computer program have been evaluated by solving a series of examples on the static response, free vibration, buckling, and nonlinear vibrations of isotropic and laminated beams. For unsymmetrically laminated beams, the nonlinear vibrations were found to have a soft spring behavior for certain boundary conditions as opposed to a hard spring behavior observed in isotropic and symmetrically laminated beams. The inplane boundary conditions were found to have a significant effect on nonlinear responses. / Master of Science